offtopic Captcha hell

Quote
roytam1

Quote
Anil
What Cloudfare protection? Cloudfare still shows that stupid 'One more step' page if you use a proxy. It would be amazing if there was a way to bypass this.

hcaptcha part may not possible to work correctly, but simple anti-ddos page can get pass (for example, https://www.ecosia.org/ )

Miracle in my ancient Opera! After years of struggling in vain with those pesky google reCaptchas, HCaptcha FINALLY allowed me to sign up somewhere, by showing a few images perfectly fine, and accepting the clicks in 1 round!! Things could be SO simple, if admins only want...
At first it was a huge struggle again, and thought in vain again as usual, but then found this recent topic:
https://news.ycombinator.com/item?id=22109969
And indeed, YES my custom useragent was culprit it seemed. Setting it to default helped.
Upon further reading, perhaps an empty string may work too, or a customized one with a Chrome UA, and perhaps custom UAs are not really the prob, only which one is used.
CATCH:
this will NOT work everywhere, since site admins can choose various difficulty levels. Guess I was just lucky that one had allowed the most basic level (probably)
Next time I'll try with KMG74...



Edited 2 time(s). Last edit at 02/25/2021 07:02AM by siria.

The mentioned problems are indeed fixed. Thank you !

Quote
siria

Quote
Voltaire
Opened on MacBook2,1 with Windows Vista ...

When opening "Preferences" this crash happens ...

roytam mentioned above that he fixed the Prefs crash already.
Try a fresh download...

Is there a newer build other than KM76.4.1-Goanna-20210220.7z created 20.02.2021 14:17 downloaded from the page we are here (uppermost link on top of the page):

(STICKY:
★ K-Meleon 76.4.1 - Goanna 3.4.6 (Build 20210220)
Updated link with current test build)

url: https://o.rthost.win/kmeleon/KM76.4.1-Goanna-20210220.7z

Btw: I downloaded it on 24.02.2021 and just now again; WinDiff shows "identical" ... The (sole and only) use was on 24.02.2021 on 14:47 (session.json 24.02.2021 14:48). I used it "out of the box" (without import of "old" browser settings etc) "portable".



Edited 4 time(s). Last edit at 02/26/2021 10:54AM by Voltaire.

New regular/weekly KM-Goanna release:
https://o.rthost.win/kmeleon/KM76.4.1-Goanna-20210227.7z

Changelog:

