Bug ID
int64 961
1.91M
| Comment ID
int64 3.98k
17.1M
| Author
stringlengths 8
48
| Comment Text
stringlengths 1
64.3k
⌀ |
|---|---|---|---|
1,786,564 | 16,042,467 |
Created attachment 9291173
Before
In 105, the extension sidebar background is no longer styled according to the installed theme.
STR:
1. Install TreeStyleTab https://addons.mozilla.org/en-US/firefox/addon/tree-style-tab/
2. Go to its preferences, under Appearance select the "Proton" style
3. Install the official Firefox Alpenglow theme https://addons.mozilla.org/en-US/firefox/addon/firefox-alpenglow/
4. Open the TreeStyleTab sidebar
Expected results:
Themed background color, see _before_ image.
Actual results:
Solid light-grey background color, see _after_ image.
|
|
1,787,386 | 16,046,726 |
Created attachment 9291684
Arrow_DarkTheme.png
**[Affected versions]:**
- Firefox Nightly 106.0a1 (Build ID: 20220825222149)
**[Affected Platforms]:**
- macOS 12.5.1
- Windows 10 x64
- Ubuntu 20.04 x64
**[Prerequisites]:**
- Have the Firefox browser installed and opened.
- Have the “browser.aboutwelcome.templateMR” pref set to true.
- Have the dark theme of the browser enabled.
**[Steps to reproduce]:**
1. Navigate to the “about:welcome” page.
2. Observe the arrow from the “Skip this step” button.
**[Expected result]:**
- The arrow is correctly displayed and visible.
**[Actual result]:**
- The arrow is black and is hardly visible.
**[Notes]:**
- This issue is also reproducible on existing users’ onboarding.
- Based on [Figma](https://www.figma.com/file/2TrMkNefxzhMCXjUqM8KIy/MR1-2022-Onboarding?node-id=4855%3A118871), the arrow should have the same color as the “Skip this step” text.
- Attached is a screenshot of the issue.
|
|
1,787,783 | 16,050,865 |
Created attachment 9292036
rec of the issue.gif
**[Affected versions]:**
- Firefox Nightly 106.0a1 - Build ID: 20220829094551
**[Affected Platforms]:**
- Windows 10 x64
- macOS 11.6.5
- Linux Mint 20.2 x64
**[Prerequisites]:**
- Have the "browser.newtabpage.activity-stream.asrouter.devtoolsEnabled" pref set to "true" in the "about:config" page.
- Have a custom theme enabled from "about:addons>Themes".
**[Steps to reproduce]:**
1. Open the browser with the profile from the prerequisites.
2. Navigate to the "about:newtab#asrouter" page and select the "Targeting" section.
3. Search for the "userEnabledActiveColorway" parameter and observe the displayed value.
**[Expected result]:**
- The "userEnabledActiveColorway" parameter is set to "false".
**[Actual result]:**
- The "userEnabledActiveColorway" targeting parameter from the ASRouter page has no value.
**[Additional Notes]:**
- If a colorway theme is applied the "userEnabledActiveColorway" parameter is set to "true".
- If a browser default theme is applied the "userEnabledActiveColorway" parameter is set to "false".
- Attached a screen recording of the issue.
|
|
1,788,035 | 16,052,178 |
User Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:104.0) Gecko/20100101 Firefox/104.0
Steps to reproduce:
I opened a pdf and I selected the text with the mouse or with ctrl + a indifferently and copied and pasted with ctrl + c and ctrl + v
Actual results:
Only a small part of the text was copied.
Expected results:
It should have copied all the selected text as chrome does.
|
|
1,788,044 | 16,052,250 |
Created attachment 9292250
rec of the issue.gif
**[Affected versions]:**
- Firefox Nightly 106.0a1 - Build ID: 20220830092750
**[Affected Platforms]:**
- Windows 10 x64
- macOS 11.6.5
- Linux Mint 20.2 x64
**[Prerequisites]:**
- Have the following prefs in the "about:config" page:
- `extensions.getAddons.langpacks.url` set to `https://mock-amo-language-tools.glitch.me/?app=firefox&type=language&appversion=%VERSION%`
- `intl.multilingual.aboutWelcome.languageMismatchEnabled` set to `true`
- `intl.multilingual.aboutWelcome.systemLocaleOverride` set to `es-CO`
- `browser.aboutwelcome.templateMR` set to `true`
- `devtools.chrome.enabled` set to `true`
- Have the "Browser Console" opened.
**[Steps to reproduce]:**
1. Open the browser with the profile from the prerequisites.
2. Focus the "Browser Console" and enter the string from [here](https://docs.google.com/document/d/16ao8Y5oZe8HcOeaFmzYOjKmxqpvoPssPXpAWfzaL2nI/edit?usp=sharing).
3. Navigate to the "about:welcome" page and click the "Skip this step" button from the first screen.
4. Observe the secondary CTA button label from the "Downloading langpack" screen.
**[Expected result]:**
- The button is labeled "Skip this step".
**[Actual result]:**
- The button is labeled "Skip".
**[Additional Notes]:**
- O the rest of the Onboarding screens the secondary CTA button is labeled "Skip this step".
- Attached a screen recording of the issue.
|
|
1,788,901 | 16,057,718 |
User Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/104.0.5112.102 Safari/537.36 Edg/104.0.1293.70
Steps to reproduce:
We've recently migrated a profile with ~2700 folders from TB 91 to TB 102. After the panacea.dat->folderCache.json migration, only ~1300 folders had an MRUTime attribute, that is, missing attribute, not attribute null.
As a consequence operations on recent folders are painfully slow, about 8 seconds to bring up Recent Folders in the folder pane.
Bug 1778888 focused on the treatment of null values, but removing the hunks from bug 1726319 and bug 1776280 worsened the situation for missing attributes.
Accessing recent folders now will again query MSF files causing noticeable delays. Fortunately, due to the fix in bug 1787609, no MSF loss should occur.
As a workaround for affected users, the folderCache.json can be edited adding a null value for MRUTime everywhere. Observed behaviour is that that makes access to recent folder instantaneous again.
|
|
1,789,315 | 16,062,982 |
User Agent: Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/103.0.0.0 Safari/537.36
Steps to reproduce:
Built bookmarks/central from https://phabricator.services.mozilla.com/rMOZILLACENTRALab10436a5d1eeaf57f7b260b8844b2e34ee557e0 with the following MOZCONFIG options:
```
ac_add_options --enable-optimize --enable-debug --enable-record-tuple --disable-jit --enable-warnings-as-errors --enable-js-shell --with-ccache=sccache
```
Actual results:
The build fails:
```
netwerk/sctp/datachannel
In file included from Unified_cpp_js_src16.cpp:20:
/home/tjc/gecko-fork/js/src/vm/Iteration.cpp:257:16: error: use of undeclared identifier 'JS_StringToId'
if (!JS_StringToId(cx, key, &id)) {
^
/home/tjc/gecko-fork/js/src/vm/Iteration.cpp:275:18: error: use of undeclared identifier 'JS_IndexToId'; did you mean 'IndexToId'?
if (!JS_IndexToId(cx, i, &id)) {
^~~~~~~~~~~~
IndexToId
/home/tjc/gecko-fork/js/src/vm/JSAtom-inl.h:109:13: note: 'IndexToId' declared here
inline bool IndexToId(JSContext* cx, uint32_t index, MutableHandleId idp) {
^
2 errors generated.
```
Expected results:
I'll submit a patch to fix this (missing include), but I thought that builds with ENABLE_RECORD_TUPLE set were running on CI as per bug 1744967 -- or does it just build the JS shell and not the entire browser?
|
|
1,789,667 | 16,065,440 |
Created attachment 9293511
rec of the issue.gif
[Windows] The background images from the "about:welcome" page are no longer visible if the "High Contrast" mode is active
**[Notes]:**
- This issue is reproducible only on Windows.
- This issue is not reproducible on the existing users Onboarding modals.
**[Affected versions]:**
- Firefox Nightly 106.0a1 - Build ID: 20220907093209
**[Affected Platforms]:**
- Windows 10 x64
**[Prerequisites]:**
- Have the following prefs in the "about:config" page:
- `devtools.chrome.enabled` set to `true`
- Have the system's "High Contrast" mode active.
**[Steps to reproduce]:**
1. Open the browser with the profile from the prerequisites and navigate to the "about:welcome" page.
2. Navigate through the "Onboarding" screens and observe the background images.
**[Expected result]:**
- The background images are visible.
**[Actual result]:**
- The background images are NOT visible.
**[Additional Notes]:**
- Attached a screen recording of the issue.
|
|
1,789,876 | 16,066,998 |
**Precondition**
* The user is connected to sync on desktop and on iOS device;
**Affected versions**
* 106.0a1 (2022-09-08)
**Tested platforms**
* Affected platforms: windows 10, macOS 10.14.6
* Unaffected platforms: Android
**Steps to reproduce**
1. On the connected iOS device, access your Firefox account from the 3 dots menu;
2. Access the “Manage account & Devices” option;
3. From the Firefox account settings page, change the password;
4. The user is disconnected from the desktop sync, the iOS device is still connected to sync;
5. Reconnect to the desktop sync, with the new password;
6. Access the Firefox view tab;
7. Observe the “Tab pickup” option;
**Expected result**
* The 3 last active tabs are displayed in ”Tab pickup” section;
**Actual result**
* The “Sync Firefox on your phone or tablet” title is displayed and the user is not seen as connected on the mobile device;
**Additional notes**
* This is not reproducible using Android devices since the “Manage account & Devices” option is not available in sync
* Signing out from the mobile device and then Signing in again, the 3 last active tabs are displayed in ”Tab pickup” section;
* Connecting the iOS device first to sync and then with the same account connecting on desktop, the 3 last active tabs are **not** displayed in ”Tab pickup” section;
* Connecting an Android device first to sync and then with the same account connecting on desktop, the 3 last active tabs **are** displayed in ”Tab pickup” section;
|
|
1,790,309 | 16,069,993 |
User Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.15; rv:104.0) Gecko/20100101 Firefox/104.0
Steps to reproduce:
1. Ensure that your accent color in system preferences is not the default blue
2. Open a pdf (https://opensource.adobe.com/dc-acrobat-sdk-docs/pdfstandards/PDF32000_2008.pdf) and highlight some text
3. Observe the highlight being the default blue
Actual results:
The text is highlighted with a blue color
Expected results:
The highlight color should match the accent color in the same way it does in web pages.
|
|
1,790,311 | 16,070,007 |
Steps to reproduce:
Cross Site Tracing (XST) used to be an effective way to bypass local controls and gain access to authorization headers and cookies protected by httponly attributes. However, for many years the main browsers have implemented limits on fetch and XMLHttpRequest that stop the direct use of the TRACE method.
Out in the real world, some servers that implement CORS simply echo back the contents of the Access-Control-Request-Headers request header in the Access-Control-Allow-Headers response header. This means that it is possible to add an X-Http-Method-Override: TRACE header, which will bypass CORS protections.
The result being that if certain conditions are met (a server has lax CORS, implements the TRACE method and the X-Http-Method-Override header), then fetch and XMLHttpRequest can still be used to bypass CORS and deliver an XST attack.
This will allow any functional XSS on the site to gain access to authorization headers and cookies protected by httponly attributes.
An example of a server with CORS that echos back the Access-Control-Request-Headers header:
fetch( 'https://realtime.www.linkedin.com/realtime/connect', {
method: 'GET',
mode: 'cors',
headers: {
'x-http-method-override': 'TRACE'
}
} );
Actual results:
The combination of issues allows an XST attack to be delivered.
Expected results:
The TRACE method should be blocked in requests from fetch and XMLHttpRequest, whether it is in the main method parameter, or in common override headers ( 'x-http-method-override, x-http-method, x-http-override )
|
|
1,790,382 | 16,072,073 |
Per :ayeddi
- "When the `about:firefoxview` is opened, the screen reader focus jumps to the onboarding callout, which is expected (I assume), but when the screen reader completes reading its content, it proceeds to read the main page content from the middle - from the Colorways section and then to the Recently closed one. This will cause users missing the top of the page, incl. the synced tabs."
- "on MacOS the dialog title is not announced by VoiceOver - the code looks good, so this is likely more general bug?"
See `featureCallout.mjs` for feature callout code.
|
|
1,790,594 | 16,073,384 |
Created attachment 9294425
HCM_LanguageMismatch.png
**[Affected versions]:**
- Firefox Nightly 106.0a1 (Build ID: 20220913092817)
- Firefox Beta 105.0 RC (Build ID: 20220912145803)
**[Affected Platforms]:**
- macOS 12.5.1
- Windows 10 x64
- Ubuntu 20.04 x64
**[Prerequisites]:**
- Have the latest version of the Firefox Nightly build installed/extracted and opened.
- Have the following prefs:
- "intl.multilingual.downloadEnabled" set to true;
- "intl.multilingual.enabled" set to true;
- "intl.multilingual.liveReload" set to true;
- "intl.multilingual.aboutWelcome.languageMismatchEnabled" set to true;
- "extensions.getAddons.langpacks.url" set to "https://mock-amo-language-tools.glitch.me/?app=firefox&type=language&appversion=%VERSION%" value;
- Have the High Contrast Mode enabled.
**[Steps to reproduce]:**
1. Restart the browser and navigate to the “about:welcome” page.
2. Click the “Skip this step” button from the first screen.
3. Hover the “Continue in [downloading browser language]" button.
4. Observe the button.
**[Expected result]:**
- The button label is visible.
**[Actual result]:**
- The button label is not visible at all.
**[Notes]:**
- Attached is a screenshot of the issue.
|
|
1,793,401 | 16,096,697 |
Created attachment 9296953
NewUsers_HCM_Windows.png
**[Affected versions]:**
- Firefox Nightly 107.0a1 (Build ID: 20221002212226)
**[Affected Platforms]:**
- Windows 10 x64
**[Prerequisites]:**
- Have the latest version of the Firefox Nightly build installed and opened.
- Have the system "High Contrast Mode" enabled.
**[Steps to reproduce]:**
1. Open the browser with the profile from the prerequisites.
2. Navigate to the "about:welcome" page.
3. Navigate through the "Onboarding" screens and observe the images' backgrounds.
**[Expected result]:**
- The background of the images is visible.
**[Actual result]:**
- Only the image assets are visible.
**[Additional notes]:**
- Attached is a screenshot of the issue.
|
|
1,793,453 | 16,097,371 |
Created attachment 9297001
Pressing F6 on an about: page
When the CFR panel is open, clicking F6 results in different behavior on `about:welcome` and `about:newtab` pages than on other web pages (see attached screen grabs). On `about:newtab` and `about:welcome` pages, it hides and reveals the CFR panel. For other pages, it navigates the user between the URL bar and the CFR calls to action.
According to [The Blind Access Journal](https://blindaccessjournal.com/2021/06/f6-is-your-friend/), `In the Google Chrome and Mozilla Firefox web browsers, F6 jumps out of the address bar and moves focus directly into the currently loaded web page.`
Behavior should be consistent across all contexts. The current behavior on `about:newtab` and `about:welcome` pages doesn't allow the user to access the CFR _or_ the page content.
We've found the same issue occurs when the URL has been altered by the user in the URL bar (and it's focused).
|
|
1,793,754 | 16,099,560 |
Created attachment 9297230
try_again_00.gif
**Found in**
* 106.0b8 (20221004185850)
**Affected versions**
* 106.0b8 (20221004185850)
* 107.0a1 (20220929014928)
**Tested platforms**
* Affected platforms: Windows 10x64, Ubuntu 20.04, macOS 11.6
* Unaffected platform: none
**Steps to reproduce**
1. Sign in to FxA account inside Firefox.
2. Open the Firefox App Menu and select the connected FxA Account > Manage Account.
3. Sign out the device that was connected.
4. Return to Firefox View and observe the Tab Pickup area.
**Expected result**
* Not sure about the correct expected result here but I think that the `Switch seamlessly between devices` message should be displayed asking the user to connect to Sync.
