Jump to content

VistaLover

Member
 View Profile  See their activity

  • Posts

    2,245

  • Joined

  • Last visited

  • Days Won

    93

  • Donations

    0.00 USD 

  • Country

    Greece

 Content Type 

Everything posted by VistaLover

  1. VistaLover
    ... Sadly, not the right screengrab linked (it's again the one for Vista SP2 x64+ExtKernel contained in your previous post ) ...
  2. VistaLover
    ... All's good! However, looking back at all this, the "decision" of one member here to "shorten" (in order to "sanitise") a reference to a, otherwise popular , website has led to so many additional posts by other members here, including me, that only remotely (if at all) touch the subject of this thread ... Posts "exploded" to all aspects of human endeavour, i.e. natural sciences, history, linguistics, particular aspects of specific languages, education, Roman numerals, coins, ethics/morality, psychic health, addictions, human relationships/[pre-]sex, anything else I neglected? Even suggestions to not "avalanche" to further OT posts were disregarded, it appears... Each and every one of the above is a very interesting topic to converse about , however I humbly think here is not the most appropriate "place"; I, as much as anybody, can live with "a few" OT comments among a majority of posts with substance/relevance to the main subject(s) discussed in a dedicated thread; more than "a few" and I picture "trains de-railed" ... At the same time, some members here wonder why "these" threads reach so quickly the 200-page threshold ... (I'm also subscribed to these threads via e-mail notifications, "OT" posts also cram my inbox (and disk space, as I use a dedicated client, not webmail), so even more personal time needed to devote on sorting out the "valuable" from the "OT" e-mails...) Again, this is only my personal view on things, I don't expect all the rest to abide by it, but I'm entitled to it, aren't I? (NB: Don't do it; don't start another "OT avalanche" due to this post of mine; "reactions" (positive, only) are there for a reason.)
  3. VistaLover
    ... Which was the exact point I raised earlier... AFAIAA, roytam1 has not (yet?) submitted any PR upstream, nor do I know whether he has any inclination to do so in the future ... And whether such an eventual PR relates to common code (between "our" UXP-fork and the "upstream" platform) or not, "upstream" do mandate that all issues & PRs submitted to their repo have been reproduced/tested on their "approved" application(s)... Some years ago, when MCP hosted their repos on GitHub, I had directly filed there a few issues discovered in NM28, with the note "I expect them to be 100% reproducible in PM, too"... All Moonchild had to do was try my STR in "his" browser; instead, I was frowned upon, my report disregarded, until I provide actual proof of the issue on PM itself ... I couldn't run PM in my OS of choice, so I then gave up ...
  4. VistaLover
    ... Actually, it wasn't me who pointed that out:
  5. VistaLover
    ... While I am notoriously "famous" here for writing (and posting!) lengthy and detailed analyses, especially in the past, time constraints induced by RL (AstroSkipper: Real Life ) meant that often now I have to convey the info I want by referencing related link(s) - the info is there, just "not in my very own words" ...
  6. VistaLover
    Actually, unrelated to what we've been discussing here... The multiple [1888] WARNING: file already exists but should not: C:\DOCUME~1\nico\LOCALS~1\Temp\_MEI18882\Cryptodome\* warnings are a manifestation of a very recent bug, which I have reported here (possibly related to the new "ways" PyCryptodome[x] is imported at yt-dlp's runtime and, thus, the "new" ways the module is being packed by PyInstaller) ... Even your very own compiles are being affected by that bug: https://github.com/nicolaasjan/yt-dlp/releases/download/2023.02.09.051430/yt-dlp_x86.exe https://github.com/nicolaasjan/yt-dlp/releases/download/2023.02.11.092312/yt-dlp_x86.exe (One needs Vista SP2+ [32 and/or 64-bit] to test these). The bug may be exclusive to the GitHub Actions compiled builds, because the ones I prepare myself locally (with CPython 3.7.16), as well as your "manual" build "yt-dlp_x86_Windows-XP.zip", do not exhibit this bug ... Sometimes I wish the yt-dlp devs left the code "at peace" for awhile, in a good-working-state... "Shouldn't fix what isn't broken" comes to mind... But what do I know (except for stumbling on bugs ) ?
