VistaLover

... I had explained this in fine detail in the past ; perhaps it was lost when the forum's database was corrupted back in June 1. First grab latest Polish language pack (for PaleMoon 28) from http://addons.palemoon.org/language-packs/ Do not install in NM28, just save to file (it should have an .xpi extension, filename=pl.xpi). 2. An .xpi file is just a renamed .zip file; if you have 7-zip installed, it can decompress .xpi files without issue; if not, rename the file to pl.zip and use your favourite archiver app to extract to a pl folder. 3. Open the folder and locate file install.rdf ; open that file in a proper text-editor (Notepad++, PSPad, etc.); if you don't have a proper text-editor installed (you should), use Window's notepad as a last resort... 4. Locate line 31 of that file; it should read: <em:maxVersion>28.0.*</em:maxVersion> 5. Change "28.0.*" to "28.1.*", save the change and close the file in your text-editor. { 5a. Optionally, you can also change line 12 to read: em:version="28.1.0" and then save the change to file } 6. Select all contents of folder pl (but not the actual folder itself) and re-zip them back to a pl.xpi file (you can also rename that to 28.1.0_pl.xpi, for telling apart from original download...) 7. Open latest NM28 (v28.1.0a1) and navigate to about:addons (addons manager); drag and drop file (28.1.0_)pl.xpi; it should offer you to install it; a browser restart would be needed... 8. For the LP to apply, you'd have to set in about:config general.useragent.locale;pl (followed by another browser restart). DISCLAIMER: The procedure outlined is just a hack; the localization may break with any future NM 28.1.x update, if new strings are added to the code by Moonchild devs for which no translated Polish strings are there in the installed pack; according to the extent of the breakage, some functions may be broken, or even the whole browser might be unable to start; in that case, start in safe mode, disable the language pack and restart in normal mode (en-US GUI). You should inquire about updated packs at the GitHub repo... @kitaro1 : That should also interest you! @roytam1: Maybe also pin this post to first page, because people seem to be unable to apply what is a simple procedure for you and me...

Hello there Since you appear to be a fellow Vista user, I decided to help and investigate... Two major issues for me: 1) I don't know any Polish, 2) (obviously) I'm not in Poland currently... I had to use a lot of Google Translate to decipher all these messages (in Polish) I was presented with and to be able to click the right on-screen buttons... I then had to find video content that wasn't geo-fenced (restricted to Polish IPs, only ); I think I settled for Ameryka-Marzen-Odcinek-12 I tried a variety of browsers on my Vista laptop, but to no avail; I suffered exactly what you reported, i.e. a black rectangle in the supposed place of their embedded player... I then kindly asked my sister to test that URL on her Firefox Quantum v61.0.2 (32-bit) copy on her Win7 SP1 64-bit laptop: she had to temporarily disable uBlock0 on that site, but, lo and behold, the programme started playing in an HTML5 player... After some further troubleshooting I did, I think I've solved the riddle: That (nasty!) Polish site performs User-Agent sniffing and simply doesn't like the Vista (NT 6.0) part of the UA, so they decline to load their player on Vista machines (didn't test, but I'm sure the same fate is reserved for XP, too, i.e. NT 5.x) The solution I chose was to apply a site-specific-UA-override (SSUAO) for "www.ipla.tv" to mimic Fx 62 (due imminently) on Win7 OS; in a "roytam1" browser (NM27/28, Serpent 52.9.0/55.0.0) open about:config and create a new string pref with name=general.useragent.override.www.ipla.tv value=Mozilla/5.0 (Windows NT 6.1; rv:62.0) Gecko/20100101 Firefox/62.0 restart browser, clear cookies for "www.ipla.tv" and reload VOD page; you are then presented with so you have to disable adblocker (ublock0 in my case) and reload ; after reloading, you have to literally sit in front of at least 3min worth of ads; if you change to another tab, the ads and time counter are paused and only resume if you focus back to that tab... Very nasty behaviour indeed Finally, after the ordeal ends, main video starts playing: The test above was conducted on a Windows Vista SP2 32-bit Home Premium laptop, browser was latest New Moon 27 offering (v27.9.1a1, 32-bit, 2018-08-24) with lav-dll-lite-mmx dlls put correctly in place...

