Jump to content
MSFN is made available via donations, subscriptions and advertising revenue. The use of ad-blocking software hurts the site. Please disable ad-blocking software or set an exception for MSFN. ×

My browser builds (part 2)


Recommended Posts

On 12/7/2020 at 1:44 PM, ArcticFoxie said:

What good is a browser that renders pages in 100 milliseconds if it takes 10 seconds for the GUI to load? 

That is what I want -- To render page and switch between tabs fast, but the loading time is not important, for I always have the browser open, even for weeks. I do not like to close it and open it again, unless it crashes (very rarely), or I need more free memory to run another big program. I'd like to know if there is any program that could measure the time of rendering page and switching tabs.

Link to post
Share on other sites

I don't know about for SWITCHING tabs.

My issue has more been about CLOSING tabs - especially for very long browsing sessions like you mention versus opening-and-closing the browser for new sessions.

Others have posted it also over the years - I really dislike clicking the X on a tab then sitting there and waiting 2 to 4 seconds for the tab to close!

This was my biggest issue with NM28 and it is one of several reasons why I prefer NM27.

I have not ran Serpent in any very long browsing session to test if this is an issue with Serpent.

I am also unaware if it is "still" an issue with NM28 - perhaps others that have reported it in the past will chime in if it is still an issue.

 

I am also open for "better" ways to test page rendering.

What I used for the below is the Status-4-Evar and PageTimerLite add-ons.

NM27 barely beats Serpent from what I see here  --  I did three page loads and posted the averages.

 

2020-12-09_9-59-59.jpg.63c23842367183f5faf2d90f02e832bb.jpg

Link to post
Share on other sites
On 12/9/2020 at 11:17 PM, ArcticFoxie said:

 

I am also open for "better" ways to test page rendering.

What I used for the below is the Status-4-Evar and PageTimerLite add-ons.

I searched the add-ons you mentioned but found PageTimerLite is nowhere to be downloaded. So I did some google and found AppTelemetry, and the treasure in the browser itself: ctrl+shift+Q.

A sample result of this page:

1.PNG p. s. The time meaningful for comparison speed of browsers, I think, is "Processing duration", which in above picture is 17184 ms, for that does not include network transmission time.

Edited by luweitest
p.s.
Link to post
Share on other sites

I do notice sluggish performance of the NM28 browser after a longer session, as if some tabs are still open in the background, consuming memory and CPU time. What sort of plugins is the "plugin container" supposed to handle? Flash is almost never used anymore, and the other plugin I have is PDF-Xchange, which also almost never open.

Link to post
Share on other sites
1 hour ago, j7n said:

I do notice sluggish performance of the NM28 browser after a longer session, as if some tabs are still open in the background, consuming memory and CPU time.

Old bugs inherited from the old Firefox codebase that appear to have been fixed in later official Firefox releases.

1 hour ago, j7n said:

What sort of plugins is the "plugin container" supposed to handle?

All NPAPI plugins, the list is on about:plugins and about:addons pages (under Plugins on the latter).

Edited by UCyborg
Link to post
Share on other sites
9 hours ago, luweitest said:

<snip>  and the treasure in the browser itself: ctrl+shift+Q.  </snip>

Sweet!  I like it, didn't know it was there!

Has proven rather interesting.  NM27 still clobbers Serpent.

Need to be careful though, some websites serve different content to different browsers so this table ensures that each browser was served the same "total requests" and "transferred size".

 

I included "Site H" for an intriguing reference (I found over a dozen sites that basically did the same thing).

The two web browsers on the same exact site had 203 requests in Serpent but only 109 requests in NM27.

You place one on one monitor and the other on the other monitor, scroll through each very carefully, and you can NOT find any DIFFERENCE between the two "renderings".

One took 40 seconds to load, the other only took 6 seconds.

One has 62 js files, the other has only 32 js files - but again NO VISUAL DIFFERENCE between the two!

One says 79 images, the other says only 42 images - but again scroll carefully and NO VISUAL DIFFERENCE between the two!

I haven't really dug deeper as to seek out any root cause for such a GIGANTIC difference when the two look IDENTICAL.

 

2020-12-10_11-25-04.jpg.0e4c01131e3d6230162bae8fafc3608a.jpg

Edited by ArcticFoxie
Link to post
Share on other sites
9 hours ago, j7n said:

What sort of plugins is the "plugin container" supposed to handle? Flash is almost never used anymore, and the other plugin I have is PDF-Xchange, which also almost never open.

... More details available below:

http://kb.mozillazine.org/Plugin-container_and_out-of-process_plugins#Plugin-container

:)

  • Like 1
Link to post
Share on other sites

New build of Serpent/UXP for XP!

