Drugwash

First you would need to find a user/service manual for your exact model/line of Deskpro since there are many versions bearing this name - I have one myself but it's a Pentium II 400MHz (Compaq Deskpro EN400 SFF). The manual should tell you the supported FSB value(s) and the maximum CPU speed as well as maximum RAM capacity and type. Alternatively, open up the case, carefully remove any dust inside and read the markings and notes on the surface of the board. There may be a dip-switch block or jumpers that set FSB, CPU speed and/or other settings and from that you can get a rough idea on board's capabilities. A better CPU is always preferrable, provided it matches the board specifications. It all depends on the purpose intended for that machine, as older software and games may not require much CPU power compared to (relatively) modern browsers and other applications. In regard to video, a GeForce3 Ti200 (or Ti400) or GeForce4 Ti4200 would work just fine but I don't know if there ever were PCI versions of those cards. Finding a good, working PCI videocard today may be quite hard. Even if the board had an AGP slot, it would probably require a SFF (Small Form Factor) videocard if you have a desktop SFF case and not a tower case. Apart from that you might wanna check for a BIOS update at the manufacture's site, if available, in case the current BIOS version has a HDD capacity limit (many/most Pentium III machines at the time didn't have LBA48 support so HDDs would be limited to 128/137GB). Can't say much about the BIOS (recovery?) partition as I never had a chance to work with one. I recall Compaq had a set of tools for that purpose that might be available at their site - look through all downloads available for your model. But that would only work if the HDD is the original one that has the hidden recovery partition. Win98SE is the last in the 98 series (not counting Millenium). There have been a lot of official updates/bugfixes/hotfixes in time but they were pulled from the MS site long ago and finding those specific to your language (Swedish?) would be quite hard. Since the unofficial packs (Auto-Patcher for Win98 and uSP3) are both exclusively for the English version of the OS, in using any of them you'd risk at least a mixture of languages if not totally bricking the OS. So you have the alternatives of going on with the system as is in case official updates are unavailable or completely wipe the HDD, find an English 98SE CD and install it, followed by any other available updates for the English version.

Yes, that is possible too but something from my hazy memory says it's not always recommended - maybe because the HDD controller may be disabled during the update and certain pending operations may fail. Of course, it totally depends on BIOS type and version and may also be a thing of the past, but it's still a risk. I've done BIOS upgrades from HDD in the past as well as from floppy - can't do it now for a 64bit AMD Gigabyte board since it won't acknowledge the USB stick, HDD is NTFS with an NT-based 64bit OS and there is no floppy drive on that machine.

Certain BIOS update utilities do have command line switches that allow saving current BIOS before updating. Indeed manufacturers provide autoexec files but those can easily be modified to add the save option. Of course, manual operation can be performed just as well. The most important thing is to pay extreme attention to the motherboard version/revision number and any other important details when downloading BIOS update, otherwise flashing the wrong file can brick the board. I did that to one of my best boards (different BIOS image for different RTC chip, got it wrong) so I've learned it the hard way.

Thanks for the addition, Doug! Indeed, many times we do forget to clean the floppy heads. I still have a very old floppy cleaner kit which is nothing but a modified floppy disk with a paper disk inside that has to be moisted with isopropyl alcohol before usage. When such kit is not available, the (handy) user can open up the floppy drive and gently wipe the heads with a piece of paper dipped in isopropyl alcohol, which should be available at the electronic parts shops (such as RadioShack, presumably).

Check/replace the floppy cable, it may be defective or it may pertain to an OEM machine that has the floppy connector mirrored or otherwise nonstandard. Alternatively, make sure the cable is connected properly and you're using the 'A' end (the one with six twisted wires) in case the cable has twin floppy capability. Also double-check that the floppy controller is enabled in BIOS. Certain newer machines can update BIOS from USB but the function must be available in current BIOS. Read the manual for your board version/revision and BIOS version. Enable legacy USB option, if available. The stick should be formatted as FAT32 unless otherwise stated in the manual. Regardless of the BIOS update, AMI remains the worst among all, maybe even worse than certain old Phoenix. It changes boot sequence on the fly after (non-)detecting boot devices so with a flaky HDD drive you may well experience random lockups or boot from a different drive than expected.

