DanR20 Posted January 10 Posted January 10 6 hours ago, LordGarfio said: Just out of curiosity, I have tested the above URL (https://www.legacy.com/us/obituaries/washingtonpost/browse) and they pass the Cloudflare's captcha in both Serpent v52.9.0 (2026-01-09) (32-bit) and New Moon v28.10.7a1 (32-bit) (2026-01-09) browsers. For now I've gone back to the 2025.12.17 basilisk 55 version since that still works right. Figure there must have some kind of change in the browser in the 2025.12.25 update.
roytam1 Posted January 10 Author Posted January 10 (edited) 8 hours ago, DanR20 said: For now I've gone back to the 2025.12.17 basilisk 55 version since that still works right. Figure there must have some kind of change in the browser in the 2025.12.25 update. thats ICU update, I wonder if upstream encounter same issue or not, or I "enbugged" when backporting pre-NT6 changes. from Bad Request URL above, some dates becoming "Invalid Date" and I need a minimal code to reproduce it. Edited January 11 by roytam1 2
roytam1 Posted 17 hours ago Author Posted 17 hours ago New build of Serpent/UXP for XP! Test binary: Win32 https://o.rthost.win/basilisk/basilisk52-g4.8.win32-git-20260117-3219d2d-uxp-aace8fa8c0-xpmod.7z Win64 https://o.rthost.win/basilisk/basilisk52-g4.8.win64-git-20260117-3219d2d-uxp-aace8fa8c0-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-g4.8.win32-git-20260117-3219d2d-uxp-aace8fa8c0-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-28.10.7a1.win32-git-20260117-d849524bd-uxp-aace8fa8c0-xpmod.7z Win32 IA32 https://o.rthost.win/palemoon/palemoon-28.10.7a1.win32-git-20260117-d849524bd-uxp-aace8fa8c0-xpmod-ia32.7z Win32 SSE https://o.rthost.win/palemoon/palemoon-28.10.7a1.win32-git-20260117-d849524bd-uxp-aace8fa8c0-xpmod-sse.7z Win64 https://o.rthost.win/palemoon/palemoon-28.10.7a1.win64-git-20260117-d849524bd-uxp-aace8fa8c0-xpmod.7z Win7+ x64 AVX2 https://o.rthost.win/palemoon/palemoon-28.10.7a1.win64-git-20260117-d849524bd-uxp-aace8fa8c0-w7plus-avx2.7z Official UXP changes picked since my last build: - Issue #1636 - Part 1 - Decouple Gio from GConf and always build it with GTK (fc78eec911) - Issue #1636 - Part 2 - Disable Gconf by default (990f256885) - Issue #1636 - Part 3 - Remove Gconf (b04e3d93c9) - Issue #2895 - Part 1: Replace XorShift128+ with Xoroshiro128++ (d4341ea07a) - Issue #2895 - Part 1b: Update license block. (d23afa84e5) - Issue #2895 - Part 2: Remove incorrect debug asserts. (5a0f190104) - Issue #2895 - Part 3: Add manual test. (75ec48bc41) - Issue #2889 - Part 0: Relocate Mozilla-specific files from Expat source. (b81b97ee47) - Issue #2889 - Part 1: Prepare Mozilla's changes and extensions for Expat 2.6.4. (9567601675) - Issue #2889 - Part 2: Import Expat 2.6.4 source. (9ade98b4fc) - Issue #2889 - Part 3: Apply Mozilla's patches to Expat 2.6.4. (0b9e221b14) - Issue #2889 - Part 4: Fix build and runtime issues. (a2f8e66162) - Issue #2889 - Follow-up: Avoid conflicts with system version of libexpat (e0f8acddc7) - Issue #2889 - Follow-up: void methods shouldn't return a value. (8956c61187) - Issue #2889 - Part 5: Rewrite patches for Expat 2.7.3. (a730f838c8) - Issue #2889 - Part 6: Import Expat 2.7.3 source. (d2acec73f7) - Issue #2889 - Part 7: Apply patches to Expat 2.7.3. (ca2ef5184a) - Issue #2889 - Follow-up: Allow generate_hash_secret_salt to be used. (acefed74fe) - Issue #2889 - Follow-up: Adjust Expat 2.7.3 local patches for removal of 3rd patch. (034eeaff97) - Issue #2889 - Follow-up: Remove fix for BZ 569229. (493169e227) - Issue #2895 - Follow-up: Fix JIT code generator for 32-bit targets. (d6e9c4d85b) - [gfx] Be more explicit when