Test binary:
Win32 https://o.rthost.win/basilisk/basilisk52-g4.7.win32-git-20201212-6d0527a-uxp-4281fcc16-xpmod.7z
Win64 https://o.rthost.win/basilisk/basilisk52-g4.7.win64-git-20201212-6d0527a-uxp-4281fcc16-xpmod.7z

source code that is comparable to my current working tree is available here: https://github.com/roytam1/UXP/commits/custom

IA32 Win32 https://o.rthost.win/basilisk/basilisk52-g4.7.win32-git-20201212-6d0527a-uxp-4281fcc16-xpmod-ia32.7z

source code that is comparable to my current working tree is available here: https://github.com/roytam1/UXP/commits/ia32

NM28XP build:
Win32 https://o.rthost.win/palemoon/palemoon-28.10.2a1.win32-git-20201212-31167a236-uxp-4281fcc16-xpmod.7z
Win64 https://o.rthost.win/palemoon/palemoon-28.10.2a1.win64-git-20201212-31167a236-uxp-4281fcc16-xpmod.7z

Official UXP changes since my last build:
- Issue #1694 - Part 1: Use scriptabledateformat for the Cookie Accept dialog. (16a1ff22a)
- Issue #1694 - Part 2: Use scriptabledateformat for Update History display. (6cb00e2cf)
- Issue #1695 - Fix socket timeout logic. (6b45065f1)
- Revert "Issue #1391 - Disable DOM Filesystem/dirpicker APIs by default." (4281fcc16)

Official Basilisk changes since my last build:
- Issue #31 - Part 1: Use nsIScriptableDateFormat in Page Info. (c7a029f)
- Issue #31 - Part 2: Use nsIScriptableDateFormat in feeds. (8a74eeb)
- Issue #31 - Part 3: Use nsIScriptableDateFormat in places library window. (5f97253)
- Issue #31 - Part 4: Use nsIScriptableDateFormat in cookie preferences. (01c2847)
- Issue #31 - Part 5: Update back-end branch pointer for toolkit changes. (6d0527a)

Official Pale-Moon changes since my last build:
- Back-end branch pointer update (unstable 2020-12-06) (31167a236)

My changes since my last build:
- Reverted Issue UXP#1694 and Basilisk#31 related changes

  • Like 4
Link to post
Share on other sites

New build of BOC/UXP for XP!

Test binary:
MailNews Win32 https://o.rthost.win/boc-uxp/mailnews.win32-20201212-355db4de-uxp-4281fcc16-xpmod.7z
BNavigator Win32 https://o.rthost.win/boc-uxp/bnavigator.win32-20201212-355db4de-uxp-4281fcc16-xpmod.7z

source repo (excluding UXP): https://github.com/roytam1/boc-uxp/tree/custom-older

* Notice: the profile prefix (i.e. parent folder names) are also changed since 2020-08-15 build, you may rename their names before using new binaries when updating from builds before 2020-08-15.

--

New build of HBL-UXP for XP!

Test binary:
IceDove-UXP(mail) https://o.rthost.win/hbl-uxp/icedove.win32-20201212-id-eed0566-uxp-4281fcc16-xpmod.7z
IceApe-UXP(suite) https://o.rthost.win/hbl-uxp/iceape.win32-20201212-id-eed0566-ia-41157bf-uxp-4281fcc16-xpmod.7z

source repo (excluding UXP):
https://github.com/roytam1/icedove-uxp/tree/winbuild
https://github.com/roytam1/iceape-uxp/tree/winbuild

for UXP changes please see above.

  • Like 1
Link to post
Share on other sites

New NewMoon 27 Build!

32bit https://o.rthost.win/palemoon/palemoon-27.9.7.win32-git-20201212-cafd4d45b-xpmod.7z
32bit SSE https://o.rthost.win/palemoon/palemoon-27.9.7.win32-git-20201212-cafd4d45b-xpmod-sse.7z
32bit noSSE https://o.rthost.win/palemoon/palemoon-27.9.7.win32-git-20201212-cafd4d45b-xpmod-ia32.7z

64bit https://o.rthost.win/palemoon/palemoon-27.9.7.win64-git-20201212-cafd4d45b-xpmod.7z