Out-of-tree changes:
* update Goanna3 to git 3b1d2cdd8..3c4ac697e:
- import changes from `dev' branch of rmottola/Arctic-Fox:
- Backed out changeset b7653e3d5f91 (bug 1174381) for widespread assertion failures. (f529b680d)
- Bug 1174381 - ServiceWorkerManager::TeardownRunnable should be called when xpcom-shutdown notification is received, r=nsm (c744d9a80)
- Bug 1134671 Keep sqlite connection open between Cache API operations. r=ehsan (43a57decc)
- Bug 1134671: Add 'override' keyword to method Context:ata::GetConnection() (in DOM Cache code). rs=ehsan (669c6eac3)
- Bug 1164100 P1 Cache API should use correct base dir even when reusing sqlite connection. r=ehsan (c1bdf85d3)
- Bug 1164100 P2 Fix defunct assertion in Cache API ActionRunnable. r=ehsan (e345b2a76)
- Bug 1161288 - Support app:// origins on Fetch API. r=baku,nsm (9c237bcdd)
- Bug 1168135 P1 Execute Cache init Action on same target thread used for other Actions. r=ehsan (30fcee443)
- Bug 1168135 P2 Add Cache Context::Init() method. r=ehsan (41dfd427a)
- Bug 1168135 P3 Cache Context should pass shared Data container to init Action. r=ehsan (2e8f19d7c)
- Bug 1169994 Fix Cache to close connection on right thread when init is canceled. r=ehsan (ca7b96b24)
- Bug 1174768 Cache should check if QuotaManager is shutting down before calling GetOrCreate. r=janv (7b06ad874)
- Bug 1110446 P1 Create marker files when Cache API context is open. r=ehsan (bb94b92ff)
- Bug 1146612 - Add a test to ensure that Cache.put() with an existing request will reorder it in the DB; r=bkelly (228ff808c)
- Bug 1162365 - Cache API does not calculate usage in QuotaClient::InitOrigin(). r=bkelly (3fa71ee24)
- Bug 1156033 - Add some missing error handling to the DOM Cache code; r=bkelly (67cd67987)
- Bug 1118298 - Client uses unknown command property session_id. r=ato (081eb8f2d) (bb3b80359)
- partly revert "Bug 1161183: Don't show the add-on version in the list view. r=dao" (3bb6519e0)
- import changes from `dev' branch of rmottola/Arctic-Fox:
- Bug 1137653: Correct the shape of the data object sent over if there are are crashes or the packet is malformed; r=jgriffin (d05804fe4)
- Bug 1121577 - Unset pref at delete_session, r=dburns (d2cefee52)
- Bug 1144907 - Restore previous context after calling marionette's gather_debug. r=davehunt (2d35526d2)
- Bug 1135846 - Expose marionette's actions code to chrome scope where applicable. r=dburns (2c91389b5)
- Bug 1118313 - newSession looks for non-conforming session_id property on Command. r=dburns (d72f256ec)
- Bug 1107706: Part 1: Change context from strings to enum (d4936c869)
- Bug 1107706: Part 2: Add error module and WebDriver error objects (95a58e4fa)
- Bug 1107706: Part 3: Add a command processor to Marionette (4ed444206)
- Bug 1107706: Part 4: Add dispatching mechanism to encapsulate connection (5a0cf4bfa)
- Bug 1107706: Part 5: Refactor Marionette chrome/content communication (a110867e3)
- Bug 1107706: Part 6: Make SpecialPowersError a prototypal Error (bebfc4e13)
- Bug 1107706: Part 7: Add timeout test for async scripts (795921b56)
- Bug 1107706: Part 8: Adapt emulator callbacks (4f0ec4ea3)
- Bug 1165395 Test Cache API after QuotaManager reset. r=janv (27ffd1273)
- Bug 1154325 P1 Add convenience routines to bind and get Blobs as strings. r=asuth (df1f9930a)
- Bug 1154325 P2 Use Blobs for CacheStorage keys to avoid encoding issues. r=ehsan (625588969)
- Bug 1179063 - Cleanup: Rename scopal back to scopeStmt. (r=me) (eaa5d8e3a)
- pointer style (39588e254)
- pointer style (7be6abefc)
- Bug 1162342 Enable sqlite WAL transactions in Cache API. r=ehsan (60a767033)
- Bug 1130452 - Remove support for the prefixMatch option in the service worker cache code; r=bkelly,baku (48d4d62f2)
- Bug 1166577 Use named sqlite params in Cache API code. r=ehsan (a0cbc5674)
- Bug 1166038 De-duplicate security info stored in the Cache API. r=ehsan (cdb8e8396)
- Bug 1166911 Modify Cache API sqlite code to use IS NULL literal when comparing an empty key. r=ehsan (a893706bf)
- Bug 1164620 - Part 1: Remove instances of #ifdef PR_LOGGING in various places. r=froydnj (69f3becd3)
- Bug 1161684 - Allow JAR channels to be intercepted by service workers. r=jdm (292f83299)
- Bug 1146169 - Check mChannel for null-ness in nsIInterceptedChannel:etSecurityInfo(); r=jdm (9ed238664)
- Bug 1168208 - Refactor the existing logic for syncing the security info between Response and channel objects into a new helper class; r=nsm,jdm,bkelly (ea8852934)
- pointer style (b6ada460f)
- Bug 1163423 - Introduce JS_HasOwnProperty. r=bhackett (c189e58f0)
- pointer style (4693171b0)
- Bug 1163423 - Use JS_HasOwnProperty in WrapperAnswer::RecvHasOwn. r=billm (5029c2787)
- Bug 1168152 P1 Use a smaller sqlite page size and a growth increment in Cache. r=ehsan (9ad7f999f)
- Bug 1120501 P1 Make it easier to use Promise::All() from C++ code. r=nsm (95d1d5e2f)
- Bug 1120501 P2 Move Cache Add/AddAll logic to child process. r=ehsan,smaug (c71f9c49b)
- Bug 1168152 P2 Use a wrapper mozIStorageConnection for shared Cache connections. r=ehsan (f930635ac)
- Bug 1168152 P3 Perform incremental vacuum at tail end of Cache db connections. r=ehsan (df76df80c)
- Bug 1168152 P4 Add a test to verify Cache incremental vacuum works. r=ehsan (320d50436)
- Bug 1168152 P5 Follow-up to use 32kb growth size as intended in previous patches. r=me (92e2d3f27)
- Bug 1144236 - Add tests for the ignoreMethod match mode in DOM cache; r=bkelly (795a31d3c)
- Bug 1150001 - Cache API should not return Response body when matching Request with HEAD method. r=bkelly (690927f85) (13b1aeb09)
- import changes from `dev' branch of rmottola/Arctic-Fox:
- Bug 1110446 P2 Cleanup stale caches/bodies if last session didn't shutdown cleanly. r=ehsan (7925cf5fa)
- Bug 1110446 P3 Add a test that forces a Cache object to be orphaned and reclaimed. r=ehsan (c61409240)
- Bug 1110446 P4 Add a test that orphanes Cache API body files. r=ehsan (97e0a6f73)
- Bug 1157670 - Fixing an incorrect assertion in QuotaManager.cpp leads to an assertion failure; r=bent (0a19eac66)
- Bug 1165119 Remove corrupt morgue directories polluting nightly profiles. r=janv (d148170d8)
- Bug 1165119 - Follow-up to address review feedback accidentally left out of last push. r=me (f7ef96873)
- Fixup to make bug 1165119 ride the trains properly. r=trivial,DONTBUILD (87d186da4)
- Bug 1162624 - Add support for restoring corrupted or missing metadata files; r=bent (57e4341e6)
- Bug 1174113 - QuotaManager: Origin initialization fails on moz-safe-about+++home; r=bent (b7673128c)
- Bug 1142694 - QuotaManager default/temporary initialization fails on some profiles; r=bent (29a286137)
- Bug 1166871 - Always force a repaint before handling a wheel event so that we don't untransform it into some other scrollframe. r=botond (28e56646d)
- Don't vertically scroll APZCs that have less than one pixel of vertical scroll range. (bug 1154134, r=kats) (1bac9c054)
- Bug 1166871 - Add a test. r=botond (45d398bb6)
- Bug 1164557 - Do not start an overscroll animation if one is already running. r=kats (287a27910)
- Bug 1163832 - Add an API to flush pending APZ repaint requests and dispatch a notification upon completion. r=botond (8b3f9e06f)
- Bug 858680 - Part 1: Perform incremental_vacuum on open databases while idle, r=janv. (715f77ad6)
- Bug 858680 - Part 2: Add idle notifications to QuotaClient, r=janv. (9f245b1bb)
- Bug 1135166 - Initialize Telemetry histogram id cache early to avoid races. r=froydnj,vladan (f0bd8278c)
- Bug 1162176, Part 1. r=mak. (f92ba4061)
- Bug 1162176, Part 2. r=janv. (f313e1cf3)
- Bug 1155634 - Move ConnectionPool creation closer to where we actually use it and at a point guaranteed to be after QuotaManager has been started. r=khuey relanding CLOSED TREE (ce489e8f4)
- Bug 1155652 - Fix two incorrect assertions r=janv (2417d91ed)
- Bug 1156063 - Intermittent application crashed [@ mozilla::dom::indexedDB::::ConnectionPool:tart] in various tests. r=janv (b1126ac71)
- Bug 1157029 - More changes to bulletproof shutdown of failed connections, r=janv. (93a425abb)
- Bug 858680 - Part 4: Perform maintenance on databases while idle, r=janv. (017d536fe)
- Bug 1130775 - Convert synchronized ops and storage registration into unified directory locks; r=bent (300f635f7)
- Bug 1130775 followup: Add missing 'override' keyword to SendResults() methods in QuotaManager.cpp. rs=ehsan (397338f5b)
- Bug 1170021 - Part 1: Merge QuotaManager with QuotaObject; r=bent (168264350)
- Bug 1170021 - Part 2: Move DirectoryLock out of QuotaManager class; r=bent (278964f88)
- pointer style (99453953c)
- Bug 1171931 - Refactor duplicated code using XRE_IsParent/ContentProcess. r=froydnj (6d1ddbff1) (3909bb5fb)
- devtools: fix parenthesizes destructuring pattern assignment targets (24bf70dac)
- devtools: check aChannel.loadInfo before use (31836d339)
- import changes from `dev' branch of rmottola/Arctic-Fox:
- Bug 1168005 - NS_QUERY_TEXT_RECT on e10s should set mString. r=masayuki (b31a73fd1)
- Bug 1166436 part.1 Create mozilla::ContentCache and TabParent should store the text to it r=m_kato (d38cd5b53)
- Bug 1166323 - Remove IME sequence number. r=masayuki,nchen (6df799fcd)
- Bug 1166436 part.2 mozilla::ContentCache should store a selection range and TabParent should use it r=m_kato (94bb8cae3)
- Bug 1166436 part.3 mozilla::ContentCache should store active composition information and TabParent should use them r=m_kato (f1d9bd4fd)
- Bug 1166436 part.4 mozilla::ContentCache should store text rects and caret rect and TabParent should use them r=m_kato (da3d2a6a8)
- Bug 1166436 part.5 mContentCache should store writing mode at selection and TabParent should use it r=m_kato (ffb7801a4)
- Bug 1166436 part.6 mContentCache should store editor rect and TabParent should use it r=m_kato (cfbd6a896)
- Bug 1166436 part.7 mozilla::ContentCache should handle WidgetQueryContentEvent r=m_kato (e0b1a5534)
- Bug 1166436 part.8 Make mozilla::ContentCache available in IPC r=m_kato (60b2a7d8f)
- Bug 1166436 part.9 PuppetWidget should have mozilla::ContentCache and send it to TabParent r=m_kato (a0f6d5c61)
- Bug 1165515 - Part 4: Add PR_LOG_INFO. r=froydnj (dc5db5c34)
- Bug 1165515 - Part 6: Add PR_LOG_VERBOSE. r=froydnj (5705657b8)
- Bug 1165515 - Part 13-1: Add log level enum class. r=froydnj (943b666ca)
- Bug 1166436 part.10 Optimize IME notification handling in PuppetWidget r=m_kato (7a29fe4be)
- Bug 1166436 part.11 Remove unnecessary public methods of mozilla::ContentCache r=m_kato (3e78f5e4b)
- Bug 1166436 part.12 Log the behavior of mozilla::ContentCache r=m_kato (8a7dbd2c5)
- Bug 1166436 part.13 mozilla::ContentCache should guess caret offset if the offset is in the range of cached text rects r=m_kato (be496f793)
- Bug 1166436 part.14 Store text rects after focus/anchor of selection r=m_kato (b6ce2e2e0)
- Bug 1166436 part.15 ContentCache::CacheTextRects() should refer TextComposition::LastData() because TextComposition:tring() may not be modified yet when it's called r=m_kato (3ab6fd703)
- Bug 1166436 part.16 Make each member of mozilla::ContentCache used by chrome process or content process r=m_kato (35be9f90a)
- Bug 1166436 part.17 ContentCache::HandleQueryContentEvent() should return caret rect if mInput.mLength of NS_QUERY_TEXT_RECT event is 0 r=m_kato (0740f40ad)
- Bug 1166436 part.18 ContentCache should guess caret rect if there is a text rect cache of the previous character r=m_kato (e2a84217f)
- Bug 1166436 part.19 ContentCache::mSelection should have invalid state and it should be checked r=m_kato (39a054b22)
- Bug 972945 - Add autoconf test to enable AltiVec on supported targets only. r=glandium (1e3eb367d) (b4ae1472a)
- import changes from `dev' branch of rmottola/Arctic-Fox:
- Bug 1147555 - Implement logging for MediaTimer. r=mattwoodrow (5cf479784)
- Bug 1160064 - Explicitly initialize logging at the right time. r=jww (fd87f4aa4)
- Bug 1163196 - Part 1: Remove instances of #ifdef PR_LOGGING in dom/security. r=froydnj (889db2d96)
- Bug 1163196 - Part 2: Wrap expensive calls in PR_LOG_TEST. r=froydnj (caefa9db6)
- Bug 1148203 - Fix build errors with combinations of enable/disable of ffmp4/ffmpeg/eme - r=cpearce,ted (27308aa76)
- Bug 1164622 - Part 1: Remove instances of #ifdef PR_LOGGING in media. r=froydnj (302eb552e)
- Bug 1164622 - Part 2: Wrap expensive calls in PR_LOG_TEST. r=ekr (808824402)
- Bug 1164556 - Part 1: Remove instances of #ifdef PR_LOGGING in toolkit. r=froydnj (2b056c1bf)
- Bug 1143197 part.1 Use current IM context at handling key events rather than active IM context r=m_kato (f4c677871)
- Bug 1143197 part.2 Assume that composition is committed if a call of gtk_im_context_reset() causes composition string becomes empty r=m_kato (59cb6d378)
- Bug 1162242 - Part 2: Remove DEBUG_TIMERS. r=froydnj (792376073)
- Bug 1162242 - Part 3: Clean up usage of PR_LOG_TEST. r=froydnj (ebab31690)
- Bug 1157343 - Protect ProgressTracker::mImage with a mutex. r=tn (5e4be06c4)
- Bug 1162751 - Part 1: Remove instances of #ifdef PR_LOGGING in image. r=froydnj (124c81181)
- Bug 1162751 - Part 2: Always disable pallete index checking. r=seth (a914e4e31)
- Bug 1155429 - Add an explicit ImageCacheKey type for the ImageLib cache. r=baku (be1d56940)
- Bug 1162751 - Part 3: Wrap expensive calls in PR_LOG_TEST. r=seth (c2af68bb8)
- Bug 1164620 - Part 2: Wrap expensive calls in PR_LOG_TEST. r=froydnj (b0470b57f)
- Bug 1163192 - Part 1: Remove instances of #ifdef PR_LOGGING in dom/xslt. r=froydnj (5f8441bd8)
- Bug 1163192 - Part 2: Wrap expensive calls in PR_LOG_TEST. r=froydnj (1ee241448)
- Bug 1160703 (Part 1) - Move ImageCacheKey out of imgLoader.h. r=baku (22f0c6e93)
- Bug 1160703 (Part 2) - Store an ImageURL pointer instead of a URI spec string in ImageCacheKey. r=baku (e36657b1e) (b4e009645)
- import changes from `dev' branch of rmottola/Arctic-Fox:
- Bug 1139804 (Part 1) - Remove references to nonexistent type imgRequestNotifyRunnable. r=baku (7ad62c7f4)
- Bug 1139804 (Part 2) - Stop accessing private imgRequest members in imgRequestProxy. r=baku (6c9e0963a)
- Bug 1139804 (Part 3) - Stop accessing private imgRequest members in imgCache* classes. r=baku (d32db20d3)
- Bug 1139804 (Part 4) - Stop accessing private imgRequest members in imgLoader. r=baku (4258c2d0e)
- Bug 1139804 (Part 5) - Stop accessing private imgRequest members in imgMemoryReporter. r=baku (630828d85)
- Bug 1139804 (Part 6) - Make imgRequest no longer friends with ProgressTracker. r=baku (7b59e3012)
- Bug 1139818 - Merge imgRequest public and private sections. r=baku (9dc33da2c)
- Bug 1135977 (Part 1) - Use modern Mozilla style for refcounting in imgLoader. r=baku (9ee774c49)
- Bug 1135977 (Part 2) - Use threadsafe refcounting for ProxyListener. r=baku (2259befb6)
- Bug 1160703 (Part 3) - Associate an imgRequest with its ImageCacheKey. r=baku (62125415a) (c6718a164)
- a very late version bump for reflecting engine changes (3c4ac697e)