**Actual result**
* `Switch seamlessly between devices` message is displayed for a split second and then the `We’re having trouble syncing` error is displayed. Clicking the Try again button does nothing.
**Regression range**
* This started with bug 1768695 when the sync error states were introduced. Also, before bug 1784969 there can be seen a page refresh.
**Additional notes**
* Attached a screen recording.
* Restarting the browser after step4 shows the `Switch seamlessly between devices` message.
|
|
1,793,905 | 16,100,663 |
User Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/105.0.0.0 Safari/537.36 Edg/105.0.1343.42
Steps to reproduce:
This bug is in reference to bug: https://bugzilla.mozilla.org/show_bug.cgi?id=1793742
We enabled the following settings via GPO/Preferences:
"dom.security_https_only_mode": { "Value": true, "Status": "locked" },
"dom.security.https_only_mode_ever_enabled": { "Value": true, "Status": "locked" }
Actual results:
The settings were enabled, but not locked (user is still able to disable them).
Expected results:
User will not be able to change https-only-mode.
|
|
1,794,081 | 16,101,833 |
Created attachment 9297536
Firefox View CFR.mp4
**[Notes]:**
- The "IMPRESSION" telemetry ping for the "Firefox View" CFR message is displayed in the "Browser Console" even if the CFR message is not displayed.
- The “DISMISS” telemetry ping is not displayed in the "Browser Console".
- I have managed to reproduce this issue only on new tabs.
**[Affected versions]:**
- Firefox Beta 106.0b8 - Build ID: 20221004185850
- Firefox Nightly 107.0a1 - Build ID: 20221006093717
**[Affected Platforms]:**
- Windows 10 x64
- Windows 7 x64
- macOS 12.6
- Linux Mint 20.2
**[Prerequisites]:**
- Have a Firefox profile older than 2 days.
**[Steps to reproduce]:**
1. Open the browser from the prerequisites.
2. Open two new tabs and focus the "Address Bar" on the last opened tab.
3. Close the unfocused tab and open a new one.
4. Repeat the previous step three times.
5. Observe the behavior.
**[Expected result]:**
- The "Firefox View" CFR message was displayed after the third tab was closed.
**[Actual result]:**
- The "Firefox View" CFR message is not displayed at all and the "IMPRESSION" telemetry ping is displayed in the "Browser Console".
**[Additional Notes]:**
- If you close regular tabs instead the CFR message appears after the “Address Bar” is unfocused.
- This issue is not reproducible if multiple tabs are closed via the option from the context menu.
- Attached is a screen recording of the issue:
|
|
1,794,403 | 16,105,535 |
Created attachment 9297825
PDF_form_00.png
**Found in**
* 107.0a1 (20221010033207)
**Affected versions**
* 107.0a1 (20221010033207)
**Tested platforms**
* Affected platforms: Windows 10x64, Ubuntu 20.04
* Unaffected platform: macOS 11.6
**Steps to reproduce**
1. Open http://foersom.com/net/HowTo/data/OoPdfFormExample.pdf.
2. Click on The Country dropdown and choose a value.
**Expected result**
* The text is properly displayed for each value.
**Actual result**
* Some values are slightly cut off.
**Regression range**
* Last good revision: 23c9a4c9b468b43106f50295a52422851d71dfda
First bad revision: d9035fae6d3c0fd5bb6253561210fa05237c68e0
Pushlog:
https://hg.mozilla.org/integration/autoland/pushloghtml?fromchange=23c9a4c9b468b43106f50295a52422851d71dfda&tochange=d9035fae6d3c0fd5bb6253561210fa05237c68e0
Potential regressor: bug 1792052.
**Additional notes**
* Attached a screenshot with the issue.
* I cannot reproduce the issue on macOS 11.5. On Windows 10x64 the issue is not so visible. The issue is more visible on Ubuntu 20.04.
|
|
1,794,717 | 16,107,893 |
Steps to reproduce the problem:
1. Choose the text annotation tool on PDF
2. Create a text box
3. Mouse cursor switches to a regular pointer instead of a caret which makes placing subsequent text boxes in the right location difficult.
|
|
1,794,956 | 16,109,553 |
Created attachment 9298334
test-case.html
User Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.15; rv:105.0) Gecko/20100101 Firefox/105.0
Steps to reproduce:
https://jsfiddle.net/am16Lhk4/
1. Set some properties to an OffscreenCanvas's 2D context, like `fillStyle`, or changing its CTM.
2. Call the OffscreenCanvas's `transferToImageBitmap()` method.
3. Perform new drawings without resetting the properties set at step 1.
Actual results:
All the properties set at step 1 where reset to their default.
Expected results:
[Per specs](https://html.spec.whatwg.org/multipage/canvas.html#dom-offscreencanvas-transfertoimagebitmap) only the bitmap buffer is supposed to get cleared by this operation, the context's properties are supposed to be preserved.
|
|
1,795,037 | 16,110,222 |
I gave Firefox View a try on my personal Nightly a week or two ago and didn't find it fit my workflow, so I turned it off using the UI. Since then, I've received multiple "You can find your recently-closed tabs in Firefox View [try it] [nah]" doorhangers after closing tabs.
Frequency: not every day, but definitely more than once a week. Perhaps once per build?
Is this UI operating as designed+desired and I'm a weirdo hitting a corner case? Or is this not desired behaviour (and you'd like some logs, environment info)?
|
|
1,795,105 | 16,110,671 |
From github: https://github.com/mozilla-mobile/android-components/issues/12892.
> Recently I've been doing some work to reduce the amount of memory used by webrender / graphics code (see bz-1780376). This was prompted by noticing that I could trivially crash Fenix on low-end devices by opening a couple dozen or so tabs, and I'm hoping it may help with issues such as mozilla-mobile/fenix#12731 as well. This has shipped in 105, and the crashing situation seems much improved. I suspect it may _help_ with the background content processes being killed, but there are probably many other factors at play there too...
>
> Anyway, whilst profiling memory usage I found that a huge amount of memory is used by android-component's tab thumbnails. See steps to reproduce and the screenshot below.
>
> I note that we do have code to trim thumbnail memory usage in response to memory pressure [here](https://searchfox.org/mozilla-mobile/source/android-components/components/browser/state/src/main/java/mozilla/components/browser/state/reducer/SystemReducer.kt), but in my experience the memory pressure event is signalled too late to prevent background processes being killed. It would be better to avoid using too much memory in the first place, where possible.
>
> cc @csadilek @jonalmeida
>
> ## Steps to reproduce
>
> 1. Build a `debuggable` (or `profileable`) build of fenix. Eg `gradlew installNightly` but with a `debuggable` line in `local.properties`
> 2. Open Android Studio's `Profiler` tab
> 3. Open Fenix on the device
> 4. Click the `+` icon next to "Sessions" in the left-hand pane, then select the parent fenix process. (eg "org.mozilla.fenix", not the ":gpu" or ":tab" processes)
> 5. Open eg ten tabs, waiting for the page to load each time. Watch the memory usage graph go up
> 6. Select "Capture heap dump" and click "Record"
> 7. Notice 100MB of `Bitmap` objects
>
> ### Expected behavior
>
> Memory usage remains at an acceptable amount.
>
> ### Actual behavior
>
> 
>
> We can see that there are a whole bunch of `Bitmap` instances, each taking up just under 10MB. I think references to these are owned by both `TabSessionState` and `ContentState` hence why those objects appear in the list too.
>
> ### Device information
>
> * Android device: Any, but I'm using a Pixel 6A
> * App (fenix, focus, ...): Fenix
> * App version: 107
>
> ┆Issue is synchronized with this [Jira Task](https://mozilla-hub.atlassian.net/browse/FNXV2-21734)
>
Change performed by the [Move to Bugzilla add-on](https://addons.mozilla.org/en-US/firefox/addon/move-to-bugzilla/).
|
|
1,795,704 | 16,116,937 |
From github: https://github.com/mozilla-mobile/fenix/issues/27298.
> ### Steps to reproduce
>
> 1. Set the language of the device or Firefox to something other than English.
> 2. Add a multilingual add-on (e.g. uBlock Origin).
> 3. View the add-on's settings screen.
>
> ### Expected behaviour
>
> If the add-on supports the selected language, it will be displayed in the selected language.
>
> ### Actual behaviour
>
> Add-on settings are only displayed in English, no matter what language it is set to.
> 
>
>
> ### Device name
>
> Sharp AQUOS R3
>
> ### Android version
>
> Android 11
>
> ### Firefox release type
>
> Firefox Nightly
>
> ### Firefox version
>
> 107.0a1 (Build #2015908203)
>
> ### Device logs
>
> _No response_
>
> ### Additional information
>
> A same issue has occurred pieviously in Fennec F-Droid.
> In both cases, Application Settings in about:support page is only en-US.
> 
>
> ┆Issue is synchronized with this [Jira Task](https://mozilla-hub.atlassian.net/browse/FNXV2-21736)
>
Change performed by the [Move to Bugzilla add-on](https://addons.mozilla.org/en-US/firefox/addon/move-to-bugzilla/).
|
|
1,782,561 | 16,012,115 |
Created attachment 9287982
ffdesync.js
Steps to reproduce:
We discovered that cookie jar overflows cause a desynchronization between the cookies listed by document.cookie and the actual cookie jar. We created a demo application to reproduce the inconsistencies. The demo is available at https://cookiedesync.minimal.blue/, full nodejs sources are attached to this report (execute with node ffdesync.js).
While creating this report we noticed a related bug at https://bugzilla.mozilla.org/show_bug.cgi?id=1761314. We decided to open a new issue and mark it as security-related due to the potential harm that this bug could cause: the reported issue allows same-origin and same-site attackers (i.e., an attacker controlling a related-domain of the target website who can set domain cookies) to fixate arbitrary cookies in document.cookie that will survive deletion attempts from the server, e.g., via the Clear-Site-Data HTTP header. This inconsistent state could introduce vulnerabilities in applications trusting the cookies read from document.cookie. Notice, for instance, that frontends often set custom HTTP headers using the values of specific cookies read via the document.cookie API. Examples are [ASP.NET](https://docs.microsoft.com/en-us/aspnet/core/security/anti-request-forgery?view=aspnetcore-6.0#javascript-1) and [Angular](https://angular.io/guide/http#security-xsrf-protection).
The issue can be manually reproduced following the steps below, please check our demo for the full set of tests.
1. Visit https://example.com/, set 181 cookies via document.cookie using the devetools Console:
for(let i=0; i<181; i++) document.cookie = `a${i}=_`;
2. Trigger a a subsequent request (ctrl-shift-r), observe the cookies attached to the request via the devtools Network tab
3. In the devtools Console, read the cookies via document.cookie
4. Go to the devtools Storage tab in the browser inspector, delete all cookies
5. In the devtools Console, read the cookies via document.cookie
Actual results:
1. Cookies are set
2. Due to the overflow, only 151 cookies a30 ... a180 (included) are attached to the HTTP request. The Storage > Cookies panel is consistently showing only these 151 cookies.
3. document.cookie shows 181 cookies, from a0 ... a180, not reflecting the actual content of the cookie jar.
4. Cookies are deleted
5. document.cookie now shows only the first 30 cookies, a0 ... a29. Notice that HTTP requests at this point do not attach any cookie. Forced page reloads do not change the output of document.cookie
Expected results:
The output of document.cookie should be consistent with the state of the cookie jar (and, consequently, with the cookies attached to HTTP requests).
|
|
1,782,564 | 16,012,139 |
User Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:103.0) Gecko/20100101 Firefox/103.0
Steps to reproduce:
I open this file: https://www.betting-und-buss.de/download/Reparatur_Auftrag.pdf
Actual results:
I can fill nearly the hole formular. But the field 'Beanstandung' is not possible to edit.
|
|
1,783,181 | 16,016,528 |
Steps to reproduce:
I ran the following script:
const length = 5;
const sampleRate = 44100;
const buffer = new AudioBuffer({ length, numberOfChannels: 2, sampleRate });
buffer.copyToChannel(new Float32Array([1, 0.5, 0, -0.5, -1]), 0);
buffer.copyToChannel(new Float32Array([-0.5, -0.25, 0, 0.25, 0.5]), 1);
const offlineAudioContext = new OfflineAudioContext({ length, sampleRate });
const audioBufferSourceNode = new AudioBufferSourceNode(offlineAudioContext, { buffer });
const stereoPannerNode = new StereoPannerNode(offlineAudioContext, { channelCount: 2 });
audioBufferSourceNode.connect(stereoPannerNode).connect(offlineAudioContext.destination);
stereoPannerNode.pan.setValueAtTime(0.5, 0);
stereoPannerNode.pan.setValueAtTime(0, 2 / offlineAudioContext.sampleRate);
stereoPannerNode.pan.linearRampToValueAtTime(1, 5 / offlineAudioContext.sampleRate);
stereoPannerNode.pan.cancelScheduledValues(3 / offlineAudioContext.sampleRate);
audioBufferSourceNode.start(0);
offlineAudioContext.startRendering().then((renderedBuffer) => {
const channelData = renderedBuffer.getChannelData(0);
console.log(Array.from(channelData));
});
Actual results:
It logs [ 0.4571067690849304, 0.2285533845424652, 0, -0.125, -0.0732232928276062 ] when running it on an M1 Mac. The last number seems to be wrong.
Expected results:
It should log [ 0.4571067690849304, 0.2285533845424652, 0, -0.125, -0.25].
It does work on an older Intel Mac in Firefox v103. It also works in Chrome or Safari on any Apple processor architecture.
|
|
1,783,542 | 16,018,310 |
Created attachment 9288718
Screencap
From Github https://github.com/mozilla-mobile/fenix/issues/26327:
> ### Steps to reproduce
>
>This happens pretty randomly so I don't have exact steps other than browsing some sites. >Although I can reproduce it consistently while searching for issues on Github (see attached >video and screenshot)
>
>It started with the nightly builds published on Google Playstore on August 3-4.
>
>I think it is coming from the space used by the keyboard app which is the part that glitches out.
>
>See attached video for an example
>
>### Expected behaviour
>
>There shouldn't be a glitch
>
>### Actual behaviour
>
>The app glitches.
>
>### Device name
>
>Nokia C03
>
>### Android version
>
>Android 10
>
>### Firefox release type
>
>Firefox Nightly
>
>### Firefox version
>
>105.0a1 20220804094607
>
>### Device logs
>
>_No response_
>
>### Additional information
>
>about:support: [aboutsupport.txt](https://github.com/mozilla-mobile/fenix/files/9267125/aboutsupport.txt)
|
|
1,784,085 | 16,023,833 |
Created attachment 9289262
Screencast 2022-08-10 17_12_14.mp4
*Affected versions:*
- Nightly 105.0a1 (2022-08-09);
*Affected platforms:*
- Windows 10 64-bit;
- Linux Ubuntu 20.04;
*Preconditions:*
- The user is already logged in to the browser with a Fx account.
- The user has gone through the connect another device flow and has few tabs opened.
- The user has activated Firefox View;
*Steps to reproduce:*
1. Be in Firefox View while disconnecting the device from the internet;
2. Click the Fx account icon and “Sync now”;
3. Reconnect to the internet;
4. Click the “Try again” button from the Tab pickup section or the “Sync now” option from browser menu;
*Expected results:*
The open pages from other devices are loaded.
*Actual results:*
Nothing happens. The “We’re having trouble syncing. Nightly can’t reach the service right now. Try again in a few moments.” messages together with the “Try again” button are still displayed.
*Notes:*
- The issue is not reproducing without step 2.
- User is blocked in without seeing any opened tabs and can unblock if restarts the browser.
|
|
1,784,260 | 16,024,940 |
**[Affected versions]:**
- Firefox Nightly 105.0a1 (Build ID: 20220810212956)
**[Affected Platforms]:**
- macOS 12.3.1
- Windows 10 x64
- Ubuntu 20.04 x64
**[Prerequisites]:**
- Have the Firefox browser installed and opened.
