Jump to content

Mathwiz

Member
 View Profile  See their activity

  • Posts

    1,858

  • Joined

  • Last visited

  • Days Won

    51

  • Donations

    0.00 USD 

  • Country

    United States

 Content Type 

Everything posted by Mathwiz

  1. Mathwiz
    Oh, come on. You cannot even enable e10s in New Moon. But you can enable it in Serpent (both 52 and 55). Did you try that? If not, you did an "apples to oranges" comparison: FF with e10s is faster than Serpent without e10s. Well, duh; FF with e10s is faster than FF without e10s too - that's why we have a whole thread on enabling it! (To be fair, FF/Serpent with e10s is quite the memory hog - and still slow - compared to Chrome 86/87 ports. So I wouldn't recommend it unless you've beefed up the RAM in your PC, preferably with 64-bit XP or later so you can get more than 4GB, or at least with an SSD for the swap file.) Technically, you're right about that. Multiprocessing is mostly a convenience for developers. But that said, it does have its advantages for us end users. Beyond the obvious (letting a 32-bit app use more RAM), if your tab crashes in FF/Serpent, and you're running in the default single-process mode, your whole browser crashes! But if that happens in multiprocess mode, you just reload the tab. <rant> And why shouldn't she be? The original idea behind HTML was supposed to be that any browser, no matter how primitive, would render a "usable" Web page, merely ignoring the tags it didn't understand. If you use a browser from 2001, it may look like a Web site from 2001, but you're still supposed to be able to use it, at least. But that's long gone, thanks largely to Google (but also others including Mozilla). Nowadays, if your browser doesn't understand the latest bullish snarkifying operator added to JavaScript in the latest Chromium release, you're likely to just get a totally blank page (or if you're "lucky," a curt message telling you to "upgrade" your browser, even if that means "upgrading" your OS to the latest monstrosity from Micro$oft, and in turn buying a new PC capable of running that monstrosity) because the Web designer used JavaScript with that bullish snarkifying operator to build the entire Web page from scratch! HTML? Who needs it? (With the side effect being you can't disable JavaScript any more, and have to rely on other, more complex means to block the spyware embedded within.) So yeah, we're angry. Not (for the most part) because we intend to use a PC that old, but because we shouldn't have to keep buying newer, more expensive PCs every few years just to keep doing the same exact things we've always done, just because some Web framework developer couldn't resist using that shiny new bullish snarkifying operator (for "security")! </rant>
  2. Mathwiz
    Well, it could be, yes. I'm just going on past experience: so far, Google has been behind far more UXP-breaking changes than Mozilla, so I figured the odds are pretty good that the next one will be Google again. (And my usual disclaimer: Not all UXP-breaking changes are "egregious;" i.e., serving little purpose other than to break older browsers. Some are quite useful to Web designers. Many do strike me as quite egregious, but I don't know enough about all this new CSS stuff to have an informed opinion yet.)
  3. Mathwiz
    So until MCP implements the "revert" CSS feature, the "best" workaround is far from ideal: toggling pref layout.css.is-where-pseudo.enabled to false, thus restoring the CSS behavior of previous UXP browser versions. Of course, by the time MCP does implement "revert," it's nearly certain Google will have come up with something else, and MCP will have to run even faster to stay in the same place.
  4. Mathwiz
    A little bold text would probably help: "New New Moon 27 Build!" Otherwise it does look kind of silly. For the beginner, I'd recommend starting with New Moon 28. If you find NM28 impossibly slow on your PC, try New Moon 27; it's less resource-intensive, but it's based on an older Firefox version so it won't work with as many Web sites.
  5. Mathwiz
    *sigh* More "features" to implement. The Red Queen's race continues....
  6. Mathwiz
    Yes, same idea. The add-on I linked to only changes the fingerprint after a user-specified amount of time has elapsed. The idea is to guard against (say) Cloudflare gathering the fingerprint twice to see if it changes. With the add-on, it will most likely be the same, so Cloudflare won't block you - but tomorrow the fingerprint will be different, and Cloudflare will start from scratch, tracking what it thinks is a whole new browser.
  7. Mathwiz
    Folks, this post was asking how to solve a specific problem with Kafan MiniBrowser - not to start an argument over which XP service pack is fastest. I have the same question. Anyone have an answer?
  8. Mathwiz
    What if you change your fingerprint periodically: https://addons.mozilla.org/en-US/firefox/addon/no-canvas-fingerprinting/?utm_source=addons.mozilla.org&utm_medium=referral&utm_content=search They can collect all the fingerprints they like, but they're quickly rendered useless.
