Leaderboard

I've been using it today, other than videos skip sometimes, works well with not errors as stated above. Looking forward to progress :)

I can confirm that MyPal 68 works absolutely fine on my end (and am typing this reply from MP68 right now). In fact, it hasn't crashed on me once (not even when I was using the beta version from before the current release). Of course, my use cases are different from those of others, and I am very grateful I've been lucky enough not to have a crash or error yet. It seems the 68ESR code base was already quite stable when Feodor decided to fork it for XP--I'm not sure if any updates, aside from security fixes, are still coming out that are applicable to that code base, though. (Of course, I know Mozilla ended support for 68ESR quite a while ago, but there are still bits and pieces of later/future code that could be used to perfect & stabilize it as much as possible.)

I'm sure he'll consider it; in the meantime, have you tried IceDove?

No problem here. Never seen this error before! I use Mypal 68 in single-process mode. Cheers, AstroSkipper

With regards to @roytam1 and @feodor2, I appreciate their builds very much. I'm sure just about everyone else here does too...they've played a big role in maintaining modern & reliable browsers for our daily use. The fact that they still support XP is an incredible bonus, and I definitely don't take that for granted as a longtime XP user. Funnily enough, there was a time when I was sticking with Feodor's builds because Roy's were updated 'too frequently', and after that there was a time when I switched over to Roy's builds because Feodor's weren't updated 'frequently enough'. Now I use browsers by both of them, and enjoy them considerably. The important thing to remember is that Roy's builds tend to be weekly or bi-weekly snapshots of where a particular code base is currently at from a maturity standpoint, while Feodor's builds (at least when he was still working with MCP code) follow the release schedule of the official browsers. It doesn't make one inherently superior or inferior to the other...it just means their approaches are different. I've never had any issues with either of them--in fact, if I'm not mistaken, there have been times when Feodor and Roy both mutually benefited from each other's work and helped one another. For all we know, Roy might end up helping Feodor with MyPal 68 & future MyPal versions, and Feodor might end up helping Roy too. Actually, I find it quite interesting to go back in the archives and try older versions of some of Roy's browsers...if you go check the gpc/files1.rt section at his website, he has an incredible treasure trove of 'alternative' browsers and other cool stuff. Early experimental builds of Serpent and New Moon, classic browsers updated with TLS 1.2 & even NSS patches to add TLS 1.2 support to older browsers open up a world of possibilities. Currently, I'm using MyPal 29.3, both the NHTPG and Humming Owl ungoogled builds of 360EEv11, and a recent Serpent 55 build. I also use other browsers (Roy's Firefox 3.6, RetroZilla and Phoenix with added TLS 1.2, Netscape 9 with updated NSS, Opera 12.18, QtWeb, etc.) for more lightweight browsing. Using a Classilla (Nokia N90) user agent keeps them quite viable even in 2022, even if their rendering capabilities are dated by today's standards...with other user agents, there are a lot of possibilities for experimentation. And of course, the vintage Geocities/Angelfire-era sites render with no problems...TheOldNet, 68k.news & FrogFind are truly amazing. I don't expect even Roy's newest build of Serpent 52/55 or New Moon 28 to be able to view every site...some, like DeepL, are starting to run into problems, but I wouldn't abandon them as lost causes just because not every site works 100% correctly. I care more about a code base being fundamentally solid, stable and reliable than anything else, and that is very true of Roy's browsers. Feodor was able to prove that he didn't need MCP code to make a good browser, and I think his 'resurrected' MyPal is quite fantastic. Roy's also done one hell of a job in maintaining mature code bases for his own browser builds. I believe we can appreciate Feodor without putting Roy down, and appreciate Roy without putting Feodor down...they've both gone above and beyond the call of duty to maintain reliable browser code bases, and by fortunate extension sustain the long-term usability of XP long after many mainstream companies started leaving XP users in the dust. Too many old computers end up in landfills...because of their work, a lot of computers that would have otherwise ended up at a dump are still in daily use (including the Pentium D-based machine that I use on a daily basis--the 'average' person would have thrown such a PC out by now, often because they'd see the 'Designed for MS Win XP' sticker on it & automatically assume it's old, slow, & useless). (I'll stop here before I end up in rant territory, if I haven't gotten there already.) My sincere gratitude to Roy, Feodor, and everyone else here at MSFN! Long live XP and all other great OSes!

Ah OK, thanks! That's a bit disappointing, but at least the potential is there to add the necessary TLS 1.3 support. Now we just need someone to do it!

I think he meant a release supporting CPUs with SSE instructions set only, lacking of SSE2 or higher.

What does this mean? A build "without" SSE? A build "for" SSE?