- Have the “browser.aboutwelcome.templateMR” pref set to true.
- Have a mobile device with a QR reader app installed.
**[Steps to reproduce]:**
1. Restart the browser and navigate to the “about:welcome” page.
2. Click the “Skip this step” button until the Mobile Cross-Promo screen is displayed.
3. Scan the QR code using the mobile device from prerequisites.
4. Observe the opened links.
**[Expected result]:**
- The “https://mzl.la/3zqmOAO” link is opened in the browser.
**[Actual result]:**
- The “https://mzl.la/3RWN0KP” link is opened in the browser.
**[Notes]:**
- This issue is also reproducible for existing users.
- Based on the [Spec](https://docs.google.com/document/d/1Zlo9PNfjIbJb2hHlRwSaHqL17y_Z_qZZGxrwA6z94g8/edit#) doc, the links should be reversed.
|
|
1,784,850 | 16,030,254 |
Steps to reproduce:
Using Nightly 105.0a1, 20220814095107, having a PDF open in a viewer on a 1366x768 laptop system, suspend the system, attach a HiDPI monitor, and resume.
Actual results:
The font of the PDF was fuzzy.
Expected results:
The font of the PDF should be sharp. Reloading the page, the screen is sharp.
|
|
1,785,218 | 16,033,366 |
User Agent: Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/103.0.5060.134 Safari/537.36
Steps to reproduce:
https://thformular.thueringen.de/thueform/cfs/eject/pdf/4300.pdf?MANDANTID=18&FORMUID=GRST-001-DE-FL
Actual results:
We are the producer of this pdf forms. The user can only fill the first page with your browser.
On Edge, Adobe, Chrome(ium) and Safari the fields on page 2 are visible and can be filled.
Expected results:
The fields on page two should be available and fillable.
|
|
1,791,352 | 16,079,834 |
```js
function f(x, y, m) {
let w = Math.log(y) && Math.min(Math.max(x / x, x), x);
if (m) {
print(w);
}
}
let z = [1, 0, 2];
for (let i = 0; i < 9; i++) {
f(1, z[i], false);
}
f(0, 1, false);
f(0, 0, false);
f(0, 2, true);
```
Output (rev 95d4708dc680, m-c latest default tip):
```
$ ./js --fuzzing-safe --differential-testing --no-threads --ion-eager testcase.js
0
$ ./js --fuzzing-safe --differential-testing --no-threads --baseline-eager --no-ion testcase.js
NaN
```
Output (rev 8b53f7a9dea2, parent of potential regressor):
```
$ ./js --fuzzing-safe --differential-testing --no-threads --ion-eager testcase.js
NaN
$ ./js --fuzzing-safe --differential-testing --no-threads --baseline-eager --no-ion testcase.js
NaN
```
```
The first bad revision is:
changeset: https://hg.mozilla.org/mozilla-central/rev/bb791e77a175
user: André Bargull
date: Mon Aug 08 17:56:26 2022 +0000
summary: Bug 1782771 - Part 5: Fold min(x, min(x, y)) and max(x, max(x, y)). r=jandem
```
Compile with `AR=ar sh ./configure --enable-debug --with-ccache --enable-gczeal --enable-debug-symbols --disable-tests`, tested on m-c rev 95d4708dc680.
Setting s-s to be safe. Andre, is bug 1782771 a likely regressor?
|
|
1,791,515 | 16,081,032 |
Created attachment 9295383
Screenshot 2022-09-20 at 10.37.05.png
**Note**
*If you try to move the drawing area ar text area on the fields, a value will be pasting.(please see recording)
**Found in**
* Fx 106.0b1
**Affected versions**
* Fx 106.0b1
**Tested platforms**
* Affected platforms: Windows 10, MacOS 11.6, Ubuntu 22
* Unaffected platforms: -
**Steps to reproduce**
1. Open Firefox
2. Access the following pdf form: https://www.irs.gov/pub/irs-pdf/f1040.pdf
3. Click on Draw/Text button
4. Try to draw or fill in on the fields
**Expected result**
* You should be able to draw and fill in on the fields of the pdf.
**Actual result**
* You are not able to draw and fill in on the fields of the pdf. Please see the recording.
|
|
1,791,520 | 16,081,084 |
The following sample will cause a nullptr dereference in Spidermonkey built from current HEAD:
```
function main() {
const v19 = async (v20,v21,v22) => {
};
for (let v26 = 0; v26 < 100; v26++) {
const v27 = v19();
const v28 = v27.finally();
for (const v30 in this) {
const v33 = "function".trimLeft(-992713.7098905144,"function");
const v34 = v33.startsWith("function");
const v37 = this.BigInt64Array;
const v38 = new v37(2);
const v39 = v38[1];
continue;
}
}
gc();
}
main();
// CRASH INFO
// ==========
// TERMSIG: 11
// STDERR:
```
Here is the stacktrace from gdb:
```
Thread 1 "js" received signal SIGSEGV, Segmentation fault.
0x0000555558176292 in IsWriteableAddress (ptr=0x0) at js/src/gc/Nursery.cpp:771
771 *vPtr = *vPtr;
(gdb) bt
#0 0x0000555558176292 in IsWriteableAddress (ptr=0x0) at js/src/gc/Nursery.cpp:771
#1 0x0000555558176175 in js::Nursery::forwardBufferPointer (this=0x7ffff771b138, pSlotsElems=0x7fffffff9aa8) at js/src/gc/Nursery.cpp:800
#2 0x000055555887cc82 in js::jit::UpdateIonJSFrameForMinorGC (rt=0x7ffff7718000, frame=...) at js/src/jit/JitFrames.cpp:1055
#3 0x000055555887ca1a in js::jit::UpdateJitActivationsForMinorGC (rt=0x7ffff7718000) at js/src/jit/JitFrames.cpp:1392
#4 0x0000555558177ad4 in js::Nursery::doCollection (this=0x7ffff771b138, reason=JS::GCReason::OUT_OF_NURSERY) at js/src/gc/Nursery.cpp:1265
#5 0x00005555581772df in js::Nursery::collect (this=0x7ffff771b138, options=JS::GCOptions::Normal, reason=JS::GCReason::OUT_OF_NURSERY) at js/src/gc/Nursery.cpp:1114
#6 0x00005555581020d5 in js::gc::GCRuntime::collectNursery (this=0x7ffff7718768, options=JS::GCOptions::Normal, reason=JS::GCReason::OUT_OF_NURSERY, phase=js::gcstats::PhaseKind::MINOR_GC) at js/src/gc/GC.cpp:4268
#7 0x00005555581056e3 in js::gc::GCRuntime::minorGC (this=0x7ffff7718768, reason=JS::GCReason::OUT_OF_NURSERY, phase=js::gcstats::PhaseKind::MINOR_GC) at js/src/gc/GC.cpp:4240
#8 0x000055555811151a in js::gc::GCRuntime::tryNewNurseryBigInt<(js::AllowGC)1> (this=0x7ffff7718768, cx=0x7ffff772a100, thingSize=16, kind=js::gc::AllocKind::BIGINT) at js/src/gc/Allocator.cpp:263
#9 0x0000555558111241 in js::gc::detail::AllocateBigInt<(js::AllowGC)1> (cx=0x7ffff772a100, heap=js::gc::DefaultHeap) at js/src/gc/Allocator.cpp:302
#10 0x000055555766ca10 in js::gc::CellAllocator::NewCell<JS::BigInt, (js::AllowGC)1, js::gc::InitialHeap&> (cx=0x7ffff772a100, args=@0x7fffffff99a7: js::gc::DefaultHeap) at js/src/gc/Allocator.h:120
#11 0x000055555766431d in JSContext::newCell<JS::BigInt, (js::AllowGC)1, js::gc::InitialHeap&> (this=0x7ffff772a100, args=@0x7fffffff99a7: js::gc::DefaultHeap) at js/src/vm/JSContext.h:267
#12 0x000055555764fa7e in JS::BigInt::createUninitialized (cx=0x7ffff772a100, digitLength=0, isNegative=false, heap=js::gc::DefaultHeap) at js/src/vm/BigIntType.cpp:146
#13 0x00005555576500f5 in JS::BigInt::zero (cx=0x7ffff772a100, heap=js::gc::DefaultHeap) at js/src/vm/BigIntType.cpp:212
#14 0x0000555557655c56 in JS::BigInt::createFromUint64 (cx=0x7ffff772a100, n=0) at js/src/vm/BigIntType.cpp:1755
#15 0x0000555557655cbb in JS::BigInt::createFromInt64 (cx=0x7ffff772a100, n=0) at js/src/vm/BigIntType.cpp:1780
#16 0x0000555558478d2d in js::jit::CreateBigIntFromInt64 (cx=0x7ffff772a100, i64=0) at js/src/jit/VMFunctions.cpp:2141
```
I don't believe this issue has security impact, but am marking it as a security issue as a precaution.
|
|
1,791,788 | 16,083,297 |
Steps to reproduce:
Install the Video DownloadHelper extension (https://addons.mozilla.org/firefox/addon/video-downloadhelper/) and its companion app (https://www.downloadhelper.net/install-coapp) on Windows 10.
Actual results:
On Firefox 105 and previous, Video DownloadHelper reports companion app installed.
On Firefox 106.0b2 (and 106.0b1 as reported by several extension users) it shows as uninstalled.
Expected results:
The native messaging companion application should report as installed whatever version of Firefox > 57.
This issue does not happen on Firefox 106 for Linux.
Note that Video DownloadHelper has 2 million active users, most of them running Windows and the companion is required for most video download operations.
|
|
1,792,014 | 16,084,626 |
Created attachment 9295782
rec of the issue.gif
**[Notes]:**
- Also, the "migrate_close" telemetry ping is not displayed in the browser console after the Import wizard is closed. (E.G. `TELEMETRY PING (about:welcome): {"experiments":{"task-continuity-sync-after-tab-change-rollout-40":{"branch":"sync-after-tab"},"tcp-rollout-beta-phase-iii-tcp-on-by-default-for-remaining-beta-profiles":{"branch":"control"}},"locale":"fr","version":"106.0","release_channel":"beta","event":"CLICK_BUTTON","event_context":"{\"source\":\"migrate_close\",\"page\":\"spotlight\"}","message_id":"FX_MR_106_UPGRADE_1_UPGRADE_IMPORT_SETTINGS","addon_version":"20220920185943","client_id":"3b147754-c319-411f-99a6-5f4c844642bd","browser_session_id":"c9c94d11-ed96-4c45-8765-7f4f144852aa"})` .
- I have not managed to reproduce this issue on macOS 11.7 and macOS 12.5.1.
**[Affected versions]:**
- Firefox Beta 106.0b2 - Build ID: 20220920185943
- Firefox Nightly 107.0a1 - Build ID: 20220921214338
**[Affected Platforms]:**
- Windows 11 x64
- Windows 10 x64
- Windows 7 x64
- Linux Mint 20.2
**[Prerequisites]:**
- Have a Firefox version older than 106 installed.
**[Steps to reproduce]:**
1. Open the browser from the prerequisites.
2. Update the browser to the latest version.
3. Navigate through the Onboarding flow until the "Import Data" modal is displayed.
4. Click the "Import from previous browser" button and complete the import flow.
5. Observe the behavior.
**[Expected result]:**
- The "Import Data" modal is dismissed and the next modal is displayed.
**[Actual result]:**
- The "Import Data" modal is still displayed.
**[Additional Notes]:**
- The "Import Data" modal is dismissed and the next modal is displayed if a second import action is performed.
- This issue is not reproducible on the "about:welcome" page.
- Attached is a screen recording of the issue:
|
|
1,792,078 | 16,085,116 |
Steps to reproduce:
After moving to the new recent 100.x series (I think it was then), Thunderbird switched to some newfangled rich text editor for calendar events. Every time I save an event and come back, newlines are added all over the place leaving everything a confusing mess.
I type some nice paragraph such as the following (not the exact text):
blah blah foo bar (and more foo bar); this is foo; this is bar
foo and bar
bar and foo
Actual results:
Thunderbird decided to randomly add line breaks, such as:
blah blah foo
bar (and more foo bar); this
is foo; this is bar
foo and
bar
bar and foo
Currently I have no way to predict where Thunderbird is going to throw in line breaks.
Expected results:
I should have got back:
blah blah foo bar (and more foo bar); this is foo; this is bar
foo and bar
bar and foo
(We're not even talking about rich formatting!!)
Why does everyone try to re-implement a word process and break things royally? If you'd just provide plain text and not screw up the lines, that would be 10 times better than this.
Or you could allow us to enter raw Markdown and then show it rendered in the view (using a common library), like Stack Overflow does.
Why oh why are you adding line breaks all over the place every time I edit a calendar event?
|
|
1,792,422 | 16,089,348 |
Created attachment 9296201
PDF_tools.gif
**Found in**
* 106.0b4 (20220925185751)
**Affected versions**
* 106.0b4 (20220925185751)
* 107.0a1 (20220925213821)
**Tested platforms**
* Affected platforms: Windows 10x64, macOS 11, Ubuntu 21.04, Windows 7x64
* Unnafected platforms: none
**Steps to reproduce**
1. Open a random PDF (http://www.pdf995.com/samples/pdf.pdf) and select the Draw tool from the PDF toolbar.
2. Open Tools menu and select Hand Tool.
3. Drag the page up and down.
**Expected result**
* No lines are drawn when dragging the page.
**Actual result**
* Dragging the page will also draw lines on the PDF page.
**Regression range**
* Will search for one ASAP if there is one.
**Additional notes**
* Attached a screen recording.
|
|
1,792,546 | 16,090,304 |
**Found in**
* Firefox 106.0b4
**Affected versions**
* Firefox 106.0b4
* Firefox 107.0a1
**Tested platforms**
* Affected platforms: Win 7, Win 10, Win 11
* Unaffected platforms: macOS, Ubuntu
**Steps to reproduce**
1. Install Firefox and open the installation folder.
2. Scroll down to the Private Window icon and pin it to the Taskbar/Start Menu.
3. Hover the icon. (or observe the title)
**Expected result**
* The title is “Firefox Private Browsing”.
**Actual result**
* The title is “private_browsing”.
**Regression range**
* If there is one I will try to find it as soon as possible.
**Additional notes**
* If there are multiple pins to the Start Menu, then multiple “private_browsing(2)” icons are displayed on the Start Menu.
* This issue affects locale builds as well.
|
|
1,792,547 | 16,090,305 |
Created attachment 9296334
normal icon.gif
**Found in**
* Firefox 106.0b4
**Affected versions**
* Firefox 106.0b4
**Tested platforms**
* Affected platforms: Windows 10, Windows 7, Windows 11
* Unaffected platforms: macOS, Ubuntu
**Steps to reproduce**
1. Install a x32 build for Firefox and open it.
2. Open a Private Window and pin it to the taskbar.
3. Uninstall Firefox and install it again without deleting anything.
4. Open a Private Window and pin it to the taskbar.
**Expected result**
* The Private Window icon is displayed on the taskbar.
**Actual result**
* A Normal Window icon is displayed on the taskbar.
**Regression range**
* If there is one I will try to find it as soon as possible.
**Additional notes**
* If the icon is clicked then a Normal Window is opened.
* On Windows 7 the issue is reproducing on x64 builds and on Windows 10 the bug is intermittent on x64 builds.
|
|
1,792,550 | 16,090,335 |
**Found in**
* Firefox 106.0b4
**Affected versions**
* Firefox 106.0b4
* Latest Nightly 107.0a1
**Tested platforms**
* Affected platforms: macOS 11.6, Windows 10, Ubuntu 18.04
**Steps to reproduce**
1. Set up a primary password
2. Restart Firefox
3. Login to FxA
4. Wait a bit and dismiss the primary password panel
5. Go to Firefox View
6. Click the Try Again button
**Expected result**
* The primary password panel appears again.