  9. Mathwiz
    OMG. If you're gonna open 20 f'ing tabs in Serpent, please read: (You do need some "real" RAM in your PC, and an SSD wouldn't hurt....) BTW, my old work PC died this week. Kevin thinks its the HDD, but it won't even boot from a CD/DVD, so I think it's the SATA controller, which means the motherboard. So I'm using my laptop, on (ugh) Win 11. No more XP, for the time being at least.... I've noticed that Serpent (55, at least) is way faster on the laptop! If only I could ditch Win 11! I'd even settle for Win 10 at this point.
  10. Mathwiz
    Sounds like you just need a theme. Any SeaMonkey theme should work, although you'll probably need to modify the install.rdf file in the .xpi archive. I think instructions on modifying install.rdf are available in one of the earlier "parts" of this thread. Or you could try IceApe, whose default theme resembles the original one.
  11. Mathwiz
    Per Wikipedia: So it's an optional mode on "modern" browsers. Also, many sites specify "strict transport security," one effect of which is to automatically upgrade http connections to https on that site, once a successful https connection has been made.
  12. Mathwiz
    There should be an option for 2FA, particularly on email sites, but it shouldn't be mandatory. Let the users decide how much security they need. OTOH, 2FA would've saved Hillary Clinton a ton of grief when Podesta got phished back in 2016, so maybe 2FA should be "encouraged," at least on non-personal accounts. The email service could require more frequent password changes if 2FA isn't used, for instance.
  13. Mathwiz
    HTTPS Everywhere, perhaps? (After all, that's what it's supposed to do.) That extension is no longer developed. "Modern" browsers don't need it, but I think it's still useful on older browsers (or browsers based on them, like ours). Usually you want the https: version if available. Luckily, it can be disabled for specific sites like this one. And after a couple of reloads, it works! Strangely, though, HTTPS Everywhere didn't force an https: connection when using the IP address; only when using the host name.
  14. Mathwiz
    That's interesting. Https has become so ubiquitous I'm surprised there are any sites left that still support plain http. In fact I just tried it on St 55 and it immediately redirects to the (non-working) https: site here. Using just the IP address (http://37.230.96.101/ for me) did the trick though. FWIW, html5test.com's tests aren't always accurate. It says St 55 doesn't support WebP images, for instance, even though it does.
  15. Mathwiz
    That has to be it. Under Bk, html5.com just won't work; under 360Chrome, it does come up with an obvious "not secure" icon where the padlock should be: This was actually under Win 7, but it is one of NHTPG's versions. I couldn't figure out how to find out exactly what was triggering the "not secure" icon in Chrome, but an expired certificate seems like the most plausible cause. I guess since it's under Win 7, I could remove the --ignore-certificate-errors flag; but then html5test.com wouldn't work under Chrome either!
  16. Mathwiz
    I just tried html5test.com for the first time in several months and found it doesn't work any more! It just shows a spinning wheel that never stops. I tried St 55, 52, IceApe, clean profile, and versions as old as May 5 of this year. Nothing worked except 360Chrome. Perhaps they changed something. Maybe I'm not looking at it right but I don't see any JS errors. Wait - I think I see the problem. There's a message: api.whichbrowser.net uses an expired security certificate. Expired on 7/12. But why does 360Chrome work? Surely that certificate is expired on all browsers!
  17. Mathwiz
    I was going to ask if @roytam1 could do that, but I figured it wasn't worth the trouble - then, it turns out, he had already done so long ago! FYI, under Win 7, Widevine 1.4.8 is recognized and appears in the "plugins" section of about:addons under St 55, but it doesn't seem to work (Bitmovin's test page reports "No DRM" under the Media Source Extensions side of the page). I suppose it's a moot point, though, since 1.4.8 has long been blacklisted anyway.
  18. Mathwiz
    Sometimes it helps to review old history, so as not to relive it! What's relevant to this forum, perhaps, is that, for a Vista user, St 55 likely lacks Vista WMF support, while UXP browsers include it. (Both have Win 7 WMF support, but of course there are many other browser choices for Win 7.) It's probably not a huge deal, though, since both browsers already include extensive (non-WMF) media support for the benefit of XP users.
  19. Mathwiz
    Aaack!
  20. Mathwiz