* Notice: the changelog above may not always applicable to XULRunner code which K-Meleon uses.

A goanna3 source tree that has kmeleon adaption patch applied is available here: https://github.com/roytam1/palemoon27/tree/kmeleon76

@roytam1 thanks!

@roytam1 thank you

@roytam1 thanks for your work!

K-Meleon still suffers from the 'unresponsive script error'. The only way out of it seems to be to kill the k-meleon process.

Great job, Roy!

But I have some bug report: KM 76.4 crashes (all versions) on loading page

https://www.worldometers.info/coronavirus/

Faulting module name: xul.dll

20210213 simply disappears with no any trace in Windows error logs.

20210220 -
Exception code: 0xc0000005
Fault offset: 0x0117be54

20210227 -
Exception code: 0xc0000005
Fault offset: 0x01182da4


All was fine in KM 76.2-76.3 (and is fine now in those builds). Thank you!



Edited 1 time(s). Last edit at 03/02/2021 10:55PM by rodocop.

Quote
rodocop
Great job, Roy!

But I have some bug report: KM 76.4 crashes (all versions) on loading page

https://www.worldometers.info/coronavirus/

Faulting module name: xul.dll

20210213 simply disappears with no any trace in Windows error logs.

20210220 -
Exception code: 0xc0000005
Fault offset: 0x0117be54

20210227 -
Exception code: 0xc0000005
Fault offset: 0x01182da4


All was fine in KM 76.2-76.3 (and is fine now in those builds). Thank you!

this bug should be fixed by rev https://github.com/roytam1/palemoon27/commit/c87f3d591



Edited 1 time(s). Last edit at 03/03/2021 02:35AM by roytam1.

