Andrew T.

From the pictures and description, this board uses an external battery pack connected to jumper J24...much easier to work around than a dead Dallas chip.

It seems like only yesterday, computers of this era were being discarded en masse and were basically free for the taking. It almost feels weird to see a bare 486 motherboard go for $30...though the original purchase price could have been 100 times that! The hardware sounds solid from the description. The capacitors aren't leaking. There aren't critical chips missing. The CMOS error ought to go away with a working battery. There's a bit of useful information on the board here. The board has an AT form factor and power supply connectors, so it won't fit in an ATX case...though some late '90s cases were designed to mount either AT or ATX motherboards. VLB or VESA Local Bus slots are the long connectors underneath the CPU sockets. It was a popular backward-compatible standard for video cards in the late 486 era; though eclipsed by PCI around 1995. Any VLB or ISA video card from the era ought to work, along with any off-the-shelf IDE CD-ROM drive. You'll also need a floppy drive to install or boot DOS and partition a hard drive. DOS 6.x and Windows 3.1x would run well (as well as Windows 3's flaky self can run, that is) and be very period-accurate. Windows 95 would also be usable, but it would be a stretch. All of these are things that would have been obvious to most computer users 20 years ago, but have been "lost" with the passage of time.

Are you looking for new laptops or used laptops? Obviously, most (all?) of the answers are going to trend towards the latter category... I run Windows 2000 on a twelve-year-old Dell Inspiron 2600 that I got for free. I seldom vouch for Dell as a company, but they still offer Win2000 driver downloads for the system.

I've rarely bothered with custom animated cursors...because in my experience, they impaired stability. Interesting that color depth may have been the issue to blame.

My experiences likely jibe with those of others. "What limitations do you face using 9x or ME in 2014?" Obviously, it's been a challenge finding compatible software and hardware for ten years and counting. There are no 95-compatible convertors for Office 2007 files that I know of. The mad and partially Apple-induced rush to replace working legacy interfaces with USB was a ploy that left 95 (and NT4) users high and dry. I use Windows 2000 for syncing up my digital camera; though not much else. "Can you still access the Internet?" Yes. You can also access the Internet with DOS or Mac OS 7; it isn't a big hurdle in itself. "Have you lost the ability to watch YouTube videos?" No; YouTube is still compatible with a patched Flash 7 on SeaMonkey 1.1 (though their browser checks attempt to throw you for a loop while trying). "Can you no longer bank and shop on the Internet through SSL?" SSL is not an issue...for shopping, anyway. I personally avoid online banking out of concerns for security. "Do certain buttons and features on websites no longer work because your version of Flash or Java are too old?" This is an issue...Flash 7 and Java 1.4 are quite a few years old, and most of the content that targets newer versions is not backwards-compatible. The biggest issues I have with websites these days isn't with browser plugins, however, but with JavaScript content: The trend of the last few years has been to link to enormous JQuery libraries and tracking scripts that bog down and break on 9x-compatible browsers, then rely on them for basic functionality so that the sites are blank or totally broken without JavaScript enabled. The Disqus comment plugin is epidemic of this: Dozens of big websites that should know better have blindly replaced their own self-hosted comment forms with that crap, and it doesn't even work on Firefox 3! "Is your web browser awfully slow and takes a long time to render webpages?" SeaMonkey 1.1 is slow (less slow with Noscript), but reliable. Opera 10 is fast, but flaky. "Do you think sometime in the future you'll be forced to retire Windows 9x or ME?" I won't make any denials here: Windows 95 isn't getting any easier to use, and I may be forced to simplify at some point and retire my 9x-era PCs in favor of a single newer computer. (Though what that computer would be or run is anyone's guess...I won't touch Windows XP or subsequent versions with a ten-foot pole; ten years of experimentation with Linux have made me dislike its conventions with a passion, and Apple lost me as a customer when the last beige G3 left the assembly line.)

Since Windows XP has finally reached its end of "support," what are the implications for Windows 9x users? I feel somewhat ambivalent about it all. I never liked Windows XP, and Product Activation is the biggest deal-breaker: I have no intention of using an operating system that binds you to phone or internet connections and the mercy of its maker, I will never run anything newer than Windows 2000 on any home computer that I care about, and that's not going to change now. Nevertheless XP is still far superior to Vista/7/8 in configurability and performance (low a bar as that may be), so it still feels a bit of a blow as the trends of popularly-endorsed technology move away from what they could be and what I wish they were. The biggest issue may be that Windows 9x becomes shunted one more step down on the obsolescence ladder. When application and hardware vendors stop targeting Windows XP and start going out of their way to use Win7/8-only APIs that break functionality with it (or use SDKs that do this for them), the task of getting stuff to work with Windows 2000...let alone a Win98/KernelEx kludge, let alone Windows 95...isn't going to get any easier.

