LLXX

I don't think ntdetect.com and ntldr take up much time of the total boot process anyway. Replacing the kernel with a 2k3 one, however, might make a difference (if it could be done... somehow)

Is there a console version of the accounts package that reads from standard input? If so, then progname < filename.txtfrom the command prompt would be all that's needed. For a Windows program... a scripting solution like AutoIt might work. That problem is that there is no concept of "standard input" in Windows.

Clear out everything in \windows\system32\dllcache and \windows\system32\prefetch, reboot to command prompt, extract explorer.exe from your XP CD and reboot. That might solve the problem.

Interesting. There are no Rootkits present on your machine, but something is suspicious with "peer guardian" would be firewall? I'm not familiar with it, but is this indicative of your machine attempting to connect to a remote site, or is another machine attempting to connect to yours? The former is very keylogger-like, the latter can be ignored so long as it's not causing any noticeable problems.I don't believe your system is so massively infected as to warrant a reformat and reinstall. If one process on your machine is initiating these connections, preventing it from starting and removing its file should be sufficient.

There is a SMART utility (I can't remember the name of it...) that lets you view more detailed information about the drive. The drive might not have failed yet, but has been detected to be very soon.

That is one huge heatsink... make sure it's mounted securely!

I don't think a new sound card would solve your problem, as you indicate... ...and... It looks as if there is a significant source of electromagnetic inteference that is inducing small voltages in the speaker cable. This small voltage gets amplified by the speaker amplifier, so that essentially the speakers become a low-frequency radio receiver.Find out what's causing that interference. There's nothing wrong with your hardware. Noisy volume control potentiometer, it's common... Don't worry about that unless you're changing the volume all the time.

No need to keep this thread... Link21 you've already created one huge thread in the 9x forum... and the results of the poll clearly indicate

You can't have two primary partitions, there is one primary then one extended partition with logical drives inside it...

It's identifying it as a TAR archive which it is not, that's why it thinks it's corrupt. It's in Microsoft compression format.

This has become quite a thread! Over 10000 views! I think this thread is actually beneficial, it's getting normally XP and 2K users to try 98se to see if the "disadvantages" and "inherently insecure" features that Link21 is claiming will manifest themselves, as a few users so far have posted... Keep this thread open!

Get her to use 98se. It's nearly the same, but has better memory management and stability. It's also much more compatible with hardware/software than 95. There are also all the Unofficial Service Packs and fixes that can be found around these forums here

I'd say the minimum is 768Mb and max is 2048Mb. On my XP Pro machine I've set it to 1024/1024 and it works fine. Fixed size pagefile is supposedly faster as it does not have to resize it and fragment it in the process.

The Rootkit Revealer results seem a bit suspicious... are you running it with other programs open at the same time? Close all open windows and then run it. Otherwise it'll give false results.

Actually, it's only the write cycles that are harmful. Read lifespan is essentially infinite, but with every erase and rewrite, the memory degrades slightly. By making it copy all the files to a ramdrive and then running from ramdrive, it is possible to prolong the life of the drive (and increase the speed greatly). I don't know if this is possible for XP, but I am aware that it can be done for DOS and 98se.If the BIOS is able to mount USB storage devices as emulated hard disks and completely hide the fact that it's a USB device, booting should be possible. Otherwise XP will interfere as it sees a new "USB storage device" attached and try to initialise it. That is most likely the source of the error message.

The stock cooler and thermal compound that come with Intel CPUs perform quite well; in most cases you can overclock them quite a bit with only the stock cooler. AMD's stock coolers seem to be less performant so their thermal compound is likely to be as well. Getting a better cooler for it wouldn't be such a bad idea. However, when you remove the cooler for cleaning, then Arctic Silver is preferred thermal compound to use on reassembly.

LiteOn are a good choice in my opinion. I have two of their burners and so far both of them have worked flawlessly for over a year.

If you read HTTP protocol document (RFC2616) you find that the "206 Partial Content" response which is how it is resumed (by requesting start and end portions of the content) is subject to availability per server. If the server doesn't support partial content ranges, you can't resume downloading partially.

Now that I see more users experiencing problems with large hard disks I will attempt to accelerate progress on a project I have been working on, to completely rewrite the ESDI_506.PDR to work with large disks properly. I will release it to the public when I'm finished. The problem becomes more and more urgent as disks become bigger and bigger...