**Actual result**
* Nothing happens.
**Regression range**
* This behaviour is from the time the Try Again button was implemented in Bug 1787619 so it is not a regression.
**Additional notes**
* Here is the output from the console
```
Sync encountered an error - see about:sync-log for the log file. policies.js:975
1664277278438 FirefoxView.TabsSetup DEBUG tryToClearError: triggering new tab sync
1664277278438 FirefoxView.TabsSetup DEBUG Handling UIState update
1664277278439 FirefoxView.TabsSetup DEBUG maybeUpdateUI, conditions not met to exit state: : error-state
1664277278439 FirefoxView.TabsSetup DEBUG maybeUpdateUI, will notify update?:: true
1664277278439 FirefoxView.TabsSetup DEBUG maybeUpdateUI, in error state:: sync-error
1664277279060 FirefoxView.TabsSetup DEBUG Handling weave:service:sync:finish
1664277279060 FirefoxView.TabsSetup DEBUG maybeUpdateUI, conditions not met to exit state: : error-state
1664277279060 FirefoxView.TabsSetup DEBUG maybeUpdateUI, will notify update?:: true
1664277279061 FirefoxView.TabsSetup DEBUG maybeUpdateUI, in error state:: sync-error
1664277279062 FirefoxView.TabsSetup DEBUG Handling UIState update
1664277279062 FirefoxView.TabsSetup DEBUG maybeUpdateUI, conditions not met to exit state: : error-state
1664277279062 FirefoxView.TabsSetup DEBUG maybeUpdateUI, will notify update?:: true
1664277279062 FirefoxView.TabsSetup DEBUG maybeUpdateUI, in error state:: sync-error
1664277279372 FirefoxView.TabsSetup DEBUG tryToClearError: triggering new tab sync
1664277279373 FirefoxView.TabsSetup DEBUG Handling UIState update
1664277279373 FirefoxView.TabsSetup DEBUG maybeUpdateUI, conditions not met to exit state: : error-state
1664277279373 FirefoxView.TabsSetup DEBUG maybeUpdateUI, will notify update?:: true
1664277279373 FirefoxView.TabsSetup DEBUG maybeUpdateUI, in error state:: sync-error
1664277280175 FirefoxView.TabsSetup DEBUG Handling weave:service:sync:finish
1664277280176 FirefoxView.TabsSetup DEBUG maybeUpdateUI, conditions not met to exit state: : error-state
1664277280176 FirefoxView.TabsSetup DEBUG maybeUpdateUI, will notify update?:: true
1664277280176 FirefoxView.TabsSetup DEBUG maybeUpdateUI, in error state:: sync-error
1664277280177 FirefoxView.TabsSetup DEBUG Handling UIState update
1664277280177 FirefoxView.TabsSetup DEBUG maybeUpdateUI, conditions not met to exit state: : error-state
1664277280177 FirefoxView.TabsSetup DEBUG maybeUpdateUI, will notify update?:: true
1664277280177 FirefoxView.TabsSetup DEBUG maybeUpdateUI, in error state:: sync-error
1664277282560 FirefoxView.TabsSetup DEBUG tryToClearError: triggering new tab sync
1664277282561 FirefoxView.TabsSetup DEBUG Handling UIState update
1664277282561 FirefoxView.TabsSetup DEBUG maybeUpdateUI, conditions not met to exit state: : error-state
1664277282561 FirefoxView.TabsSetup DEBUG maybeUpdateUI, will notify update?:: true
1664277282561 FirefoxView.TabsSetup DEBUG maybeUpdateUI, in error state:: sync-error
```
|
|
1,792,814 | 16,092,419 |
Created attachment 9296596
fx_view_page_00.gif
**Found in**
* 106.0b5 (20220927185813)
**Affected versions**
* 106.0b5 (20220927185813)
* 107.0a1 (20220929014928)
**Tested platforms**
* Affected platforms: Windows 10x64, Ubuntu 20.04, macOS 11.6
* Unaffected platform: none
**Preconditions**
* Have the Home Page button added to the toolbar.
**Steps to reproduce**
1. Open the Firefox View page by clicking the Pinned icon.
2. Click on the Home button from the toolbar.
3. Click on the Firefox view button.
**Expected result**
* The Firefox View page is displayed.
**Actual result**
* Nothing happens.
**Regression range**
* Not a regression. I can reproduce the issue starting with bug 1762903 when the Firefox view tab was implemented.
**Additional notes**
* Attached a screen recording.
* Clicking the Back arrow, will display the Firefox View page. Also restarting the browser or moving the Firefox View button to another part of the browser (e.g toolbar) will make it function again.
|
|
1,792,815 | 16,092,421 |
**Found in**
* Fx 107.0a1
**Affected versions**
* Fx 107.0a1
* Fx 106.0b5
* Fx 105.0.1
**Affected platforms**
* Windows 10
* Ubuntu 20
* macOS
**Steps to reproduce**
1. Launch Firefox.
2. Open a Private Browsing window.
3. Click the websearch bar type a word and hit enter.
4. Go to about:telemetry -> Raw JSON -> Raw Data and check for the search_counts under keyed.histogram
**Expected result**
* There is an 'urlbar-handoff' telemetry ping recorded.
**Actual result**
* Only a 'urlbar' telemetry ping is recorded.
**Regression range**
* This is not a recent regression. Will look into a regression range ASAP.
|
|
1,779,005 | 15,980,716 |
User Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:102.0) Gecko/20100101 Firefox/102.0
Steps to reproduce:
1. Set proxy settings to "Automatic Proxy configuration URL"
2. Set proxy configuration URL to "http://intranethost/proxy.pac" (use reachable URL)
3. Save changes
4. Access any website
5. Switch to network where proxy.pac and there defined proxys are not reachable
6. Reload website or access any website
Actual results:
Firefox shows the error "Unable to find the proxy server."
If you reload the page the error repeats *mostly* but not strictly!
If you restart Firefox you can access any website with direct connection.
Expected results:
Until Version 101 when proxy is not reachable anymore, Firefox instantly uses direct connection.
Btw: The in bug 1207798 described settings network.proxy.failover_direct or network.proxy.fallback_to_direct have no impact on this, not in 102 nor in 101!
|
|
1,779,359 | 15,983,391 |
User Agent: Mozilla/5.0 (X11; Linux x86_64; rv:91.0) Gecko/20100101 Firefox/91.0
Steps to reproduce:
1. Open the Web Developer Tools -> Network Monitor tab
2. Navigate a web site
3. Right click on an HTTP request item and choose "resend" menu
4. Change something in the 'message body' text box
5. Click on the 'send' button
Actual results:
1. The content-length in 'message header' box stay unchanged when modifying the 'message body'.
2. No message was sent when the send button is clicked, I think it may be caused by the mismatch between the content-length field of the header and the actual length of the message being sent.
Expected results:
In Firefox 91, the content-length will change automatically when the request body is changed.
|
|
1,779,728 | 15,986,174 |
Created attachment 9285613
rec of the issue.gif
**[Affected versions]:**
- Firefox Release 102.0.1 - Build ID: 20220705093820
**[Affected Platforms]:**
- Windows 10 x64
- Windows 7 x64
**[Prerequisites]:**
- Have the "nimbus.debug" pref set to "true" in the "about:config" page.
**[Steps to reproduce]:**
1. Open the browser from the prerequisites.
2. Navigate to the "about:studies?optin_slug=import-data-onboarding-time-saved&optin_branch=treatment-d&optin_collection=nimbus-preview" URL
3. Navigate to the "about:welcome" page and click the "Not now" button until the "Import data" screen is displayed.
4. Observe the primary button's label.
**[Expected result]:**
- The "Load from [previous browser]" label is displayed.
**[Actual result]:**
- The "Import from [previous browser]" label is displayed instead.
**[Notes]:**
- According to the [Figma specifications](https://www.figma.com/file/CjnCrvdbCIPO42YcHYFEuM/Import-Experiment?node-id=0%3A1) the "Load from [previous browser]" label should be displayed on the primary button of the Import screen from the "Treatment D" branch.
- Attached a screen recording of the issue:
|
|
1,779,750 | 15,986,290 |
Created attachment 9285621
image_2022_07_15T11_49_21_300Z.png
**[Affected versions]:**
- Firefox Release 102.0.1 - Build ID: 20220705093820
**[Affected Platforms]:**
- Windows 10 x64.
- Windows 7 x64
**[Prerequisites]:**
- Have the Firefox Release 103 browser installed.
- Have the "nimbus.debug" pref set to "true" in the "about:config" page.
**[Steps to reproduce]:**
1. Open the browser from the prerequisites.
2. Navigate to the "about:studies?optin_slug=import-data-onboarding-time-saved&optin_branch=treatment-a&optin_collection=nimbus-preview" URL.
3. Navigate to the "about:welcome" page and click the "Not now" button until the "Import data" screen is displayed.
4. Observe the primary button's label.
**[Expected result]:**
- The "Import from [previous browser]" label is displayed.
**[Actual result]:**
- The "Import your data" label is displayed instead.
**[Notes]:**
- Based on the [Mana Page](https://mana.mozilla.org/wiki/pages/viewpage.action?spaceKey=FPS&title=Import+Data+Onboarding+Step+-+Emphasize+time+saved+%28First+run%2C+new+users%29-+Fx+103), the "Import from [previous browser]" label should be displayed on the primary button of the Import screen from the "Treatment A" branch.
- Attached is a screenshot of the issue.
|
|
1,780,093 | 15,990,381 |
From github: https://github.com/mozilla-mobile/fenix/issues/26066.
> ### Steps to reproduce
>
> I remember text selection magnifier was added few months ago but since last few weeks I am observing this is not May I know why the feature removed?
>
>
> ### Expected behaviour
>
> On selecting text magnifier should show
>
> ### Actual behaviour
>
> Magnifier no longer showing
>
> ### Device name
>
> asus rog 3
>
> ### Android version
>
> android 12
>
> ### Firefox release type
>
> Firefox Nightly
>
> ### Firefox version
>
> latest
>
> ### Device logs
>
> _No response_
>
> ### Additional information
>
> _No response_
>
> ┆Issue is synchronized with this [Jira Task](https://mozilla-hub.atlassian.net/browse/FNXV2-21036)
>
Change performed by the [Move to Bugzilla add-on](https://addons.mozilla.org/en-US/firefox/addon/move-to-bugzilla/).
|
|
1,780,215 | 15,991,244 |
User Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.15; rv:101.0) Gecko/20100101 Firefox/101.0
Steps to reproduce:
compile spidermonkey to wasm and execute:
```
for (let i = 0; i < 5; i++)
print(Math.random());
```
The first two calls have the same first 3 or 4 digits
This is because js::GenerateXorShift128PlusSeed wants two seed values and uses js::GenerateRandomSeed() to get those values, js::GenerateRandomSeed() tries to get values from mozilla::RandomUint64 but that is returning Nothing(), causing js::GenerateRandomSeed() to fallback to using the current time as it's return value.
Having the same value for both seeds seems to have the unintended result that the first two calls to the PRNG have the same first 3 or 4 digits.
This can be reproduced on this page which uses spidermonkey compiled to wasm https://jandem.github.io/sm-wasi/?source=for%20(let%20i%20%3D%200%3B%20i%20%3C%205%3B%20i%2B%2B)%0A%20%20%20%20print(Math.random())%3B%0A
Actual results:
The first two calls have the same first 3 or 4 digits
Expected results:
The first two calls should not have the same first 3 or 4 digits
|
|
1,780,445 | 15,993,298 |
Both ASRouter and RemoteImages attempt to fetch Remote Settings attachments and look up the server URL by a pref that no longer exists.
|
|
1,781,269 | 16,002,871 |
I can reproduce this fairly reliably (> ~50% of the time) in debug builds on Nightly. It seems more rare in opt builds (but I haven't done as much testing there). Sometimes I get an error in the Browser Console when closing it, after which it will not re-open (no error shown when attempting to re-open, though).
The error is not always the same, here are the ones that I've captured:
```
JavaScript error: resource://devtools/client/webconsole/webconsole-ui.js, line 390: TypeError: can't access property "resourceCommand", this.hud is null
```
```
JavaScript error: resource://devtools/client/webconsole/webconsole-ui.js, line 145: TypeError: can't access property "dispatch", this.wrapper.getStore() is null
JavaScript error: , line 0: NotFoundError: No such JSWindowActor 'DevToolsFrame'
JavaScript error: , line 0: NotFoundError: No such JSWindowActor 'DevToolsFrame'
JavaScript error: , line 0: NotFoundError: No such JSWindowActor 'DevToolsFrame'
JavaScript error: , line 0: NotFoundError: No such JSWindowActor 'DevToolsFrame'
JavaScript error: , line 0: NotFoundError: No such JSWindowActor 'DevToolsFrame'
JavaScript error: , line 0: NotFoundError: No such JSWindowActor 'DevToolsFrame'
JavaScript error: , line 0: NotFoundError: No such JSWindowActor 'DevToolsFrame'
console.error: (new TypeError("can't access property \"top\", props.toolboxDoc.defaultView is null", "resource://devtools/client/shared/components/menu/MenuButton.js", 111))
console.error: ({})
```
(The `JSWindowActor` thing appears to be unrelated, or at least not directly at fault - I sometimes get it without the Browser Console breaking.)
|
|
1,782,186 | 16,007,999 |
Created attachment 9287614
egy kor.pdf
User Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:102.0) Gecko/20100101 Firefox/102.0
Steps to reproduce:
Tried to open a pdf file.
- open attached file
- password is 'Hello'
- observe that the screen is empty
now, if you exchange the 2 variables in the pdf (with notepad++ for example)
/StmF /StdCF -> /StmF /Identity
/StrF /Identity -> /StrF /StdCF
then it works. -- a circle appears as intended
All other pdf reader apps that i tried work. (acrobat, sumatra)
Actual results:
Didn't open
Expected results:
Should have opened
|
|
1,782,250 | 16,008,447 |
Steps to reproduce:
I have an email address in the verizon.net domain, which for several years now has been supported by aol.com. Since last December I have been using OAuth2 as the authentication method. After upgrading to TB 102.0, "Get Messages" fails to connect to pop.aol.com used by one of my accounts. It tells me OAuth2 is not supported for that server.
Actual results:
TB fails to get messages and displays a pop saying that OAuth2 is not supported by pop.aol.com.
This is incorrect, so TB seems to be enforcing invalid constraints on pop.aol.com.
Expected results:
Before the upgrade to TB 102.0, OAuth2 connection to pop.aol.com worked fine. And it still works from my iPhone. So why does TB not allow it?
Note that prior to TB 102.0, TB would not let me directly create a server config using OAuth2 for pop.aol.com. However, I was able to create one for another server and then change it to pop.aol.com. As of TB 102.0, there seems to be a hard check that prevents trying this connection. So I get the impression that TB previously had an invalid "config" for pop.aol.com and now that invalid config is being enforced. Wht does TB not allow OAuth2 via pop.aol.com?
|
|
1,797,345 | 16,130,317 |
Created attachment 9300176
Exemption tables.png
User Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:108.0) Gecko/20100101 Firefox/108.0
Build ID: 20221025094808
### Steps to reproduce
1. Go to multiple sites and add exemptions of affected areas (see attached screenshot)
2. Go to about:preferences#privacy and click their respective `Exemptions` buttons to view exemption tables.
### Actual result
(See attached screenshot) Values in 2nd+ column do not align nicely. They are probably affected by the length of the value in the previous column.
Here is a non-exhaustive list of affected tables:
1. Warn you when websites try to install add-ons (This one contains pre-defined entries in a new profile so it's quicker to verify the bug/fix)
2. Block pop-up windows
3. Enable HTTPS-Only Mode in all windows
4. Cookies and Site Data > Manage exemptions
5. Certificates > View Certificates > Servers
### Expected result
Every column lines up properly.