To be honest, I'm disappointed by what I see with each update
Always the same
Some pages (80%) don't load or go blank
Some css interfaces don't display as they should
Some html5 functions don't work
WebGL does not work
There are more bugs/errors
I use this browser because it uses less resources than other browsers and because I have designed 6 macros for my use
I will be posting those macros soon as I am really disappointed with the quality of each update
If something is not done soon, this browser will be unusable by the progress of the web pages
Either you change the engine or I don't know what you are going to do
Anyway, it is appreciated at least that updates are released

Quote
Shiro Naomi
To be honest, I'm disappointed by what I see with each update
Always the same
Some pages (80%) don't load or go blank
Some css interfaces don't display as they should
Some html5 functions don't work
WebGL does not work
There are more bugs/errors
I use this browser because it uses less resources than other browsers and because I have designed 6 macros for my use
I will be posting those macros soon as I am really disappointed with the quality of each update
If something is not done soon, this browser will be unusable by the progress of the web pages
Either you change the engine or I don't know what you are going to do
Anyway, it is appreciated at least that updates are released

If your macros are so good for KMeleon quality, you should post them to @roytam1. :s
By the way, 80% of unloaded/blank pages? Really? For me only 10% have problems. :mad:

Quote
J.G.
By the way, 80% of unloaded/blank pages? Really? For me only 10% have problems. :mad:

Perhaps you don't visit many pages that contain new html5, css and js features

About the bugs/errors:

Error de lectura XML: entidad no definida Ubicación: jar:file:///C:/Users/Shiro/Desktop/Escritorio/%C2%A0/KM-Goanna/browser/omni.ja!/chrome/kmeleon/content/kmeleon/netError.xhtml Número de línea 282, columna 42:
        <h1 id="et_unknownProtocolFound">&unknownProtocolFound.title;</h1>
-----------------------------------------^

Quote
Shiro Naomi

Such XML errors I've seen getting fixed, but RT fixes only the native english version, the other language packs in omni.ja need to be fixed by KM locale maintainers. Sadly hardly anyone around anymore, but you're lucky - for the spanish version! It would probably help if you post an example URL where this happens. And please verify if the same happens in english?

New regular/weekly KM-Goanna release:
https://o.rthost.win/kmeleon/KM76.4.2-Goanna-20210306.7z

Changelog:

In-tree changes:
- app: GetKeyboardMayHaveIME API is removed (as of NM27 rev 7d12d066b) (d5ea5f1c)
- update XULRunner patch for KM76 (as of NM27 rev 7d12d066b) (dd737fc7)
- bump KM version for API change (as of NM27 rev 7d12d066b) (132db6b6)