source repo: https://github.com/roytam1/palemoon27

repo changes since my last build:
- import changes from `dev' branch of rmottola/Arctic-Fox:
 - Bug 1159056. Don't store a value in CSS pixels in a variable with type nscoord in ScrollFrameHelper::SetCoordAttribute. r=dholbert (432f06158)
 - Bug 1138442 - Disable reftests on Mulet in parity with B2G/B2G Desktop. r=ahal (8b5c55da7)
 - Don't clip in ComputeFrameMetrics if the scrollframe won't be layerized. (bug 1148582 part 2, r=tn) (088b6ef29)
 - Bug 1157579 - Make sure the visual scroll port size includes the area of scrollbars when used in TryLayout. r=tn (c35244374)
 - Bug 1139306: Fix margin sides for right-to-left scrollbars depending on the actual position of the scrollbar as determined by layout.scrollbar.side, r=tn (34328312d)
 - Fix some full zoom FrameMetrics changes not propagating to APZ. (bug 1147648, r=botond) (91f009ac4)
 - Bug 1159405 - Fix "Found a non-root APZ with no handoff parent". r=botond (09b0ddf07)
 - Bug 1161040 - Don't set a scrollparent on the root content document in the child process that points to itself. r=tn (d0dfbffb0)
 - Bug 1161040 - Add some assertions to guard against infinite loops in the handoff chains. r=botond (e6bae82e8)
 - Bug 1153539 patch 1 - Improve comments for nsLayoutUtils::ComputeSuitableScaleForAnimation. r=roc (1b2bc00cd)
 - Bug 1153539 patch 2 - Compute scale for rasterizing off-main-thread transform animations based on ratio to display size in addition to maximum and minimum scale. r=roc (97d632d0f)
 - Bug 1160250 - Extract a helper method from the triplicated comp-bounds function to determine if we need to deflate the scrollbars. r=botond,tn (8d79c845f)
 - Bug 1160250 - Extract a helper function from the triplicated composition-bounds-updating code. r=botond,tn (5b3f71329)
 - Bug 1160250 - Collapse the code from CalculateRootCompositionSize into UpdateCompositionBounds. r=botond,tn (db0f081d6)
 - Bug 1160250 - Put in a missing transfomToAncestor scaling factor. r=botond,tn (14837aa3a)
 - Bug 1160250 - Fix up the rect being used for the composition size calculation. r=tn (0ba58021f)
 - Bug 1139575 - Reset velocity on axes when going from a pan to a pinch. r=botond (2e03d04ab)
 - Bug 1136971 - Draw layer borders correctly in the presence of multi-FrameMetrics. r=kats (e72f5b171)
 - Bug 1148871 - Only apply checkerboarding background color to opaque layers. r=botond (20e070b84)
 - Bug 1157327 - Don't cull out container layers, because their prepared data might be null and cause crashes. r=nical (3d06c925f)
 - Bug 1157327 follow-up to add a comment. r=me and DONTBUILD (4cd74c0a1)
 - Bug 1160566 - Make FrameMetrics.mCompositionBounds private and add a getter/setter for it. r=botond (277dd9afd)
 - Bug 1160566 - Change some return values to be reference-to-const. r=botond (2ce17b63f)
 - Bug 1158933 - Prevent scrollbars from jumping while zoomed in the B2G browser. r=botond (eca8aae7d)
 - Bug 1158933 - Replace the compositedHeight/scrollableHeight (and width) term with the adjusted scrollThumbRatio. r=botond (236af4c4a)
 - Bug 1156401 - Fix painting when building fennec with apzc. r=kats, r=botond (be18764b9)
 - Bug 1159305 - patch 1 - Provide logical accessors for nsStylePosition and nsStyleSides fields. r=dbaron (7a90c519d)
 - Bug 1159305 - patch 2 - Clean up layout code by using the new accessors. r=smontagu (926eb7dd5)
 - Bug 1164406 - Expose scrollbar-area deflation in nsLayoutUtils. r=kats (42e5e8784)
 - Bug 1139895 - Add fuzz for 1062792-1.html. r=tn (e38d0ac73)
 - Bug 1139893 - Disable reftest bugs/1062108-1.html on mulet. r=jfkthame (930590d73)
 - Move the APZ scroll frame clip onto FrameMetrics. (bug 1148582 part 3, r=mstange,tn) (c8e05f8d7)
 - Fix PostprocessRetainedLayers to not test occlusion with asynchronous clips. (bug 1148582 part 5, r=tn) (cc71cafe2)
 - Bug 1146626 - Refresh the reftest snapshot after adjusting the async scroll offsets. r=roc (1f041955d)
 - Bug 1160642. Add reftest-async-zoom to apply an async zoom before taking snapshot. r=dbaron (1e8e54d8d)
 - Bug 1160285 - Add a staticruntime version of mfbt and link some stuff against it. r=glandium (c733c49c7)
 - Bug 786520 - Install things to $(DIST)/branding from moz.build instead of manual rules in Makefile.ins. r=mshal (42d38eb26)
 - Bug 1155776 - move USE_EXTENSION_MANIFEST to moz.build; r=mshal (a87ecc2ca)
 - Bug 1155816 - part 0 - remove NO_JS_MANIFEST from testing/mochitest/Makefile.in; r=mshal (af90a362c)
 - Bug 1155816 - part 1 - move NO_JS_MANIFEST to moz.build; r=mshal (3a55e99a0)
 - Bug 978591 - Remove per-directory uses of MOZ_CHROME_FILE_FORMAT. r=glandium (5084d62b5)
 - Bug 991983 - Set GARBAGE for GeneratedSources in the recursivemake backend. r=gps (b8ae57c14)
 - Bug 991983 - Emit absolute paths for UnifiedSources. r=gps (4100640d0)
 - Bug 991983 - Emit absolute paths for other sources. r=gps (f30ddd773)
 - Bug 991983 - Remove commented code in gyp_reader.py. r=gps (0685eca6a)
 - Bug 1167474 - Remove ELFHACK_BUILD hack. r=mshal (b9d56c0b2)
 - Bug 1097804 - Part 1 - Create a library containing nsISocketTransportService and nsIDNS that can be used to support standalone WebRTC. r=mcmanus (d2f83eb1c)
 - Bug 1097804 - Part 2 - Added xpcomrt build of library containing unicode util functions. r=gps (fa9306f34)
 - Bug 1162852 - Remove EXTRA_COMPILE_FLAGS. r=gps (340312bc6)
 - Bug 1162852 - Remove EXTRA_ASSEMBLER_FLAGS. r=gps (f124c902c)
 - Bug 1162852 - Remove the VARIABLES method of GypContext now it is doing nothing. r=gps (b6c1c0246)
 - accidental commit (a1bbb1ada)
 - Bug 991983 - Define SOURCES as SourcePath. r=gps (678065bce)
 - Bug 991983 - Use objdir-relative SOURCES instead of GENERATED_SOURCES. r=gps (b54db0304)
 - Bug 991983 - Make TEST_HARNESS_FILES use the *Path classes instead of a separate set of methods to resolve paths. r=gps (025fe1ca9)
 - Bug 991983 - Add a deprecation hint for GENERATED_SOURCES. r=gps (67c75f4ab)
 - Bug 1172800 - Fixup after bug 991983. r=gps (ce6e8c7a3)
 - Bug 1164218 - Allow running individual mochitests and reftests in chaos mode. r=roc,froydnj (f2ec1d9d8)
 - Bug 1154231 - Part 1. Use LRU to manage remote layers. r=mattwoodrow (1fe47e64f)