### Additional notes
Here is a non-exhaustive list of tables that are **not affected**:
1. Location/Camera/Notifications Settings: 2nd column are drop-down buttons
2. Enhanced Tracking Protections/Logins and Passwords > Manage Exceptions: table has only 1 column
3. Cookies and Site Data > Manage Data: multi columns but all align correctly
4. Certificates > View Certificates > Authorities
## Regression
Last good Nightly: 2022-10-19
First bad Nightly: 2022-10-20
pushlog: https://hg.mozilla.org/mozilla-central/pushloghtml?fromchange=a1297d435b3fddcb752141a4d6e95693b12734a6&tochange=59fa65e9da080d96cf90d9c99a458372b8d58b4d
Bisecting autoland builds:
https://hg.mozilla.org/integration/autoland/pushloghtml?fromchange=60a8d6f734e902102d87b8b1bd15b62bb4ba5ebe&tochange=6c60375237864880161d4a8915408e68ec28d1e3
This is regressed by [D159690 - Use modern flex on in-content pages.](https://phabricator.services.mozilla.com/D159690) from bug 1779695.
|
|
1,797,963 | 16,134,578 |
Created attachment 9300765
FIrefoxViewSpotlight_impressionCount.gif
**[Affected versions]:**
- Firefox Nightly 108.0a1 (Build ID: 20221027215515)
- Firefox Beta 107.0b6 (Build ID: 20221027185833)
- Firefox Release 106.0.2 (Build ID: 20221025065831)
**[Affected Platforms]:**
- Windows 10 x64
- macOS 12.5.1
- Linux Mint 20.2 x64
**[Prerequisites]:**
- Have the Firefox browser installed/extracted and opened.
- Have the “browser.newtabpage.activity-stream.asrouter.devtoolsEnabled” pref set to true.
**[Steps to reproduce]:**
1. Navigate to the “about:newtab#devtools” page.
2. Search for “FIREFOX_VIEW_SPOTLIGHT” and observe that its impression is “0”.
3. Click the Firefox View tab button.
4. Make sure that the “Firefox View” spotlight is displayed.
5. Switch back to the ASRouter page and observe its impression.
**[Expected result]:**
- The message impression is counted as “1”.
**[Actual result]:**
- The message impression is incremented with 2.
[Notes]:
- This issue is also reproducible when navigating to the “about:firefoxview” page to see the spotlight.
- I didn't manage to reproduce this issue with other messages.
- Attached is a screen recording of the issue.
|
|
1,798,014 | 16,134,862 |
User Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:108.0) Gecko/20100101 Firefox/108.0
Steps to reproduce:
Open any website or any page longer than the screen size (example: www.repubblica.it). Verified also in the about:support page.
Actual results:
The mouse wheel scrolling has no effect.
If I restart the browser in safe-mode the issue doesn't happen.
Tried with a laptop with video card Intel(R) UHD Graphics 620 (0x8086,0x3ea0) with driver 31.0.101.1999 and also the former 30.x.
Expected results:
The mouse wheel scrolling should work as it worked until version 107.
|
|
1,798,842 | 16,142,574 |
Created attachment 9301711
Broken page with URL
User Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.15; rv:106.0) Gecko/20100101 Firefox/106.0
Steps to reproduce:
1. Click on the Site information (Lock icon)
2. Click on "Connection secure" arrow (when "Connection verified by a certificate issuer that is not recognized by Mozilla.")
3. In the Connection security, click on "Learn More". (in the message 'Mozilla does not recognize this certificate issuer. It may have been added from your operating system or by an administrator. Learn More')
4. The hyperlink on Learn More text is pointing to https://support.mozilla.org/1/firefox/106.0.3/Darwin/en-US/enterprise-roots which in-turn redirects to https://support.mozilla.org/en-US/kb/enterprise-roots?as=u page saying "Page Not Found"
Actual results:
The "Learn More" link should is pointing to a resource which is unavailable and the it loads a "Page Not Found" page
Expected results:
The "Learn More" link should have redirected to the correct resource to learn more about using certificates which are not recognized by Mozilla
|
|
1,799,229 | 16,145,733 |
Crash report: https://crash-stats.mozilla.org/report/index/f479ef91-0c35-4a28-902b-e918f0221103
Reason: ```EXCEPTION_ACCESS_VIOLATION_WRITE```
Top 10 frames of crashing thread:
```
0 xul.dll JS::loader::ModuleLoaderBase::DisallowImportMaps js/loader/ModuleLoaderBase.h:294
0 xul.dll nsContentSink::ProcessLinkFromHeader dom/base/nsContentSink.cpp:334
1 xul.dll nsContentSink::DoProcessLinkHeader dom/base/nsContentSink.cpp:255
2 xul.dll mozilla::detail::RunnableMethodArguments<>::applyImpl xpcom/threads/nsThreadUtils.h:1147
2 xul.dll mozilla::detail::RunnableMethodArguments<>::apply xpcom/threads/nsThreadUtils.h:1153
2 xul.dll mozilla::detail::RunnableMethodImpl<nsMemoryReporterManager*, nsresult xpcom/threads/nsThreadUtils.h:1200
3 xul.dll mozilla::RunnableTask::Run xpcom/threads/TaskController.cpp:538
3 xul.dll mozilla::TaskController::DoExecuteNextTaskOnlyMainThreadInternal xpcom/threads/TaskController.cpp:851
4 xul.dll mozilla::TaskController::ExecuteNextTaskOnlyMainThreadInternal xpcom/threads/TaskController.cpp:683
4 xul.dll mozilla::TaskController::ProcessPendingMTTask xpcom/threads/TaskController.cpp:461
```
Null deref crash. The line we're crashing on in nsContentSink::ProcessLinkFromHeader() is this: `mDocument->ScriptLoader()->GetModuleLoader()->DisallowImportMaps();`
I'm guessing that something in that long chain of dereferences is null, so some null checks might be in order. This is happening off a runnable, so maybe something went away in the meanwhile.
The first instance I see of this is on build 20221010214639, but maybe the signature just changed around then.
|
|
1,812,508 | 16,251,923 |
Created attachment 9314167
116642.zip
We've received a report in https://github.com/webcompat/web-bugs/issues/116642 where some elements are not rendered correctly on https://manage.opensrs.com. The site is using [can.js](https://canjs.com) library.
While the site requires login, I've came up with a slightly reduced test case that javascript code on the site and attached an archive with sample code (can be run with `python3 -m http.server 8000`).
To reproduce:
1) Open 116642.html in release (109) or recent Nightly
2) Scroll to the bottom of the page and observe the list towards the end
**Actual:**
[object Object] is displayed (as `item.title` here is a reference to `item`)
**Expected:**
Titles are displayed
This looks like a timing issue since adding debugger statements makes it work correctly. It doesn't happen in Chrome or Firefox prior to bug1798365. Also it depends on the size/contents of the processed array.
The library does custom processing of arrays/objects to create observable objects. I did some analysis and added logs to the library source code to try and debug it.
The objects are stored in "cache" (basically another object) with a certain key and this cache is cleared periodically. At some point of looping through the list of items a key becomes `undefined` and an item is stored in cache with that key, which is I believe is causing the circular reference (during some later processing/deep copy) as it keeps referring to the same object in cache until it's cleared again.
|
|
1,813,542 | 16,258,302 |
Created attachment 9314852
SignInLink_TreatC.png
**[Affected versions]:**
- Firefox Beta 110.0b7 (Build ID: 20230129190147)
**[Affected Platforms]:**
- Windows 10 x64
**[Prerequisites]:**
- Have the latest version of Firefox Beta 110 installed and opened.
- Have the “nimbus.debug” and “devtools.chrome.enabled” prefs set to “true” in the “about:config” page.
- Have the browser’s Dark theme enabled.
**[Steps to reproduce]:**
1. Force enroll in the Treatment C branch of the experiment using the following link: about:studies?optin_slug=window-modal-vs-tab-modal&optin_branch=treatment-c&optin_collection=nimbus-preview
2. Focus the browser console and run the “Cc["@mozilla.org/browser/clh;1"].getService().wrappedJSObject.firstRunProfile = true; Cc["@mozilla.org/browser/browserglue;1"].getService().wrappedJSObject._maybeShowDefaultBrowserPrompt()” syntax.
3. Observe the “Sign in” link from the first modal of the Onboarding.
**[Expected result]:**
- The “Sign in” link is visible.
**[Actual result]:**
- The “Sign in” link is barely visible.
**[Notes]:**
- This issue is only reproducible in the Treatment C of the experiment due to the gradient opaque overlay.
- Also, it is not reproducible when the browser’s Light theme is enabled.
- Attached is a screenshot of the issue.
|
|
1,813,556 | 16,258,372 |
Created attachment 9314860
reprl.py
Steps to reproduce:
The following bug is a bit annoying to reproduce, sorry for that. During fuzzing, I encountered an assertion violation at js/src/gc/GC.cpp:5044
Unfortunately, the I only managed to reproduce the crash when supplying 2 separate files via the reprl interface (standard for Fuzzilli fuzzing); neither concatenating nor `-f a.js -f b.js` worked. I wrote a small python script which instanciates the js-shell in reprl mode and feeds the 2 snippets.
First, there will be an assertion violation at `Assertion failure: rt->supportedImportAssertions.ref().empty(), at js/src/vm/Modules.cpp:45`. This one in probably unrelated; I removed this assert as it triggers occasionally during fuzzing.
The second assert, `!tc->isMarkedGray()` seems more interesting.
Bisecting the GC issue points to commit 8d993476b0fca3d388d139a207d8fe838aeb2187 related to bug 1749298. Setting s-s because one other regressor of bug 1749298 was set to sec-high.
Tested on commit fb1e6d6e5735dcf12d96fde70351aca305961b53. The option `--enable-js-fuzzilli` is required in .mozconfig for enabling the reprl interface.
Note that debugging the issue under rr might not work out of the box because (on my system). rr uses one of the hard-coded fuzzilli file-descriptors
(100, used by fuzzilli for REPRL_CRFD). Changing `REPRL_{CRFD, CWFD, DRFD, DWFD}` in the attached script and js.cpp allows reproduction under rr.
```
#0 0x000055555859559d in js::gc::detail::AssertCellIsNotGray (cell=0x1d6a25a67988)
at js/src/gc/GC.cpp:5044
#1 0x000055555752bb30 in JS::AssertCellIsNotGray (maybeCell=0x1d6a25a67988)
at obj-x86_64-pc-linux-gnu/dist/include/js/RootingAPI.h:423
#2 0x000055555754f511 in JSContext::enterRealmOf (this=0x7ffff6b2f100, target=0x1d6a25a67988)
at js/src/vm/JSContext-inl.h:301
#3 0x000055555759cac6 in js::AutoRealm::AutoRealm<JSFunction*> (this=0x7fffffffe680, cx=0x7ffff6b2f100,
target=@0x7fffffffe690: 0x1d6a25a67988) at js/src/vm/Realm-inl.h:62
#4 0x0000555557520cd9 in MaybeRunFinalizationRegistryCleanupTasks (cx=0x7ffff6b2f100)
at js/src/shell/js.cpp:1115
#5 0x00005555574f8abe in RunShellJobs (cx=0x7ffff6b2f100)
at js/src/shell/js.cpp:1162
#6 0x00005555574ea836 in Shell (cx=0x7ffff6b2f100, op=0x7fffffffea58)
at js/src/shell/js.cpp:11333
#7 0x00005555574e541c in main (argc=8, argv=0x7fffffffed18)
at js/src/shell/js.cpp:12459
```
|
|
1,814,345 | 16,261,655 |
Created attachment 9315321
Screen Recording 2023-02-01 at 11.19.35.mov
**Notes**
- This does not reproduce in Chrome. Please see the attached recording for more details.
**Found in**
* Nightly 111.0a1;
**Affected versions**
* Nightly 111.0a1;
**Tested platforms**
* macOS 12;
* Windows 10;
* Ubuntu 22;
**Affected platforms**
* macOS 12;
* Windows 10;
* Ubuntu 22;
**Unaffected platforms**
* N/A;
**Steps to reproduce**
1. Launch Nightly and go to http://mail.google.com .
2. Login with a valid account.
3. Click on the “Compose” button to open a new e-mail window.
4. From the editor, select any text color.
5. Write something in the e-mail.
**Expected result**
* Text color selection is respected.
**Actual result**
* Text color selection is not respected.
**Regression range**
* First bad: 2023-01-26
* Last good: 2023-01-25
* Pushlog: https://hg.mozilla.org/integration/autoland/pushloghtml?fromchange=2e626352a4f6f0a7cb598d35d3ace53d4be7f284&tochange=0659120fb8f2386bbab75f04527a2854e0e6465e
* Potentially regressed by: bug 1811161
|
|
1,814,360 | 16,261,786 |
Created attachment 9315336
rec of the issue.gif
**[Affected versions]:**
- Firefox Beta 110.0b8 - Build ID: 20230131190033
**[Affected Platforms]:**
- Windows 10 x64
**[Prerequisites]:**
- Firefox is not installed.
- Have the Firefox "[setup.exe](http://ftp.mozilla.org/pub/firefox/releases/110.0b8/win64/en-US/Firefox%20Setup%20110.0b8.exe)" downloaded to your PC.
- Have the "[user.js](https://drive.google.com/file/d/1CxnCpHB9VswWuEGpZ0jebNaQDznxF6DM/view?usp=share_link)" file saved to your PC.
**[Steps to reproduce]:**
1. Install Firefox using the "setup.exe" from prerequisites and uncheck the "Pin to taskbar" option from the install flow.
2. Create a new Firefox profile but do not open it.
3. Navigate to the profile folder and paste the "user.js" from the prerequisites.
4. Open the browser using the previously created profile and the "--first-startup" command.
5. Navigate to the "about:studies" page and observe the displayed studies.
**[Expected result]:**
- The "Window Modal vs. Tab Modal" experiment is not displayed.
**[Actual result]:**
- The "Window Modal vs. Tab Modal" experiment is displayed.
**[Additional Notes]:**
- Attached is a screen recording of the issue:
|
|
1,814,401 | 16,262,336 |
Created attachment 9315379
SetToDefaultModal_WrongImage.png
**[Affected versions]:**
- Firefox Beta 110.0b8 (Build ID: 20230131190033)
**[Affected Platforms]:**
- Windows 10 x64
- Windows 11 x64
**[Prerequisites]:**
- Have the latest version of Firefox Beta 110 installed and opened.
- Have the “nimbus.debug” and “devtools.chrome.enabled” prefs set to “true” in the “about:config” page.
**[Steps to reproduce]:**
1. Force enroll in the Treatment C branch of the experiment using the following link: about:studies?optin_slug=window-modal-vs-tab-modal&optin_branch=treatment-c&optin_collection=nimbus-preview
2. Focus the browser console and run the “Cc["@mozilla.org/browser/clh;1"].getService().wrappedJSObject.firstRunProfile = true; Cc["@mozilla.org/browser/browserglue;1"].getService().wrappedJSObject._maybeShowDefaultBrowserPrompt()” syntax.
3. Observe the first modal.
**[Expected result]:**
- The “Set to Default” modal is correctly displayed and has the correct image.
**[Actual result]:**
- The “Set to Default” modal has the wrong image.
**[Notes]:**
- This issue is also reproducible in the Treatment A and B branches of the experiment.
- This issue is not reproducible on the Control branch of the experiment.
- Attached is a screenshot of the issue.
|
|
1,814,405 | 16,262,358 |
Created attachment 9315382
MobileLink_WrongParameters.gif
**[Affected versions]:**
- Firefox Beta 110.0b8 (Build ID: 20230131190033)
**[Affected Platforms]:**
- Windows 10 x64
- Windows 11 x64
**[Prerequisites]:**
- Have the latest version of Firefox Beta 110 installed and opened.
- Have the “nimbus.debug” and “devtools.chrome.enabled” prefs set to “true” in the “about:config” page.