  7. VistaLover
    In that linked GH issue, you were at PyInstaller v4.10 (custom "yt-dlp" build); have you given their most-up-to-date version 5.3 (linked in my previous comment) a shot? The idea is that a custom build has a unique bootloader not shared with the rest of the EXEs compiled with the widely available PyPI edition of PyInstaller, thus raising less red flags (false detections) with (mostly) low-quality AV suites ... And, in any case, you did say in that comment that: So, you stand to lose nothing by trialing v5.3... My 2c, of course...
  8. VistaLover
    Latest PyInstaller on PyPI is at version 5.7.0 ; yt-dlp "org" have used custom compiles of PyInstaller v 4.5.1->4.9->4.10->5.2 and are now (probably) using v5.3 ... The 32-bit wheels of those custom compiles can be seen in below branch: https://github.com/yt-dlp/Pyinstaller-Builds/tree/gh-pages/i686 Direct link for latest v5.3: https://github.com/yt-dlp/Pyinstaller-Builds/raw/gh-pages/i686/pyinstaller-5.3-py3-none-any.whl If you so wish, you yourself can switch over to this custom yt-dlp compile of PyInstaller: python -m pip uninstall pyinstaller (will uninstall v5.7.0 fetched from PyPI) python -m pip install "path-to-downloaded/pyinstaller-5.3-py3-none-any.whl" or even: python -m pip install "https://github.com/yt-dlp/Pyinstaller-Builds/raw/gh-pages/i686/pyinstaller-5.3-py3-none-any.whl" ... then build yt-dlp_x86.exe with that ...
  9. VistaLover
    Thank you for this most excellent read ; lengthy, but still excellent! Considering it was written 7 years ago, I wonder what the author would've said about today's "abomination" of the web ...
  10. VistaLover
    ... The UXP forks (NM28/St52) offered here by roytam1 (infinite thanks BTW ) are based 99.5% on the same application platform (UXP) that official Pale Moon (and Basilisk) is based on ... Like "our" forum, "theirs" is also inhabited by numerous very knowledgeable members; because the platform is more-or-less common, if issue "A" affects PM, there's a strong likelihood same issue "A" affects NM28 and/or St52... If solution to issue "A" is offered inside the PM forums, then good chances are the same solution will work for NM28/St52, too... Case in point, the recent "discourse" breakage... Fortunately, one can visit the official PM forums and read most content there without a mandatory registration - the site is safe, renders fine in "our" browsers and, frankly, is also a chest full of invaluable knowledge on "legacy" browsers, "legacy" extensions, etc. When I, @UCyborg or some other person here includes a link to the PM forum, should be assumed by the frequenters here that it links to content that is also relevant to "our" browsers , or, at the most extreme, it links to general discussions with tangence/interest to "us", too ... IOW, you don't have to use Pale Moon browser to follow a suggested link to their forum - FWIW, I don't use PM myself, for the simple reason it's incompatible with my OS, however that fact doesn't stop me from visiting their forum for "consultation" on various issues/problems (mind you, the forum is less "hostile" now to read, with "you-know-who" banned from posting ) ... I understand my opinion(s) might not resonate well with everybody here, "c'est la vie" , and, certainly, I don't want to sound unpleasant to some of you, but...
  11. VistaLover
