Jump to content

My Browser Builds (Part 2)


Recommended Posts


On 11/7/2019 at 5:11 AM, Kwasiarz said:

@VistaLover I have a similar problem with GMail freezing the whole browser whenever I close a tab with it open. Even reported it here but was ignored, and no, sadly creating a new profile doesn't help in this case.

Thanks for trying a new profile and confirming that it doesn't help.

I don't think it's a matter of you and @kitaro1 being ignored, so much as the fact that no one here has a solution to the problem. I haven't tested it, but I strongly suspect "official" Pale Moon and Basilisk (on Win 7+) will have the same issue. I seriously doubt this is a bug that both @roytam1 and @feodor2 have both independently managed to introduce into their respective PM forks.

So unless @looking4awayout's UOC patch helps, we're at the mercy of MCP to come up with a fix. Unfortunately the browsers (that @kitaro1 mentioned) that don't have this issue with GMail simply don't run on XP (or Vista), and can't be made XP-compatible.

If that's the case, you know the drill: confirm the issue with official PM on a supported OS (Win 7 or later) and report it to MCP directly via PM's Help / Submit Feedback link.

 

On 11/7/2019 at 11:54 AM, siria said:

It's a real Killer-Trap that different browsers share the same profile, without any warning. Even advanced users need a LOT of experience to even consider such a possibility before trying a new browser. Even when fully aware it's a fork.

I agree; this has bitten me too when trying multiple versions of, say, Firefox, on the same PC. @VistaLover has the right answer - install at most one non-portable version of each browser - but it's easy to forget and not always practical.

Luckily some combinations do work: Basilisk/Serpent 52, Serpent 55, PM/NM, BNavigator, Seamonkey, and FF all use different profile folders, so one of each of those can coexist peacefully without using "portable" versions.

On 11/7/2019 at 11:54 AM, siria said:

IMO at the very least a confirmation/warning should pop up when installing/unzipping a new browser and it adopts an already existing profile.

That's a bit much to ask of, say, 7-Zip; how's it supposed to know that what you're unzipping is a browser with this particular potential issue? But I'll look at adding a message of that sort to my "Download Latest Browser" batch file.

 

On 11/7/2019 at 6:23 PM, caliber said:

e10s = multi process on NM28 ?

AFAIK e10s/multiprocess only works on FF and Serpent, not NM; trying to enable it on NM just crashes it. But @roytam1 is customizing NM by locking some e10s-related preferences, so inadvertently installing the "wrong" UOC patch from @looking4awayout won't crash the browser.

Link to comment
Share on other sites

Why msfn no potrable paf exe browsers?
Unpacking is just the beginning.
I install a lot of portable programs. In the case of the mypal installer, which launches the application / extraction to created applications, it also adds items to regedit and firewall
I also use portable paf programs. Example: https://sourceforge.net/projects/portableapps/files/Mozilla Firefox%2C P.E. Legacy/Mozilla Firefox%2C Portable Edition Legacy52 52.9.0/
It is a convenient and installed / extracted browser based on the firewall approval operation.

Edited by kitaro1
Link to comment
Share on other sites

2 hours ago, Mathwiz said:

AFAIK e10s/multiprocess only works on FF and Serpent, not NM; trying to enable it on NM just crashes it. But @roytam1 is customizing NM by locking some e10s-related preferences, so inadvertently installing the "wrong" UOC patch from @looking4awayout won't crash the browser.

it doesn't crash here it just has no real effect but I admit I once saw  2 opened processes in the taskbar manager. :yes:

screenshot.3.png

Link to comment
Share on other sites

On 11/4/2019 at 5:10 AM, jumper said:

Startpage.com gives me a "No cypher overlap" error.

Not sure if KM 74 has the pref security.ssl3.ecdhe_rsa_aes_256_sha, but setting that pref true is probably your best bet for Startpage.com.

(I have it set false on mine because SHA1 suites are considered weak, but I've enabled other cipher suites that Startpage.com can use.)

