jaclaz

Good (which means bad) your test seems to lead to an actual file error. I was wondering if you could (via a command line tool) test all files for integrity, 7-zip is not suitable because it is very verbose with the "t" (test) switch. There is a tool (cabextract) that would do nicely, as it just prints the MD5 of the (valid) file but it seems like not supporting wildcards in the names of files, so you would need to make a batch to scan each file one by one. I tested the (yes I know) standalone DOS version that seems to run just fine on XP: https://www.bttr-software.de/ports/ The Windows version on the Author's site: https://www.cabextract.org.uk/ is Cygwin, so right now I cannot test it. jaclaz

Personally I wouldn't use Winiso (or any other of the "advanced/automagic" tools to recreate the .iso, let alone a "writing to .iso tool". "Iso writing" may or may not work, of the two I would prefer rebuilding the whole .iso from "source". But if you are using <some way> to mount the .iso stored on a USB thumb drive, and not actually burning it to a CD/DVD and booting from it, it is possible that the little quirks/artifacts that a writing to .iso tool may create would be irrelevant, i.e. it is possible that you can recreate a .iso which is good enough for your use but that may fail if used to burn a CD. I would be curious to know what happens with yours (corrupted) compressed files when you expand them manually. A "normal" XP should have the command line EXTRACT[1] (in Vista and later should have become EXPAND) program, but also 7-zip should be able to expand the compressed files . While expanding you might have some errors, or it may succeed to create the expanded file, in this latter case (since you have another copy of the original expanded file) it would be interesting if there are differences. It is possible (improbable but possible) that your .iso image corruption is not at file level, but somewhere else, and while you can get the file from the mounted .iso, the setup when it is running somehow fails in retrieving it. jaclaz [1] Sometimes as EXTRAC32.EXE, or you may need to find the "support tools" if you want the MS original: https://msfn.org/board/topic/24670-windows-xp-service-pack-2-support-tools/

Why (the heck)? Did you believe to be helpful, useful, cool, what? If you think it is a "right thing" to do it would be IMHO only decent to state that you were pasting verbatim a ChatGPT answer and - additionally - quote the exact prompt you gave to ChatGPT. jaclaz

Yes, most files in a XP .iso are compressed. The program(s) used to compress them are COMPRESS and MAKECAB (and possibly also CABARC). I would try to use first compress (the version that is in the Windows 2003 Resource Kit which can be found on archive.org): https://ss64.com/nt/compress.html You need of course exactly the same version of the files (both the compressor and the actual file to be compressed) so that the compressed file is exactly the same size of the original. Then it is relatively easy to replace the files (actually the bytes representing the files to be replaced) on the .iso image, it is not straightforward, but on a .iso all files should be contiguous so it should be just a matter of locating the position of the file inside the ,iso image and then overwrite with dd or similar the appropriate extent. It has to be checked but either DMDE or ISOBUSTER should be capable of doing this. Otherwise (if you actually need the .iso file) you will need to rebuild the .iso (using either CDIMAGE or MKISOFS) and this may be more difficult (both utilities have lots of settings that need to be used correctly), but it is of course doable. I don't want to put you down in any way, but if you have some corruption in the files, there may be other corruption in other files that you simply couldn't notice because (due to the errors you had) you did not go far enough in the install, so it might take more time and effort than what you may expect. jaclaz

Neither. It doesn't exist an "alignment sector size". The device will expose either a 512 Byte sector size or (maybe, actually I have not yet seen a SSD that has the larger size ) a 4096 byte sector size. You cannot change how the device is seen. Internally the SSD will have a "page size" that is usually a multiple of 4096 bytes. Normally the "right" alignment for a partition would be 1 Mbyte. (1024*1024)=1,048,576 bytes) In the case of NTFS, the default cluster size is 4096 bytes and in NTFS "everything is a file", so a NTFS filesystem is inherently aligned to the partition start, for FAT (16 or 32) there are special considerations to make the filesystem aligned to a multiple of the device page size (though on a fast device like a SSD normally is, it won't make a noticeable difference). The "normal" (even if "wrong") partition alignment on XP is 63 sectors, no matter if HDD or SSD because that is what the XP disk management expects (and will normally create). You can align the partition to 1 Mbyte (as it happens normally on Vista or later), you won't be able to do that with XP disk manager. you will need a third party tool BUT the first time you will use Disk Management even if for changing another thing (such as the active status of a partition) you will likely lose ALL volumes in Extended partition. See: http://reboot.pro/index.php?&showtopic=9897 So, if you plan to align the partition to 1 Mbyte you'd better have ONLY primary partitions OR never use the XP disk manager. jaclaz