If it's an AMI BIOS, you better check boot sequence twice - AMI is the most stupid BIOS logic I've ever encountered, which killed me many hours and brain cells. Alternatively check with Mr. Loew for more advices.

In regard to uSP3, you don't have to install the whole package - only the core components that are mostly MS updates/bugfixes/runtimes. The others are more or less optional components. Since you said that last time the computer came with newer versions of software, most likely the guys at Gateway Ireland had also updated certain system components. Core components in uSP3 may just be what's needed. It's an option to consider, not pushing or anything. Either way, good luck!

Found a version that says 1.1.0.0/2.0.2.0 here. It's 6.74MB, has 9x/2k/XP/XP64/Vista/Vista64 subfolders inside. Download requires completion of a "modern" captcha (one of the most stupid things invented by man) so certain domains may have to be unblocked in browser's addons (for whoever uses Firefox with NoScript, Ghostery, RequestPolicy, GoogleDisconnect and/or similar). No idea if the driver works or not - I don't have anything wireless in the house and not even planning to.

Here you are: KB328310

Very kind of you, thank you very much! EDIT: After downloading and a binary compare, the old and new files are identical. File version says (in both): 2.10.00.03 But that's fine, at least we got alternative links to download from.

Can't remember which version I've installed nor if there were any issues at install time (I hate my failing memory!) but I know it does work. EDIT: Looking in my 'drivers' folder I see only hpspm98.exe, which would be v2.10. As long as it works it doesn't really matter if it's an older version. Maybe it needs specific compatibility mode (such as 98SE) if you're using KernelEx. ISRT, just as a wild guess, could stand for 'InstallShield RunTime' but could also mean something else.

Honestly I don't know - I'm not familiar (anymore) with DOS utilities and commands. The HP package adds the ability to create TCP/IP ports, which Windows 9x natively lack in. The trick of using a printer name such as LPT1 - LPT4 allows a more straightforward way of connecting to the network printer, as the LPT* naming convention is reserved and prioritised. However, if the motherboard has a parallel port built in and is enabled in BIOS, that name cannot be assigned to a network port. Also, other applications may have created similarly-named ports, so you'll have to check and see which name is available within the 1-4 range (since 5-9 don't work - in DOS windows - according to Multibooter's quote above). If you want to pursue the 'net' solution you may try to duplicate the registry entry using LPT2-LPT4 for printer name/path, since LPT1 may be hardcoded for the built-in parallel port. Alternatively disable the parallel port in the BIOS and use that entry for the network printer. Just make sure you backup the registry before modifying anything.

Darn right! But do think of the fact that it all depends on the point of view. From here, we see empty hands - from behind those glasses-like contraption they might see the whole Universe. Problem is, we all get what we see…

USB malfunction may happen for different reasons: - incorrect/defective/mismatched drivers - incorrect/incomplete driver installation - missing required components such as DirectX, VC Runtime (msvcrt, msvcp/msvcr-60/70/71/etc) and/or many others Uninstalling current USB drivers and installing an updated package such as Unofficial SP3.x with NUSB 3.6 may help.

Opens fine here at the time of posting. Try with http instead of https, maybe you have issues with secured connections. Alternatively, the IP/domain may be blocked by some tool you have installed or by your ISP. Try a ping/tracert. Try a proxy.

Try WinWGet.

Have you tried the above? I have a Lexmark E352dn connected through a switch to my home LAN, named it LPT4, assigned it a fixed IP (this is important since the DHCP server would assign random IPs on reboot and 9x machines don't like that), installed the required drivers and it works perfectly.

Maybe wget?