**[Steps to reproduce]:**
1. Force enroll in the Treatment C branch of the experiment using the following link: about:studies?optin_slug=window-modal-vs-tab-modal&optin_branch=treatment-c&optin_collection=nimbus-preview
2. Focus the browser console and run the “Cc["@mozilla.org/browser/clh;1"].getService().wrappedJSObject.firstRunProfile = true; Cc["@mozilla.org/browser/browserglue;1"].getService().wrappedJSObject._maybeShowDefaultBrowserPrompt()” syntax.
3. Navigate through the Onboarding flow until the “Mobile” modal is shown.
4. Click the “Send yourself a download link instead” button.
5. Observe the URL’s parameters of the opened link.
**[Expected result]:**
- The URL contains a specific "utm_term "parameter for the experiment.
**[Actual result]:**
- The URL doesn’t contain a specific "utm_term" parameter for the experiment.
**[Notes]:**
- This issue is also reproducible in the Treatment A and B branches of the experiment.
- This issue is not reproducible on the Control branch of the experiment.
- Attached is a screen recording of the issue.
|
|
1,814,593 | 16,263,662 |
Created attachment 9315559
SingInLink_WrongParameters.gif
**[Affected versions]:**
- Firefox Beta 110.0b8 (Build ID: 20230131190033)
**[Affected Platforms]:**
- Windows 10 x64
- Windows 11 x64
**[Prerequisites]:**
- Have the latest version of Firefox Beta 110 installed and opened.
- Have the “nimbus.debug” and “devtools.chrome.enabled” prefs set to “true” in the “about:config” page.
**[Steps to reproduce]:**
1. Force enroll in the Treatment C branch of the experiment using the following link: about:studies?optin_slug=window-modal-vs-tab-modal&optin_branch=treatment-c&optin_collection=nimbus-preview
2. Focus the browser console and run the “Cc["@mozilla.org/browser/clh;1"].getService().wrappedJSObject.firstRunProfile = true; Cc["@mozilla.org/browser/browserglue;1"].getService().wrappedJSObject._maybeShowDefaultBrowserPrompt()” syntax.
3. Click the “Sign in” button from the top-right corner of the page.
4. Observe the URL’s parameters of the opened link.
**[Expected result]:**
- The URL contains a specific "utm_term" parameter for the experiment.
**[Actual result]:**
- The URL doesn’t contain a specific “utm_term” parameter for the experiment.
**[Notes]:**
- This issue is also reproducible in the Treatment A and B branches of the experiment.
- This issue is not reproducible on the Control branch of the experiment.
- Attached is a screen recording of the issue.
|
|
1,814,745 | 16,264,665 |
When attempting to use the WebGPU API (entry point `navigator.gpu.requestAdapter`, to be precise), a fresh Firefox instance with the `dom.webgpu.enabled` pref. set to `true` behaves as expected (see attached `expected-webgpu-enabled-from-start.png`). That is, calls to `navigator.gpu.requestAdapter().await` in a JS console returns expected results. Attempting to use `requestAdapter` while the pref. is disabled from start also yields expected behavior (see attached `expected-webgpu-disabled-from-start.png`).
Now, the problem: when changing `dom.webgpu.enabled` to `true` after it was `false` on start-up doesn't work! A restart of Firefox can work around this issue, but we should fully support dynamically enabling or disabling WebGPU via the `dom.webgpu.enabled` pref.
Steps to reproduce:
1. Run a fresh session of Firefox Nightly with the `dom.webgpu.enabled` pref. set to `false` from start.
2. Open `about:config` and set `dom.webgpu.enabled` to `true`.
3. Open a JS console (i.e., via `F12`), and evaluate the expression `navigator.gpu`. Observe that we get a `GPU` object back, implying that the WebGPU API is available for usage.
3. Using the console again, evaluate the expression `navigator.gpu.requestAdapter()`. Observe that we get a `Promise` rejection with the message `WebGPU is not enabled!`, which is unexpected. See also the attached `actual.png`.
|
|
1,801,756 | 16,169,121 |
User Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:109.0) Gecko/20100101 Firefox/109.0
Steps to reproduce:
This started happening a few days ago
Using the latest versions of Firefox Nightly
When I go to the GMail web site
The web page is presented
Nothing on the web page can be spoken by NVDA using mouse over
Fortunately, it still works fine using the regular Firefox
Actual results:
I use the mouse to navigate / hover over content on the web page and nothing is spoken by NVDA
I have tried to restart NVDA, the computer, and update Firefox Nightly but it does not help
Actually, it also seems to be a problem with other Google web pages
Google contacts, voice, calendar, drive, documents...
Expected results:
Usually when i use the mouse and hover over content on the GMail web page the content is spoken by NVDA
|
|
1,802,888 | 16,177,454 |
Created attachment 9305571
StarFinderSuperCharacterSheet.pdf
Steps to reproduce the problem:
1. Open the pdf
2. Enter 20 in the score field for Dexterity
The total values in Armor Class section are wrong: they should be 15, 15 and 23 when they're 5010, 150 and 158.
The values used to compute the total are concatenated because one of the field used to compute the sum has a string value.
The field type are guessed thanks to the action they've and then when the value is set we convert it into a number.
But here, the Misc Mod field doesn't give any hint that the expected value is a number so it must infer from the value itself!
In the same pdf, the values in the Carrying Capacity (page 2) are wrong (4, 5, 10) compared to the ones we've in Acrobat (5, 6, 11): it's due to a rounding issue in `util.printf`.
|
|
1,803,041 | 16,178,453 |
User Agent: Mozilla/5.0 (X11; Linux x86_64; rv:109.0) Gecko/20100101 Firefox/109.0
Steps to reproduce:
Open a youtube video
Actual results:
The time area only displays the number of seconds,
Expected results:
Should display via Orca (screen reader) the time, eg. 1 minute, 2 seconds of 3 minutes, 2 seconds. 2022-11-14 did it
|
|
1,803,117 | 16,178,882 |
### Steps to reproduce
1. Enable Cookie banner protection in private browsing
2. Open Twitter.com in Private tab
### Expected behavior
3. Twitter.com is opened
### Actual behavior
3. Cookie banner protection clicks on the app banner and open in app prompt is shown
### Device information
* Firefox version: Nightly/Main
* Android device model:
* Android OS version:
### Any additional information?
|
|
1,803,600 | 16,182,557 |
User Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:107.0) Gecko/20100101 Firefox/107.0
Firefox for Android
Steps to reproduce:
I have a php / javascript based web-application that is running since years very well.
On some pages I am using geolocation, mainly on Android-smartphones. Nothing has been changed, in this year.
Additional Informations:
My location: Darmstadt / Germany.
The issue is tested on 2 android smartphones and a PC.
The geolocation in Firefox on the pc is normal fast, less than 5 seconds.
Actual results:
Since some weeks only in Firefox (Standard, Beta and Nightly) the geolocation is very slow or I get a timeout.
The timeout-option is set to 60 sec.
The application has no problems on Chrome, Opera, DuckDuckGo or the MI-Browser from Xiaomi, so I am sure that will be a bug in Firefox 107.2.0 for Android and not in my Application.
Expected results:
Get the position in less than 5 seconds.
|
|
1,815,196 | 16,269,130 |
Created attachment 9316096
hidden_field_is printed.pdf
User Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/104.0.5112.102 Safari/537.36
Steps to reproduce:
I opened a PDF file in a firefox tab. This PDF contains a form with hidden buttons (option Hidden (Exclude from Layout)).
Mozilla/5.0 (Macintosh; Intel Mac OS X 10.15; rv:109.0) Gecko/20100101 Firefox/109.0
Mozilla/5.0 (Macintosh; Intel Mac OS X 10.15; rv:109.0) Gecko/20100101 Firefox/111.0
Actual results:
These hidden button are printed as grey boxes.
Expected results:
These hidden button should not be printed.
Extract from https://help.adobe.com/en_US/livecycle/11.0/DesignerHelp/WS107c29ade9134a2c583558f12a7dc955d9-8000.2.html:
Hidden (Exclude from Layout) The object is not visible on-screen, not visible in the printed form, and does not occupy any space in the form layout.
Chrome does not print theses hidden fields.
|
|
1,815,476 | 16,270,960 |
Created attachment 9316370
j2.pdf
Attach (recommended) or Link to PDF file here:
Steps to reproduce the problem:
1. Open the attached PDF in Firefox. See that it has some data
2. Open the PDF in some other PDF viewers.
What is the expected behavior?
Data shows
What went wrong?
In Chromium and Evince, text fields show only when focused. In Adobe PDF Viewer they don't show at all, I don't think (that's what my brother who was filing the PDF told me).
|
|
1,815,865 | 16,273,722 |
Created attachment 9316752
Screenshot 2023-02-09 at 12.34.10.png
**Found in**
* Nightly 111.0a1
**Affected versions**
* Nightly 111.0a1
**Tested platforms**
* macOS 12
* Windows 10
* Ubuntu 22
**Affected platforms**
* macOS 12
* Windows 10
* Ubuntu 22
**Unaffected platforms**
* N/A
**Steps to reproduce**
1. Launch Nightly.
2. Go to about:addons and change to Dark Theme.
3. In a new tab, trigger an Intervention Card(eg. type 'refresh firefox', 'cache firefox' in the Address Bar)
4. Hover over the 'Refresh Firefox' button.
**Expected result**
* Hover highlights properly and the text is readable.
**Actual result**
* Hover makes the text barely readable.
**Regression range**
* First bad: 2023-02-08
* Last good: 2023-02-07
* Pushlog: https://hg.mozilla.org/integration/autoland/pushloghtml?fromchange=602aa16c20d47216f2e4a3b8877c3e34ca947f33&tochange=d271704f07ad34f25bea298c69e6c02cbcf6bea8
* Potentially regressed by: bug 1815423
|
|
1,816,867 | 16,281,452 |
### Preconditions:
Set local language to en-US or en-CA.
### Steps to reproduce
1. Open the app. Go to Home Screen.
2. Go to "Stories by topic" section and select all 9 topic.
### Expected behavior
The earliest selection should be deselected.
### Actual behavior
All topics can be selected
### Device information
* Firefox version: Beta 111.0b1 + Nightly 112.0a1 (2023-02-15)
* Android device model: all
* Android OS version: all
### Any additional information?
|
|
1,810,653 | 16,238,488 |
The test [feature_callout_second_screen_positioned_left_of_element](https://searchfox.org/mozilla-central/rev/d39a17381a14606032f7b8e82789bf281beb1241/browser/components/firefoxview/tests/browser/browser_feature_callout_position.js#42-72) was originally written to test `arrow_position: "end"` but with the Colorways callouts removed, there are no messages with that position. So it's actually testing a screen with `arrow_position: "bottom"` but it's still comparing the left positions of the callout and parent, and still using the parent selector for the old tour screen 3 (the colorways card, which is now hidden). We should update this and maybe add another test that artificially sets `arrow_position: "end"` so it can be tested, while such messages are absent from the tree.
|
|
1,810,800 | 16,239,563 |
STR:
- open a pdf
- click outside Firefox to remove any focus
- press ctrl
- move the mouse on the viewer and play with the wheel.
The document is a way too much zoomed/unzoomed.
|
|
1,811,466 | 16,243,746 |
Created attachment 9313233
Video of the bad behaviour and the correct one
User Agent: Mozilla/5.0 (Windows NT 6.3; Win64; x64; rv:109.0) Gecko/20100101 Firefox/110.0
Steps to reproduce:
(I've confirmed this in a fresh profile, video of the behaviour, and a correct behaviour from previous version is attached)
When you have the "menu bar" enabled (the one with File, Edit, View etc) move your mouse over any of them, don't click, just move the mouse over it, then move it anywhere else, again without clicking it.
Actual results:
The item you hovered on will be lit as if you were still hovering over it, and no firefox keyboard shortcuts will work (ctrl + t, ctrl + shift + p etc) and all key inputs will be interpreted as if you really are focused on the menu bar (left and right arrows will move you left and right on the menu bar items, down arrow will open the curently selected item etc) until you either press the alt key (which does normaly focus / defocus the menu bar) to defocus the bar, or click anywhere else in any window (this is new in 110.0b3, previously in 110.0b2 the only way was to click any menu item).
Expected results:
The focus should happen on mouse over, but when the mouse leaves, it should defocus it (as it was before and should still be, based on normal Windows window behaviour)
|
|
1,811,507 | 16,244,149 |
[Affected versions]:
- Firefox Nightly 111.0a1 - Build ID: 20230120093625
[Affected Platforms]:
- Windows 10 x64
- macOS 12.4
- Linux Ubuntu 20.04 x64
[Prerequisites]:
- Have Firefox en-US build installed and open.
- Have a VPN client connected to a US server (for people located outside the US).
- Have the `browser.urlbar.weather.featureGate` pref set to `true`.
[Steps to reproduce]:
1. Open the browser from the prerequisites.
2. Open a new tab.
3. Click inside the Address Bar.
4. Make sure a Weather result is displayed (repeat previous steps if not).
5. Hover the cursor over the Weather suggestion.
6. Observe the behavior.
[Expected result]:
- The Weather suggestion is highlighted.
[Actual result]:
- Nothing happens, the Weather suggestion doesn’t show any UI changes.
[Notes]:
- The background highlight is also missing when focusing the result using the keyboard navigation.
- Here is a screen recording of the issue: [link](https://drive.google.com/file/d/1_H6rGZk-7A_ZQvZLowAt-sR3ZRfb0O3Y/view?usp=sharing).
[Regression]:
- Based on mozregression, it seems that the bug that caused this regression is Bug 1801297.
- Push log: https://hg.mozilla.org/integration/autoland/pushloghtml?fromchange=6b9c280f186bdf9338b45df9047465b80008c33d&tochange=95efc81f5d0da6252772c8a344388de3df68001e
- 17:15.50 INFO: Last good revision: 6b9c280f186bdf9338b45df9047465b80008c33d
- 17:15.50 INFO: First bad revision: 95efc81f5d0da6252772c8a344388de3df68001e
|
|
1,811,510 | 16,244,156 |
Created attachment 9313276
test-firefox.pdf
User Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/109.0.0.0 Safari/537.36
Steps to reproduce:
Open or Print a PDF that has a text field for numbers with the format of "1.234,56".
That field has "5,25" as its content (filled out not via firefox)
Note:
this is a new issue introduced in v109 as it worked correctly in v108
Actual results:
When the field is not selected (for editing) it shows "5,00".
if you select the field it shows "5,25"
It does show "5,25" if the content is "5.25".
Expected results:
In both cases, selected or not selected, it should show "5,25".
|
|
1,811,627 | 16,244,766 |
Created attachment 9313375
ffbug_2080.htm
`ThrottleInputStream::Read()/ReadSegments()` (`netwerk/base/ThrottleQueue.cpp` ) ignore the caller's requested read size `aCount`, potentially causing lower layers to read data into memory beyond the end of a caller's buffer.
`ThrottleInputStream` is used only by `ThrottleQueue`. I have found one user (the throttling control on the Developer Tools panel) which can be used to observe usage of the incorrect read size, but which doesn't suffer a WBB (write beyond bounds) because the read size is not used to buffer data, only to send data out a socket. I am not sure whether there are other users, though `nsHttpTransaction::Init()` has the ability to use the class via its `eventSink` argument.
The bug is that `ThrottleInputStream::Read()/ReadSegments()` don't minimize the amount of data potentially available in the underlying stream (`realCount`) against the requested read size `aCount`. (While `Available()`, line 110, takes `aCount`, it doesn't use it):
```
102: NS_IMETHODIMP
103: ThrottleInputStream::ReadSegments(nsWriteSegmentFun aWriter, void* aClosure,
104: uint32_t aCount, uint32_t* aResult) {
...