 - Bug 1128037 - Minimalist support for wildcards in jar manifests. r=gps (1b07a6ffc)
 - Bug 1162569 - default engine files should be in the omni.ja file, r=markh,glandium. (b07e54881)
 - Bug 1157279. Escaping CSS identifiers should use lowercase letters for hex digits, not uppercase ones. r=dbaron (6f220936e) (729a46113)
- import changes from `dev' branch of rmottola/Arctic-Fox:
 - Bug 1077651 Measure frame uniformity by synthesizing native events. r=kats,mrbkap (95a796145)
 - Bug 1154231 - Part 2. Reclaim cached resources when memory-pressure occurs. r=mattwoodrow (cff6ac1f9)
 - fix build (15112d396)
 - Bug 1155649 - XFlush at the end of frames when OMTC is enabled on Linux. r=jrmuizel (8bc22dd42)
 - Bug 1158120 - Edit include and comments that contained gfxIntSize and nsIntSize. r=nical (fc25c0c96)
 - Bug 1167235 - Part 1: Add code exposing a PersistentBufferProvider. r=nical (8340f8ede)
 - Bug 1134251 - Disable broken B2G Desktop tests on Mulet. r=jmaher, r=jgilbert (db4d4253c)
 - Bug 1134271 - Skip conformance/canvas/buffer-offscreen-test.html on Android; r=jgilbert (2fe8d8623)
 - Bug 1124996 - Remove expected fail on OSX 10.10. r=jgilbert (388438a7e)
 - Bug 1145492 - Update FramebufferTexture2D to allow binding mipmaps. r=jgilbert (5e4c6bf95)
 - Bug 1158089 - Remove LAYERS_D3D10 enum value since it's unused. r=Bas (720d3b3b4)
 - Bug 1156058 - Null pointer check. r=jgilbert (f9b9c6ca6)
 - Bug 1131463 - Report AtomicRefCounterWithFinalize doing the wrong thing with AddRef and Release in release build as well. r=sotaro (7c009766e)
 - Bug 1142071 - Re-add WaitForBufferOwnership() r=jgilbert (0753bcd25)
 - partial revert of patch not found... (4ed1e76f1)
 - Bug 1143979 - Use RAII local instead of useless temporary. - r=kamidphish (d8a50143f)
 - Bug 1017865 - Refactor attach/detach for FB attachments. - r=kamidphish (9c7e41065)
 - Bug 1144906 - Add accel E10S backend for WebGL compositing. - r=jrmuizel,mattwoodrow,nical,sotaro (559ab767f)
 - Bug 1167235 - Part 2: Add support for the basic buffer provider to CanvasLayer. r=nical (4ac399c16)
 - Bug 1155252 - Add a pref to control the maximum canvas 2d size and set it to 0x7ff. r=jrmuizel (7ae6cde3f)
 - Bug 1167235 - Part 3: Switch CanvasRenderingContext2D to use the new BufferProvider API. r=nical (f678c14ab)
 - Bug 1192159: Do not forget about the transform when not using an active target. r=jrmuizel (a4bd28a75)
 - Bug 1167235 - Part 4: Remove DrawTarget as a possible means of initializing Canvas layers. r=nical (dbd153cfb)
 - Bug 1167235 - Part 5: Make CanvasLayerComposite ImageHost type agnostic. r=nical (df15bd85d)
 - Bug 1167235 - Part 6: Fix up HasInternalBuffer return value on TextureHostDirectUpload. r=nical (437cd1680)
 - Bug 1091851 - Fix a race condition in Sqlite.jsm shutdown. r=mak (8b6ac8848) (3b276a539)
- import changes from `dev' branch of rmottola/Arctic-Fox:
 - Bug 1168630. Part 1. Add layers id to hit test node. r=botond (710ee191d)
 - Bug 1168630. Part 2. Only return an APZC from GetAPZCAtPoint that has the same layers id as the layer that was hit. r=botond (61adedf3c)
 - Bug 1168630. Part 3. Add IsLayersIdRoot to frame metrics. r=botond (4ebd296b3)
 - Bug 1151663 - Extract some helper functions for writing APZ mochitests into a separate file. r=kats (eacc622a4)
 - Bug 1168630. Part 4. Rename AsyncPanZoomController::IsRootForLayersId to HasNoParentWithSameLayersId. r=botond (0e543aa5c)
 - Bug 1168630. Part 5. If GetAPZCAtPoint hit a layer but did not find an APZC on the ancestor chain with the same layers id then return the root APZC of the same layers id. r=botond (40462b612)
 - Bug 1168630. Part 6. Modify the test for bug 1119497 based on the new assumption of always having a root APZC. r=botond (4ae4a4631)
 - Use the correct scrollframe clips with nested scrollframes in APZ. (bug 1148582 part 6, r=botond) (15c7c447d)
 - Fix asynchronously scrolling containerful subframes. (bug 1148582 part 6.1, r=tn) (abab9c2ca)
 - Bug 994541 - Skip 789933-1.html on Linux debug due to frequent crashes. (20854625f)
 - Bug 1184349 - Add lr to SavedNonVolatileRegisters on ARM64. r=efaust (b6aa8e624)
 - pointer style (01aaf5ffd)
 - Bug 1186122 - Part 1/2 - Remove StackPointer on ARM64. r=efaust (ac1171b98)
 - Bug 1185757 - Fix manual manipulation of StackPointer for JSOP_IN. r=efaust (eff86b56b)
 - Bug 1186122 - Part 2/2 - Remove BaselineStackReg from ARM64. r=efaust (c64f767ad)
 - Bug 1188650 - Part 2/2 - Make shared/LOpcodes-shared.h. r=nbp (fcd067fe3)
 - Bug 1182936 part 2 - IonMonkey: MIPS32: Move mips macros to mips32. r=nbp (dc17ea527)
 - Bug 1134744 - Reftest for baseline of radio button and checkbox. r=smontagu (9fd9e79ca)
 - Bug 1136557 - Reftests with nested spans in vertical mode. r=smontagu (7971e5104)
 - Bug 1142928 - Make 1090168-3.html fuzzy on Mulet as on B2G. r=jfkthame (6797807c9)
 - Bug 1134849 - Reftest for orthogonal inline elements. r=dbaron (d44daa3ef)
 - Bug 1082249. Invalidate background-attachment:fixed table-part backgrounds when scrolled. r=mattwoodrow (2df3218b3)
 - Bug 1144080 - Disable reftests on Mulet for TaskCluster. r=ahal (2a26fa749)
 - Bug 1145218 followup - Mark 1136557-*-nested-spans reftests fuzzy on Windows. no_r=me (833fab4ca)
 - Bug 1130907 - patch 1 - Account for writing-mode when incorporating border widths into intrinsic size. r=smontagu (a3b0dfa1c)
 - Bug 1130907 - patch 2 - Account for writing-mode in nsContainerFrame::DoInlineIntrinsicISize. r=smontagu (33f5c2333)
 - Bug 1130907 - patch 3 - Reftest for contribution of border widths to intrinsic size. r=smontagu (cafc14f18)
 - Bug 1130907 - patch 4 - Reftest for container-frame intrinsic size involving borders. r=smontagu (5b2898c3e)
 - Bug 1145936 - Reftest for sizing of foreignObject with vertical writing mode. r=smontagu (28ce89062)
 - Bug 1047973 - pattern with patternUnits=objectBoundingBox and preserveAspectRatio=... meet stretches to the size of a shape. r=dholbert (bb40de381)
 - Bug 1142982 - implement dominant-baseline=middle properly. r=heycam (20c0c9d66)
 - Bug 932958 - Use the Skia Software backend by default for canvas on OS X. r=jrmuizel (08c78d8dd) (cafd4d45b)