The video at least confirmed that the P7 (or P10) connector is with all 3.3V. As said I personally never heard of a PFC (Power Factor Correction) PSU giving issues, but that doesn't mean much, the catching fire could be true or only a metropolitan legend for all I know. From what I know the PFC is essentially circuitry on the "high voltage" side of the PSU (AC mains) and whether a PSU has (or has it not) is (should be) irrelevant on the low (DC) voltage side, the whole idea is to transform an reactive load on the grid into a (almost) resistive one, thus old dansdata article explains it: http://www.dansdata.com/gz028.htm jaclaz

Never heard of non-original PSU's catching fire or exploding though (surely the old AT ones but I seem to remember also early ATX ones) some old ones didn't "like" to be powered on without a load (in the good ol' times we used a tail light bulb 21/5 W connected to the 12V and to the 5V rails to give it a minimal load when testing). In those cases the PSU would "whine" loudly and if you didn't power it off/disconnect from mains quickly it would let out the magic smoke. Again, with Dell everything is possible, but if there are (besides the el-cheapo ones) specific adapters by reputable suppliers for the specific Dimension models, such as the one I posted a link earlier: https://www.atxpowersupplies.com/Dell-P10-connector-adapter.php it should mean that the reports you found were for other models. jaclaz

The limit is in the partitioning not in the NTFS. The field (in MBR style partitioning LBA addressing) is 32 bits, so you cannot have more than 2^32 sectors (-1) i.e. 4,294,967,296-1=4,294,967,295 The "usual" or "traditional" size of hard disk sectors being 512 bytes, you can access at most 4,294,967,295x512=2,199,023,255,040 which is the 2 (or 2.2) TB limit . BUT some new hard disks are "native 4K" (i.e. they have sectors sized 8x512=4096 bytes) or some external disks are mapped (in the USB bridge) to look as if they had 4 KB sectors. Suddenly you can address with MBR 4,294,967,295x4096=17,592,186,040,320 bytes. jaclaz

Interesting vulnerability. (CVE-2019-8286). No idea if it injects "malicious java code", but the specific vulnerability is essentially: "IF the user accesses (possibly through a phishing link) a specially crafted website, it is possible to retrieve the Kaspersky Product ID" CVSS v3.0 score of 4.3 is actually a medium value (on the low side): https://en.wikipedia.org/wiki/Common_Vulnerability_Scoring_System really nothing to worry about, I would expect the reasons why the product was banned by US, Germany and more[1] to be much more serious than hypothetically revealing a product ID (and thus - still in theory - allowing some sort of de-anonimization). jaclaz [1] BTW there are different levels of these, some countries have "only" banned its use on government computers, some have also issued warnings about it to private companies and citizens.

Well, you provided a link to a page on which there is a "fact-checked" stamp, an "official fact checked" certification doesn't exist (AFAIK) anywhere, very likely the good people @ cybernews.com are honest and accurate, but it is not like that green checkmark has any actual meaning, the Authors and the Fact Checkers on that site are interchangeable[1], even if there is no such a stamp on zdnet or (say) BBC that doesn't mean that the info in them is automatically worse or less accurate. Only for the record, cybernews.com is actually a rather mysterious site, it came out of nowhere only a few years ago, it is based in Lithuania, they have rapidly gained international popularity (and increased income): https://rekvizitai.vz.lt/en/company/adtech_lt/ but besides the generic/boilerplate "How good we are" it seems like a normal (BTW successful) news site, ads and/or links sponsored. The same company has also https://healthnews.com/ and https://cooltechzone.com/ this latter seemingly dedicated to VPN's.(which are also often the theme of articles on cybernews.com). jaclaz [1] i.e. the same person once is the Author and on another article he/she is stated a the fact-checker and viceversa