109: uint32_t realCount;
110: nsresult rv = mQueue->Available(aCount, &realCount);
111: if (NS_FAILED(rv)) {
112: return rv;
113: }
114:
115: if (realCount == 0) {
116: return NS_BASE_STREAM_WOULD_BLOCK;
117: }
118:
119: rv = mStream->ReadSegments(aWriter, aClosure, realCount, aResult);
120: if (NS_SUCCEEDED(rv) && *aResult > 0) {
121: mQueue->RecordRead(*aResult);
122: }
123: return rv;
124:
125: }
```
[This function doesn't use `aRemaining`:]
```
267: NS_IMETHODIMP
268: ThrottleQueue::Available(uint32_t aRemaining, uint32_t* aAvailable) {
269: MOZ_ASSERT(OnSocketThread(), "not on socket thread");
270: TimeStamp now = TimeStamp::Now();
271: TimeStamp oneSecondAgo = now - TimeDuration::FromSeconds(1);
272: size_t i;
273:
274: // Remove all stale events.
275: for (i = 0; i < mReadEvents.Length(); ++i) {
276: if (mReadEvents[i].mTime >= oneSecondAgo) {
277: break;
278: }
279: }
280: mReadEvents.RemoveElementsAt(0, i);
281:
282: uint32_t totalBytes = 0;
283: for (i = 0; i < mReadEvents.Length(); ++i) {
284: totalBytes += mReadEvents[i].mBytesRead;
285: }
286:
287: uint32_t spread = mMaxBytesPerSecond - mMeanBytesPerSecond;
288: double prob = static_cast<double>(rand()) / RAND_MAX;
289: uint32_t thisSliceBytes =
290: mMeanBytesPerSecond - spread + static_cast<uint32_t>(2 * spread * prob);
291:
292: if (totalBytes >= thisSliceBytes) {
293: *aAvailable = 0;
294: } else {
295: *aAvailable = thisSliceBytes;
296: }
297: return NS_OK;
298: }
```
You can verify that `ThrottleInputStream::ReadSegments()` can call the lower layer with a byte count exceeding `aCount` by doing the following:
1. Start FF and attach a debugger to it.
2. Set a BP in `ThrottleInputStream::ReadSegments()` on the call to `mStream->ReadSegments()`.
3. Open FF's Web Developer Tools panel, then the network tab.
4. Check "Disable Cache" and select "WiFi" from the throttling control.
5. Put ffbug_2080.htm on a webserver. If the server is not running on 127.0.0.1, modify ffbug_2080.htm appropriately to do the GET request to the appropriate server.
6. Load ffbug_2080.htm in FF.
7. When the BP fires, examine `realCount`. Notice that it far exceeds `aCount` (in my test, `realCount` was `0x1e0000` and `aCount` was `0x8000`.
8. Step into `mStream->ReadSegments()`. Control should end up in `nsStringInputStream::ReadSegments()`. Proceed to the line beginning
```
nsresult = aWriter(...);
```
and notice that `source` (as limited by this function's `aCount` and `maxCount`) contains the initial HTTP request to fetch ffbug_2080.htm. The request isn't very long, so it doesn't exceed the original `aCount` (0x8000).
9. Proceed, take the BP again, step into `mStream->ReadSegments()` again, and proceed as in step 8. Notice that this function's `aCount` is now `0` because `maxCount` is `0`.
10. Proceed and take the BP again, and step into `mStream->ReadSegments()` again, and proceed as in step 8. Notice that `source` contains the long `GET` request that the script in ffbug_2080.htm generated, and that the function's `aCount` and `maxCount` are both `0xc877`. This is much greater than the `aCount` that `ThrottleInputStream::ReadSegments()`'s caller requested (`0x8000`).
11. You can trace all the way into `nsHttpConnection::OnReadSegment()` and find that `0xc877` bytes do get written to the socket.
|
|
1,811,694 | 16,245,098 |
Created attachment 9313413
Screenshot_20230121_111327.png
User Agent: Mozilla/5.0 (X11; Linux x86_64; rv:109.0) Gecko/20100101 Firefox/109.0
Steps to reproduce:
Attempted to fill out IRS form W-9 in Firefox 109.0 (Linux vanilla Mozilla build).
Form located at: https://www.irs.gov/pub/irs-pdf/fw9.pdf
Enter a social security number under "Part I" that has leading zeros in one or more sections. Fake example number: "023-45-0789".
Actual results:
The leading zeros are dropped when tabbing to the next field. Applies to all three parts of the social security number field. Resulting number displayed after entering "023-45-0789" is "23-45-789".
Expected results:
The leading zeros in these fields should have been retained. The only workaround is to use the letter "O" instead of "0" (zero).
|
|
1,812,001 | 16,248,754 |
Steps to reproduce:
The following sample crashes the js shell on commit cf3af6bb6657278880f8baf38435eeb8f2d5d86c in `GetIteratorImpl<true>` when invoked as: `obj-x86_64-pc-linux-gnu/dist/bin/js --fast-warmup --fuzzing-safe --enable-iterator-indices --ion-warmup-threshold=100 --differential-testing crash.js`
Not setting s-s because iterator-indices are disabled by default.
```
for (let v0 = 0; v0 < 100; v0++) {
function F1() {
this[-65537] = 0;
this[-1869467081] = 0;
for (const v6 in this) {
this[v6];
}
}
const v8 = new F1();
}
```
|
|
1,812,387 | 16,251,312 |
STR:
1. Enable HTTPS-only mode in about:preferences (under Privacy & Security)
2. Open http://example.com
3. Click the identity icon on the urlbar
4. Click the HTTPS-only mode menulist
AR:
Nothing happens
ER:
popup should open
|
|
1,804,844 | 16,192,527 |
### Steps to reproduce
1. Install Firefox Nightly on an old Android (here 6.0.1.)
2. Try to sync Firefox Account via QR code from your computer
### Expected behavior
Successful sync.
### Actual behavior
java.lang.NoSuchMethodError: No virtual method sort(Ljava/util/Comparator;)V in class Ljava/util/ArrayList; or its super classes (declaration of 'java.util.ArrayList' appears in /system/framework/core-libart.jar)
at com.google.zxing.qrcode.detector.FinderPatternFinder$$ExternalSyntheticApiModelOutline0.m(R8$$SyntheticClass:3)
at com.google.zxing.qrcode.QRCodeReader.decode(QRCodeReader.java:544)
at com.google.zxing.MultiFormatReader.decodeInternal(MultiFormatReader.java:14)
at mozilla.components.feature.qr.QrFragment.decodeSource$feature_qr_release(QrFragment.kt:22)
at mozilla.components.feature.qr.QrFragment$imageAvailableListener$1$onImageAvailable$1.invokeSuspend(QrFragment.kt:23)
at kotlin.coroutines.jvm.internal.BaseContinuationImpl.resumeWith(ContinuationImpl.kt:9)
at kotlinx.coroutines.DispatchedTask.run(DispatchedTask.kt:107)
at kotlinx.coroutines.scheduling.CoroutineScheduler$Worker.run(CoroutineScheduler.kt:82)
Suppressed: kotlinx.coroutines.DiagnosticCoroutineContextException: [StandaloneCoroutine{Cancelling}@4755318, Dispatchers.Default]
### Device information
* Firefox version: 109.0a1 (Build 2015920297)
* Android device model: Lenovo Tab TB2-X30F_160816 (Stock firmware, after factory reset)
* Android OS version: 6.0.1
### Any additional information?
I just wanted to see if that old thing still runs. The stock firmware uses such an old playstore services library that I had to update it manually via apk. After that it seemed to recover well and it updated all the rest of the apps, but the OS remained the old one. I can run further tests on this device, if needed (though I'd totally accept if we just said "unsupported old Android" somewhere).
|
|
1,805,387 | 16,197,931 |
Created attachment 9307983
ss of the issue.png
**[Affected versions]:**
- Firefox Beta 109.0b1 - Build ID: 20221212143511
**[Affected Platforms]:**
- Windows 10 x64
- macOS 12.6
- ubuntu 20.04 x64
**[Prerequisites]:**
- Have a Firefox Beta 109 "fr" locale installed.
- Have a new Firefox profile with the "nimbus.debug" pref set to "true" in the "about:config" page.
**[Steps to reproduce]:**
1. Open the browser with the profile from the prerequisites.
2. Navigate to `about:studies?optin_slug=pdfjs-feature-callout&optin_branch=treatment-b&optin_collection=nimbus-preview`.
3. Navigate to an URL ending in ".pdf" (e.g. https://www.africau.edu/images/default/sample.pdf).
4. Observe the PDF callout message.
**[Expected result]:**
- All the elements from the PDF callout message are visible.
**[Actual result]:**
- The callout title overlaps with the "X" button.
**[Additional Notes]:**
- This issue is also reproducible on Firefox "de" locale.
- Attached a screenshot of the issue.
|
|
1,805,413 | 16,198,083 |
Created attachment 9308008
PDFcallouts_Infobar.png
**[Affected versions]:**
- Firefox Beta 109.0b1 (Build ID: 20221212143511)
**[Affected Platforms]:**
- Windows 10 x64
- macOS 11.7.1
- Ubuntu 22.04 x64
**[Prerequisites]:**
- Have the latest version of Firefox Beta 109 installed.
- Have created and opened a Firefox profile between 6 weeks and 3 months old.
- Have the Firefox browser not set as default.
- Have the “nimbus.debug” pref set to true in the “about:config” page.
- Have the "Always check if Firefox is your default browser" option from the “about:prefererences” unchecked.
**[Steps to reproduce]:**
1. Force enroll in the Treatment A branch of the experiment using the following link: about:studies?optin_slug=pdfjs-feature-callout&optin_branch=treatment-a&optin_collection=nimbus-preview
2. Open a new tab and restart the browser.
3. Ensure that the “Set as default” infobar is displayed.
4. Navigate to a PDF URL (e.g. https://www.clickdimensions.com/links/TestPDFfile.pdf) in the same tab.
5. Observe the top-right part of the page.
**[Expected result]:**
- The first message of the PDF Feature Callout message points to the Text icon.
**[Actual result]:**
- The first message of the PDF Feature Callout message points to the info bar.
**[Notes]:**
- This issue is also reproducible in the Treatment B of the experiment.
- The second message of the PDF callout also points to the info bar.
- Attached is a screenshot of the issue.
|
|
1,805,626 | 16,199,554 |
Steps to reproduce:
GLBlitHelper fails to correctly save and restore the buffer for vertex attribute bindings. In line 580, it saves the VBO currently bound into GL_ARRAY_BUFFER. Then, in line 596, it binds the saved vertex attribute to that VBO. This is wrong, since the VBO associated with the vertex attribute is a property of the VAO, and distinct from the current GL_ARRAY_BUFFER binding.
What it needs to do is use GL_VERTEX_ATTRIB_ARRAY_BUFFER_BINDING to get the actual VBO bound to the VAO, like it does for all the other VAO properties.
On an Apple M1 Mac running Asahi Linux with the mesa-asahi-edge drivers:
1. `apitrace trace firefox https://app.plex.tv/desktop/`
2. `apitrace dump firefox.trace`
Actual results:
The Plex website background is magenta, the system stutters, and the kernel reports GPU timeouts.
This happens because the incorrect save/restore in GLBlitHelper ends up binding the VAO to a 0/NULL VBO, and the GPU then cannot fetch vertex attribute data and faults.
Expected results:
The website renders properly and no GPU timeouts occur.
|
|
1,805,921 | 16,201,301 |
**Steps to reproduce**
1. With `devtools.custom-formatters` and `devtools.custom-formatters.enabled` set to true
1. Open the console in this page
1. Evaluate the following snippet
```js
window.devtoolsFormatters = [{
header: (obj, config) => {
if (obj.hasOwnProperty("customFormatHeader")) {
localStorage.setItem("test", 12);
return [
"span",
{
},
"custom formatted header",
];
}
return null;
},
}
];
({
customFormatHeader: true
})
```
**Expected results**
The regular object is printed (`Object { customFormatHeader: true }`) on the console, as the `header` function is effectful (calling `localStorage.setItem`)
**Actual results**
The custom formatted version is printed (`custom formatted header `), meaning the side-effect-free debugger failed to bail.
---
We're missing something here.
The side-effect-free debugger has a `onNativeCall` hook that should handle those native calls (https://searchfox.org/mozilla-central/rev/17aeb39742eba71e0936ae44a51a54197100166d/devtools/server/actors/webconsole/eval-with-debugger.js#397-416)
_But_ the doc states that
> SpiderMonkey only calls onNativeCall hooks when execution is inside a debugger evaluation associated with the debugger that has the onNativeCall hook. Such evaluation methods include Debugger.Object.executeInGlobal, Debugger.Frame.eval, and associated methods.
(See https://github.com/mozilla/gecko-dev/blob/43cb6eca1c3069d46d589d52ab4949257e630d19/js/src/doc/Debugger/Debugger.md#onnativecallcallee-reason)
The issue is that we're using `DebuggerObject#call` to call the custom formatter hook functions (https://searchfox.org/mozilla-central/rev/17aeb39742eba71e0936ae44a51a54197100166d/devtools/server/actors/object.js#791), and so the `onNativeCall` hook is not triggered, which makes our side-effect-free setup quite inefficient.
We would need to either:
- trigger `onNativeCall` hooks when calling `DebuggerObject#call`/`DebuggerObject#apply`
- or find a way to call the custom formatters via `executeInGlobal`
|
|
1,806,411 | 16,206,073 |
Created attachment 9308929
share_image_with_link.mp4
### Steps to reproduce
1. Tap and hold on any image until pop menu appears.
2. Tap share image.
3. Select an app to share with.
4. Observe along with the image, there is the image link.
### Expected behavior
Do not share the additional image link.
### Actual behavior
When sharing an image, the link of the image is also shared.
### Device information
* Firefox version: Nightly 110.0a1
* Android device model: OnePlus 6
* Android OS version: Android 11
### Any additional information?
|
|
1,806,831 | 16,209,037 |
Created attachment 9309356
21-12-2022_14-27-10.png
User Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/108.0.0.0 Safari/537.36 Edg/108.0.1462.54
Steps to reproduce:
tries to close firefox
Actual results:
continues to run under background processes
Expected results:
he should have ended himself completely
|
|
1,807,594 | 16,214,389 |
From github: https://github.com/mozilla-mobile/fenix/issues/25753.
> ### Steps to reproduce
>
> 1. Go to "Settings" -> "Search" and ensure that "Show search suggestions" is turned off. Exit the Settings.
> 2. Start typing in the address bar.
> 3. Click the "Search engine" button, and then click on a different search engine in the menu.
> 4. Continue typing in the address bar.
>
> ### Expected behaviour
>
> Search suggestions _should not_ appear because they are turned off in the settings.
>
> ### Actual behaviour
>
> Search suggestions _do_ appear during step 4.
>
> ### Device name
>
> Google Pixel 3a
>
> ### Android version
>
> Android 12
>
> ### Firefox release type
>
> Firefox
>
> ### Firefox version
>
> 102.1.1
>
> ### Device logs
>
> (I saved the logcat on my computer, but before sharing it, I'd need to review it for sensitive information, and I don't want to spend the time to do that unless/until the developers actually need the logcat, since the issue seems to be so easy to reproduce.)
>
> ### Additional information
>
> [Screen recording](https://mattmccutchen.net/private/firefox-search-suggestions-bug-20220624-recording.mp4) (hosted on my own web server because the file is too big for GitHub).
>
> The problem does not occur with Firefox 101.2.0, so this seems to be a regression. I consider it a privacy problem, so I plan to stay on Firefox 101.2.0 until it's fixed.
>
> ┆Issue is synchronized with this [Jira Task](https://mozilla-hub.atlassian.net/browse/FNXV2-20806)
>
Change performed by the [Move to Bugzilla add-on](https://addons.mozilla.org/en-US/firefox/addon/move-to-bugzilla/).
|
|
1,808,540 | 16,222,133 |
Created attachment 9310658
Screencast from 2023-01-04 15-14-23.webm
User Agent: Mozilla/5.0 (X11; Linux x86_64; rv:108.0) Gecko/20100101 Firefox/108.0
Steps to reproduce:
Logged in to Firefox Relay yesterday, then tried to generate a mask today.