... You are not to blame, but you simply misunderstood The green Verified button on the GitHub repository: has nothing to do with extension signing on AMO (addons.mozilla.org) (see here). The legacy (XUL) version of uBlock0 on GitHub is not signed, hence it couldn't be verified by Firefox for installation; in fact, if you navigate to the installation instructions for the XUL flavour, it is stated that:

You can easily change the name string of installed Java version to "Java 8 Update 181" with CCleaner: Tools Tab => Uninstall => Select "Java 8 Update 152" in the "Programs to Remove" column, then on the top right press the blue "Rename" button; but this will only change the name string (hence an aesthetic change), not the actual version string (for 8u181, it should be 8.0.1810.13)... [tested only in Vista SP2, YMMV with XP SP3]

According to FileHippo (Technical tab), it was June 12, 2018; same thing is suggested by the vendor themselves at: https://helpx.adobe.com/shockwave/release-note/release-notes-shockwave-12.html

Highly unlikely ... Slimjet uses Chromium's engine (webkit), in which every tab loads its own (content) process; you can inspect Slimjet's memory consumption by launching (SHIFT+ESC) its task manager; Slimjet comes with its own adblocker too, add to that various other addons you may have installed and the RAM goes up quickly... It is no secret that Chromium-based browsers achieve higher rendering speeds at the expense of your RAM; on the other hand, Mozilla based browsers, especially of the pre-Quantum type, use less memory (with the exact same tabs loaded) compared to Chromium browsers; this would be even lesser for Moonchild's browsers, where e10s (electrolysis) is by default either not supported (Pale Moon 27) or disabled (Pale Moon 28, Basilisk52, Basilisk55). Have you tried PM28XP yet (built on the same UXP platform like Serpent 52.9.0) ? I have been following this thread recently and not many other people have come forth with reports about "sluggishness" in FirefoxESR 52/Basilisk 52/Basilisk 55 in combination with MSE; only you and @heinoganda, both using Virtual Machines; it would be interesting to hear from someone not on an XP VM... The stable (release) channel is currently on version 61.0.1 (Firefox Quantum) and the new ESR channel is on version 60.1.0 (Firefox Quantum); 62.0 is the current version of the beta channel of Fx Quantum. ESR channels (usually) get a major version bump every 7 release cycles, e.g. FxESR 38 => FxESR 45 => FxESR 52. The next ESR channel after (the XP/Vista compatible) 52 would've been FxESR 59 (with Fx 52.8.x being the EOL release of ESR 52), but Mozilla shifted the next ESR major update to a Quantum 60.0 based one, hence FxESR 52 got an additional cycle's reprieve (FxESR 52.8.1 => FxESR 52.9.0) ... Same here; my Kaspersky Antivirus subscription ends in 3 months' time... FWIW, KAV behaves normally with referenced browsers (initial CPU increase when the browser is first launched and previous session restored, but afterwards things run smoothly; no sluggishness, apart from very "heavy" sites like facebook and instagram, which I practically never visit) ...

Thanks; we're already aware, thanks to @Vistapocalypse:

... Thankfully, a member of the Moonchild dev team has done that some 2.5 hours ago... #679

Thanks, but as I said this bug is present in all NM28 builds I tested over the last month; as for the build ID string, you, of course, should know better, as being the person compiling the builds, but I thought the string was unique to each build (and certainly different between the two architectures for the same source snapshot); of the files listed in your server (o.rthost.cf/palemoon) I haven't yet come across two different files with the same buildID ... Of course, that would be file "palemoon-28.0.0b4.win32-git-20180721-b4f748a7c-xpmod.7z"; what would be more important for me to know is whether you (or anyone else for that matter) are able to verify reported behaviour? No, in the end all of my initially composed message was submitted, it just would not get through in one go! I had to temporarily copy the content off the forum's editor to my local text editor (PSPad) and then, as said, post it in small installments... Another manifestation of the forum's software malfunctions is the fact, despite me having opted to "Notify me of replies", I wasn't notified by e-mail of the last two replies above this one (@siria's & @roytam1's), while previous replies in this thread were successfully announced via e-mail notifications...