Link to post
Share on other sites

New regular/weekly KM-Goanna release:
https://o.rthost.win/kmeleon/KM76.3.1-Goanna-20201212.7z

Changelog:

Out-of-tree changes:
* update Goanna3 to git fa9ed1e11..cafd4d45b:
- import changes from `dev' branch of rmottola/Arctic-Fox:
 - Bug 1159056. Don't store a value in CSS pixels in a variable with type nscoord in ScrollFrameHelper::SetCoordAttribute. r=dholbert (432f06158)
 - Bug 1138442 - Disable reftests on Mulet in parity with B2G/B2G Desktop. r=ahal (8b5c55da7)
 - Don't clip in ComputeFrameMetrics if the scrollframe won't be layerized. (bug 1148582 part 2, r=tn) (088b6ef29)
 - Bug 1157579 - Make sure the visual scroll port size includes the area of scrollbars when used in TryLayout. r=tn (c35244374)
 - Bug 1139306: Fix margin sides for right-to-left scrollbars depending on the actual position of the scrollbar as determined by layout.scrollbar.side, r=tn (34328312d)
 - Fix some full zoom FrameMetrics changes not propagating to APZ. (bug 1147648, r=botond) (91f009ac4)
 - Bug 1159405 - Fix "Found a non-root APZ with no handoff parent". r=botond (09b0ddf07)
 - Bug 1161040 - Don't set a scrollparent on the root content document in the child process that points to itself. r=tn (d0dfbffb0)
 - Bug 1161040 - Add some assertions to guard against infinite loops in the handoff chains. r=botond (e6bae82e8)
 - Bug 1153539 patch 1 - Improve comments for nsLayoutUtils::ComputeSuitableScaleForAnimation. r=roc (1b2bc00cd)
 - Bug 1153539 patch 2 - Compute scale for rasterizing off-main-thread transform animations based on ratio to display size in addition to maximum and minimum scale. r=roc (97d632d0f)
 - Bug 1160250 - Extract a helper method from the triplicated comp-bounds function to determine if we need to deflate the scrollbars. r=botond,tn (8d79c845f)
 - Bug 1160250 - Extract a helper function from the triplicated composition-bounds-updating code. r=botond,tn (5b3f71329)
 - Bug 1160250 - Collapse the code from CalculateRootCompositionSize into UpdateCompositionBounds. r=botond,tn (db0f081d6)
 - Bug 1160250 - Put in a missing transfomToAncestor scaling factor. r=botond,tn (14837aa3a)
 - Bug 1160250 - Fix up the rect being used for the composition size calculation. r=tn (0ba58021f)
 - Bug 1139575 - Reset velocity on axes when going from a pan to a pinch. r=botond (2e03d04ab)
 - Bug 1136971 - Draw layer borders correctly in the presence of multi-FrameMetrics. r=kats (e72f5b171)
 - Bug 1148871 - Only apply checkerboarding background color to opaque layers. r=botond (20e070b84)
 - Bug 1157327 - Don't cull out container layers, because their prepared data might be null and cause crashes. r=nical (3d06c925f)
 - Bug 1157327 follow-up to add a comment. r=me and DONTBUILD (4cd74c0a1)
 - Bug 1160566 - Make FrameMetrics.mCompositionBounds private and add a getter/setter for it. r=botond (277dd9afd)
 - Bug 1160566 - Change some return values to be reference-to-const. r=botond (2ce17b63f)
 - Bug 1158933 - Prevent scrollbars from jumping while zoomed in the B2G browser. r=botond (eca8aae7d)
 - Bug 1158933 - Replace the compositedHeight/scrollableHeight (and width) term with the adjusted scrollThumbRatio. r=botond (236af4c4a)
 - Bug 1156401 - Fix painting when building fennec with apzc. r=kats, r=botond (be18764b9)
 - Bug 1159305 - patch 1 - Provide logical accessors for nsStylePosition and nsStyleSides fields. r=dbaron (7a90c519d)
 - Bug 1159305 - patch 2 - Clean up layout code by using the new accessors. r=smontagu (926eb7dd5)
 - Bug 1164406 - Expose scrollbar-area deflation in nsLayoutUtils. r=kats (42e5e8784)
 - Bug 1139895 - Add fuzz for 1062792-1.html. r=tn (e38d0ac73)
 - Bug 1139893 - Disable reftest bugs/1062108-1.html on mulet. r=jfkthame (930590d73)
 - Move the APZ scroll frame clip onto FrameMetrics. (bug 1148582 part 3, r=mstange,tn) (c8e05f8d7)
 - Fix PostprocessRetainedLayers to not test occlusion with asynchronous clips. (bug 1148582 part 5, r=tn) (cc71cafe2)
 - Bug 1146626 - Refresh the reftest snapshot after adjusting the async scroll offsets. r=roc (1f041955d)
 - Bug 1160642. Add reftest-async-zoom to apply an async zoom before taking snapshot. r=dbaron (1e8e54d8d)