Link to comment
Share on other sites

Palemoon28 lagging and GMail freezing, interesting read:
https://forum.palemoon.org/viewtopic.php?f=3&t=20553
https://forum.palemoon.org/viewtopic.php?t=20843
https://forum.palemoon.org/viewtopic.php?f=3&t=20573

Regarding profiles again, if anyone likes to choose parallel profiles at browser start too:
in older Gecko versions (but perhaps in current forks too? no idea) that setting was stored here:
in the current profile's parent folder, in a "profiles.ini" file, this line:
[General]
StartWithLastProfile=0

Edited by siria
Link to comment
Share on other sites

On 11/7/2019 at 11:04 AM, looking4awayout said:

Try using the UOC Patch and the Enforcer, if you are using an old computer. See if it makes a difference.

@3dreal Why don't you try FF45 ESR SSE alongside the UOC Patch and the Enforcer? ;)

I did really finally install UOC Patch but didnt know about Enforcer. But if i dont err i installed in PM/NM27 for SSE. Whats next?
UPDATE:
have found the UOC-thread. site already closed.

and found out: wrong UOC-patch Firefox 45 to PM 27-sse used. how to rectify before right installation?

I didnt see any improvements. PM-crashes are Facebook-related before FB made this terrible update. cannot be used anymore but crippled with the m.facebook-trick see roytam1s trick by preplacing www. to m. better forgetting FB with palemoon on SSE.  maybe its my machine only.
So UOC only for FF45 ESR SSE? and Enforcer on your site?


I am using Kmeleon for FB and MSFN-board since when i surf here, not logged in yet-i am getting a blue wall. very few visible. cannot use the board with PM-SSE then. Cannot switch to alternative drive since there i have a login-problem and other drive are boot-dead. only could save back older backup to new drive. or new winxp-setup(dont have time now).

MS and FB are about to kill PM! thats clear, no doubt about it. Yahoo is killing its groups all data will be deleted soon.....! Geocities free was killed 2009 etc.

Beside this:I will go back to analoge(photography) being independent. And since i am a lucky pensionist i dont need to deliver anything to anyone but to me....
Update: cannot use PM and MSFN anymore its so buggy. switching back and forth between correct screen and blue wall. i am now on Kmeleon which seems to crash also from time to time, wheras PM_SSE(if i dont err will observe) is crashing in FB mainly. Yes i have a hardware-problem. not enough space etc.
----------------------------------------
UPDATE: PM/NM 27-SSE + FB
Surprised that i could see a user-video-was life-replay here. and almost perfect reg. sound.
Didnt yet change anything reg. prefs.
Will do the KMeleon-profile-workaround. but i am still unsure since i cannot install it. Do i have to find profiles at different locations?Every update has its own folder from where i am starting KM. its on second partition. maybe i should try saving next update in system partition? to make installation working.

Till now i did only adjust PM/NM 27 when updating where it was needed. Adjustments must be made after upates and resets. Very strange: since some weeks search in about:config is not working. i mean if there is an entry present already still search doesnt display it. so then i always have to check to be sure. i will observe.
 

Edited by 3dreal
Video in FB working again-PM/NM 27-SSE crappy Pc
Link to comment
Share on other sites

17 hours ago, siria said:

@Kitaro1
I'm absolutely with you here. It's a real Killer-Trap that different browsers share the same profile, without any warning. Even advanced users need a LOT of experience to even consider such a possibility before trying a new browser. Even when fully aware it's a fork.
Am just not sure how difficult it would be to change this behavior. Or if there's any chance that at least a new "parallel" profile is created? (Absolutely love being able to choose from parallel profiles at startup, a native Mozilla function since ages, no idea if still possible in 'modern' versions now)
But IMO at the very least a confirmation/warning should pop up when installing/unzipping a new browser and it adopts an already existing profile.

@3dreal:
if not noticed yet, mobile facebook URL is "mobile.facebook.com"
Some sites don't respect mobile URLs though and redirect to full desktop size, if the useragent is not a mobile string too. If you like a GUI to change UAs in KM, I recommand my useragents2018-macro ;-) It makes full use of the native engine functions, toggling only prefs, doing nothing else, and contrary to other macros allows to fake the global default UA too.

I have replaced www. by m. but FB is so rudementarily then its better changing  to KMeleon. the only FB-browser for SSE.  do i need an addon there.
What about alternative to Nocript in PM? Https Everywhers-since advised-also deactivated. Complete YT saver nonworking-once worked. Is not important now. and some videos on WMP are buggy too. will use an alternative player. or change MOBO or PC finally. enought lost lifetime!