I thought all the "about:" references were the culprit, however in the end they got through... Another suspect was the embedded imgur image link, or links native to the forum itself; however, the forum software errors sometimes do not follow a defined pattern and the full composed post just won't be submitted, for no apparent reason ; I had to do it by a real trial-and-error method, submitting single paragraphs (or even sentences), checking they went on-line and then editing post by adding another paragraph (or sentence), until my whole report was finally through... It was really a nerve endurance test...

I am running into a minor New Moon 28 bug, which has been persisting over several weeks; I can't tell how old it is, cause I started testing NM28 from v28.0.0a4; I was able to reproduce on a new clean profile, with no extensions added, using latest build 28.0.0b4 32-bit (buildID 20180720233208) STR In a new clean profile (happens in an old, "dirty" one, too), navigate to about:support; the "Application Basics" top section contains clickable links (in blue) to some internal (about:) pages; middle-clicking on all but one of these links (the "about:buildconfig" one) results in a blank (empty) new tab; so does right-clicking and selecting Open in a new tab, but left-clicking works as expected, opening the internal "about:" page in the same tab... I think this issue should be reported to Moonchild and be hopefully fixed; I do have a GitHub account myself, however, being on Vista, I can't run the official MC PM28 builds (Win7+) ; FWIW, Serpent 52.9.0, built on the same UXP platform, does not present this issue... [OT: I had a heck of a time trying to post this report, I did it literally line by line, as the forum software always responded with a Forbidden error message (must have received more than 20 of those...); I know this is a known situation, however it's getting worse to the point one can no longer post without being driven mad... ]

See also this I posted over at the Vista subforum...

It may well look as if Malwarebytes Anti-Exploit (MBAE) 1.12.1.90 be the last version to run on Windows Vista SP2; this was reported by a Vista user on their forum, several hours ago ; MBAE is a perpetual beta free software with all premium features turned ON for volunteers/testers; the new auto-update to v1.12.1.97 brought with it, alongside several serious issues with Chromium-based browsers on Win7+), loss of functionality on the Vista OS (XP too?), because of new functions missing in Vista's kernel32.dll Version 1.12.1.90 can be fetched navigating to vendor's official page here ; in the event that page changes, direct link to the setup is https://malwarebytes.box.com/s/hrxfh20wtbre7tfaimf56i5chlodia4z Special provision has to be taken not to trigger auto-update to the non-compatible 1.12.1.97 version, see: https://www.wilderssecurity.com/threads/malwarebytes-anti-exploit.354641/page-152#post-2770670 (courtesy of MSFN member @Sampei.Nihira)

Can you spot the oxymoron in the above statement with the following practice: IOW, you have excluded a browser process (FirefoxESR 52.9.0) from your Anti-Virus/Security Suite (MSE), which leaves you open for possible attacks while you browse/download (as an OS browser is the main app one accesses the web with and a perfect candidate vector for infection); I wouldn't do this myself, nor recommend it to others ; but, of course, you're the master of your own box ...

Hi . It appears the Tycho platform for Pale (New) Moon 27.x.x has been EOL'd, as per 5da550a So PM 27.9.4 will be (probably?) the last official stable (release) version on that platform (and the last to officially support Vista SP2 ). Since you were building New Moon 27.x.x not from the release branch 27.9_RelBranch but from the master branch, perhaps just recompile one final time, but you yourself update appVersion on the master branch to 27.9.5a1 (see c674a10) to make it more evident, beyond any doubt, that the code snapshot is some commits ahead of the last stable 27.9.4 release; just a humble suggestion for next weekend's eventual updates... I didn't have the slightest what TenFourFox is, so after searching I found out it is a fork of Mozilla Firefox for Power Macs that has significantly diverged from Mozilla code after the 45 ESR milestone; for transparency, are those changes the ones designated as M* and where exactly their details are to be found (somewhere inside https://github.com/classilla/tenfourfox perhaps?) ? Thanks for your ongoing hard work

According to this PAs.com comment : Electron framework requires Win7+ by definition ... and according to that thread, further code was removed by M$, hindering considerably stealth portabilisation of Skype 8 even on electron compatible OSes Nice find, many thanks!