(Note that the attached screencast shows that a Nightly update is available as this was yesterday's version, but I get the same after an update.)
Actual results:
I got a 403 error. I was still logged in to my Firefox account, but my session on the Relay website had apparently expired - I was no longer logged in there.
The "Learn more" link just took me to the Relay website.
Expected results:
I guess Firefox could have asked me to log in to Relay again? Or have a token that is valid for longer?
The error message could also be more helpful, as well as the "Learn more" link.
|
|
1,809,080 | 16,225,958 |
In Nightly, pinching on the MacBook trackpad leads to zoom all the window content (pdf.js toolbar included) when it must only zooms the pages.
In using mozregression, I got:
https://hg.mozilla.org/integration/autoland/pushloghtml?fromchange=db51149fe9596f05600937729cb381b9a3ac390b&tochange=46dae1c56dade459258373ddfed526da69c68c49
|
|
1,809,221 | 16,228,904 |
From github: https://github.com/mozilla-mobile/fenix/issues/28258.
> ### Steps to reproduce
>
> Visit any webpage in the browser
>
> ### Expected behaviour
>
> USER-AGENT: Mozilla/5.0 (Android 13, Mobile; rv:110.0_)Gecko/110.0 Firefox/110.0
>
> ### Actual behaviour
>
> USER-AGENT: Mozilla/5.0 (Android 13, Tablet; rv:110.0_)Gecko/110.0 Firefox/110.0
>
> ### Device name
>
> Google Pixle 6
>
> ### Android version
>
> Android 13
>
> ### Firefox release type
>
> Firefox Nightly
>
> ### Firefox version
>
> 110
>
> ### Device logs
>
> _No response_
>
> ### Additional information
>
> I do not know where to turn for this issue.
>
> Ive been using Firefox Nightly on my Pixel-phones for a year or two now. A couple months ago, Firefox Nightly did an update, and a lot of web pages I visit have been returning strange on my phone. Either the text is way too small, almost like Desktop mode is enabled, or graphs and texted are skewed.
>
> I have tried disabling my apps, which there are only 3: Dark Reader, Bitwarden, uBlock Origin. But it does not solve the issue.
>
> Last night I started to research browser request data and a few websites returned some output on my Pixel 6 phone
>
> USER-AGENT: Mozilla/5.0 (Android 13, Tablet; rv:110.0_)Gecko/110.0 Firefox/110.0
>
> My work phone, a Samsung S21FE, Returns the following on Nightly
>
> USER-AGENT: Mozilla/5.0 (Android 13, Mobile; rv:110.0_)Gecko/110.0 Firefox/110.0
>
> When I switch my Firefox Nightly to Desktop Mode on this webpage, the USER-AGENT returns as
>
> Mozilla/5.0 (x11: Linux x86_64; rv:110.0) Gecko/20100101 Firefox/110.0
>
> Switching back to mobile mode thinks the phone is a tablet.
>
> When I open normal Firefox or Firefox Focus, my user-agent returns as USER-AGENT: Mozilla/5.0 (Android 13, Mobile; rv:108.0_)Gecko/110.0 Firefox/108.0
>
> So everything is fine, but Firefox Nightly thinks my phone is a Tablet... WHY!!!
>
> I have tried looking online from Mozilla white-papers and resetting the user-agent in the browser does not work, nor does it apply to phone browsers. I cant find anything about this online.
>
> Hoping someone here can really help me out. Thank you
>
> ┆Issue is synchronized with this [Jira Task](https://mozilla-hub.atlassian.net/browse/FNXV2-22167)
>
Change performed by the [Move to Bugzilla add-on](https://addons.mozilla.org/en-US/firefox/addon/move-to-bugzilla/).
|
|
1,809,264 | 16,229,017 |
From github: https://github.com/mozilla-mobile/fenix/issues/15202.
> Hi team, 👋
> I'm using [Android App Links](https://developer.android.com/training/app-links/); the official way of doing deep linking since Android API 23. In that scenario, the domains are previously verified by Android at -user application- install time and their intent-filters are approved or not. Once verified, visiting a link that matches those intent-filter criteria, would invoke their app without asking the user for choosing one app.
>
> When visiting a callback URL such as "https://myapp.com/callback" the Chrome browser redirects successfully without further action. On Firefox, that's not the case. Because the URL doesn't have any content, as it's meant to be used only for registering that redirect, Firefox displays a "Not found." text and white page in the custom tab, and unless I select "Open in app" from the menu, I won't be redirected to the app.
>
> 
>
> The question for you is, **how is this menu option intuitive for someone that just installs the Firefox browser?**.
>
> We have customers that are losing conversion rates because of this design decision. Their users don't understand how to log in to their applications after they come across this white page. And I couldn't find any official documentation on it that I could share with them.
>
> ## Steps to reproduce
> 1. Create an android application targeting as min API android version 23.
> 2. Register an intent filter that matches against a callback URL you can redirect to from your site.
> 3. Set `autoVerify=true` so the domain verification happens at install time.
> 4. Host the assets links file in the well-known directory of your domain, using the hash for the key you are using to sign the APK.
> 5. Install a release in the device, making sure it was signed with the same key from the previous step. (All this is detailed in the android docs linked at the top).
> 6. Run your app, call a custom tab that opens a site with a button that when clicked would invoke that callback URL you've registered in the intent-filter.
> 7. Watch how firefox ignores the redirect when you click the button.
> 8. Click the "Open in app" menu option and watch how it redirects successfully.
>
> I'm testing this on a sample app using the [Auth0 SDK](https://github.com/auth0/Auth0.Android).
>
> ### Expected behavior
> Redirect back to the app automatically, **without** additional user action.
>
> ### Actual behavior
> The page shows content loading error "Not found." and the user has to click the overflow menu `...` and select "Open in app" for the app to successfully receive the intent.
>
> What I'd expect from Firefox is to at least check those apps that have officially registered their intent-filters using App Links and not thinking twice about opening it on that same browser tab or in an external app. The domain has been already verified by the OS at install time using the hosted `.well-known/assetslinks.json` file, so the deep link should be trusted.
>
> I've also tested first enabling the "Open links in apps" but that didn't change anything, at least when used from a Custom Tab.
>
> 
>
>
> ### Device information
> I tried this on a Pixel 2 emulator with Google Play Services, running Android 28. I've installed the latest Firefox app from the [Playstore at that time](https://play.google.com/store/apps/details?id=org.mozilla.firefox&hl=en) (yesterday). I don't think that's the "nightly" version, but rather the classic. The version name according to the Playstore listing is `80.1.3`.
>
> * Android device: Pixel 2, with Google Apps installed. Android API 28.
> * Fenix version: ?
>
> Related links:
> - https://support.mozilla.org/en-US/kb/set-firefox-android-open-links-native-apps
> - https://support.mozilla.org/en-US/questions/1297628
>
> ┆Issue is synchronized with this [Jira Task](https://mozilla-hub.atlassian.net/browse/FNXV2-7102)
>
Change performed by the [Move to Bugzilla add-on](https://addons.mozilla.org/en-US/firefox/addon/move-to-bugzilla/).
|
|
1,809,923 | 16,233,177 |
Created attachment 9311999
Firefox - UrlExtension Dangerous File Extension.mp4
Chromium Bug Reference: https://bugs.chromium.org/p/chromium/issues/detail?id=1307930
While visiting a webpage if the download initiate for .lnk/ .local these file types are changed to .download by firefox. But i have found a way to download .url File while lead to Arbitrary File Read
Note: .local/ .lnk are already restricted and .url is missed
We can use Blob Method with type text to download the .url file without changing it to .download
FIREFOX VERSION : 108.0.2 (64-bit)
REPRODUCTION CASE:
.Url Download POC
1. Open the Below Attached HTML
2. You can See the .Url file Downloaded
.Url Download Arbitrary File Read POC
1. Open the Below Same Attached HTML
2. You can See the .URL File Downloaded
3. Upload the same file in the Upload Button
4. You can See the content of system.ini file
I have also attached the POC Video for your reference
|
|
1,799,638 | 16,150,731 |
User Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:106.0) Gecko/20100101 Firefox/106.0
Steps to reproduce:
Past Daylight Saving Time (DST, Summer time) has wrong start time. DST should not start at "hh:59:59.001", but at "hh:00:00".
This issue has 2 ways to reproduce.
Reproduce by Firefox:
1. Set your computer time zone to "Asia/Tokyo".
2. Start F12 Developer Tools Javascript Console.
3. Execute following Javascript cases. (The last 2 lines are wrong)
```
console.log((new Date("1948-05-01T23:59:59"))); // Date Sat May 01 1948 23:59:59 GMT+0900 (Japan Standard Time)
console.log((new Date("1948-05-01T23:59:59")).getTimezoneOffset()); // -540
console.log((new Date("1948-05-01T23:59:59.000"))); // Date Sat May 01 1948 23:59:59 GMT+0900 (Japan Standard Time)
console.log((new Date("1948-05-01T23:59:59.000")).getTimezoneOffset()); // -540
// following case is NG
console.log((new Date("1948-05-01T23:59:59.001"))); // Date Sun May 02 1948 00:59:59 GMT+1000 (Japan Daylight Time)
// following case is NG
console.log((new Date("1948-05-01T23:59:59.001")).getTimezoneOffset()); // -600
console.log((new Date("1948-05-02T00:00:00"))); // Date Sun May 02 1948 01:00:00 GMT+1000 (Japan Daylight Time)
console.log((new Date("1948-05-02T00:00:00")).getTimezoneOffset()); // -600
```
4. Set your computer time zone to "America/New_York".
5. Start F12 Developer Tools Javascript Console.
6. Execute following Javascript cases.
```
console.log((new Date("1966-04-24T01:59:59"))); // Date Sun Apr 24 1966 01:59:59 GMT-0500 (Eastern Standard Time)
console.log((new Date("1966-04-24T01:59:59")).getTimezoneOffset()); // 300
console.log((new Date("1966-04-24T01:59:59.000"))); // Date Sun Apr 24 1966 01:59:59 GMT-0500 (Eastern Standard Time)
console.log((new Date("1966-04-24T01:59:59.000")).getTimezoneOffset()); // 300
// following case is NG
console.log((new Date("1966-04-24T01:59:59.001"))); // Date Sun Apr 24 1966 02:59:59 GMT-0400 (Eastern Daylight Time)
// following case is NG
console.log((new Date("1966-04-24T01:59:59.001")).getTimezoneOffset()); // 240
console.log((new Date("1966-04-24T02:00:00"))); // Date Sun Apr 24 1966 03:00:00 GMT-0400 (Eastern Daylight Time)
console.log((new Date("1966-04-24T02:00:00")).getTimezoneOffset()); // 240
```
Reproduce by SpiderMonkey:
1. Execute following commands
```
$ TZ=Asia/Tokyo ./mach run
0:00.33 /home/user/moz/mozilla-unified/obj-x86_64-pc-linux-gnu/dist/bin/js
js> console.log((new Date("1948-05-01T23:59:59")))
Sat May 01 1948 23:59:59 GMT+0900 (Japan Standard Time)
js> console.log((new Date("1948-05-01T23:59:59")).getTimezoneOffset())
-540
js> console.log((new Date("1948-05-01T23:59:59.000")))
Sat May 01 1948 23:59:59 GMT+0900 (Japan Standard Time)
js> console.log((new Date("1948-05-01T23:59:59.000")).getTimezoneOffset())
-540
js> console.log((new Date("1948-05-01T23:59:59.001")))
Sun May 02 1948 00:59:59 GMT+1000 (Japan Daylight Time)
js> console.log((new Date("1948-05-01T23:59:59.001")).getTimezoneOffset())
-600
$ TZ=America/New_York ./mach run
0:00.32 /home/user/moz/mozilla-unified/obj-x86_64-pc-linux-gnu/dist/bin/js
js> console.log((new Date("1966-04-24T01:59:59")))
Sun Apr 24 1966 01:59:59 GMT-0500 (Eastern Standard Time)
js> console.log((new Date("1966-04-24T01:59:59")).getTimezoneOffset())
300
js> console.log((new Date("1966-04-24T01:59:59.000")))
Sun Apr 24 1966 01:59:59 GMT-0500 (Eastern Standard Time)
js> console.log((new Date("1966-04-24T01:59:59.000")).getTimezoneOffset())
300
js> console.log((new Date("1966-04-24T01:59:59.001")))
Sun Apr 24 1966 02:59:59 GMT-0400 (Eastern Daylight Time)
js> console.log((new Date("1966-04-24T01:59:59.001")).getTimezoneOffset())
240
$ hg log
changeset: 714960:cf9bf33343df
bookmark: autoland
tag: tip
user: Mozilla Releng Treescript <[email protected]>
date: Tue Nov 08 06:58:10 2022 +0000
summary: no bug - Bumping Firefox l10n changesets r=release a=l10n-bump DONTBUILD CLOSED TREE
```
Actual results:
Actual results are described as above.
Expected results:
For "Asia/Tokyo":
```
console.log((new Date("1948-05-01T23:59:59"))); // Date Sat May 01 1948 23:59:59 GMT+0900 (Japan Standard Time)
console.log((new Date("1948-05-01T23:59:59")).getTimezoneOffset()); // -540
console.log((new Date("1948-05-01T23:59:59.000"))); // Date Sat May 01 1948 23:59:59 GMT+0900 (Japan Standard Time)
console.log((new Date("1948-05-01T23:59:59.000")).getTimezoneOffset()); // -540
// following case shows expected result
console.log((new Date("1948-05-01T23:59:59.001"))); // Date Sat May 01 1948 23:59:59 GMT+0900 (Japan Standard Time)
// following case shows expected result
console.log((new Date("1948-05-01T23:59:59.001")).getTimezoneOffset()); // -540
console.log((new Date("1948-05-02T00:00:00"))); // Date Sun May 02 1948 01:00:00 GMT+1000 (Japan Daylight Time)
console.log((new Date("1948-05-02T00:00:00")).getTimezoneOffset()); // -600
```
4. Set your computer time zone to "America/New_York".
5. Start F12 Developer Tools Javascript Console.
6. Execute following Javascript cases.
```
console.log((new Date("1966-04-24T01:59:59"))); // Date Sun Apr 24 1966 01:59:59 GMT-0500 (Eastern Standard Time)
console.log((new Date("1966-04-24T01:59:59")).getTimezoneOffset()); // 300
console.log((new Date("1966-04-24T01:59:59.000"))); // Date Sun Apr 24 1966 01:59:59 GMT-0500 (Eastern Standard Time)
console.log((new Date("1966-04-24T01:59:59.000")).getTimezoneOffset()); // 300
// following case is NG
console.log((new Date("1966-04-24T01:59:59.001"))); // Date Sun Apr 24 1966 01:59:59 GMT-0500 (Eastern Standard Time)
// following case is NG
console.log((new Date("1966-04-24T01:59:59.001")).getTimezoneOffset()); // 300
console.log((new Date("1966-04-24T02:00:00"))); // Date Sun Apr 24 1966 03:00:00 GMT-0400 (Eastern Daylight Time)
console.log((new Date("1966-04-24T02:00:00")).getTimezoneOffset()); // 240
```
Node.js returns correct getTimezoneOffset result.
Interestingly, Chrome returns correct getTimezoneOffset result, but returns incorrect DST notation (wrong "Japan Standard Time" or "Eastern Daylight Time").
FYI
https://github.com/eggert/tz/blob/2022f/asia#L2135
Rule Japan 1948 only - May Sat>=1 24:00 1:00 D
https://github.com/eggert/tz/blob/2022f/northamerica#L352
Rule NYC 1921 1966 - Apr lastSun 2:00 1:00 D
This issue seems to be related to IANA tz database "D" LETTER/S.
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.