Link to comment
Share on other sites

@3dreal What are the specs of your machine? I do not have Facebook, but I believe that Facebook should have no issues with Firefox 45 ESR SSE by Roytam1. Back when I was on Facebook (a very short time, as I don't like social networks) it ran perfectly fine in mobile mode with the UOC Patch. If I wanted to use Facebook Messenger, I used the Facebook Messenger plugin for Pidgin. It worked wonderfully and perhaps there's a similar plugin for Miranda NG as well.

About Youtube videos, you can just use YT 2 Player to automatically redirect videos to VLC, or if you really want to watch them in browser, just download GreaseMonkey and the Viewtube script, in order to replace the HTML5 player with a VLC window (once you choose VLC, of course). I watch Youtube videos at 576p without issues. Previously, with the overclocked 6800GT, I watched them at 480p. Don't forget to use UAControl to block the user agent so Youtube will switch to the old, more lightweight layout. It scrolls like butter on my Tualatin.

Keep in mind you need at least a DirectX 9.0C and a 128 bit buss (256 bit is preferred) graphics card in order to enjoy the benefits of the UOC Patch.

Give it a try before dumping your old computer. If I can comfortably browse the web on a Pentium III-S, you can do pretty much the same. Provided you have the suitable parts and you meet the requirements of the Patch.

 

On 11/8/2019 at 5:39 AM, 3dreal said:

I did really finally install UOC Patch but didnt know about Enforcer. But if i dont err i installed in PM/NM27 for SSE. Whats next?
UPDATE:
have found the UOC-thread. site already closed.

and found out: wrong UOC-patch Firefox 45 to PM 27-sse used. how to rectify before right installation?

I didnt see any improvements. PM-crashes are Facebook-related before FB made this terrible update. cannot be used anymore but crippled with the m.facebook-trick see roytam1s trick by preplacing www. to m. better forgetting FB with palemoon on SSE.  maybe its my machine only.
So UOC only for FF45 ESR SSE? and Enforcer on your site?


I am using Kmeleon for FB and MSFN-board since when i surf here, not logged in yet-i am getting a blue wall. very few visible. cannot use the board with PM-SSE then. Cannot switch to alternative drive since there i have a login-problem and other drive are boot-dead. only could save back older backup to new drive. or new winxp-setup(dont have time now).

MS and FB are about to kill PM! thats clear, no doubt about it. Yahoo is killing its groups all data will be deleted soon.....! Geocities free was killed 2009 etc.

Beside this:I will go back to analoge(photography) being independent. And since i am a lucky pensionist i dont need to deliver anything to anyone but to me....
Update: cannot use PM and MSFN anymore its so buggy. switching back and forth between correct screen and blue wall. i am now on Kmeleon which seems to crash also from time to time, wheras PM_SSE(if i dont err will observe) is crashing in FB mainly. Yes i have a hardware-problem. not enough space etc.
 

Why do you keep sticking to Pale Moon 27? Use Roytam1's Firefox 45 ESR SSE. Do not mix the codebases of the UOC Patch, they are different because they are not compatible for every codebase. ;)

Always start with a new profile when you install the UOC Patch and Enforcer. The old profile might be corrupted and so, cause issues.

In order to replace the wrong version you have installed in Pale Moon 27, just download the right version (38 ESR) and replace it the same way you have installed it first. Remember that you cannot use the 45 ESR version of the patch on Pale Moon 27 and New Moon 27 because they use the codebase of Firefox 38 and so, they are not compatible with the 45 ESR version of the UOC Patch and Enforcer. You will have to use the 38 ESR one. The reasons are many, but the most important ones are that PM27/NM27 do not support Asynchronous Pan and Zoom, and Tiled Compositing is bugged.

But as I said, I recommend you to use Firefox 45 ESR SSE. It does support Async pan/zoom and tiled compositing works.

Sorry for the double post, I haven't noticed your reply before. :)

Edited by looking4awayout
Link to comment
Share on other sites

Reading about such killer probs on "full" facebook and even here on MSFN is rather scary. Am stuck on a yet older system and browser version, used to such pains on 'modern' sites and wasting much time finding workarounds, but didn't imagine that younger browsers are already so very badly broken too.