For me Mypal 68 is working great too. No crashes and an overall impressive performance on hard-to-load websites. But regarding to the development of higher versions I would prefer completing missing functions in Mypal's settings and fixing the immature addon support at the moment. Generally I think it's an early state of development and there is some work necessary to get a final version of Mypal 68. Anyway, I am glad to have a browser like Mypal 68 and very interested in coming releases. Cheers, AstroSkipper

It works great here. Never got a crash here either. Awaiting Mypal versions 78 and 91. Mainly 91 since there will be an even better and modern experience.

StartAllBack 3.4 Underlined are the changes since 3.3.9d

OK, have not done that. From your query, I take it the button icons are actually a customized character font, much the same as many Web sites (including MSFN) use to add small icons to their sites. Let me fire up the 32-bit version of St 55, which still shows the bug on my home PC, and see if a similar bug affects the icons at the top of the MSFN page. Hmm.... MSFN's icons appear normal, even though St 55's icons are fouled up: Still, I think you're onto something. Just can't quite figure it out yet.

Is anyone able to use the latest release without getting this error two minutes in? Nothing peculiar about my system... Serpent works fine...

@roytam1 : Success indeed! : File "./basilisk/omni.ja" was de-optimised and unpacked (with 7-zip) to omni dir; compiled file "./omni/jsloader/resource/gre/modules/GMPUtils.jsm" was DELETED; source file "./omni/modules/GMPUtils.jsm" was edited according to the "fix"; contents of omni dir were compressed back to an omni.ja (zip) archive; modified omni.ja file was optimised and put in the place of the original (that came with the 2022-05-12 release).

No problem with this maypal, other than being quantum. As per roytam's builds, I have been using them for 5 or more years, and all issues (within reason) in those builds have always been solved, fast, well and kindly. Just sayin'.

I have to disagree. I do test my builds, but only preliminarily tested. I just can't test every aspects each time I build.

Dark theme used to be have a following by people who thought it was cool, dating back to the "High Contrast Black" days of 3.x and NT. It wasn't originally meant to be anything other than an accessibility option. Then sometime during the late 90s some studies were run and it was found that using the dark themes (the High Contrast Black OS theme) was beneficial to CRT display longevity and even became a recommendation in computing magazines (and even in web design) up through the mid 2000s. However the thing that changed was the move away from CRT to LCD panels and it was learned that the dark themes had no benefit any longer. Unfortunately it was too late to remove the people's want of a dark theme they didn't actually need any longer and we still have it as an option to this day.

Just checking in and see feedback has dropped for this project last Month - any progress? I realize this is early phase testing, but hopefully he is still working on it to increase our options. Hope everyone involved is managing as best as possible where there is no peace in the world right now.

Mathwiz wrote: >I second that motion. The change log is usually quite long and very technical; > hard to understand even for those of us who dabble in programming. > Putting it in a spoiler tag would let us refer to it if need be but save our scroll wheels if not... No, please not SPOILER! That tends to be hidden completely for old browsers, and not sure if also hindering page-search or archiving. Perhaps something else - "code"? But that increases the size yet more, using ultrawide Courier-fonts. Another...? (That said, I'd be surprised if RT changes anything at all only for reading convenience) But being a selfhelper, couldn't resist playing a bit now ;-) Tweak-1: This CSS filters all comments written by RT, then limits all single P-elements to a max height and if too long, shows a scrollbar: body.ipsApp article.ipsComment div[data-quotedata*='roytam1'] div.cPost_contentWrap p{ max-height: 400px !important; overflow-y: auto !important; } CSS-variations for example: remove author-requirement, or limit complete-comment height instead of single P-elements Tweak-2: This JAVASCRIPT (bookmarklet?) filters all comments written by RT, but (unlike above) takes the complete comment text. Then checks if the TEXT "- Bug " is contained more often than 8x. Only THOSE comments get limited with a scrollbar: var c="body.ipsApp article.ipsComment div[data-quotedata*='roytam1']"; var j, el, lg0, lg2, x, LIST=document.querySelectorAll(c); for (j=0; j < LIST.length; j++){ el=LIST[ j].querySelector("div[data-role='commentContent']"); lg0=el.textContent.length; lg2=el.textContent.replace(/- Bug /gi, '').length; x =(lg0 - lg2)/6; if (x > 8) { el.style.maxHeight='400px'; void(el.style.overflow='auto'); /* just for TESTING! Remove line: */ void(el.style.backgroundColor='yellow'); }} JS-variations: remove author requirement, choose another max-height, choose another TEXT, choose another hits-limit .

Can confirm, it runs on Windows 2000 with KernelEX after modifying mozglue.dll and xul.dll. Here are the instructions I used.