Out-of-tree changes:
* update Goanna3 to git 3c4ac697e..7d12d066b:
- import changes from `dev' branch of rmottola/Arctic-Fox:
- Bug 1147722 - Part 2. Remove keyboardMayHaveIME method. r=wmccloskey (c0c7c8e85)
- Bug 1120487 - Implement shim before moving security checks into AsyncOpen; adding wrapper (r=sicking,sworkman) (d2eb4b169)
- Bug 1120487 - Implement shim before moving security checks into AsyncOpen, ioservice changes (r=sicking,sworkman) (41aca5645) (eb7ab47f8)
- import changes from `dev' branch of rmottola/Arctic-Fox:
- Bug 1165515 - Part 1: Convert PR_LOG to MOZ_LOG. r=froydnj (034b6056f)
- Bug 1165515 - Part 3: Convert PR_LOG_TEST to MOZ_LOG_TEST. r=froydnj (38739377e)
- Bug 1165515 - Part 5: Convert instances of PR_LOG_ALWAYS. r=froydnj (867725f77)
- Bug 1165515 - Part 7: Convert PR_LOG_DEBUG + 1 to PR_LOG_VERBOSE. rs=froydnj (afe55d0b8)
- Bug 1165515 - Part 8: Convert log level 6 to PR_LOG_VERBOSE. r=jesup (d01127f2c)
- Bug 1165515 - Part 9: Remove instances of using numeric log levels 15. rs=froydnj (2ff8b0056)
- Bug 1165515 - Part 10: Convert mtransport/logging.h to use PR_LOG levels. r=ekr (a0334c607)
- Bug 1165515 - Part 11: Align CSFLogLevel with PR_LOG levels. r=jesup (ca57ae88a)
- Bug 1165515 - Part 12: Convert nsPluginLogging to use PR_LOG levels. r=bsmedberg (1f3226ca8)
- Bug 1165515 - Part 13-2: Replace usage of PRLogModuleLevel and PR_LOG_*. rs=froydnj (96db7f2e9)
- Bug 1165515 - Part 14: Undef PR_LOG macros when using mozilla/Logging.h. r=froydnj (81d9dc8e5)
- fix some bustage after Bug 1165515 - Part 14, stuff missed in part 13 (80c83c78d)
- Bug 1165515 - Part 15: Reduce log level of WebRTC during testing. r=jesup (dda33e173) (3ea1cc0dd)
- import changes from `dev' branch of rmottola/Arctic-Fox:
- Bug 1172466 part.1 Make helper classes to notify IME nested classes of IMEContentObserver r=smaug (95efd78dc)
- Bug 1162818 part.1 nsEditor shouldn't release/forget mComposition becuase it should be handled by it after reframing r=ehsan (8adf7d49b)
- Bug 1162818 part.2 Make it possible to set IME selection from outside of IMETextTxn r=ehsan (ea30e4b7b)
- Bug 1162818 part.3 nsEditor should store actual composition string length in it r=ehsan (dd64d5d4e)
- Bug 1162818 part.4 Restore IME selection at initializing selection of the editor r=ehsan (5c3971072)
- Bug 1162818 part.5 The offset of nsEditor::InsertTextImpl() should be minimum offset of IME selections if there is r=ehsan (0ddabbc7b)
- Bug 1162818 part.6 Don't notify IME of anything during reframing the editor r=smaug (63362909f)
- Bug 1162818 part.7 Add test for reframing focused editor when it has composition r=smaug (0e8a59014)
- Bug 1172466 part.2 Use runnable event for notifying IME of focus r=smaug (698bc5257)
- Bug 1172466 part.3 Create an abstruct class which is a base class of classes notifying IME r=smaug (93f543ab7)
- Bug 1172466 part.4 Don't notify IME during reflow r=smaug (873102a98)
- Bug 1175392 part.1 IMEContentObserver and TabParent should use IMEStateManager::NotifyIME() r=smaug (3f13c391f)
- Bug 1175392 part.2 IMEStateManager should manage notifications to IME especially delayed notifications from remote process r=smaug (32f0ab730)
- Bug 1176954 part.1 Child process should notify its parent process when it dispatches composition or selection event into the DOM tree r=smaug (7466055f6)
- Bug 1171810 ContentCache should store union text rect of whole selected text r=m_kato (6d1265a41)
- Bug 1171814 ContentCache should store first character rect because Yosemite's Japanese IME sometimes tries to query it r=m_kato (966fcbca1)
- Bug 1173678 ContentCache should return union rect even if some character rects are not cached but the first character of the range is cached r=m_kato (e979292dd)
- Bug 1177388 Create ContentCacheInParent and ContentCacheInChild for making their purpose clearer r=m_kato (1418b32cc)
- Bug 1176954 part.2 ContentCache in parent process should manage how many events are sent but not yet received r=smaug (0219843ee)
- Bug 1176954 part.3 Don't send selection change, text change nor composition update notification to IME from TabParent until all events sent to the child process is received by it r=smaug (9a5195af7)
- Bug 492394 part.1 NS_QUERY_CHARACTER_AT_POINT should also return tentative caret offset for the point r=smaug, sr=smaug (24fe80f6b)
- Bug 492394 part.2 Implement ITextStoreACP::GetACPFromPoint() r=emk (b0df3db89)
- Bug 1172219 part.1 nsTextStore shouldn't notify IME while events are being dispatched and until they have not been handled yet r=emk (00149b33d)
- Bug 1172219 part.2 Notify TSF of layout creation and destruction r=emk (c2a3bcbbf)
- Bug 1172219 part.3 Modify nsTextStore::mSelection with new selection in OnSelectionChangeInternal() r=emk (47ebf2d2b)
- Bug 1176954 part.4 Define IMENotification::TextChangeDataBase and whose member names should be same as IMEContentObserver::TextChangeData's r=smaug (ea0e0f8a1) (43d0e566c)
- commonjs: fix parenthesizes destructuring pattern assignment targets (3f0d8da0b)
- import changes from `dev' branch of rmottola/Arctic-Fox:
- Bug 1176954 part.5 IMENotification::TextChangeDataBase should have a state which indicates that it's not initialized r=smaug (5b79eff32)
- Bug 1176954 part.6 IMEContentObserver should use IMENotification::TextChangeData r=smaug (f5ba0bc3e)
- Bug 1176954 part.7 IMENotification::TextChangeData should have a method to merge with another instance and IMENotification::MergeWith() should use it r=smaug (6978581c7)
- Bug 1151186 part.1 Don't reset IME enabled state after focus is changed by JS during setting focus r=enndeakin (37d70fa0b)
- Bug 1151186 part.2 nsHTMLEditor should accept composition events even while there is no active editing host but it has composition r=ehsan (9edcefa44)
- Bug 1053053 part.1 Active TabParent should be managed by IMEStateManager r=smaug (65e2222dd)
- Bug 1053053 part.2 Notify child process's IMEStateManager of that it should stop manageing IME state r=smaug (acab307e3)
- Bug 1053053 part.3 Notify child process's IMEStateManager of menu key event listener being installed r=smaug (8c3dd208f)
- Bug 1053053 part.4 Make IMEStateManager::sContent StaticRefPtr r=smaug (481701599)
- Bug 1053053 part.5 Make IMEStateManager::sActiveIMEContentObserver StaticRefPtr r=smaug (880a49eb4)
- Bug 1153518, allow drops on editors when the source is a remote browser, r=ehsan (1cc64b832)
- Bug 1179122 TextComposition should manage a composition which is even in a child process r=smaug (b2166e8f6)
- Bug 1176955 part.1 TextComposition should guarantee that WidgetSelectionEvent should be handled by same content as the target of composition events when there is a composition r=smaug (d4c6847c8)
- Bug 1176955 part.2 NOTIFY_IME_OF_SELECTION should have a flag which indicates if it's caused by a selection event r=smaug (902137ddd)
- Bug 1176955 part.3 nsTextStore shouldn't notify TSF of selection change which is caused by selection set event r=m_kato (398229e93)
- Bug 1184004 part.1 Give a name to each struct in IMENotification r=smaug (851e43638)
- Bug 1184004 part.2 Make all structs in IMENotification IPC-aware r=smaug (158e073a3)
- Bug 1176954 part.8 ContentCacheInParent::OnEventNeedingAckReceived() should log the event message r=smaug (94a53ba41)
- Bug 1184004 part.3 IPC methods to notify IME should use IMENotification for its argument r=smaug (71c8907c3) (7fc7f2d2a)
- import changes from `dev' branch of rmottola/Arctic-Fox:
- Bug 1159497 - Fix intermittent browser_bug427559.js failures r=MattN (7b3ea3f61)
- Bug 1156110 - Make nsView::mWindow an nsCOMPtr; r=roc (463a0d551)
- Bug 1167085 - Initialize nsBaseWidget::mLayerManager to nullptr. r=roc (a35cadb61)
- Bug 1157941 - If the current PresShell is suppressed, paint the old one if it is available r=tn,Enn (a50bcd9e3)
- Bug 1176959 nsGtkIMModule shouldn't commit composition when a selection change is notified but it was caused by selection set event r=m_kato (eae9672d1)
- Bug 1184449 part.1 IMENotifiation:electionChangeData should store selected string r=smaug (a34eddb64)