    ... Well , if only members here had been following more closely the recent exchange between UCyborg and me, and, more importantly, if link(s) to the official Pale Moon forum had been actually clicked on, then there would have been no room left for disambiguation of the abbreviation "PH" UCyborg used in his reply to me : ... thus: ... It appears (drawing from a past recollection of mine) that the "crowd" here is hesitant to click links pointing to MCP's forum, for whatever "deserved" or "undeserved" reasons ...
  12. VistaLover
    https://repo.palemoon.org/MoonchildProductions/UXP/commit/1f32c17fcf767a366b1547f51fccc7dcb13ff719 https://repo.palemoon.org/MoonchildProductions/UXP/commit/ffdba3d91e0ca7df003a97f564451fb0ae97fa8e ; hopefully will arrive in coming weekend's roytam1 builds ...
  13. VistaLover
    ... Actually Fx v79.0 , as per the linked documentation ; it was Opera (adjacent to the Fx column) that implemented "??=" in its v71.0... Since 79 is even further from 68 (than 71), it probably makes things for a native implementation in Mypal68 more difficult ...
  14. VistaLover
    Regarding the recent discussion about discourse-based forums being BROKEN under UXP-based browsers (NM28/St52/possibly also St55/moebius), I found a related thread in the official PM forums: https://forum.palemoon.org/viewtopic.php?f=70&t=29327 The point in the thread where the recent breakage happened (i.e. implementation of the "??=" operator by Discourse) is this one ; the approach savvy members there took to address the breakage is to use JustOff's Modify-HTTP-Response (legacy) extension below: https://github.com/JustOff/modify-http-response/releases/tag/1.3.8 Once installed, go to "about:addons" and access that extension's options; locate the "Filters" input field; in the initially empty field, paste the code below: [["/global\\.discourse-cdn\\.com|community\\.(frontrowcrew|cartalk)\\.com|forum\\.(manjaro|openwrt)\\.org|forum\\.italia\\.it/",["/browser-detect-/",["/.*/g",""]],["/vendor-/",["/(t\\.discourse\\.hoisted|t\\[e\\]|r)(\\?\\?|\\|\\|)=(\\{\\}|\\[\\])/g","$1$2($1=$3)","n??=[]","n||(n=[])"]],["/discourse-/",["e.draft||=t.draft","e.draft||(e.draft=t.draft)","/(t\\.__registry__\\._typeInjections\\.service|[ne]|f\\[e\\])(\\?\\?|\\|\\|)=(\\[\\]|\\{\\})/g","$1$2($1=$3)"]]]] NB: You'd better use a proper code-editor to copy/paste, to avoid any errors... Then "tick" the "Enable" setting above the "Filters" input field... @msfntor : If you now visit https://community.brave.com/ in your UXP-based browser, (hopefully) the forum will load OK (it does here, with St52): I do hope you're happier now ... The inner-workings of the method is that the extension intercepts the UXP-incompatible JS code sent by Discourse and then transpiles it on-the-fly, based on the Search-and-Replace RegExp filter specified... All credit for the filter code belongs to PM-Forum members Kris_88 and adoxa ... @Art7220 : The procedure I detailed above seems to also work for your "own" Discourse-based forum: https://forums.mst3k.com/
  15. VistaLover
    Yes I did; please understand I don't harbour a "negative disposition" towards you, nor am I out there to insult you ... But what was the point in the context of this recent discussion to mention that the homepage https://www.sitepoint.com/ loads (scrolls) OK in DCB? "Discourse" is a "forum" platform, so I had expected you to link/refer to their actual "community" page, https://www.sitepoint.com/community/ which is indeed broken in DCB (I assume? ) and UXP...
  16. VistaLover
    ... Wrong again ; their Homepage doesn't use "Discourse", but their Forum does : https://www.sitepoint.com/community/
  17. VistaLover
    ... As you have already discovered, it's the same issue as this one ; my suggestion to you in my previous post on this thread still stands :
  18. VistaLover
    ... And: https://community.brave.com/ (courtesy of @msfntor ) ...
  19. VistaLover
    ... What it says on the tin: a) We (intend to) ONLY support "last week's" iteration of the Chromium engine ... b) You can forget "the previous 10 years of the Internet" , thus ALL browser engines that are still compatible with that Internet "snapshot" ...
  20. VistaLover