But if you say K-Meleon (76.2?) works better there, that means NewMoon27 must be able to work the same way, since both use the same engine. Unless there are yet more probs involved?
But probably just a matter of fiddling with different useragent-strings or css-tricks. Some special useragent strings for stubborn sites come already predefined in the browsers, and need newer strings occasionally. You could check those on about:config page, filtering just for 'override' or 'facebook'.
Although those prefs only work if no "Addon" is installed, since those I've looked at closer all overrule that native useragent system with their own one.
You can also open those problematic sites and check the really sent UA on scratchpad or whatever those non-KM browsers have, by a little scriptlet:
alert(navigator.userAgent);

Link to comment
Share on other sites

14 hours ago, looking4awayout said:

@3dreal What are the specs of your machine? I do not have Facebook, but I believe that Facebook should have no issues with Firefox 45 ESR SSE by Roytam1. Back when I was on Facebook (a very short time, as I don't like social networks) it ran perfectly fine in mobile mode with the UOC Patch. If I wanted to use Facebook Messenger, I used the Facebook Messenger plugin for Pidgin. It worked wonderfully and perhaps there's a similar plugin for Miranda NG as well.

About Youtube videos, you can just use YT 2 Player to automatically redirect videos to VLC, or if you really want to watch them in browser, just download GreaseMonkey and the Viewtube script, in order to replace the HTML5 player with a VLC window (once you choose VLC, of course). I watch Youtube videos at 576p without issues. Previously, with the overclocked 6800GT, I watched them at 480p. Don't forget to use UAControl to block the user agent so Youtube will switch to the old, more lightweight layout. It scrolls like butter on my Tualatin.

Keep in mind you need at least a DirectX 9.0C and a 128 bit buss (256 bit is preferred) graphics card in order to enjoy the benefits of the UOC Patch.Give it a try before dumping your old computer. If I can comfortably browse the web on a Pentium III-S, you can do pretty much the same. Provided you have the suitable parts and you meet the requirements of the Patch.

sorry now correct. had a kmeleon-crash again when editing. new data lost... should have saved before restarting.
better editing outside and posting then. here the old message:

1. Videocard is strong enough but has a driver-problem. resolution reduced and screen squeezed then. MOBO has multiple problems due to capacitor-plague. Its not that grave. cannot or do not want to switch MOBO since second PC (SSE2) is not setup again. same capacitor plague problem have a huge thread in sysopt-forum, better not reading since complicated, need time and strong nerves. but i learnt a lot in last 15 months about how pcs are really working.
Gigabyte GA-7N400 Pro 2-universal SATA-IDE-Raid MOBO. SSE   4/200 other drives boot-dead and second has a loginui.exe problem. must urgently do a full back, systenpartition first. not enough space on systempartition. 3 partitions. second= copy of systempartition(was a mistake since now not enough space) yes will not change partition. could be suicide now.
2nd SSE-2 PC. Asrock Alive -Dual eSata 2. was wonderfully working had w2k(still have drives), Winxp and Win7-drives. everythink ok. 16/2TB. Have now multiple MOBOs of both brands.  quickest possible processor.
some capacitors are ill, the one who is resposible to hold cmos-data is ill/dead. cannot switch PSU totally offline or i must reenter setting during every boot.
2. did now a smartdefrag2-using also perfectdisk 11 and defraggler. i am switching. should defrag more often.
3. there are/were some strange things now in about.config. search was nonworking anymore.

Drive seems clean and temperatures ok. replaced ram and videocard. pc full open outside case. yet i am cooling HDD from outside(with multi-protected ice-brick! plus fan-rig.
I am here with Kmeleon
Will first backup and when in 1TB drive i will tryout what you suggested. systempartition must be enlarged.
Best i will do a complete new setup. First i want to reactivate second PC. Or i will install i7 beforehand. One bank is forcing us to use win10 for ebanking at 2020. but i have netbook and phablet. also there not powerful enough but lightweight netbook. ok to surf and leaving data alone.
4. yes never touched profiles did that in earlier Firefoxes.
Pls give me hints/link how to renew profiles for all browsers i can use: FF 45 ESR, Kmeleon and PM 27SSE.

Edited by 3dreal
Link to comment
Share on other sites

12 hours ago, looking4awayout said:

In order to replace the wrong version you have installed in Pale Moon 27, just download the right version (38 ESR) and replace it the same way you have installed it first.