converting to cairo formats. (f6d18c7d78) - [gfx] Use better types for cairo image surface manipulations. (6ca8cf782b) - [DOM] Check worker script size (d8f7202951) - [Auth] Only set GSS_C_DELEG_FLAG flag when we confirm server supports delegation. (5321363a1b) - [NSS] Update NSS (464b24c608) - [XPCOM] Add hashkey helper classes for future use. (7b9eda3f69) - [gfx] Add some additional sanity checks to cairo surface creation. (3aff60e443) No official Pale-Moon changes picked since my last build. No official Basilisk changes picked since my last build. My changes picked since my last build: - application: remove gio from MOZ_EXTENSIONS_DEFAULT (69c17afbe0) - htmlparser: break when infinite loop happens (8201dadf41) Update Notice: - You may delete file named icudt*.dat and icu63.dll inside program folder when updating from old releases. * Notice: From now on, UXP rev will point to `custom` branch of my UXP repo instead of MCP UXP repo, while "official UXP changes" shows only `tracking` branch changes. 4
roytam1 Posted 17 hours ago Author Posted 17 hours ago New build of BOC/UXP for XP! Test binary: MailNews Win32 https://o.rthost.win/boc-uxp/mailnews.win32-20260117-40a79c75-uxp-aace8fa8c0-xpmod.7z BNavigator Win32 https://o.rthost.win/boc-uxp/bnavigator.win32-20260117-40a79c75-uxp-aace8fa8c0-xpmod.7z source repo (excluding UXP): https://github.com/roytam1/boc-uxp/tree/custom * 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.win32-20260117-id-656ea98-uxp-aace8fa8c0-xpmod.7z IceApe-UXP(suite) https://o.rthost.win/hbl-uxp/iceape.win32-20260117-id-656ea98-ia-c642e3c-uxp-aace8fa8c0-xpmod.7z source repo (excluding UXP): https://github.com/roytam1/icedove-uxp/tree/winbuild https://github.com/roytam1/iceape-uxp/tree/winbuild 2
roytam1 Posted 17 hours ago Author Posted 17 hours ago New build of post-deprecated Serpent/moebius for XP! * Notice: This repo will not be built on regular schedule, and changes are experimental as usual. ** Current moebius patch level should be on par with 52.9, but some security patches can not be applied/ported due to source milestone differences between versions. Test binary: Win32 https://o.rthost.win/basilisk/basilisk55-win32-git-20260117-aa5eeba9b-xpmod.7z Win64 https://o.rthost.win/basilisk/basilisk55-win64-git-20260117-aa5eeba9b-xpmod.7z repo: https://github.com/roytam1/basilisk55 Repo changes: - ported from UXP: Issue #1636 - Part 1 - Decouple Gio from GConf and always build it with GTK (fc78eec9) (522108f71) - ported from UXP: Issue #1636 - Part 2 - Disable Gconf by default (990f2568) (cb5b1cf63) - ported from UXP: Issue #1636 - Part 3 - Remove Gconf (b04e3d93) (c7c7666f8) - import from UXP: - Issue #2895 - Part 1: Replace XorShift128+ with Xoroshiro128++ (d4341ea0) - Issue #2895 - Follow-up: Fix JIT code generator for 32-bit targets. (d6e9c4d8) - Issue #2895 - Part 1b: Update license block. (d23afa84) - Issue #2895 - Part 2: Remove incorrect debug asserts. (5a0f1901) - Issue #2895 - Part 3: Add manual test. (75ec48bc) (1ba1395ca) - import from UXP: - Issue #2889 - Part 0: Relocate Mozilla-specific files from Expat source. (b81b97ee) - Issue #2889 - Part 1: Prepare Mozilla's changes and extensions for Expat 2.6.4. (95676016) - Issue #2889 - Part 2: Import Expat 2.6.4 source. (9ade98b4) - Issue #2889 - Part 3: Apply Mozilla's patches to Expat 2.6.4. (0b9e221b) - Issue #2889 - Part 4: Fix build and runtime issues. (a2f8e661) - Issue #2889 - Follow-up: Avoid conflicts with system version of libexpat (e0f8acdd) - Issue #2889 - Follow-up: void methods shouldn't return a value. (8956c611) - Issue #2889 - Part 5: Rewrite patches for Expat 