The off-line windows 32-bit installer (file jre-8u181-windows-i586.exe) continues to function properly on this old Vista SP2 32-bit laptop; I suppose this is of no consolation to XP users of this thread, just reporting it as a FYI here; the GUI during installation (which itself took quite long, ca. 5min on this old, under-resourced, machine) was slightly redesigned to be of a more elongated rectangular shape; but it did include a link to the following Oracle article: Release Notice a message the XP users are unlikely to see, due to the installer being broken on that OS Regards

You do realise you can't/shouldn't stay put for ever on a specific browser code snapshot just for the sake of an extension, however valuable it might be for you . Pale Moon 28, at this stage of its development, is constantly evolving - it is expected that some things will break (including extensions) while other things (e.g. page rendering) and/or various bugs are being ironed out... In fact, even when PM28 reaches a stable/release status (early autumn 2018?), we don't expect full 100% compatibility with the Pale Moon 27 ecosystem ; this is due to huge changes in the underlying platform (Tycho [Mozilla 38] -> UXP [Mozilla 52]). I only made the jump from PM27 to NM28 last week and spent several days to tailor it according to my liking; some of the extensions that were working on 27 quit on 28, so upgrades/replacements had to be found; most of all, I mourned over the breakage of my favourite complete theme, FT Deep Dark v9.5.4, but then I moved on to a similar, but 28 compatible, one, Dark Moon 2.0.3 ... But I digress In what exact way does dTa! work badly? More importantly, what version have you got installed? Please see relevant PM forum thread: https://forum.palemoon.org/viewtopic.php?f=44&amp;t=19494 If still on 2.0.18.1, try upgrading to the stable 3.0 branch (i.e. v3.0.8): https://addons.mozilla.org/en-US/firefox/addon/downthemall/versions/?page=1#version-3.0.8 You can also try the nightly 3.0 branch (v3.0.9pre): https://code.downthemall.net/nightly/dta-3-0/ or even the nightly 3.1 branch (3.1.1pre): https://code.downthemall.net/nightly/trunk/ You need Moon Tester Tool Pale Moon extension to install above dTa! versions... (As a precaution when testing new things, always back up your NM profile first ) Regards OT: Nice to see the forum is usable again , after 3 days of serious issues...

@kitaro1 The Polish Language Pack I linked to targets New Moon v28.0.0b2 and newer: <em:minVersion>28.0.0b2</em:minVersion> <em:maxVersion>28.0.*</em:maxVersion> You are still on an older a4 version, as per your screenshot... Please update to the latest b2 build kindly offered by @roytam1 and the pack will install (but may break with coming weekend's release... ): (And BTW, 95% of what I post about has been proof-read and tested as being valid at the time of posting... )

Duplicate post deleted I have run into serious issues today trying to browse/login/post/edit posts in this forum Links clicked take ages to load, more often than not giving a "500 Error" page... Login in can require two or three attempts, writing a post seems to work OK, but when clicking the "Submit/Save" button it just stalls there ad infinitum (and then you may end up with a duplicate post...). I have no script-blocking extensions installed, all cookies are allowed for the "msfn.org" host , fully disabling uBlock Origin doesn't seem to matter... While it's a high probability this is site-related (let's hope it gets sorted soon without other loss of data), I couldn't help noticing that the browser I was using, NewMoon 28.0.0b2 (2018-06-29) (32-bit) was more prone to producing those forum errors compared to, say, FirefoxESR 52.9.0

Draft language packs for Pale Moon 28 (on UXP) have been announced here: https://forum.palemoon.org/viewtopic.php?p=144777#p144777 which takes us at: https://github.com/JustOff/pale-moon-localization/releases/tag/28.0.0_RC1 As noted, These are of course installable on @roytam1's New Moon 28 builds (PM28XP pre-release builds), but be warned that because Pale Moon 28 is still undergoing many GUI changes, they (Language Packs) are prone to break with any future release . After the installation, remember to modify the value of pref general.useragent.locale accordingly and restart the browser, in order for the pack to be used... In the event of a future break 1. If the browser is able to start and AOM is accessible,, disable the language pack (in about:addons) and restart, to go back to the default (en-US) localisation. 2. If the browser GUI is totally broken and won't start, then start in safe mode and permanently disable/remove the culprit LP. Or locate in Windows Explorer your NM28 profile and manually delete file "./extensions/langpack-[x]@palemoon.org.xpi", where [x] is the code for your language (e.g. en-GB), then start the browser normally... You should keep an eye on the linked GitHub page for newer, fixed, releases... @kitaro1 : Polish LP for NM28