 - Bug 1160285 - Add a staticruntime version of mfbt and link some stuff against it. r=glandium (c733c49c7)
 - Bug 786520 - Install things to $(DIST)/branding from moz.build instead of manual rules in Makefile.ins. r=mshal (42d38eb26)
 - Bug 1155776 - move USE_EXTENSION_MANIFEST to moz.build; r=mshal (a87ecc2ca)
 - Bug 1155816 - part 0 - remove NO_JS_MANIFEST from testing/mochitest/Makefile.in; r=mshal (af90a362c)
 - Bug 1155816 - part 1 - move NO_JS_MANIFEST to moz.build; r=mshal (3a55e99a0)
 - Bug 978591 - Remove per-directory uses of MOZ_CHROME_FILE_FORMAT. r=glandium (5084d62b5)
 - Bug 991983 - Set GARBAGE for GeneratedSources in the recursivemake backend. r=gps (b8ae57c14)
 - Bug 991983 - Emit absolute paths for UnifiedSources. r=gps (4100640d0)
 - Bug 991983 - Emit absolute paths for other sources. r=gps (f30ddd773)
 - Bug 991983 - Remove commented code in gyp_reader.py. r=gps (0685eca6a)
 - Bug 1167474 - Remove ELFHACK_BUILD hack. r=mshal (b9d56c0b2)
 - Bug 1097804 - Part 1 - Create a library containing nsISocketTransportService and nsIDNS that can be used to support standalone WebRTC. r=mcmanus (d2f83eb1c)
 - Bug 1097804 - Part 2 - Added xpcomrt build of library containing unicode util functions. r=gps (fa9306f34)
 - Bug 1162852 - Remove EXTRA_COMPILE_FLAGS. r=gps (340312bc6)
 - Bug 1162852 - Remove EXTRA_ASSEMBLER_FLAGS. r=gps (f124c902c)
 - Bug 1162852 - Remove the VARIABLES method of GypContext now it is doing nothing. r=gps (b6c1c0246)
 - accidental commit (a1bbb1ada)
 - Bug 991983 - Define SOURCES as SourcePath. r=gps (678065bce)
 - Bug 991983 - Use objdir-relative SOURCES instead of GENERATED_SOURCES. r=gps (b54db0304)
 - Bug 991983 - Make TEST_HARNESS_FILES use the *Path classes instead of a separate set of methods to resolve paths. r=gps (025fe1ca9)
 - Bug 991983 - Add a deprecation hint for GENERATED_SOURCES. r=gps (67c75f4ab)
 - Bug 1172800 - Fixup after bug 991983. r=gps (ce6e8c7a3)
 - Bug 1164218 - Allow running individual mochitests and reftests in chaos mode. r=roc,froydnj (f2ec1d9d8)
 - Bug 1154231 - Part 1. Use LRU to manage remote layers. r=mattwoodrow (1fe47e64f)
 - Bug 1128037 - Minimalist support for wildcards in jar manifests. r=gps (1b07a6ffc)
 - Bug 1162569 - default engine files should be in the omni.ja file, r=markh,glandium. (b07e54881)
 - Bug 1157279. Escaping CSS identifiers should use lowercase letters for hex digits, not uppercase ones. r=dbaron (6f220936e) (729a46113)
- import changes from `dev' branch of rmottola/Arctic-Fox:
 - Bug 1077651 Measure frame uniformity by synthesizing native events. r=kats,mrbkap (95a796145)
 - Bug 1154231 - Part 2. Reclaim cached resources when memory-pressure occurs. r=mattwoodrow (cff6ac1f9)
 - fix build (15112d396)
 - Bug 1155649 - XFlush at the end of frames when OMTC is enabled on Linux. r=jrmuizel (8bc22dd42)
 - Bug 1158120 - Edit include and comments that contained gfxIntSize and nsIntSize. r=nical (fc25c0c96)
 - Bug 1167235 - Part 1: Add code exposing a PersistentBufferProvider. r=nical (8340f8ede)
 - Bug 1134251 - Disable broken B2G Desktop tests on Mulet. r=jmaher, r=jgilbert (db4d4253c)
 - Bug 1134271 - Skip conformance/canvas/buffer-offscreen-test.html on Android; r=jgilbert (2fe8d8623)
 - Bug 1124996 - Remove expected fail on OSX 10.10. r=jgilbert (388438a7e)
 - Bug 1145492 - Update FramebufferTexture2D to allow binding mipmaps. r=jgilbert (5e4c6bf95)
 - Bug 1158089 - Remove LAYERS_D3D10 enum value since it's unused. r=Bas (720d3b3b4)
 - Bug 1156058 - Null pointer check. r=jgilbert (f9b9c6ca6)
 - Bug 1131463 - Report AtomicRefCounterWithFinalize doing the wrong thing with AddRef and Release in release build as well. r=sotaro (7c009766e)
 - Bug 1142071 - Re-add WaitForBufferOwnership() r=jgilbert (0753bcd25)
 - partial revert of patch not found... (4ed1e76f1)
 - Bug 1143979 - Use RAII local instead of useless temporary. - r=kamidphish (d8a50143f)
 - Bug 1017865 - Refactor attach/detach for FB attachments. - r=kamidphish (9c7e41065)
 - Bug 1144906 - Add accel E10S backend for WebGL compositing. - r=jrmuizel,mattwoodrow,nical,sotaro (559ab767f)