And the laundromat - read 'cloud' - is freakin' broken too. It's gonna be a dirty future for us…

Maybe they can get the old Windows team in as leave replacements? The Windows 98 team, maybe. But if those guys - assuming they're still alive - still have the slightest shadow of a decency, they'd refuse to be associated again with the 21st century M$. FormFiller is right and I've said it too: Windows is now nothing but a Linux, and a bad one at it. And for the life of me, I simply don't understand why both teams must copy from one-another only the worst (un)features, discarding exactly what is required to truly personalize and improve each and every user's experience. Is there something in the air that affects human judgement, or what?

They've been working hard, they all deserve a break. Say, for a thousand years or so. Would that be enough (for the world to get back to normal)…?

I've recently seen the same behavior. However, I almost always use FlashGet, exactly for fear of random stream failures. I've used it for many years, since it used to display that advertising banner. As I mentioned a few times around, I have a metered connection that allows me 5GB download monthly at full speed (that'd be about 512kB/s for a 7.2Mb/s GPRS connection), thereafter dropping to about 15kB/s for the rest of the month. The temporary stream disconnections are somewhat harder to notice when speed is capped, but at full speed there are clear gaps, repeating at a consistent rate. I'd bet on the DASH system, possibly combined with a quick server switch to maintain an equal load. Remember YouTube also uses the DASH system which makes certain players unable to playback mp4 files downloaded from there. Since most servers support the 'resume' feature, these gaps shouldn't be a problem when using a capable download manager such as FlashGet.

Oh that's what they call it? They had to find an appealing name for a trojan horse-like "feature". Who's to say it will be triggered only and exclusively when the device was stolen? Any such "feature" can be abused of. Imagine, say, a M$ employee pi$$ed off by the fact that one user disabled (almost) all spying abilities in SpywareOS 10 and they desperately want that user's data. If they can't get it then nobody should have it, so they trigger the "anti-theft". Bang!, the user looks at a dead screen. He wasn't an arab terrorist, he wasn't Michael Hastings, he wasn't Edward Snowden - he was just an ordinary guy who didn't like third party snooping around, wasting his Internet traffic and using up machine's resources.

Some time ago I read an article about Intel's i5 CPU that had built-in wireless capability and an autonomous internal power source (or just the stand-by 5V power - can't remember) plus a control code that allowed somebody in the know to effectively kill the CPU, allegedly only when the owner would be suspected of something serious. The article marched on the idea that the computer could have been killed even when not powered on. I've no idea what exactly is the truth here but there must be something. And I'm positive many, many "smart" devices have secret control codes that can be input manually (cell phones come to mind first) and/or wireless that perform certain tasks and in case of sensible devices there may very well be this self-destruct capability. Obviously, common people would/should not find out about any of this, unless someone from the inside spills the beans.

What if suddenly none of the newer computers would accept anything other than Win10? They have the power to coerce hardware manufacturers as they did with blocking the development of 9x drivers for new hardware. Want a 'ready for…' logo? Ditch 9x compatibility. Manufacturers obeyed - it's all about money, users - people - don't count in the least. I see here people losing their former XP/Vista/7/8.x/etc licenses if their older computers were OEMs with serials stuck on the side or bottom. They won't be able to revert unless they purchase new licenses. That is, if M$ would still sell old version licenses. I'd bet they wouldn't. A crazy thought just came to mind, probably too crazy: what if they're actually looking for something on people's computers (beyond personal data that they may already have)? All this free OS upgrade plus spyware features plus hardware trade-in leads to the idea that they're desperately looking for something - no idea what. Or I'm just too paranoid. Also, it looks like they want to be in total control of the whole computer world. Which means, if at one point they decide to kill all computers, they may just have the means to do it and nobody would be able to escape, thanks to the most recent "features" implemented in CPUs, chipsets, motherboards and so on, together with the evergrowing wireless network of the IoT. Sudden IT blackout sending us directly back to Stone Age.