- Bug 1184449 part.2 nsIMM32Handler should store selection range as far as possible r=m_kato (d318ad109)
- Bug 1167105 - Part 1. Set temporary range until OnSelectionChange is called. r=masayuki (6aff5bd9d)
- Bug 1167105 - Part 2. Improve OnSelectionChange implementation. r=masayuki (4cb1f297f)
- Bug 1184449 part.3 Fix the bustage of Mac OS X r=myself (8e35b4aa2)
- Bug 1184986 NOTIFY_IME_OF_POSITION_CHANGE should be put off until all sending events are recieved by child process r=smaug (a1909600a)
- Bug 1192156 Separate IME specific structs, classes and enums from nsIWidget.h to IMEData.h r=roc (53a9f5f4c)
- Bug 1128607 - Add freshness check for OneCRL (r=keeler) (149e6a57f)
- Bug 1128607 - Test the freshness check for OneCRL (r=keeler) (c2408b646)
- Bug 1175382 TextRangeArray::TargetClauseOffset() and TextRangeArray::TargetClauseLength() shouldn't be public and same name methods of WidgetCompositionEvent should be used r=m_kato (5010f8363)
- Bug 1181240 - Part 1: Copy methods from gfx3DMatrix,r=vlad (dbb277cb2)
- Bug 1150704 - Use the same dest rect calculation for nsDisplayImage::GetOpaqueRegion and nsDisplayImage::GetDestRect. r=dholbert (60853571d)
- Bug 1182929 - Work around bug 803703 by refusing to turn heavily-downscaled images into image layers. r=seth (ecf9f5c9e)
- Bug 1177587 - Detect use of <img> elements for animation and use sync decoding to reduce flicker. r=tn (463ca0fd6)
- Bug 1151359 (Part 1) - Predict the size of nsImageFrame images before drawing. r=tn (c0cd67ee0)
- Bug 1151359 (Part 2) - Add support for detecting transform property changes that affect scale to ActiveLayerTracker. r=mstange (69094936c)
- Bug 1122526 patch 1 - Factor in the scale from ancestors when computing scale for layer with OMT animation of transform. r=roc (75378e2c8)
- Bug 1122526 patch 2 - Only do OMTA transform scale choosing for layers that are for transform display items. r=roc (299d3644b)
- Bug 1122526 patch 3 - Correctly account for transform:none when finding maximum scale for a transform animation. r=roc (6b6e190ad)
- Bug 1184200 - Update a check for APZ enabled. r=dvander (01c9f72df)
- Bug 1122526 patch 4 - Use the nearest widget size as the maximum size for an animated layer, in case it's a popup larger than the toplevel window. r=roc (941979bcb)
- Bug 771367 - Support compositor animation for pseudo-elements. r=dbaron (625ad9140)
- Bug 1151359 (Part 3) - Treat nsImageFrames subject to scale animation as having an identity scale when predicting size. r=tn,birtles (708fb4323)
- Bug 1238337. If the intrinsic size of the image hasn't changed then we don't need to do a new predictive image decode. r=mats (6b2bbe118)
- Bug 1181240 - Part 2: Replace gfx3DMatrix with Matrix4x4 in layout,r=vlad (e99b4dae5)
- Bug 1147279 - Get correct translation when layer optimize away. r=roc (a0bbf3c91)
- Bug 1176411 - Stop parenthesizing array/object pattern destructuring assignment targets in addon SDK code. r=jsantell, a=KWierso (remaining parts of rev 3f0d8da0b) (91a80210e)
- Bug 1146136 - Fix most in-tree tests that parenthesize destructuring pattern assignment targets, as ES6 forbids such parenthesization. (was rev 24bf70dac) (55d943c12)
- Bug 1145764 - Remove unused functions. r=nical (6bb54ad02)
- Bug 1150549 - Simplify TiledContentHost. r=jrmuizel (b800fa87d)
- Bug 1169339 - Fix incorrect tile storage order in TiledContentHost. r=jrmuizel (618e476bd)
- Bug 1145764 - Add logging for tile retention inside a TiledLayerBuffer. r=nical (113c84fbe)
- Bug 1170189 - Use TilesPlacement to handle tile coordinates conversion. r=BenWa (c3f082a21) (5fe82ff0f)
- partly imported Bug 1183896. Don't assume that relatively positioned table cells actually have positioned offsets computed, because in the case of position:sticky they don't. r=dholbert (c87f3d591)
- import changes from `dev' branch of rmottola/Arctic-Fox:
- Bug 766345 - Part 1 - Implement DEAA Antialiasing for transformed layers (v5 Patch). r=vladimir, r=djg (a5b4175d8)
- Bug 766345 - Part 2 - Implement Matrix4x4::TransformAndClipRect (v2 Patch). r=vladimir (fba290751)
- Bug 766345 - Part 3 - Implement correct clipping of DEAA edges for quads that cross the w=0 plane (v3 Patch). r=vlad (1574cd5e8)
- Bug 766345 - Part 4: Adjust reftests (v3 patch). r=vlad (8a78cc3c4)
- Bug 1158120 - Replace nsIntSize by mozilla::gfx::IntSize in gfx/layers. r=nical (39b6ad475)
- Bug 1170189 - Simplify TiledContentHost's render loop. r=BenWa (040aac295)
- Bug 1170189 - Simplify the client-side tiling code. r=BenWa (069820cf1)
- Bug 1170189 - Remove the TiledLayerComposer interface. r=BenWa (843da52ff)
- Bug 1158122 - Remove some of the occurences of nsIntRect in gfx. r=nical (600608f14)
- Bug 1181240 - Part 3: Replace gfx3DMatrix with Matrix4x4 in gfx,r=vlad (e27a99500)
- Bug 1181240 - Part 4: Remove gfx3DMatrix,r=vlad (36138a985)
- Bug 997709 part 1 - nsComboboxDisplayFrame can't be split so its reflow status is always NS_FRAME_COMPLETE. r=heycam (15696aad1)
- Bug 997709 part 2 - Prevent the nsComboboxDisplayFrame from being blockified if the parent context has display:flex/grid. r=heycam (7bedb6d69)
- Bug 1140216. Remove asserts that stuff that we never create is not null, since it clearly is null. r=jwatt (aad9d387b)
- Bug 997709 part 3 - tests. (50f3e3a16)
- Bug 1113206: Make nsComboboxControlFrame and nsListControlFrame use logical coordinates and support vertical writing modes - patch by smontagu with additions by jfkthame. r=jfkthame,smontagu (aaaff31e9)
- Bug 1181890 - Center children of ruby content frame if necessary after the bidi reposition. r=jfkthame (8b3b5fdb2)
- Bug 1123284 - pt 1 - Make nsTextBoxFrame somewhat aware of vertical writing mode, to allow <input type=file> to display properly. r=smontagu (b395ba174)
- Bug 1123284 - pt 2 - Fix Get{Min,Pref}ISize in nsLeafBoxFrame for vertical mode. r=smontagu (4a828e0ab)
- Bug 1079151 - patch 1 - Update constraint calculations in nsHTMLReflowState to work with logical coordinates. r=smontagu (ce11e8f6b)
- Bug 1079151 - patch 2 - Convert nsAbsoluteContainingBlock to use logial coordinates. r=smontagu (ad2b0000c)
- Bug 1079151 - patch 3 - Remove failure annotations from the vertical abs-pos reftests. r=smontagu (d95333230)
- Bug 1079151 - patch 4 - Handle unconstrained inline-size when computing constraints for an orthogonal absolutely-positioned block. r=smontagu (99ffede00)
- Bug 1079151 - Updated test files with corrected comments and references as needed. r=jfkthame (c655a98d0)
- Bug 1175492 - unpref transform-origin percentage handling for SVG elements r=jwatt (cda2fffa7)
- Bug 1175094 - Include borderPadding in the combobox's containerWidth, so that dropdown arrow is positioned correctly in RTL. r=smontagu (e93ccaa34)
- Bug 1177614 - Provide a utility method on nsHTMLReflowState to return the computed size including border-padding, for use as a container for logical coordinate conversions, or zero if unconstrained. r=dholbert (89feb2d8a)
- Bug 1180178: Add writing-mode and logical coordinates to frame dumps, r=jfkthame (786a6d878)
- Bug 1157569 - part 12 - Convert physical values from BCPropertyData to logical when returning from Get[Included]OuterBCBorder. r=roc (2d5790f2a)
- Bug 1157569 - part 13 - More conversion of physical to logical terminology in border-collapse calculations. r=dholbert (a85618bf4)
- Bug 1176523 - Convert Get/SetContinuousBCBorderWidth in nsTableColFrame and nsTableColGroupFrame to logical coordinates. r=dholbert (6198c6eb3)
- Bug 1157569 - part 14 - Finish conversion of border-collapse code in nsTableFrame to logical coordinates. r=dholbert (700bbf09b)
- Bug 1157569 - Reftest for RTL table with border-collapse. r=roc (aa4fc0092)
- Bug 1157569 - Followup to address review nits (renamings, comment updates) from parts 13 and 14. (6cccb95df)
- Bug 903135 - Multi platform MAR verification updater support. r=rstrong (7643732df)
- Bug 991993: Disable NSS for updater on OSX and enable native APIs. r=smichaud,rstrong (1490dead8) (7d12d066b)