No, I am not putting words in anyone's mouth. I asked (I believe politely/amicably) for a reason for this change of mind as it seemed unrelated to a cited event that happened more than one year ago. Then, since someone started nitpicking, accusing me of (non existing) provocations/accusations, I tried (vainly) to explain how that was how I understood (or misunderstood) your post. Now you accuse me of putting words in other people's mouths. Rest assured it is not the case. Still, I am both a comedian and and a clairvoyant, so if I actually wanted to put words in someone else's mouth, I would have been allowed to do that and also aspire to some success with it. Anyway, I see that in this new MSFN it is not possible anymore to have a normal, plain, conversation, so I'll declare this thread an SEP and move somewhere else. jaclaz

No it is not a provocation, nor an accusation, nor anything else, I simply listed the main things forbidden by MSFN Rules, and couldn't see which one of them has even a feeble connection to the possible reason why Astroskipper changed his mind. The political/religious debate (Rule #2.b) is prohibited, but there is an exception carved out explicitly for technology topics, and a simple statement, like the one Astroskipper just posted, is anyway more than enough, there is nothing to debate about it. What perplexed me in the original announcement was the apparent non sequitur, as I read it, essentially: "The national authority of my country issued a warning in March 2022, I didn't believe it for some 15 months, today I changed my mind." So I asked if there was some relevant news that triggered this re-evaluation. jaclaz

Re-checking, the link I posted: https://www.atxpowersupplies.com/Dell-P10-connector-adapter.php and other pictures of the specific for Dell adapter: https://www.ebay.com/itm/235000267788 seem to have ALL the same colour and thus be ALL 3.3V (besides some re-mapping of the 20 pin connector), on the same page: https://www.atxpowersupplies.com/Dell-P10-connector-adapter.php there is a warning/foot note that this is NOT the "standard" 6 pin AUX connector (which is what you have since definitely the one in your photos has orange-orange-red). IF the original one you have has three wires white-blue those should be all at the same voltage (3.3V) and it would be compatible with this pinout info (which is NOT the "standard" aux one): https://old.pinouts.ru/Power/dell_atxaux_pinout.shtml So DO NOT use the adapter you have "as is", you need to either re-map it or procure yourself a "proper" one, jaclaz

No intention to speak for others, I was clearly expressing my personal opinion (I don't think ...). Anyway, I cannot understand which of our Forum Rules may prevent you from explaining what made you change your mind (if you actually want to share it, of course), warez?, political or religious debate?, racism, pornography, threats, profanity, or excessive vulgarity?, flaming other members? No problem however, I can cope with this question remaining unanswered. jaclaz

Sure, you are perfectly free to change your mind, but I don't think anyone is interested in knowing that you changed your mind, whilst maybe someone may be interested in the reasons why. The German warning is some 15 months ago, roughly at the same time the US FCC issues a similar one: https://www.fcc.gov/supplychain/coveredlist (the US banned Kaspersky use in government computers much earlier, 2017 if I recall correctly). So, unless you overlooked those signals, what made you change your mind now? jaclaz

I see, those are not the (more common) two rows Molex type connectors, they are the linear 6 pin ones that some (old) ATX power supplies had., these ones: https://allpinouts.org/pinouts/connectors/power_supply/atx-aux-power/ https://pinoutguide.com/Power/atxaux_pinout.shtml https://www.informit.com/articles/article.aspx?p=31105&seqNum=4 The colour coding of the adapter seems right, black for 0/COM, Orange for 3.3V and Red for 5V. The notch(es) on both the original and the adapter seem "wrong" compared to the reference above, but yes, they can normally be removed with a cutter. The colour coding of the (original) connector seems (even if different) coherent, black for 0/COM, Blue for 3.3V and white/blue for 5 V, but it would be nice to find a reference DELL pinout to make sure. Anyway this adapter is mapped similar to yours and it is explicitly mentioned as being compatible with Dimension 4100: https://www.atxpowersupplies.com/Dell-P10-connector-adapter.php what is "strange is that it is called "P10" whilst your original one is marked "P7" (but with DELL anything non-standard or confusing is to be considered "normal"). It is likely given the blue and white/blue cables (but remember it is a DELL) that your particular machine has an auxiliary connector which is compatible, some (earlier?) models had only 3.3V on that connector. Also I believe you need to check also the actual 20 Pin connector, some DELL's had something different in them, see: https://www.vogons.org/viewtopic.php?f=46&t=59959 jaclaz