1. I don't need more than one processor. The 4.17GHz Pentium IV (overclocked) I'm using right now is much more than "very sufficient" to run the OS and the programs I have.2. Nor do I need one CPU that pretends to be two. 3. I have successfully run 2Gb of memory. I am still awaiting a third 1Gb stick so I can test 3Gb. Not that I need that much anyway, since 512Mb is currently what I'm using and it's perfect. 4. Adds extra bloat to the filesystem and is unnecessary for a single-user machine. 5. Security is unimportant for a single-user machine. 6. See #4 and #5. If you want multiuser, run a Client e.g. Novell Netware and a dedicated Server. That is *designed* for multiuser. The NT series were originally designed for *server* applications. That is why they have much unnecessary features that really do not apply to the context of a client. They are essentially server OSs. Explain how I can boot an NT kernel from a floppy and use a command prompt, then run WIN.COM to load the GUI from the hard drive, when the NTLDR itself is larger than the entire DOS 7.10 kernel You do not know what you're talking about. I've been writing system-level code ever since the mid 1980s. I've read most of the series of Intel x86 and IA-32 Technical References. I've memorised most of the DOS kernel21 calls. I know PC hardware and software much more than many users here. DOS was a good OS and it still is. Thousands of embedded 86+ systems still use it due to its versatility and simplicity combined with low system requirements. It's been around for a long time, so it's become a very mature OS. Just compare the number of "service packs" and "patches" that were released in the DOS times (0) with those of Windows (hundreds). DOS was developed nearly perfectly before its release. When it was released, it was already nearly completely free of any bugs. Now, especially with 2K and XP, Microsoft coding standards have dropped horribly. They have to "correct" bugs in their code that should've been corrected before it was even released.DOS *can* do multitasking with appropriate software, e.g. Desqview, DOS SHELL SWAPPER, and DOS+MTX. DOS 7.10 is actually a 32-bit OS. You don't believe? Just try running it on a 286 - it won't work. It needs at least a 386. Same goes for Windows. If it was not 32-bit it would've run on a 286 (16-bit PM) and had a 16M physical memory limitation (24-bit address). In fact, Windows 3.11 was already 32-bit, else how can you explain why it supports up to 256MB of RAM? The 9x systems are certainly built on the DOS kernal. It's a 32-bit protected-mode GUI on top of a 32-bit real-mode kernel. That is what blesses 9x systems with their generous backward compatibility, flexibility, and versatility. Not exactly. In 9x systems, the "kernel" proper is the Virtual Machine Manager (VMM), while "system" DLLs such as kernel32.dll are actually at ring3. Being able to read/write to kernel32.dll's memory space is actually reading/writing to usermode. The VMM32's own memory space is protected from writing but readable - if you attempt to write to it without setting the correct page access rights, the "Access Violation" error is trapped by the OS. Also, being able to read/write to the "user-kernel" is a virtue of flexibility, as it enables extensions and patches to be easily applied to the kernel.98se does not need security updates, not even a virgin install. A virgin install of XP is much more vulnerable to remote exploitation. A virgin install cannot be exploited remotely because it simply does not have the in-built network services that NT series have. If you install a trojan or virus, that's error of the user, not the OS. Remember the WMF vulnerability a few weeks back? All the NT-series were vulnerable to the exploit simply by visiting a web page, while all of us that used 98se stayed completely immune. On my system, I even downloaded an infected WMF file from a website that was listed as being dangerous and prohibited from visiting. Attempting to open it had no effect at all, whereas if I was on 2K or XP the mere act of visiting the site would've triggered an automatic execution even with a secured Internet Explorer. What's more "inherently insecure"? That is a rhetorical question.

But if you use Intel CPU, you don't notice any difference, because the microcode is already updated by the BIOS and Windows doesn't update it.

Either boot into command prompt and do it there, or use the Task Manager to kill all the explorer.exe process, run cmd.exe (get the command prompt) and do it there, then restart explorer. You can't replace it while it's in use (i.e. your GUI is operating). ...and they're supposed to be back\slashes. expand f:\i386\explorer.ex_ c:\windows\explorer.exe

Always check the source code of the page when you see something wrong... this is obvious isn't it? FireFox doesn't seem to like https:// links. If you see in the FireFox, all the links that are displayed have normal http:// in front of them, but all the rest https:// are missing. That is the cause of the problem, I don't know what the solution is because I don't use FireFox.

There might be a rootkit present as well, which is hiding some registry keys and startup entries. Get the Rootkit Revealer (also from Sysinternals) to check.

The thermal limiter on most Intel CPUs activates around 68-70°C. If you're idling at 60°, then it will definitely reach the limit when loaded. But, with good cooling, if you can get the temperatures to stay down in the low 50s under load, it should be possible to overclock a bit.However, if you overclocked it to e.g. 3.6GHz and it works, but is now running closer to the thermal limiter setpoint, under load it may go down to 1.8GHz or 900MHz which is far worse than the 3.2GHz you started with. I can tell this from experience. I have a 2.4GHz Celeron that I overclocked to 3.7 - it was able to boot into Windows and otherwise seemed to work, but under benchmarking it performed equivalent to 850MHz without crashing or freezing, and the temperature remained steady at 68°C. This was because the thermal limiter was activated and in fact the CPU wasn't running at 3.7GHz, it was at 850MHz.