
Mathwiz
MemberContent Type
Profiles
Forums
Events
Everything posted by Mathwiz
-
My Browser Builds (Part 5)
Mathwiz replied to roytam1's topic in Browsers working on Older NT-Family OSes
The only reason I knew about it is that it came up a few months ago when St started doing it in single-process as well as multiprocess mode, and someone came up with that setting as a workaround and posted it in this thread. I guess St was fixed later on, but the fix only works in single-process mode? Edit: It was actually in the previous "generation" of this thread: https://msfn.org/board/topic/184051-my-browser-builds-part-4/?do=findComment&comment=1228833 -
My Browser Builds (Part 5)
Mathwiz replied to roytam1's topic in Browsers working on Older NT-Family OSes
There's a fix for that. Set dom.keyboardevent.keypress.dispatch_non_printable_in_content to true in about:config. -
My Browser Builds (Part 5)
Mathwiz replied to roytam1's topic in Browsers working on Older NT-Family OSes
Google brought me to that page even before you did, but I swear I tried disabling everything that the article said "safe mode" disabled, and I still couldn't get it to work! BTW there's a contradiction between the St 52 dialog box and the Mozillazine article: The dialog box says custom settings are disabled, the Mozillazine article says they are not. My experience tells me the Mozillazine article is correct, at least for St 55. (Don't know if MCP changed that in UXP.) At any rate, I just gave up. I copied my St 55 profile from work to my home PC with the help of a thumb drive. That fixed several of my problems; e.g., M$'s Windows Update Catalog works again. Chase.com is very cantankerous, but can be made to work with enough patience (although TBH, I usually just go to Edge for that site now). I'll probably never figure out the core issue. I can only tell you it isn't multiprocess mode, nor is it most of my add-ons. -
My Browser Builds (Part 5)
Mathwiz replied to roytam1's topic in Browsers working on Older NT-Family OSes
I must admit, this one has me baffled. If I go to about:profiles and click "Restart with Add-Ons Disabled," the Windows Update Catalog loads fine. But if I disable every add-on manually and restart, the site still hangs at the "loading" state. I even tried switching back to the default theme, on the off-chance that it was causing the issue, but it didn't help either. Clearly, "Restart with Add-Ons Disabled" is doing more than disabling all add-ons and custom themes, but it's not at all clear to me what to try next. -
My Browser Builds (Part 5)
Mathwiz replied to roytam1's topic in Browsers working on Older NT-Family OSes
WTF now, M$? Github.com comes up fine in Edge or even 360EE. Edit: For information, the following pref is at its default value: network.http.accept-encoding;gzip, deflate, vary Edit 2: Hmm.... Works in St 52, not St 55. Could it be a profile issue? BRB.... Edit 3: Yep, it's a profile issue. Works in a clean profile, or even in my default profile with add-ons disabled. So, I just have to figure out which of my many add-ons is causing this Edit 4: Found it! It was Palefill 1.28, which, ironically, I don't think we even need any more! (I've instead been injecting polyfills with Violentmonkey as needed.) Unfortunately, disabling Palefill 1.28 didn't fix the other M$ site that no longer loads in Serpent: the Windows Update Catalog. I'll keep working on that. -
The number of open issues probably has a lot to do with how long each browser has been available. R3dfox has only been around for about 2 months. (This thread is only 4 pages long so far.) So it hasn't had much time to accumulate issue reports. Also, I think it's harder to modify modern browsers to work on XP/Vista, at least without "help" from OCAPI or the Vista extended kernel, than it is to modify them to run on 7/8/8.1. Support for the latter was removed only recently, so there's not as much code to change - thus, not as much opportunity for Murphy.... Edit: Speaking of XP/Vista, I saw on the R3dfox author's Github page a project called R3dfox Classic, a fork of Waterfox Classic. No work has been done on it AFAICS, but the readme says it'll be an XP/Vista version, I assume using techniques similar to what @feodor2 does to make MyPal 68 work on XP. Waterfox Classic was forked from FF 56, so if this comes to pass it'll probably look and act a lot like @roytam1's Serpent. The bad news, though, is that Waterfox Classic hasn't been updated since Nov. 2022, so there's a lot of catching up to do beyond making it XP compatible.
-
Getting back on topic, I think all this confusion was due to using the word "real" in the thread title and first post: I'm more familiar with Ungoogled Chromium than Brave, but in either case, I read "real" as meaning "developed by the same authors." So "real" Ungoogled Chromium would have to be developed by the Ungoogled Chromium team. @win32 could develop Ungoogled Supermium, but it wouldn't be "real" Ungoogled Chromium. If that's what the OP means, then the answer is most likely no. AIUI, the last version of Ungoogled Chromium compatible with Win 7/8/8.1 was version 109 (same as the "Googled" version). The Ungoogled Chromium team has shown no interest in backporting newer versions to older OSes. "Ungoogling" Chromium is hard enough work!