2.7.3. (a730f838) - Issue #2889 - Part 6: Import Expat 2.7.3 source. (d2acec73) - Issue #2889 - Part 7: Apply patches to Expat 2.7.3. (ca2ef518) - Issue #2889 - Follow-up: Allow generate_hash_secret_salt to be used. (acefed74) - Issue #2889 - Follow-up: Adjust Expat 2.7.3 local patches for removal of 3rd patch. (034eeaff) - Issue #2889 - Follow-up: Remove fix for BZ 569229. (493169e2) (ee15f2005) - application: remove gio from MOZ_EXTENSIONS_DEFAULT (d7440fac8) - htmlparser: break when infinite loop happens (c016f9cc6) - import from UXP: [gfx] Be more explicit when converting to cairo formats. (f6d18c7d) (7094d3f22) - import from UXP: [gfx] Use better types for cairo image surface manipulations. (6ca8cf78) (efc7747c2) - import from UXP: [DOM] Check worker script size (d8f72029) (a480b4db2) - import from UXP: [Auth] Only set GSS_C_DELEG_FLAG flag when we confirm server supports delegation. (5321363a) (4149b3514) - import from UXP: [NSS] Update NSS (464b24c6) (3381dabad) - import from UXP: [XPCOM] Add hashkey helper classes for future use. (7b9eda3f) (c627d5f28) - import from UXP: [gfx] Add some additional sanity checks to cairo surface creation. (3aff60e4) (aa5eeba9b) 1
raddy Posted 3 hours ago Posted 3 hours ago 13 hours ago, roytam1 said: Win32 https://o.rthost.win/palemoon/palemoon-28.10.7a1.win32-git-20260117-d849524bd-uxp-aace8fa8c0-xpmod.7z Failed for me. If i visit https://www.bing.com/translator or click some internal links on rutracker.org - new moon make me "hdd stress test" and fill ram up to 2GB. I'm return to previous palemoon-28.10.7a1.win32-git-20260110-d849524bd-uxp-3e9ca11775-xpmod.7z
roytam1 Posted 3 hours ago Author Posted 3 hours ago (edited) 1 hour ago, raddy said: Failed for me. If i visit https://www.bing.com/translator or click some internal links on rutracker.org - new moon make me "hdd stress test" and fill ram up to 2GB. I'm return to previous palemoon-28.10.7a1.win32-git-20260110-d849524bd-uxp-3e9ca11775-xpmod.7z heard similar report on different websites but both of them are not reproducible here. EDIT: try translating in bing translator, and it does happen: > mozjs.dll!js::jit::BacktrackingAllocator::tryAllocateRegister(js::jit::BacktrackingAllocator::PhysicalRegister & r, js::jit::LiveBundle * bundle, bool * success, bool * pfixed, mozilla::Vector<js::jit::LiveBundle *,4,js::SystemAllocPolicy> & conflicting) Line 1395 C++ [Inline Frame] mozjs.dll!js::jit::BacktrackingAllocator::tryAllocateNonFixed(js::jit::LiveBundle *) Line 1221 C++ mozjs.dll!js::jit::BacktrackingAllocator::processBundle(js::jit::MIRGenerator * mir, js::jit::LiveBundle * bundle) Line 1291 C++ mozjs.dll!js::jit::BacktrackingAllocator::go() Line 835 C++ mozjs.dll!js::jit::GenerateLIR(js::jit::MIRGenerator * mir) Line 1928 C++ mozjs.dll!js::jit::CompileBackEnd(js::jit::MIRGenerator * mir) Line 1998 C++ mozjs.dll!js::HelperThread::handleIonWorkload(js::AutoLockHelperThreadState & locked) Line 1554 C++ mozjs.dll!js::HelperThread::threadLoop() Line 1918 C++ [Inline Frame] mozjs.dll!js::detail::ThreadTrampoline<void (__cdecl&)(void *),js::HelperThread *>::callMain(mozilla::IndexSequence<0>) Line 233 C++ mozjs.dll!js::detail::ThreadTrampoline<void (__cdecl&)(void *),js::HelperThread *>::Start(void * aPack) Line 227 C++ but javascript related changes only happened in https://repo.palemoon.org/MoonchildProductions/UXP/issues/2895 which doesn't look like what shows in malloc-trace shown above. and when memory usage is high (but not too high to be crashed), about:memory shows high heap-unclassified and later it gets reclaimed. hmm, so strange. EDIT2: another one to be suspected is expat library update: https://repo.palemoon.org/MoonchildProductions/UXP/issues/2889 Edited 2 hours ago by roytam1