    ... Leave that envvar empty/don't use it at all, so that your ISP connection is used DIRECTLY for the cert revocation check (which is performed over plain HTTP); in any case, only the secure connections the game attempts should be redirected to the TLS proxy, i.e. ONLY the HTTPS_PROXY envvar should be used... Just my 2c, of course...
  21. VistaLover
    ... Their forum is based on the discourse platform, but "discourse" have recently implemented another UXP-exterminator, the operator called "nullish coalescing assignment ("??=") : Your only hope under XP is to use 360EEv13.x/minibrowser to load that forum, because that operator was first implemented in Chromium 85 (Firefox 79) ... Most sadly , "discourse" is being used by many forums/communities, so this is just going to only escalate in the coming days ...
  22. VistaLover
    HULU (a US-only service, BTW) uses DRM exclusively for all of its streams... Unlike Netflix, where a fallback to using the Silverlight NPAPI plugin for "legacy" browsers is in place (correct me if wrong , at least that's my recollection of things on Netflix ), HULU demands the WidevineCDM (owned by Google ) to properly function... WidevineCDM is currently incompatible with the Windows XP/Vista OSes (and in the far past was only compatible under XP with the Google Chrome web browser); if on Win7+, you must use one of the latest Chromium variants and/or latest Firefox to watch HULU ... BTW, NM28 does not support DRM on any OS, while St52 supports DRM on Vista+; however, the version of WidevineCDM it ships with (and supports), v1.4.9.1088, has been deprecated by Google, i.e. it can't acquire decryption keys from the dedicated Widevine lic servers ... @Art7220 : Got an Android phone? If yes, download their app and watch HULU's DRM streams there ... As you can see, Google not only control which webpages you can successfully load in your (non-Chrome) browser, they also control which rich content is available there, too (and also demand you update your OS to watch it) ... A true dictatorship, if you ask me...
  23. VistaLover
    Respectfully, I beg to differ ... Google do come up with new JS drafts and are very swift to implement them into their own browser monopoly ... Adherence to well-established web standards is a lesser concern to them - they practically control themselves the W3C board, so everything "new" and "fancy" they devise will find its way into a "revised"/updated Web Standard ... TL:DR: Google aren't better at "complying" to web standards; they "own" "www" with their monopoly and can do "as they please" with web standards... As the majority of current mainstream browsers are Chromium-based, what Google do dictates what the rest are forced to do... Mozilla are funded by Google, their "current" Gecko engine aspires to be a Chromium fork; so, Firefox also adopts what Google have already implemented a while back into their browser (a policy known as "Chrome-parity") ... That leaves ALL the rest browser choices (non-Chromium/non-Firefox) lagging behind in Web Compatibility...
  24. VistaLover
    ... You were not paying attention, were you:
  25. VistaLover
    ... Well, those builds use a CPython 3.7.1 version "hacked"/patched to work specifically on WinXP, ONLY (whereas official CPython 3.7.x, by the PSF, requires WinVista as minimum [client] WinOS) ... The PSF EoS'ed WinXP SP3 with the CPython 3.4.x branch; this patched py3.7 compilation borrows modified "system-like" files from either Wine/ReactOS/OneCore API (not sure exactly which ) projects (files bcrypt.dll, kernelXP.dll, ntext.dll, psapi.dll, ws2_xx.dll), which specifically target NT 5.1 and aim to backport missing functions from NT 6.0 (where py3.7 runs natively ) . It's the same (type of) CPython 3.7.x "hack" that is incorporated into the latest ProxHTTPSProxy versions (maintained/released by MSFN member AstroSkipper); this nice project runs successfully ONLY under Windows XP; for Vista [and higher, though Win7 updated to EoS shouldn't need ProxHTTPSProxy ], the embedded Python has to, somehow, be switched to one of the official py3.7 releases...
×
×
  • Create New...