-
Probably not. I'm pretty sure there are applications where AVX instructions can offer considerable performance improvement. It's just that those applications aren't browsers and email clients. Well, at least, not yet. I'm sure Google is working on new CSS transitions or something that will be dog-slow without AVX. One can never waste use enough CPU/GPU cycles on flashy graphics!
-
Probably because if you're a Windows programmer, then you're probably running Windows 11, or at least 10, anyhow; so you see no need for one. Folks like @roytam1 who have modern equipment but still care about supporting older OSes are a rare breed. You're right - but it's because the software we're trying to backport was changed to require the latest OS versions, often intentionally. No one is forcing Google or Mozilla to call functions only available in Win 10+. For that matter, no one is forcing MCP to call functions only available in Win 7+. That was a choice they all made. I realize that said choice was often made implicitly, by moving to the latest (C/Rust/whatever) compilers, which by default, target the latest OSes. But it was still a choice to move to the latest compilers, and it was also a choice not to configure them to support older OSes. This is often done in the name of "performance," but (as with some hardware improvements like AVX) the performance gains are minimal - unless the idea is to force users to buy new hardware to run the new OS required to run the latest browsers. Of course your browser will run faster if you make all your users go out and buy new PCs!
-
True enough. As I said, it's a judgement call. But I don't think we need to get lost in the weeds of exactly what constitutes a "knock-off." From the thread title it's clear to me that the OP is looking for a recent version of Ungoogled Chromium and/or Brave ported to Windows <10. I suspect many other MSFN members are looking for the same. In theory it doesn't sound hard to do. Start with, say, Supermium/Thorium source, apply the Ungoogled Chromium patches, and build. But you have to know how to build Chromium, what to do if it doesn't compile or doesn't work - and of course you need compilers that can target your desired OS version and hardware to run them on. Not many of us have the needed hardware, software, and expertise - not to mention interest in tackling the project. I know I don't. I suspect that's why it hasn't been done yet.
-
I understood the term to refer to "name-brand" or "well-known" browsers (Firefox,Chromium/Chrome/Edge,IE) vs. Mozilla/Chromium forks intended to remove possibly unwanted features (Ungoogled Chromium) and/or to extend support to older OSes (R3dfox,Supermium/Thorium) That makes it a bit of a judgement call. Is Brave a "name-brand" browser, or is it a "knock-off" of Chromium? The OP seems to consider it name-brand, but others may differ. Maybe others are reading it differently. So far I'm the only one who's voted in favor of "knock-offs."
-
My Browser Builds (Part 5)
Mathwiz replied to roytam1's topic in Browsers working on Older NT-Family OSes
Sounds a lot like mine! Mine has a faster clock but otherwise very similar. It's about 10 years old but it still runs Serpent quite smoothly (at least in multiprocess mode, which isn't available in PM/NM/Basilisk). Switching topics, I think the problem with Serpent's fill login not working at Chase.com is due to Chase recently implementing custom elements on their login screen. They don't have <input> tags any more. Custom element support was added to UXP (& 55 by Roytam) but I don't think the fill login function was updated to work with custom elements. -
My Browser Builds (Part 5)
Mathwiz replied to roytam1's topic in Browsers working on Older NT-Family OSes
The irony is, the original idea behind Pale Moon, waaay back before XP even, at least as I understood it, was that it was supposed to just be Firefox optimized for Windows, so it was supposed to be (and seemed to be, on my ancient Win 98 PC) faster than vanilla Firefox. No new or different features, just better performance. This was like PM 3.6 or so, but it sure seems like they went astray somewhere along the line. And enabling AVX or whatever in 64-bit builds isn't going to make up for it. I tried disabling multiprocess mode on St 55 on Win 7 and had a similar experience. Couldn't even type a post on MSFN at 10-20 seconds per letter, with one CPU core maxed out! So, even with a single core, you might have had better luck with e10s forced on. It's my opinion that the OS version makes little difference in performance, assuming the application (browser or whatever) will run on both OSes. The app might be faster if optimized for a newer version, but in that case it's unlikely to run on the older version at all. It's mostly the hardware, rather than the OS, that provides good performance. -
My Browser Builds (Part 5)
Mathwiz replied to roytam1's topic in Browsers working on Older NT-Family OSes
I'm not sure that's true. Maybe it's true of some, but I often read what's going on "over there" when @VistaLover or @UCyborg links to a thread about a technical issue that affects UXP browsers (and "UXP-like" browsers such as St 55) generally. That said, there's often an arrogance "over there" that their decisions are the only "correct" decisions. It's not merely hating on MCP to recognize that there's still a lot of unnecessary tension between them and @roytam1 (not to mention @feodor2). It's important to recognize where we stand. I'm perfectly happy with MCP and @basilisk-dev refusing to support Vista and older OSes. You have to draw the line somewhere. But that doesn't make it the correct decision for everyone on the planet! Some folks prefer to stick with XP, some Vista, etc. I happen to prefer Windows 7 without Aero, which many folks think is weird because it makes it look like Win 98! But I think it's best to adopt a "live and let live" attitude, and not try to force my preferences on anyone else. Everyone is trying to make the best decision for themselves, and I think it's better to support whatever folks are trying to do than to belittle them because they made a different choice than I might have. -