I think this means hell hath frozen over.

My question: Would Chrome have worked on Windows 9x out-of-the-box if Google's developers had used a different compiler?

User-agent spoofing? That sometimes works in a pinch, but it rarely goes far when the issue at stake is a plugin or the page coding and scripting itself.There is a custom compile of SeaMonkey 1.1.20pre here that incorporates a few Firefox 3-era rendering tricks. It's the end of the road as far as 9x/NT3-4 compatible browsers go without kernel extensions...I used it for a while, though sadly these days it doesn't seem like enough. I wouldn't be surprised if this was actually because of an oversight. With big sites chasing after Chrome, Safari, Firefox, and IE willy nilly, Opera and SeaMonkey often end up in the statistical noise...and the browsers get overlooked in "we only support" lists and useragent-sniffing scripts alike! I tend to think that mobile websites are makeshift bandages over deeper functional and technological problems. The sniffing routines that redirect people to them are often shoddy and introduce problems of their own. (Facebook redirected all SeaMonkey users to their mobile site a few years back, even though the normal site worked fine.) Plus too many sites are severely crippled in functionality relative to their non-mobile counterparts, so I'm left nonplussed by the trend. While I've expounded on the plusses of 95 OSR2 time and time before (cough), a good bit of it boils down to this: No browser integration with IE, no DLL-commandeering by IE, no help files in IE windows, no file managers in IE windows, no disk space wasted by IE, no gaping security holes introduced by IE, and no performance issues induced by IE. If you take Windows 95 and iron IE 4 or 5.5 over it, you've defeated the benefits of using Windows 95.That said, I don't use it exclusively: My secondary computer runs Windows 2000, so I'm not confined when the limits are reached.

Google have been trying to push Chrome (and by extension, Windows XP+ and all it entails) on everyone. They've been dropping annoying "upgrade" banners on all their properties; even in places where they don't make a lick of sense. They've also used useragent sniffing to withhold tools or redirect users of not-up-to-the-second browsers to older versions of sites: Don't try looking for the "Search by Image" option for Google Images on Firefox 3, since it won't be there. 3.6.28 is the last version of Firefox I'll use on any platform, but for me the frustrations started kicking in three to five years ago when sites like Flickr and Slashdot began breaking and barfing on Firefox 1.5.0.12. I avoided the 2.0 upgrade since it didn't work on Win95 out of the box, and eventually moved to Opera 10 chasing the ever-moving illusion of popular compatibility. As far as I'm concerned, there is no reason why any site shouldn't render decently on any standards-compliant, CSS- and PNG-capable browser of the last decade...Firefox, Opera, SeaMonkey, Camino, Konqueror, Netscape 7.2, or anything else you might fancy...but sadly, that hasn't been the case. Whatever happened to the hallmarks of the post-browser war web I hoped we were heading into ten years ago; a web with graceful degradation, device-neutral delivery of content, and standards like XHTML 1.0 Strict? Lost in the cloud of HTML5 trendiness, kludgy JQuery/AJAX abuse, bloated codebases, crass advertiser-coddling, and user-tracking; no doubt.

I touched on some of those considerations years ago:http://www.msfn.org/board/topic/104812-why-use-win-98seme/#entry694890 http://www.msfn.org/board/topic/134161-windows-95-or-windows-98-se/#entry858039

Thanks for the feedback and investigation work! The "handle random port DNS" pronouncements read like incoherent gibberish. My first guess was that it had something to do with systems without the Winsock 2 update...but that's been around for almost as long as Windows 95 itself, so no excuses there. I'd like to think that users of Linux and sub-XP Windows have a fair amount in common...namely, an enterprising spirit and a disenchantment with the state and direction of Microsoft's commercial software development. But I digress... It'd be challenging to do useragent experimentation, that's for sure. It's also difficult to discover where these sites exist without being blocked by them first! "There are no windows 95 machines that can access the net anymore" is a total hoot. I almost wish I could browse the SpambotSecurity site for more gems of trivially-rebuttable idiocy...alas, I'm limited to the Google cache for that. Well, I'm posting here now!

In between Flash 8, kludgy scripts, and restrictive browser checks, I've experienced quite a few frustrations using Windows 95 on the web over the last few years. Only rarely have I encountered downright hostility, though...but this takes the cake. Over the last year or two, I've stumbled across sites that run some server-side PHP package called Spambot Security ZB Block. If you access one of these sites from a Windows 95 computer (or any browser with a Win95 useragent), this arrogant and absolutely obnoxious message appears: They're not kidding. If I try to access any page on the site from any computer from the same IP afterward (regardless of OS), a plain screen with a "Error 503 : Service Temporarily Unavailable" message is all that results. Thankfully ZB Block seems to be pretty rare, but I have encountered it here, here, and here. It also afflicts the vendor's own site...including this confrontation thread, ironically enough. Has anyone else encountered this?