... This isn't so... Cyberfox 52.9.0 (officially supporting Win7+, but can be hacked to run on Vista SP2, too...) had just been released mere days ago, the buildID of the x86 binary was 20180626222034; but they made a big blunder there : Cyberfox is distributed together with a full set of language packs (91 to be exact), but they had forgotten to change the maxVersion string in their install.rdf files (from 52.8.0 - there never was a 52.8.1 release, unlike FxESR - to 52.9.0), so their browser was only available in en-US locale... The Australian developer was forced to rush release a new "fixed" version with working (embedded) language packs, hence the version bump: https://github.com/InternalError503/cyberfox/commit/7fd07cc BuildID of new x86 binary is 20180629141022; security wise, Cyberfox 52.9.1 is on par with FirefoxESR 52.9.0. If/when there'll be a FirefoxESR 52.9.1 chemspill release, Toady will follow soon after, but, of course, never release in advance of Mozilla...

I had been otherwise occupied all day long, so only recently skimmed through the numerous MSFN e-mail notifications on this subject... My input: Moonchild wrote: Really? English is not my mother tongue, but doesn't "should continue" imply they are currently able to run on Vista? This is the furthest thing from the truth! Lots of places inside their forum; I did a quick search and I came up, amongst others, with: https://forum.palemoon.org/viewtopic.php?f=1&amp;t=15806 This was when Moebius/UXP-T1 was forked from a Mozilla Platform 53.0a1 code snapshot; that snapshot was already devoid of any residual XP/Vista OS support, as the good () guys over at Mozilla had already excised all relevant code... MC team did absolutely nothing to restore at least partial Vista support... https://forum.palemoon.org/viewtopic.php?p=114825#p114825 https://forum.palemoon.org/viewtopic.php?f=61&amp;t=18253 (Basilisk and Windows Vista; I urge you to read the entire thread , but I've selected the following part:) Officially released Basilisk 55.x.x.x was never compatible with Vista; by pure luck (or negligence on their part), the compiler flags were such that lowering the subsystem value of the executables (6.1 -> 6.0) would allow for them to run on the Vista kernel, but OS integration was suboptimal; non-existing code targeting Vista resulted in WMF (and possibly other parts, e.g. EME, aka Widevine DRM plugin) being broken! https://forum.palemoon.org/viewtopic.php?p=132054#p132054 And when in the start of spring this year they ditched Moebius in favour of UXP-T2 (now just UXP), they forked a Mozilla 52 ESR platform with full XP+Vista support and what was the first thing they did? Meticulously removed all that support (as it's always easier to delete existing code than write new one...). Are we to believe that MC has just now had a sudden change of heart and he's willing to devote precious coding time to undoing the Vista massacre? I'm not that gullible... What's worse, in official Basilisk 52 + Pale Moon 28, they have amended their compiler flags/optimisations to fully ignore NT 6.0 (Vista/Server 2008); this has been already documented previously in this thread by esteemed member @WinClient5270, but it was also the result of my own tests weeks ago, when the first official builds were publicly released... Unfortunately, it's more than those... Inspecting latest (official) PaleMoon_unstable-28.0.0b1.en-US.win32[buildID=20180625093249] package with dependency walker, I am disheartened; focusing on just xul.dll module, this has calls to 6 missing functions in kernel32.dll GetCurrentProcessorNumberEx K32EnumProcessModules K32GetModuleFileNameExW K32GetProcessImageFileNameW K32GetProcessMemoryInfo K32QueryWorkingSet and to 2 missing functions in shell32.dll SHGetPropertyStoreForWindow SetCurrentProcessExplicitAppUserModelID Main executable (palemoon.exe) has lesser issues, of course I wasn't bothered to check other DLLs... Similar results are obtained with official Bk52 releases... And I emphasise again, it isn't only compilation configuration that needs to be changed to target Vista, it's actual browser code that needs to be rewritten to accommodate a fully functional, Vista compatible, application... Will Moonchild deliver? I think not... (and until PM 28 "stable" gets released in the final quarter of 2018, Vista user-base will have dropped even lower to merit his support considerations...) @WinClient5270, if memory serves right, I recollect you having registered previously in the Pale Moon forums (about an issue with Visual Studio 2013 dlls affecting PM 64-bit, that I had identified for you...); perhaps the best way to clear the FUD once and for all would be to post in the linked thread (by Jody) and ask plain and simple what MC's definitive decisions are; not his thoughts/intentions, but if he's actually determined to support Vista in PM28 (at least in the same level he supported Vista in PM27); please don't hesitate to convey some of my reservations/points I raised here ; full WMF+MSE support on both 32-bit/64-bit Vista architectures would be a must; UXP (unlike Tycho) is suitable for that, if they're willing to undo the damage they did to it concerning Vista; these days, noone wants a browser that won't play back embedded audio-visual content... Here's hoping...