One thing to watch out for: the file names are different (e.g., UOC_Patch_38.js vs. UOC_Patch_45.js). So besides copying the _38.js files, you also need to delete the *_45.js files you copied earlier; otherwise NM 27 will load both and the result will probably not be what you want!

Link to comment
Share on other sites

rmottola upstream pushed lots of commits of front-end side, so some merge commits issued and new test build is uploaded: http://o.rths.ml/gpc/files1.rt/arcticfox-27.9.18.win32-20191109.7z

- Bug 1090583 part 2 - Fix the property count check in MonitorAssign to check the actual count instead of the capacity. r=bhackett (3f1daa802)
- Bug 1090583 - Fix a pre-existing issue with MLoadElementHole and negative index checks. r=bhackett (1278f0b1a)
- Bug 1141190 - Remove docshell/test/moz.build; r=mshal (641a1716f)
- bug 1161402 assert that some public methods are called on reader task queue r=cpearce (c8355d774)
- bug 1161402 assert that public methods are not called after shutdown r=cpearce (07db4d8d0)
- cleanup and update (0c236d56a)
- Bug 1137906 - Shutdown cached children when shutdown the document. r=surkov (45e3894eb)
- Bug 1141676 - Reftest for sizing of synthetic sub/superscript glyphs r=jdaggett (eab4327fb)
- Bug 1141748 - Do not use non-standard flag argument of String.prototype.replace in CSSUnprefixingService.js. r=dholbert (631be2090)
- Bug 1134537 - Delete GnomeVFS extension. r=roc (cd03d727a)
- Bug 1063880 - Allow a --disable-compile-environment to go through a full mach build. r=mshal (67ef7e874)
- Bug 1063880 - Do not emit Linkable-derived objects when COMPILE_ENVIRONMENT is not set. r=mshal (7e4b1f82a)
- Bug 1141660 - Always export memalign from jemalloc, even on Windows. r=njn (b625580bf)
- Bug 1141731 - Avoid problems caused by the inclusion of mozmemory.h in mozalloc.cpp by including mozmemory_wrap.h instead. r=njn (ad25d296d)
- Bug 1135354 - Crash/Hang when clicking URL with a huge length r=mcmanus (84b653c37)
- update (3a1d896a3)
- remove errneously commited patch (39668bbff)
- Merge remote-tracking branch 'upstream/master' into fix-winbuild (30889728d)
- Bug 1138293 - Use malloc/free/realloc/calloc instead of moz_malloc/moz_free/moz_realloc/moz_calloc. r=njn (620eb32f2)
- Bug 1138293 - Remove moz_malloc/moz_free/moz_realloc/moz_calloc. r=njn (62f747129)
- Bug 968520 - Explicitly specify the allocator for nsTArray_base functions. r=froydnj (8e11c1a0c)
- Bug 968520 - Add fallible variant of nsTArray::SetCapacity. r=froydnj (3a33bc53e)
- Bug 968520 - Add fallible variants of nsTArray::InsertElementsAt. r=froydnj (d52030d81)
- Bug 968520 - Add fallible variant of nsTArray::EnsureLengthAtLeast. r=froydnj (7f22637b6)
- Bug 968520 - Add fallible variant of nsTArray::SetLength. r=froydnj (a6f43b1b0)
- Bug 968520 - Add fallible variants of nsTArray::ReplaceElementsAt. r=froydnj (e17169b21)
- Bug 968520 - Add fallible variants of nsTArray::InsertElementAt. r=froydnj (0ed62209a)
- Bug 968520 - Add fallible variants of nsTArray::AppendElements. r=froydnj (c4f4d2a1e)
- Bug 968520 - Add fallible variants of nsTArray::AppendElement. r=froydn (3fbf72f7a)
- Bug 1165729 - Remove unnecessary template declaration for fallible nsTArray::AppendElement. r=froydnj (f72cc5efd)
- Bug 968520 - Make nsTArray::InsertElementSorted support moves. r=froydnj (217178725)
- Bug 968520 - Add fallible variants of nsTArray::InsertElementSorted. r=froydnj (f99e300b9)
- Bug 968520 - Add more fallible variants of nsTArray::InsertElementsAt. r=froydnj (218d4f025)
- update with patch that I could not apply without breaking things (9312c3780)
- Bug 1072814 - Convert trivial Promise.defer() usages to new Promise() r=yoric (475e1ce95)
- Bug 1072814 - Fix SessionFile.jsm to use new Promise() API r=yoric (ad7699876)