 - Bug 1167235 - Part 2: Add support for the basic buffer provider to CanvasLayer. r=nical (4ac399c16)
 - Bug 1155252 - Add a pref to control the maximum canvas 2d size and set it to 0x7ff. r=jrmuizel (7ae6cde3f)
 - Bug 1167235 - Part 3: Switch CanvasRenderingContext2D to use the new BufferProvider API. r=nical (f678c14ab)
 - Bug 1192159: Do not forget about the transform when not using an active target. r=jrmuizel (a4bd28a75)
 - Bug 1167235 - Part 4: Remove DrawTarget as a possible means of initializing Canvas layers. r=nical (dbd153cfb)
 - Bug 1167235 - Part 5: Make CanvasLayerComposite ImageHost type agnostic. r=nical (df15bd85d)
 - Bug 1167235 - Part 6: Fix up HasInternalBuffer return value on TextureHostDirectUpload. r=nical (437cd1680)
 - Bug 1091851 - Fix a race condition in Sqlite.jsm shutdown. r=mak (8b6ac8848) (3b276a539)
- import changes from `dev' branch of rmottola/Arctic-Fox:
 - Bug 1168630. Part 1. Add layers id to hit test node. r=botond (710ee191d)
 - Bug 1168630. Part 2. Only return an APZC from GetAPZCAtPoint that has the same layers id as the layer that was hit. r=botond (61adedf3c)
 - Bug 1168630. Part 3. Add IsLayersIdRoot to frame metrics. r=botond (4ebd296b3)
 - Bug 1151663 - Extract some helper functions for writing APZ mochitests into a separate file. r=kats (eacc622a4)
 - Bug 1168630. Part 4. Rename AsyncPanZoomController::IsRootForLayersId to HasNoParentWithSameLayersId. r=botond (0e543aa5c)
 - Bug 1168630. Part 5. If GetAPZCAtPoint hit a layer but did not find an APZC on the ancestor chain with the same layers id then return the root APZC of the same layers id. r=botond (40462b612)
 - Bug 1168630. Part 6. Modify the test for bug 1119497 based on the new assumption of always having a root APZC. r=botond (4ae4a4631)
 - Use the correct scrollframe clips with nested scrollframes in APZ. (bug 1148582 part 6, r=botond) (15c7c447d)
 - Fix asynchronously scrolling containerful subframes. (bug 1148582 part 6.1, r=tn) (abab9c2ca)
 - Bug 994541 - Skip 789933-1.html on Linux debug due to frequent crashes. (20854625f)
 - Bug 1184349 - Add lr to SavedNonVolatileRegisters on ARM64. r=efaust (b6aa8e624)
 - pointer style (01aaf5ffd)
 - Bug 1186122 - Part 1/2 - Remove StackPointer on ARM64. r=efaust (ac1171b98)
 - Bug 1185757 - Fix manual manipulation of StackPointer for JSOP_IN. r=efaust (eff86b56b)
 - Bug 1186122 - Part 2/2 - Remove BaselineStackReg from ARM64. r=efaust (c64f767ad)
 - Bug 1188650 - Part 2/2 - Make shared/LOpcodes-shared.h. r=nbp (fcd067fe3)
 - Bug 1182936 part 2 - IonMonkey: MIPS32: Move mips macros to mips32. r=nbp (dc17ea527)
 - Bug 1134744 - Reftest for baseline of radio button and checkbox. r=smontagu (9fd9e79ca)
 - Bug 1136557 - Reftests with nested spans in vertical mode. r=smontagu (7971e5104)
 - Bug 1142928 - Make 1090168-3.html fuzzy on Mulet as on B2G. r=jfkthame (6797807c9)
 - Bug 1134849 - Reftest for orthogonal inline elements. r=dbaron (d44daa3ef)
 - Bug 1082249. Invalidate background-attachment:fixed table-part backgrounds when scrolled. r=mattwoodrow (2df3218b3)
 - Bug 1144080 - Disable reftests on Mulet for TaskCluster. r=ahal (2a26fa749)
 - Bug 1145218 followup - Mark 1136557-*-nested-spans reftests fuzzy on Windows. no_r=me (833fab4ca)
 - Bug 1130907 - patch 1 - Account for writing-mode when incorporating border widths into intrinsic size. r=smontagu (a3b0dfa1c)
 - Bug 1130907 - patch 2 - Account for writing-mode in nsContainerFrame::DoInlineIntrinsicISize. r=smontagu (33f5c2333)
 - Bug 1130907 - patch 3 - Reftest for contribution of border widths to intrinsic size. r=smontagu (cafc14f18)
 - Bug 1130907 - patch 4 - Reftest for container-frame intrinsic size involving borders. r=smontagu (5b2898c3e)
 - Bug 1145936 - Reftest for sizing of foreignObject with vertical writing mode. r=smontagu (28ce89062)
 - Bug 1047973 - pattern with patternUnits=objectBoundingBox and preserveAspectRatio=... meet stretches to the size of a shape. r=dholbert (bb40de381)
 - Bug 1142982 - implement dominant-baseline=middle properly. r=heycam (20c0c9d66)
 - Bug 932958 - Use the Skia Software backend by default for canvas on OS X. r=jrmuizel (08c78d8dd) (cafd4d45b)