    Sounds like they originally "targeted" Vista (didn't use any 7-specific functions) but changed it at the last minute. Did the alpha version (base for St 55) run on Vista? So it runs, but it runs like it runs on XP (if patched with, e.g., the old XomPie tool): no WMF support, so you Vista users are better off with FF ESR 52.9 than with "hacked' FF 53 (except perhaps for somewhat better WebExtension add-on support in 53).
  21. Mathwiz
    Doesn't FF 53 run on Vista? I actually do that intentionally with Serpent, in order to change the Help / About Serpent pop-up window. You can replace/improve the logo and text: I don't bother to re-optimize, which undoubtedly slows down launching the browser, but I don't do that often enough to care! (Image at left courtesy of @dencorso; unfortunately we couldn't confirm it was free of copyright restrictions, so it was never made generally available. But it looks cool so I'm using it on my own personal copy .)
  22. Mathwiz
    Yes, the Wayback Machine is an excellent resource. And thanks for finding it and posting the link. I'm still disgusted with dotPDN's attitude, though. Perhaps I'll have better luck with the above link, but I tried the portableapps version, but couldn't get the JPEG XL plug-in to work with it, even though I have a 64-bit system. Speaking of which.... I still rely on a few 16-bit apps, which of course don't work on 64-bit Windows, because Micro$not has decreed that only the two latest "bitness" levels shall be supported. The need for a 32-bit OS to run those apps was a major reason for installing "XP Mode" on Win 7 when I upgraded back in 2015 or so!
  23. Mathwiz
    Probably, the extension uses JavaScript to decode JPEG XL, which is probably a lot slower than a JPEG XL decoder written in C or C++. Yes, 4.3.12 was the version I was looking for. I did stumble across it at portableapps.com, and wondered if they had the full version vs. just an "online" installer that wouldn't work. But portableapps.com has their own "PortableApps Platform" which looked like yet another complication. Turns out you don't have to use it, but I also figured a "portable" app wouldn't upgrade my existing app or keep my settings. After all, by definition a "portable" app is designed to be installed on and run from a thumb drive. Besides, I was already disgusted with dotPDN's attitude. I didn't mind that they stopped supporting Win 7, but what really irked me was that they went out of their way to erase their last Win 7 version from as much of the Internet as possible, so if you didn't upgrade in the days between 4.3.12 and 4.4, you were SOL. Reminded me of too many folks' attitude towards XP/Vista: "We've decided not only to stop supporting it, but that you shouldn't be using it, so we're going to do as much as we can to pressure you to 'upgrade.'" Oh - and Paint.net 4.3.12 doesn't come with the JPEG XL plugin - you have to add that later yourself. Switching to Gimp was a lot less work.
  24. Mathwiz
    Serpent may be a somewhat different beast. It (well, at least St 55) has a second pref, network.http.accept.image (that I guess MCP added), but I don't know what it does as compared to the pref inherited from FF. But in any case, the FF 52 experiment showed that Bing isn't smart enough not to send WebP to a browser that indicates it doesn't want WebP. Extremely bad practice, indeed. Back to JPEG XL. I tried to update Paint.net on my Win 7 system last night to a version that supports JPEG XL (which happens to be the last Paint.net version for Win 7), and discovered the author meanly removed all but the latest (Win 10+) version from his GitHub page. I tried to find the needed version, but the only installer I could find is an "online" installer and it kept getting a stupid SSL error. I don't have time for that kind of nonsense, so I just uninstalled Paint.net and switched to Gimp. Gimp isn't as user-friendly, but it still runs on Win 7 and it includes a JPEG XL plugin. I loaded in a big 3.3MB .PNG image and exported it as .JXL. The two images look identical to my (admittedly untrained) eyes, but the .JXL one was about 1/4 the size of the .PNG. Wow - no wonder folks want to switch!
  25. Mathwiz
    It probably works for the very specific case of Bing on IE, because Microsoft wrote IE and knows what will (and won't) work with it. As for FF 52, by default it just sends Accept: */* for images, which just tells the Web page to send whatever it thinks is best, so Bing sends WebP which doesn't work. There's a pref in FF-based browsers to control what image formats it tells the Web page to use: image.http.accept. Perhaps changing the pref from */* to something like image/png,image/jpeg,image/*;q=0.3,*/*;q=0.1 would get Bing to send FF 52 the correct format. In the latest Serpent 55, that pref defaults to image/webp,image/jxr,image/png,image/*;q=0.1,*/*;q=0.1 (Note that image/jxr is JPEG extended range, not a typo for JXL.) So if @roytam1 adds JPEG XL support to Serpent, he should probably add image/jxl to the front of that default value string. Reference: https://developer.mozilla.org/en-US/docs/Web/HTTP/Content_negotiation/List_of_default_Accept_values#values_for_an_image
×
×
  • Create New...