* Notice: the changelog above may not always applicable to XULRunner code which K-Meleon uses.

A goanna3 source tree that has kmeleon adaption patch applied is available here: https://github.com/roytam1/palemoon27/tree/kmeleon76

Quote
Shiro Naomi
To be honest, I'm disappointed by what I see with each update
Always the same
Some pages (80%) don't load or go blank
Some css interfaces don't display as they should
Some html5 functions don't work
WebGL does not work
There are more bugs/errors
I use this browser because it uses less resources than other browsers and because I have designed 6 macros for my use
I will be posting those macros soon as I am really disappointed with the quality of each update
If something is not done soon, this browser will be unusable by the progress of the web pages
Either you change the engine or I don't know what you are going to do
Anyway, it is appreciated at least that updates are released

bug won't go away itself if there is no one reporting it in a reproducible way.

But... there isn't any localized Goanna version...., right?

If you are doing it, Shiro Naomi, don't overwrite with locales not designed for K-meleon Goanna!!!

You aren't overwriting the Goanna's omni.ja with some K-meleon Gecko earlier localized version, right?

Bad idea...

Quote
JohnHell
You aren't overwriting the Goanna's omni.ja with some K-meleon Gecko earlier localized version, right?

No

Quote
roytam1
bug won't go away itself if there is no one reporting it in a reproducible way.

I have done it several times, but I am still ignored as far as I can see
But if I have to do it again, its ok
But at least I hope to be read

Quote
siria
Such XML errors I've seen getting fixed, but RT fixes only the native english version, the other language packs in omni.ja need to be fixed by KM locale maintainers. Sadly hardly anyone around anymore, but you're lucky - for the spanish version! It would probably help if you post an example URL where this happens. And please verify if the same happens in english?

I going to download the latest version and test it

Quote
Shiro Naomi
I have done it several times, but I am still ignored as far as I can see
But if I have to do it again, its ok
But at least I hope to be read

I did read your "reports" but I can't reproduce it so there will not able to get it fixed.

For bug reporting, you have to ensure that the bug can be reproducible on unmodified copy of distribution with new profile, and list out procedures for reproducing the bug, like:

- Start K-Meleon browser
- Open settings dialog
- it crashes / application quits silently

- Start K-Meleon browser
- it crashes upon main window is opening

- Start K-Meleon browser
- Goto https://example.com
- it shows XML entity error dialog

etc.



Edited 3 time(s). Last edit at 03/07/2021 10:42AM by roytam1.

Will the Tastedive.com search bug be fixed soon?

Quote
roytam1
I did read your "reports" but I can't reproduce it so there will not able to be fixed.

For bug reporting, you have to ensure that the bug can be reproducible on unmodified copy of distribution with new profile, and list out procedures for reproducing the bug, like:

- Start K-Meleon browser
- Open settings dialog
- it crashes / application quits silently

- Start K-Meleon browser
- it crashes upon main window is opening

- Start K-Meleon browser
- Goto https://example.com
- it shows XML entity error dialog

etc.

Ok
I will take this into account in future reports
lately I replace the old files with the new ones from the latest update that you give to us
I don't touch the profile folder since there is the history, cookies, bookmarks and the cache of the icons I have on each page of the bookmark.

Quote
nerdyguy1990
Will the Tastedive.com search bug be fixed soon?

17:31:16.552 SyntaxError: invalid for/in left-hand side 1 vendor.3e1771a25f2d5aad586c.chunk.js:2:57581

for/in syntax will be hard to fix

(_HOWTO_ create multiple profiles howto)

Quote
Shiro Naomi
I don't touch the profile folder since there is the history, cookies, bookmarks and the cache of the icons I have on each page of the bookmark.[/i]

Edit > Configuration > Profile Directory
Make a backup copy of whole profileS folder.
Then:

Edit > Manage Profiles...
Check "Ask at startup", click button "NEW" and proceed. Open homepage, then restart again.

Now you have a new alternative, parallel profile, and can choose between 2 profiles (old or new) at startups. Or yet more profiles, how many a user wants.
In KM such "new" profiles aren't entirely clean though, since most users customize KM in root folders, put macros in root/macros/ instead of profile folder, change toolbar buttons in default toolbars.cfg etc.

But getting a completely fresh KM installation is easy too, since it's portable by default if the root folder contains an empty file "profile.ini". Just unzip a KM build into a completely new folder, with everything being original there.



Edited 1 time(s). Last edit at 03/06/2021 01:30PM by siria.

Quote
Shiro Naomi

Quote
J.G.
By the way, 80% of unloaded/blank pages? Really? For me only 10% have problems. :mad:

Perhaps you don't visit many pages that contain new html5, css and js features
About the bugs/errors:

Error de lectura XML: entidad no definida Ubicación: jar:file:///C:/Users/Shiro/Desktop/Escritorio/%C2%A0/KM-Goanna/browser/omni.ja!/chrome/kmeleon/content/kmeleon/netError.xhtml Número de línea 282, columna 42:
        <h1 id="et_unknownProtocolFound">&unknownProtocolFound.title;</h1>
-----------------------------------------^

Please, may you post the website that causes this error?
By the way, @roytam1, thanks for the last update!
Note: I use the spanish locale of KMeleon Goanna.



Edited 4 time(s). Last edit at 03/06/2021 08:15PM by J.G..

Thank you for update, Roy!

https://www.worldometers.info/coronavirus/ bug is gone now!!!

New regular/weekly KM-Goanna release:
https://o.rthost.win/kmeleon/KM76.4.2-Goanna-20210313.7z

Changelog:

Out-of-tree changes:
* update Goanna3 to git 7d12d066b..371e98abd:
- import changes from `dev' branch of rmottola/Arctic-Fox:
- Bug 1180166 - Ensure all elements of tableBorderReset[] are properly initialized. r=dholbert (0ce218eea)
- Bug 1177690 - part 0 - Tidy up some erratic whitespace. r=dholbert (4367c5ffd)
- Bug 1177690 - part 1 - Bevel the correct ends of horizontal collapsed-border segments in vertical-rl writing mode. r=dholbert (0efa53fb2)
- Bug 1177690 - part 2 - Tidy up the use of DrawTableBorderSegment similarly in BCInlineDirSeg:aint. r=dholbert (fef2b0ebb)
- Bug 1153140 - Include borderPadding width in the containerWidth for grid item placement. r=mats (ceaf03493)
- Bug 77999 - Part 1: Add nsDocumentRuleResultCacheKey. r=dbaron (8d3599b2b)
- Bug 1159772. Don't let nsContainerFrame:yncWindowProperties make widget calls that can result in synchronous painting during reflow or frame construction. r=mats (395ec1340)
- Bug 1009214 part 1 - [css-grid] Don't wrap placeholders in an anonymous grid item. r=dholbert (51f5db33f)
- Bug 1009214 part 2 - [css-grid] Update the reference for a few tests that wrapped placeholders in anonymous grid items. (48d8f2363)
- Bug 1151316 part 1 - [css-grid] Implement relative positioning for grid items. r=roc (470c7beb6)
- Bug 1153140 - dir=rtl versions of a couple of the grid reftests. r=mats (8942648a6)
- Bug 1079164 - Initial testcases for CSS grid in vertical writing mode. r=mats (667ec9ce2)
- Bug 1149555 - Update resize event firing to follow the specs, dispatch right before rAF callbacks, r=dbaron (2bb998687)
- ckout Bug 1149555 because of tresize regression, a=backout (8e9a9c614)
- Bug 1141867 - patch 1 - Use the proper writing mode to return the inline-size from FloatMarginISize. r=smontagu (af6a62275)
- Bug 1141867 - patch 2 - If FloatMarginISize returns unconstrained, we need to do an early reflow on the orthogonal float to find out its real size. r=smontagu (683e2f813)
- Bug 1141867 - patch 3 - Reftests for contiguous orthogonal floats. r=smontagu (b67e31c51)
- Bug 1141867 - Add OSX 10.10 fuzz to the newly-added tests. a=merge (d8587d21e)
- Bug 1157691 - Explicitly enable writing-mode support for reftests that require it. r=smontagu (17890dd40)
- Bug 1151316 part 2 - [css-grid] Tests for relative positioning of grid items. (ef8c924af)
- Bug 1171773 - Handle reflowing a grid item with different writing mode from its grid container. r=mats (1b775f1c1)
- Bug 1171773 - Reftest for grid item with different direction from its container. r=mats (84aa60f67)
- Bug 1173662 part 1: Rename local variable 'wm' in nsNumberControlFrame::Reflow. r=jfkthame (2a409480b)
- Bug 1173662 part 2: Use logical sizes in nsNumberControlFrame::Reflow. r=jfkthame (c408e7b67)
- Bug 1173662 part 3: Use a LogicalPoint for child position in nsNumberControlFrame::Reflow. r=jfkthame (815f9c70f) (0f991de4b)
- import changes from `dev' branch of rmottola/Arctic-Fox:
- Bug 1155478 - Use smart pointers for three members of nsIPresShell; r=roc (5ebb5e433)
- Bug 1150614: place floats correctly in vertical writing modes with dir="rtl", r=jfkthame (020ca8893)
- Tests for bug 1150614, r=jfkthame (edba1ffeb)
- Fix test regressions from bug 1150614, r=jfkthame (a68607adc)
- Bug 1162485 - Ensure the vertical-text.enabled pref is set for reftests that involve vertical writing modes. r=smontagu (948cb3a80)
- Bug 1167930 - Handle direction:rtl in vertical modes when converting a LogicalMargin to physical. r=smontagu (0ca0afe6e)
- Bug 1180643 - Fix the SetY() method on LogicalPoint to update the correct field. r=smontagu (3071253fd)
- Bug 1181087 - part 1 - Eliminate the few callers of physical-coordinate setters on the logical-coordinate geometry classes. r=smontagu (3462b41a8)
- Bug 1181087 - part 2 - Remove the unused physical setter methods on logical-coordinate classes. r=smontagu (0715086a8)
- Bug 1145218 - Make the float manager's writing-mode field debug-only, and assert that it matches what callers are passing in. r=smontagu (ee625acd2)
- Bug 756984 - Collapse the selection on the last text node on the line, skipping br and inline frames when clicking past the end of line; r=roc,ehsan (c4dcea1d6)
- Bug 1131451 part 1 - Replace containerWidth with containerSize in logical-coordinate classes and APIs, frame classes, etc. r=dholbert (ad4b55deb)
- Bug 1173646 part 1: Add reftest with flex items whose writing-mode differs from their flex container. r=mats (3dd3d7f97)
- Bug 1173646 part2: Add reftest variants with different writing-mode values on flex container. (no review) (0c82aad87)
- Bug 1173646 part 3: Make FlexItems cache their WritingMode, for quick access. r=mats (7398c33a6)
- Bug 1173646 part 4: Convert isize/bsize into flex container's writing-mode in flexbox macros that aim to pick a main-axis or cross-axis value. r=mats (80ebd414e)
- Bug 1155488 - Mark nsIPresShell::mHiddenInvalidationObserverRefreshDriver as MOZ_UNSAFE_REF; r=roc (10c98ad8f)
- Bug 1156008 - Convert CapturingContentInfo::mContent to StaticRefPtr; r=roc (9ffee3ef4)
- Bug 1177690 - part 3 - Reftests for bevelled corners on collapsed table border. r=dholbert (26e3aff00)
- Bug 1131451 part 2 - Respect the container height when converting vertical-RTL inline-direction coordinates. r=dholbert (594340514)
- Bug 1131451 followup - Remove one more bidi-in-vertical-mode hack that is no longer required. r=dholbert (696b0194f)
- Bug 1183896. Don't assume that relatively positioned table cells actually have positioned offsets computed, because in the case of position:sticky they don't. r=dholbert (remaining parts of rev c87f3d591) (f868b8a64) (8cd81a861)
- import changes from `dev' branch of rmottola/Arctic-Fox:
- Bug 1184388 - Try to use uniform groups for multidimensional constant arrays and structures, r=jandem. (9a8686e50)
- Bug 1184388 - Don't try to reallocate inline elements buffer of unboxed arrays, r=jandem. (8cae2ec7c)
- Bug 1184388 - Fix transposed arguments to GiveObjectGroup, r=jandem. (d417f04c1)
- Bug 1169514 - Part 3: Use a shared preference style sheet across all pres contexts. r=jwatt (f0eb9653b)
- Bug 1169512 - Add and remove quirk.css rather than enable and disable a clone of it. r=jwatt (537bf9210)
- Bug 1176075 - Add a missing check for nursery objects during Ion compilation, r=jandem. (84da20541) (833190af3)
- import changes from `dev' branch of rmottola/Arctic-Fox:
- Bug 1180608 - Remove bogus assert, r=jandem. (b91acfeb8)
- Bug 1159495 - Change Windows to Unix line endings in GMPUtils.cpp. r=whitespace (7d895d348)
- Bug 1168053 - Unified build fix in dom/media/gmp. r=jwwang (dd78a0aa4)
- Bug 470143 - Part 1/2 - Implement an IM IC for typeof nosuchvar. r=jandem (9e7678d6d)
- Bug 1145442 - Add documentation for JIT strategies. r=shu (caacb732e)
- Bug 1145442 - Update JIT Optimization docs patch. r=shu (6995cada4)
- Bug 1145442 - More jit coach optimization outcome docs. r=shu (84deebdc0)
- Bug 470143 - Part 2/2 - TrackedOptimization changes for TypeOfNoSuchVar. r=shu (fccfb49fa)
- Bug 1178653 - Fix a few more tests for modified error messages. r=bustage in a CLOSED TREE (e14d56dc0)
- Bug 77999 - Part 2: Add RuleProcessorCache. r=dbaron (f79158f64) (798773343)
- import changes from `dev' branch of rmottola/Arctic-Fox:
- Bug 77999 - Part 3: Gather document rules and produce an nsDocumentRuleResultCacheKey in nsCSSRuleProcessors. r=dbaron (0f356122a)
- Bug 77999 - Part 4: Mark nsCSSRuleProcessors as ready for expiration from the RuleProcessorCache once no nsStyleSets are using them. r=dbaron (f7ccd7106)
- Bug 1170888 - Restyle the document in EnsureSafeToHandOutCSSRules if we previously cloned sheet inners outside of that method. r=bzbarsky (9b9b8d5e2)
- Bug 77999 - Part 5: Cache eAgentSheet and eUserSheet rule processors in the RuleProcessorCache. r=dbaron (8e6bb4ee4)
- Bug 77999 - Followup build bustage fix. (a127001fd)
- Bug 77999 - Remove stray printf. (281aabfd3) (371e98abd)

* Notice: the changelog above may not always applicable to XULRunner code which K-Meleon uses.

A goanna3 source tree that has kmeleon adaption patch applied is available here: https://github.com/roytam1/palemoon27/tree/kmeleon76

@roytam1 thanks

@roytam1 thanks!

@roytam1 thanks!

Images from aliexpress pages still can not be shown.