- Bug 1072814 - Emulate old Promise.defer() API for non-trivial cases r=yoric (60b0f5ac9)
- Merge remote-tracking branch 'upstream/master' into fix-winbuild, nsTArray*.h are replaced with one in https://hg.mozilla.org/releases/mozilla-esr45/file/31e36b1649741834a9775b8e2fca02164f002145/xpcom/glue (4616d3107)
- Bug 1218882 - lz4.js should be usable outside of workers, r=Yoric. (360bb7839)
- Merge remote-tracking branch 'upstream/master' into fix-winbuild (2d279bfd4)
- Bug 1140264 part 1 - Avoid expanding auto-hidden ruby annotations. r=dbaron (5b6ddba10)
- Bug 1140264 part 2 - Move and merge line breaking code in rbc reflow to reduce useless computation and fix undesired break point. r=dbaron (0d3aaad79)
- Bug 1140264 part 3 - Rename IsInlineDescendantOfRuby to ShouldSuppressLineBreak and exclude rbc/rtc from it. r=dbaron (e929b296b)
- Bug 1140264 part 4 - Reftest for this bug. r=dbaron (3aa19d44e)
- Bug 1141842 - Add LastOptionalBreakPriority to nsLineLayout to simplify some code. r=roc (81cc1dcd0)
- Bug 1137875 - Have the SharedFrameMetricsData hold a copy of the layers id. r=nical (81c657510)
- Bug 1133237 - When transitioning from a shadow layer tree that has plugins to a tree that does not, make sure the old plugins get hidden properly. r=roc (11ff16c17)
- Bug 1137875 - Have the LayerTransactionChild hold a copy of the layers id. r=nical (f9a090aad)
- Bug 1137875 - Clear shared frame metrics objects associated with a PLayerTransaction when it is destroyed. r=nical (c5037ee27)
- Bug 1119878 Part 1: Change SandboxTarget to hold sandbox target services to provide functions. r=aklotz, r=glandium, r=cpearce (7378d13e4)
- Bug 1119878 Part 2: Change IPC code to hold ProcessID instead of ProcessHandle. r=billm, r=dvander, r=aklotz, r=cpearce (fa002f559)
- Bug 1149971 Part 1: Follow-up to bug 1119878 to fix broken IPDL tests. r=dvander (76b843d43)
- Revert Bug 1125848 - Consolidate PCompositor's creation-destruction logic because of 10.7 MacOS bustage (ceda5a133)
- adapt assertion to be as introduced in Bug 1125848 (877071282)
- Bug 1160190 followup. Make ServiceWorker actually disabled on mulet so we can reopen the CLOSED TREE. r=bkelly (1a03ee7c2)
- Bug 1123846 - Restrict some activities to be provided by the system app r=ferjm (c7ca76805)
- fix header include order (01b1289df)
- Bug 1151644 - Don't disallow the basic compositor backend. r=jrmuizel (b70633afd)
- Merge remote-tracking branch 'upstream/master' into fix-winbuild (ecb5c656e)
- Bug 1155823 - Properly shutdown the CompositorVsyncDispatcher. r=kats (a3dee13e8)
- add gfxCrashReporterUtils as of 1180688  2015-07-13 (433fa6bdb)
- Bug 1029673 - Correctly report OMTC compositing in crash reports - r=Bas (87fc22936)
- Bug 1180688 - Detect whether the widget will be able to present frames with BasicCompositor on Mac. r=mstange (842ed309f)
- Merge remote-tracking branch 'upstream/master' into fix-winbuild (e1d15a47f)
- Bug 1125848 - Consolidate PCompositor's creation-destruction logic. r=sotaro (bd9905d37)
- Bug 1180688 - Reset the appropriate widget states if creating a compositor fails. r=mstange (461339a0c)
- missing part of Bug 1119878 Part 2: Change IPC code to hold ProcessID instead of ProcessHandle. (631e112dc)
- Bug 1065785 - [e10s] Use session restore to reload crashed tabs. r=smacleod. (5c1a2cc21)
- Merge remote-tracking branch 'upstream/master' into fix-winbuild (a9fdac034)

Link to comment
Share on other sites

1 hour ago, Mathwiz said:

One thing to watch out for: the file names are different (e.g., UOC_Patch_38.js vs. UOC_Patch_45.js). So besides copying the _38.js files, you also need to delete the *_45.js files you copied earlier; otherwise NM 27 will load both and the result will probably not be what you want!