My Browser Builds (Part 5)
Mathwiz replied to roytam1's topic in Browsers working on Older NT-Family OSes
Thanks for posting, @Jody Thornton. To me, the thread looked like just a straw man attack on Win XP: what happens if I turn off the firewall and connect my PC directly to the Internet without an AV program or even a router for protection? Oh, wow; bad things! Well, obviously the fault must be Win XP, not the fact that I turned off the firewall and connected my PC directly to the Internet without an AV program or even a router for protection! As for New Moon, there was just that one comment - and he seemed mostly angered by those who ask for NM support at the PM forum, despite the fact that @roytam1 and the rest of us constantly tell anyone who'll listen not to do that! Certainly not @roytam1's fault. -
My Browser Builds (Part 5)
Mathwiz replied to roytam1's topic in Browsers working on Older NT-Family OSes
Interesting. Unfortunately I don't see a specific answer to who owned the x.com domain, which (I assume) wasn't a big thing before Musk decided to rebrand Twitter as X (leading me to refer to the two domains collectively as Xitter.com). -
My Browser Builds (Part 5)
Mathwiz replied to roytam1's topic in Browsers working on Older NT-Family OSes
Off-topic: I wonder who owned x.com before Musk bought it? Surely someone did - I bet they made a bundle. -
My Browser Builds (Part 5)
Mathwiz replied to roytam1's topic in Browsers working on Older NT-Family OSes
Even if we got HEVC pass-thru working (on PCs with hardware HEVC decoding) it wouldn't help. You'd need the newest Widevine version to watch DRM-protected streams. No realistic chance of that happening on a browser targeting older OSes. -
My Browser Builds (Part 5)
Mathwiz replied to roytam1's topic in Browsers working on Older NT-Family OSes
You have to log in to view that forum. Could you summarize for those of us unwilling to do that? But even without reading the above, my theory is, Tobin trained them to hate us - how dare we presume to do anything differently than them - don't we understand they're perfect and all their decisions, including refusing to support XP/Vista, are correct? And that feeling lingers, even with Tobin out of the picture. -
My Browser Builds (Part 5)
Mathwiz replied to roytam1's topic in Browsers working on Older NT-Family OSes
OK, now things have gotten beyond ridiculous. Chase.com is one thing, but now I can't even open the Micro$oft Update Catalog in Serpent 55! It starts, but gets stuck in the "loading" state; never finishes (not even with a blank screen). User agent overrides appear to have no effect, although I obviously haven't tried every possibility. Thing is, I can open M$ Update Catalog even in 360EE, so it doesn't require a particularly new browser to load it. Edit: Never mind. M$ update catalog loads just fine in a clean profile. (FWIW, Chase.com works more reliably in a clean profile too, although "fill login" is still broken in Serpent. I don't even need my StructuredClone polyfill any more!) Obviously one of my add-ons or about:config tweaks is interfering. Someday I'll learn to check with a clean profile before posting an issue. -
My Browser Builds (Part 5)
Mathwiz replied to roytam1's topic in Browsers working on Older NT-Family OSes
You'd think if that were the problem, the forum could at least be a little bit less opaque and just say, "you can't say '<whatever>' here!" There were no swear words, so next time it happens, I'll just post a screen shot and let someone else explain to me what I said that was 403 forbidden. BTW, on my Chase.com issues, I discovered that, sometimes, it does work in Serpent, although that's rare on my home PC. I now suspect Chase has some kind of timeout, and since my home PC is slower than my work PC, the timeout rarely happens at work. (So now Chase is trying to require not only a modern browser, but also a modern-enough PC to run its Javascript-addled site quickly enough for its taste!) -
My Browser Builds (Part 5)
Mathwiz replied to roytam1's topic in Browsers working on Older NT-Family OSes
Something is wrong with the site - I can't seem to post more than one line without getting "403 Forbidden." -
My Browser Builds (Part 5)
Mathwiz replied to roytam1's topic in Browsers working on Older NT-Family OSes
I just wish Chase would quit "fixing" things that ain't broke... -
My Browser Builds (Part 5)
Mathwiz replied to roytam1's topic in Browsers working on Older NT-Family OSes
test -
Off-topic, but this is a pet peeve of mine. Why doesn't Micro$oft (or anyone else, for that matter) print huge numbers like that with commas (or periods for Europe) - or, since we don't really need to know the RAM usage right down to the byte - just abbreviate it to 2.57 GB? It's virtually impossible to quickly gauge the magnitude of a long string of digits like that - instead you have to stare at it and count digits from right to left! It's as if they've forgotten that computers were invented to do drudgery like that for us!