modnar Posted 2 hours ago Posted 2 hours ago (edited) 1 hour ago, roytam1 said: hmm, so strange. Quite. See here - a newly opened Serpent 52.9 2026.01.16 (latest): only went to majorgeeks.com and then to msfn.org Forum and basilisk.exe in TaskManager is wildly going up and down 1 full CPU core usage and memory from 700MB to 1.5GB and back down to ~700 then CPU usage stops. First I noticed that also in my XP VM which I use to get me fresh prefs.js to compare with my normal one outside VM (all of 1.3GB of dedicated VM RAM went missing while on sapphiretech.com). Edit: in the before session imgur.com, there weren't any swings, just normal behaviour; RAM usage going up till ~1.2GB (even somewhat lower than with the previous version of 1.3GB). Edit2: Could it have something to do with: - htmlparser: break when infinite loop happens (8201dadf41) ? Edit3: On videocardz.com and again on msfn.org (new session after videocardz crash) Serpent crashed with stated mozjs.dll as the point of breaking. Faulting application basilisk.exe, version 4.8.7.8956, faulting module mozjs.dll, version 0.0.0.0, fault address 0x00293c73. Both times. Edit4: Again on videocardz.com, now with different reason - xul.dll and then again on msfn Forum the upper mentioned mozjs.dll. Faulting application basilisk.exe, version 4.8.7.8956, faulting module xul.dll, version 4.8.7.8956, fault address 0x00063a79. Edited 1 hour ago by modnar additional happenings
Ben Markson Posted 2 hours ago Posted 2 hours ago Using: basilisk52-g4.8.win32-git-20260117-3219d2d-uxp-aace8fa8c0-xpmod.7z Crash report: AppName: basilisk.exe AppVer: 4.8.7.8956 ModName: mozglue.dll ModVer: 4.8.7.8956 Offset: 0000fba6 After several crashes while using eBay I've gone back to: basilisk52-g4.8.win32-git-20260110-3219d2d-uxp-3e9ca11775-xpmod.7z Ben.
LordGarfio Posted 1 hour ago Posted 1 hour ago Maybe DeLeaker could be useful ?. DeLeaker 2025.16 (Profiler for C++, C#, .NET and Delphi) C++ memory leak detection Deleaker is an extension for all major IDEs and a standalone application for memory leak detection - memory, GDI, and handles so far. Even the most stable of Windows applications are not immune to resource leaks. And of all the bugs and issues, memory leak detection tends to be the most difficult, especially when found in GDI objects and menus. And the golden rule of thumb is that the sooner bugs are found and dealt with, the less expensive they prove to be. While there's no shortage of tools and add-ons to help track down memory leaks, few are capable of tracking GDI resource leaks that can devastate Windows performance. Deleaker is one of the few tools capable of this, and will have a minimal impact on your application's performance. Deleaker is a memory leak detector that integrates with all major IDEs: Visual Studio, Delphi, C++ Builder, and Qt Creator. 1
roytam1 Posted 1 hour ago Author Posted 1 hour ago (edited) reverting changes in #2895 seems help. now 32bit builds are being compiled and same filename archives will be replaced later. EDIT: all today's 32bit build archives are updated. Edited 1 hour ago by roytam1 1
modnar Posted 1 hour ago Posted 1 hour ago 19 minutes ago, roytam1 said: reverting changes in #2895 seems help. now 32bit builds are being compiled and same filename archives will be replaced later. Thank you! Mighty fine! When it worked, this Serpent really ate less RAM than any version to date.
Recommended Posts
Please sign in to comment
You will be able to leave a comment after signing in
Sign In Now