Generally speaking, connectors like those can be re-mapped, using two thin steel nails or a couple of small "spectacles" screwdrivers it is possible to bend a little the little wings that hold the single connector inside the plastic one, then the brass terminal can be re-inserted in the "right" position, as well with a small cutter it is possible to carve a little the plastic of the plugs to make them "round" and fit in a "square" receptacle. The real issue is to have a reliable source for the wiring/pinout. Maybe, it is this one: https://www.dell.com/community/XPS-Desktops/XPS-8940-ATX-6-pinout/td-p/7779556 the colour of wires on the original might help. jaclaz

Is anything on this thread really needed? I can only see senselessly whining and bickering. jaclaz

BUT it needs to run on 9x/Me? I am asking because - generally speaking - this kind of software needs lots of memory/RAM, otherwise it will often swap to hard disk, slowing down each and every operation, and the typical machine running 9x/Me will have some memory limits and relatively slow hard disk(s). jaclaz

So, what you really need is a way to boot the PE at a time when the normal "system" disk is connected. Essentially you are saying that when you hot-connect the disk to the PE, it cannot see it, it may be a limitation of the WinPE (of all WinPE's or just your particular build) or it may be something connected to one of your BIOS/UEFI settings. Usually there are settings in BIOS (or UEFI) that allow booting from an external disk (USB) with the internal disk(s) normally connected. Which motherboard is it? Are you booting BIOS (or CSM) or UEFI? It would be really strange that you don't have such an option, in case there really isn't one there is a possible workaround adding an entry for a bootloader to your existing BCD, but depending on your PC booting BIOS or UEFI and other factors it may be complex. Before anything else, have you tried re-scanning with diskpart or devcon [1] after having connected the disk? See here: https://answers.microsoft.com/en-us/insider/forum/all/windows-10-does-not-detect-a-hotswap-hard-drive/4bab657c-8c62-4092-99e1-ef7c83466cc0?page=1 jaclaz [1] diskpart is surely in your WinPE, devcon may or may not be there

Unfortunately you cannot create Amperes without using Watts, if we go for: 30A@3.3V 30A@5V 40A@12V those imply 30*3.3+30*5+40*12=99+150+480=729 Watts, so a 750 W would be more suitable, but newer 750W ones generally speaking don't have as well more juice on the 3.3V and 5V than the mentioned 20-25A, so if the requirements are confirmed, OP has no other option than finding an used PSU , it seems like the extra Wattage on recent PSU's goes all on the 12 V rail. See also here: https://www.vogons.org/viewtopic.php?t=34076 jaclaz

The 98lite and the other projects on litepc.com are AFAIK not much more than a hobby project, the Author is a biologist/ecologist consultant with an academic background, I don't think he is getting much from the sales: https://www.washingtonpost.com/archive/business/1999/05/29/researcher-cuts-the-link-to-explorer/ec3e78d4-3078-4734-b962-9bc17d4be3c4/ https://brooks.eco/profile jaclaz

I see, then, I don't think that any of the mentioned tools may be useful. You need/want something more like an integrate/slipstream tool. Try looking around (on msfn) for topics related to msbatch.inf and/or infinst.exe and for "unattended installs", I cannot remember a tool for doing that, only some (complex) manual methods, though. There is a (relatively) new kid on the block, however (that may or may not be useful to your goal): https://www.vogons.org/viewtopic.php?t=69068 https://github.com/rishooty/Q9xS jaclaz

A current 600 or 650W PSU should have *around" (and roughly): 20-25A @3.3V 20-25A @5V 40-50A@12V older PSU's (current at the time of that motherboard) were - if I recall correctly - a little beefier on the 3.3V and 5V lines and a little feebler on the 12V one. jaclaz

Without making an exact calculation (but you can find several different PSU calculators online), that should be something that could run off a 450 or maybe 500 W PSU, here is an old test made with a 460 W PSU, JFYI: https://www.phoronix.com/review/245/6 Personally I would look for something in the 600 to 650 W range, power supplies tend to work better if they are a bit oversized. That Chieftec model you mention appears to be a rather expensive PSU (used). Again personally I wouldn't go for an used PSU, in my (little) experience PSU tend to fail for no apparent reason[1] and I never saw a real difference in their life correlated to their cost (and supposed quality/brand name), so, while a "no-name" one might be risky, many cheap, known brands would do. jaclaz [1] I suspect it all boils down to hours powered on + number of switch on/off + years of age, the only thing you can (often, not always) check on a used PSU is the manufacture date, but you have no way to know if it was inside a machine working 24/7 for years or a machine powered once weekly for a couple hours to check e-mail.