I wouldn't be surprised if Opera (like Firefox) required some library updates that a fresh Win95 installation wouldn't have. Keeping track of all the DCOMs, Winsock 2s, and whatnot after so many years is easier said than done; though there's a decent list here. As far as I know, 10.10 is the last compatible-out-of-the-box Opera version and 10.63 is the last hex-editable one.

I tried 1 and 2 early on. 3 would be impossible without a time machine. More observations: If I install Opera 10.10 (or 9.64), it works. If I overwrite 10.10 with some 10.63 files...even if none of these are profile files, and none of these have been modified since 2010...I get the error. I think I'll just revert to 10.10. It has a better UI, and it isn't as though 10.63 is much more current these days to be worth the effort.

Another day, another software problem! This one's a bit more significant than the Skype stuff, though. For several years I've been regularly using a lightly hex-edited copy of Opera 10.63 on my Win95 system. Today, it crashed (which honestly wasn't unusual). But then when I tried to run it afterwards, this ridiculous "Opera crashed while trying to show the crash dialogue" box appeared, and the software refused to load: The crash logs themselves invariably begin with the lines "OPERA-CRASHLOG V1 desktop 10.63 3516 windows C:\PROGRAM FILES\OPERA10\OPERA.EXE 3516 caused exception C0000005 at address 00000002 (Base: 400000)," followed by a stack and memory dump. Although Opera has always felt a bit flaky compared to the Mozilla browsers, this is the first time I've ever experienced this behavior. Even more confoundingly, nothing I've tried except for replacing it with version 9.64 has had any effect: I tried wiping the profile as well as uninstalling and reinstalling in a different directory, but the "crashed while trying to show the crash dialogue" dialogue kept popping up afterwards every time! This seems to be a pretty obscure issue: Searching the web brought up nothing except for a few isolated leads caused by people running Opera in circumstances even more unusual than mine (e.g., in Wine on Linux). Would the Opera community forum be of any help, or are pre-XP topics there good as dead these days?

I tried your link, but the executable threw an error that OLEACC.DLL was not found. That library was apparently distributed with the Microsoft Active Accessibility 1.3 add-on for Win95, which MS no longer has available to download. Oh, what fun!

Nowadays the Skype 1.4 installer throws out the error in my first post and refuses to install, just like all the other 1.x versions. I'm surprised you were able to run it on your configuration (at some point in time, anyway!). Maybe the APIs changed between 0.9 and 1.4?

Well anyway...from the info here, I suspected that the 98/2000-only FlashWindowEx function was something that could be hex edited around. "FlashWindowEx" appears nowhere in the Skype 0.90 executable, though. So...I give up.

Hi folks! Sorry about the lapse; I'm back from a rather trying offline week. With that over, here's a brief summary of what's been deduced so far. * No version of Skype was ever promoted as being compatible with Windows 95. * There's a chance that the earliest versions might have been able to work on the OS nevertheless, but finding out is difficult because the application installer "calls home," expects a specific online response, and refuses to run. * The Skype 1.x installer is built with a program called Inno Setup. * Interestingly, a subfolder containing "SkypeVersionChecker.dll" is generated in c:\windows\temp\ each time the installer is run. The DLL doesn't contain a visible URL, though. But, would a pre-1.0 version have worked? As it turns out the 0.90 installer was built with a different setup program, and it installs successfully without "calling home!" That doesn't mean Skype will run, though. This error appears if I try to do that: FlashWindowEx API is presumably a 98-only API. Here's a thread that covers that sort of issue. (I wonder if there's a way to hack around this?) So what about skipping the official client and trying GWConnect? Its installer is as snippy as Skype 1.x's about not installing, since on Windows 95 (and others?) it throws this error in your face: I treat this entire subject, by the way, as an exercise in curiosity. If something that might work doesn't work, or doesn't install that should install...my natural instinct is to find out why. Of course to actually be functional, Skype would have to install, run, and connect...and 3.2 on Windows 2000 could be what's necessary these days to clear all three hurdles.

Yep. That just appears to make the installer hang and require a force quit when it does its automatic checking. Now I'm dying to hack the installer, though. I poked around it with Resource Hacker and a hex editor, but didn't see anything that was an obvious version check.

Does anyone know if a Windows 95-compatible version of Skype exists...or ever existed, for that matter? I do know that the early 1.x versions were compatible with Windows 98. No luck experimenting with those, though. If I so much as run the installer, this charming dialogue is the first thing to appear:

Surprisingly, YouTube seems to be back to working with Flash 7 as of today.

I think the OP brought up a legitimate question; though probably not in the way he intended. What's the benefit of running Windows 9x on a new computer, when it's easier to run it on an older one instead? It isn't as though a Pentium III isn't fast enough.

That's a bummer; something about the frame previews they added must have broken things. It was fun while it lasted.