I have no Palemoon-folder in Programme(german). I explained installer non-working. palemoonx86 is empty so is 64 bit version. maybe created during my installation-trials. so i must surf to palemoon-folders which will be changed after every update? i am extracting updates into seperate folders. as i said installer non-working.
will first install patch for FF 45ESR where i didnt yet patch UOC. and remove the wrong one beforehand in Palemoon-folder. i have a FF-folder in Programme but its from earlier FF-installations. So where to install the patches? yes i know the workaround but must run FF and PM outside programs aka programme-folder.

i have these folders: I removed wrongly set FF-patch in PM-folder. maybe i didnt backup prefs.js beforehand.
so i must extract/copy UOC Patch to prefs-folder beside channel-prefs?

pm27sse_uocpatch.png

Moonchild Mozilla.png

Edited by 3dreal
Link to comment
Share on other sites

New build of Serpent/UXP for XP!

Test binary:
Win32 https://o.rths.ml/basilisk/basilisk52-g4.4.win32-git-20191109-4a0c734e1-xpmod.7z
Win64 https://o.rths.ml/basilisk/basilisk52-g4.4.win64-git-20191109-4a0c734e1-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.rths.ml/palemoon/palemoon-28.8.0a1.win32-git-20191109-4a0c734e1-xpmod.7z
Win64 https://o.rths.ml/palemoon/palemoon-28.8.0a1.win64-git-20191109-4a0c734e1-xpmod.7z