Hello ! UAS isn't needed, strictly speaking, because Basilisk (both 55/moebius and 52/UXP), unlike Firefox ESR 52.x.x, has support for Site-Specific-User-Agent-Overrides (SSUAO); for example, if you set the following "string" preference (in about:config), general.useragent.override.addons.mozilla.org;Mozilla/5.0 (Windows NT 6.1; rv:52.0) Gecko/20100101 Firefox/52.0 when visiting AMO they'll think you are using Fx 52 (the deprecated 52.0 or ESR 52.x.x); but if you feel more comfortable adding one more extension, be my guest ; in any case, most of these installation problems from AMO can be overcome, for the time being, by navigating to the "classic desktop site" version of the extension page (scroll all the way down, then click on "View classic desktop site"; a session cookie is set (mamo); you should be again able to install from the "legacy" installation page) ... I have toyed with various UAOs for the "userstyles.org" site in Bk55, and the issue I reported doesn't seem to be UA related (but, of course, I might be wrong...). This reply was made by using latest Bk 55.0.0 (2018-06-23) (32-bit) on WinVista SP2 x86; as you can see, I had no problem quoting excerpts from your two recent posts; please start in safe mode or (better) create a new clean profile to test; if everything works as expected, it's some extension or user preference that interferes; only you know how you've configured things; on my side, things are working pretty fine . Regards

Hello again Roy I did update back to the latest build, Basilisk (Serpent) 55.0.0 (2018-06-23) (32-bit) (buildID=20180623000710) and toggled the mentioned pref: SUCCESS All previously non-functional userstyles now work as expected! Cheers for finding the culprit! And I saw you've already pushed the necessary change to your forked Bk55 repo: https://github.com/roytam1/basilisk55/commit/81899e5 But (sadly, there's always a but... ) : ... this issue is still there, in an annoying fashion . Userstyles.org have made it all the more difficult to install a userstyle they host, if 1) you're not running one of the major browser brands, notably latest versions of Chrome/Quantum, and 2) you're not running the latest webextension version of Stylish. They do not provide an interface to install a specific style on its page, as I said the "Install with Stylish" button is just a lure to force Stylish WE 3.1.1 down your throat . With New Moon and Serpent browsers (Stylish 2.1.1), I have found clearing the browser cache and reloading the page several times will finally get you the "install style" pop-up prompt... But the detection of an installed style is a very handy feature that doesn't work in Bk55; if I visit https://userstyles.org/styles/94667/selected-tab-blue-font-color-for-ft-deepdark with latest Bk55 (buildID=20180623000710) with that userstyle installed, I get: In the case of latest Basilisk (Serpent) 52.9.0 (2018-06-23) (32-bit) (buildID=20180623010357), with the same userstyle installed and active, detection succeeds: Possibly some javascript code which only works in Bk52/UXP but not in Bk55/Moebius Can you please come up with a solution for this? FWIW, it would appear that the detection of an installed userstyle also fails in both NM 27 [27.9.1a1 (2018-06-15) (32-bit)] and NM 28 [28.0.0a4 (2018-06-23) (32-bit), so this bug isn't unique to Bk55! I think you meant a working WMF+EME (Widevine) Vista implementation - thanks for not throwing in the towel beforehand I see this has already been taken care of in https://github.com/roytam1/basilisk55/commit/7042385 We'll have to wait for your next builds over the weekend to test As ever, lots of Regards