* 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

Link to post
Share on other sites

Notice: Since renovation of my home is completed, the relocation of my server will be undo in a day or two(I don't have exact time of action right now). During the period the connection to my server will be unavailable.

Link to post
Share on other sites
On 12/11/2020 at 6:13 PM, roytam1 said:

New build of Serpent/UXP for XP!

Test binary:
Win32 https://o.rthost.win/basilisk/basilisk52-g4.7.win32-git-20201212-6d0527a-uxp-4281fcc16-xpmod.7z

source code that is comparable to my current working tree is available here: https://github.com/roytam1/UXP/commits/custom

NM28XP build:
Win32 https://o.rthost.win/palemoon/palemoon-28.10.2a1.win32-git-20201212-31167a236-uxp-4281fcc16-xpmod.7z

Official UXP changes since my last build:
- Issue #1694 - Part 1: Use scriptabledateformat for the Cookie Accept dialog. (16a1ff22a)
- Issue #1694 - Part 2: Use scriptabledateformat for Update History display. (6cb00e2cf)

Official Basilisk changes since my last build:
- Issue #31 - Part 1: Use nsIScriptableDateFormat in Page Info. (c7a029f)
- Issue #31 - Part 2: Use nsIScriptableDateFormat in feeds. (8a74eeb)
- Issue #31 - Part 3: Use nsIScriptableDateFormat in places library window. (5f97253)
- Issue #31 - Part 4: Use nsIScriptableDateFormat in cookie preferences. (01c2847)
- Issue #31 - Part 5: Update back-end branch pointer for toolkit changes. (6d0527a)

My changes since my last build:
- Reverted Issue UXP#1694 and Basilisk#31 related changes

Many thanks for this new batch of UXP-based forks! :wub:
I, for one, am not taking these builds of yours for granted, they do require dedication and considerable effort on your part (despite "upstream" :angry: constantly belittling your offerings as being just "hackjobs" ... ).

Be that as it may, might I also kindly ask why the official UXP issue #1694,

https://repo.palemoon.org/MoonchildProductions/UXP/issues/1694

and official Bk issue #31,

https://repo.palemoon.org/moonchildproductions/basilisk/issues/31

were backed-out from your custom UXP branch? :dubbio:

The thing is I was actually following closely the original report in the official forums,

https://forum.palemoon.org/viewtopic.php?f=61&t=25728

and immediately thought that would be a favourable change to implement; after all, MCP would just be restoring what was already extant in Mozilla v51.0 and later broken by Mozilla devs :realmad: in v52.0 of their platform...

E.g. my (custom) date/time format configuration in my system is "dd/MM/yyyy HH:mm"; New Moon 28 respects that setting,

iewZiYQ.jpg

because, while the platform code is a Mozilla v52.6 fork, the application code itself is a Firefox < 52.0 fork, so not affected...

OTOH, latest Serpent 52.9.0 does not respect my custom date/time format OS configuration,

QlQEdxB.jpg

because both app+platform code are Mozilla v52.6 forks and inherit the Mozilla caused breakage,,, :(

As a result, Serpent displays date/time in a non-user-configurable US+12h clock format "M/d/yyyy, h:mm tt" :( ; me personally, I would have liked uniformity between NM28 and St52; what do other members here think?

For the record, Mozilla, in later versions of their Firefox browser, tied date/time display to browser locale being used, but even then, the display format is fixed/non-configurable... :} With Serpent 52.9.0 (and now, sadly, NM28 too...) being only an en-US localised app, this is a moot point...

This is just a thought, but perhaps issues UXP#1694 + Bk#31 could be implemented in our tree behind a user (i.e. about:config) pref? ...

Kindest, warmest greetings! :)

  • Like 1
  • Upvote 1
Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
  • Recently Browsing   0 members

    No registered users viewing this page.

×
×
  • Create New...