Official repo changes since my last build:
- Issue #1064: Don't get certificate details synchronously. (3252e2200)
- No issue: Clean up `exceptionDialog.js` (a63272b53)
- MoonchildProductions#1251 - Part 1: Restore initial Solaris support, fixed up. (afc187cc3)
- MoonchildProductions#1251 - Part 2: Make the mozconfig loader POSIX-compliant. (9d449ce61)
- MoonchildProductions#1251 - Part 3: Finally end the long tradition of casting getpid() to int. (64e03d014)
- MoonchildProductions#1251 - Part 4: Core build system changes, lots of libevent/IPC junk. (4105ebb6e)
- MoonchildProductions#1251 - Part 5: Fix POSIX compliance issue in process_util.h. (f105b741e)
- MoonchildProductions#1251 - Part 6: Solaris needs an audio implementation. (57bfda37a)
- MoonchildProductions#1251 - Part 7: All the posix_m* memory-related stuff, gathered together. (3647f42c2)
- MoonchildProductions#1251 - Part 8: Align pointer for char_16t. (c0d814c1d)
- MoonchildProductions#1251 - Part 9: Look for hypot in the math library (libm). (cf75ede0d)
- MoonchildProductions#1251 - Part 10: ipc_channel_posix.cc should use IOV_MAX. (52f2321cb)
- MoonchildProductions#1251 - Part 11: libpng uses C89 now to avoid buffer overflows. (7966d22df)
- MoonchildProductions#1251 - Part 12: Add Solaris/illumos support to WasmSignalHandlers. (76c55f747)
- MoonchildProductions#1251 - Part 13: Redefining abort in C++ requires extern "C" (5c28f10c1)
- MoonchildProductions#1251 - Part 14: libstagefright build should avoid _GLIBCXX_OS_DEFINES. (e3fb8bd24)
- MoonchildProductions#1251 - Part 15: fdlibm should provide definition for u_int32_t and u_int64_t. (4f6639a1b)
- MoonchildProductions#1251 - Part 16: Resolve namespace conflicts with dbm on Solaris. (fca7c45a6)
- MoonchildProductions#1251 - Part 17: All the libffi and libxul.so issues, resolved. (beea314ff)
- MoonchildProductions#1251 - Part 18: (Hopefully) final version of build system fixes. (7e5ff857e)
- MoonchildProductions#1251 - Part 19: Make the unpreprocessed file script work on Solaris. (e51afbcc2)
- MoonchildProductions#1251 - Part 20: Add atomic.h to system-headers. (db34ef993)
- Fix a bunch of dumb typos and omissions. (575f51a27)
- MoonchildProductions#1251 - Part 21: Fix Flash player and some extensions being incompatible. (5a4a4990c)
- MoonchildProductions#1251 - Part 22: Remove some unused type declarations from IPC process_util. (7d65eb2b3)
- MoonchildProductions#1251 - Part 23: Allow AMD64 build to work. (2f4488521)
- MoonchildProductions#1251 - Part 24: Remove temporary GNU M4 workaround. (9cd92944e)
- MoonchildProductions#1251 - Part 25: Fix link paths. (21146d0d5)
- MoonchildProductions#1251 - Part 26: Oracle Solaris gsed/ICU fix. (e4f90cbb5)
- MoonchildProductions#1251 - Part 27: Fix ifdef style. (687a798e6)
- Fix nits. (a9dc528a4)
- Merge pull request #1262 from athenian200/solaris-work (21b3f6247)
- Issue #146 - Part 1: Draw each table's background on their own display list items. (145527207)
- Issue #146 - Part 2: Remove custom table painting component. (0de40040f)
- Issue #1265 - Expose sndio as a build option for any supporting system (873f2cf8c)
- Issue #1267 - Part 1: Update libcubeb to a1200c34. (d162ecbaf)
- Issue #1267 - Part 2: fix libcubeb bindings in dom (22b35fa8e)
- Merge pull request #1270 from g4jc/libcubeb (b7c6c779e)
- Merge pull request #1269 from g4jc/sndio (6ee3467a6)
- Issue #146 - Part 3: Create nsDisplayTableFixedPosition to avoid display list collisions when processing the background image of a table. (db98e3eff)
- Issue #146 - Part 3: Create nsDisplayTableBackgroundImage to avoid display list collisions when processing the background image of a table. (ee663e293)
- Issue #146 - Part 4: Adjust tests for fixes. (f1b043af1)
- Issue #146 - Part 5: Treat table row groups as containing blocks. (ce11d5cae)
- Issue #146 - Part 6: Allow `position: sticky` on table elements. (5ab2da700)
- Issue #146 - Part 7: Remove no longer relevant reftest. (76052e837)
- Revert "Issue #1267 - Part 2: fix libcubeb bindings in dom" (732fe3624)
- Revert "Issue #1267 - Part 1: Update libcubeb to a1200c34." (1960d6e08)
- Bump Goanna version. (ee8c5d387)
- Issue #1271 - Fix build failure in current in-tree libcubeb sndio module (ff0f4b11a)
- Merge pull request #1272 from g4jc/libcubeb (22b410f31)
- Merge branch 'table-sticky-work' (bbc2206a0)
- Issue #1252 - Remove fragile fast path. (082670ecb)
- Merge branch 'master' into certexception-work (c5c44d120)
- Issue #1064 - Part 2: Fix shorthand and services module import. (5f37447ac)
- Issue #1064 - Part 3: Fix notifyObservers() call. (6df8aa495)
- Issue #1274 - Part 1: Adopt the cairo version as our own. (7efd71616)
- Issue #1274 - Part 2: Update Windows prerequisite defines. (45f665b59)
- Issue #1274 - Part 3: Update libpixman licensing (620fdbdf4)
- Issue #1274 - Part 4: Remove some unnecessary checks for old Windows versions. (7b8a065ea)
- Issue #1274 - Part 5: Misc fixes and version update. (5f80c286c)
- Issue #1275 - Add pref to disable warning prompts for SuperfluousAuth and AutomaticAuth, and default to not prompting. (7f5d38150)
- Merge branch 'master' of https://github.com/MoonchildProductions/UXP (aff468a48)
- Issue #1263 - Part 1: Remove DiskSpaceWatcher (60c83971f)
- Issue #1263 - Part 2: Remove DiskSpaceWatcher tests (eb35cd108)
- Merge pull request #1214 from MoonchildProductions/certexception-work (8bb208397)
- Merge pull request #1276 from MoonchildProductions/cairo-work (0f3635dc8)
- Issue #1254 - Update SQLite to 3.30.1 (4a0c734e1)

My changes since my last build:
- Reverted Issue #1274 - Part 2: Update Windows prerequisite defines.
- Reverted Issue #1274 - Part 4: Remove some unnecessary checks for old Windows versions. (7b8a065ea)

Link to comment
Share on other sites

Guest
This topic is now closed to further replies.
  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...