jaclaz

I am also not convinced that the errors you got come from the added SATA card, the riser or some other combination of hardware /drivers. They seem to me too "specific" to be attributed to hardware issues. I have seen similar reports by people that were accessing NTFS with this (or that) version of filesystem drivers from Linux, related (directly or indirectly) to short name generation, but XP and 10 should not produce errors for this reason, other similar errors have been reported involving "NTFS resident" data, files smaller than around 780 bytes, but your files are "normal" ones. The paths you have shared are long, but not long enough to hit this (or that) limitation (that I know of), 102 characters for the path + an even longish filename won't ever touch the 255 character limit. And I don't think that the Paragon GPT driver is involved either, it should work at a "lower" level giving access to the volume, so from then on it is (NTFS) filesystem drivers that are in control. On the other hand the backup tool you have, if it uses some kind of caching/flow control or similar may (evidently in extremely rare cases) produce some kind of race condition that leads to the errors. BUT it is still not too convincing as a thesis, as - once the mis-writes of the relevant bytes happened - the error should be detected by BOTH the Windows 10 AND the XP CHKDSK. So, if it is detected on only one of the two OS, and this behaviour is consistent, it should mean that XP *somehow* writes differently from Windows 10 (and Windows 10 doesn't *like* the way XP writes in NTFS) which makes very little sense, as we would have seen tens, hundreds or thousands of similar error reports. I am really perplexed.

Well, what is new is also usually old, or at least new things are not necessarily safer than old ones, unexpectedly face recognition just got a hit (in Thailand and Vietnam) JFYI: https://therecord.media/hackers-target-southeast-asia-bank-accounts-trojans-facial-recognition jaclaz

But this is AudioPhysicalFingersFingerprint, it is an entirely new field. jaclaz

The paper is here: https://www.ndss-symposium.org/wp-content/uploads/2024-618-paper.pdf Of course everyone will start swiping with knuckles, scared by the possibility (actually extremely remote) that something like what is described in the paper actually works in the real world. A much more dangerous attack, with almost 100% success : https://www.dailymail.co.uk/news/article-4069090/Mommy-shopping-Six-year-old-Arkansas-girl-used-sleeping-mom-s-fingerprints-unlock-iPhone-buy-13-Pokemon-gifts-Amazon-total-250.html jaclaz

Good now reboot to XP and run CHKDSK on it. Post results. Then reboot to 10 and run it again on 10. Post results. jaclaz

Oww, comeon, I was joking, as you well know it is the approach used on italian television to vote at (say) Ballando con le stelle or (still say) the Festival di Sanremo, but also at the Primarie del PD, Possible? Yes. Probable? No. jaclaz

Well, the idea was to run CHKDSK without parameters (it doesn't mean "with the /f parameter"). Even without the /F parameter the CHKDSK (which in theory, without parameters is running in "read only" mode) does make some corrections BUT (mostly) only lists the issues and asks to be re-run with the /F parameter. Running it without parameters on one OS, then re-rerunning it without parameters on the other should give an idea about these issues being "real" or "artifacts" of the one (or the other) CHKDSK/NTFS version. If you run with the /F parameter and it fixes the issue, the issue is not there anymore, the other OS should not detect it, and when you go back to the first it should not be able to find it (the same one) anymore (as it was fixed). If each OS fixes the issues (differently) you risk to enter an endless flip-flop. I would expect that the SAME issue (uncorrected) would be detected on both OS. jaclaz

@Dave-H I don't know, the errors your CHKDSK reported are too "narrow" to be (IMHO) attributed to "generic" connection/transmission issues, they are all about cross-linked files, they sound more like some issues at filesystem level, if it was a cable (let's say the riser) picking up some interferences you would probably have had "different" errors, a cross-linked file, a wrong entry in $MFT, an actually corrupted file, a truncated file, etc. jaclaz

Sure: https://www.imdb.com/title/tt0079470/quotes/?item=qt0471984&ref_=ext_shr_lnk jaclaz

No. But that looks very good on your curriculum vitae, should you look for a job. jaclaz

Who tells you I didn't gave some thought about it? It's easy, money . 1 US dollar (or 1 Euro or 1 British Pound, we will simplify using a 1:1 fixed rate of exchange) of donation to found MSFN.ORG expenses every 5 likes. You buy in advance a carnet of 10 likes, (which will be called £ik€$), that you can "spend" on other members posts. Essentially, put your money where your mouth is. I would guess that creating voting rings (if any exist) will be discouraged. You see, I am actually a genius in marketing (and also modest, not everywhere, but in spots) jaclaz.

@Andalu In your case, whatever issue(s) CHKDSK detected (and corrected when running with the /F parameter) seems to be "trifling" (they are not even listed, unlike the ones Dave-H got). They appear (at first sight) more like something (minor) different between XP and 10 versions of NTFS, in the case of Dave-H, which seem all related to cross-linked files/clusters. There is a "strange" KB/doc by Paragon: https://kb.paragon-software.com/article/1004 that essentially states that XP CHKDSK is more "rigid" than later OS version, maybe this is part of the issue. jaclaz

Yes, make a normal GPT style disk. The first partition (within the 2.2 TB) will be seen by XP AND it can be used by both Windows XP AND Windows 10. The second partition (beyond the 2.2 TB) will be seen by XP as well, only don't use it for writing when under XP, or you could decide (under XP only) to remove the drive letter assigned to it. Though whether it will make any difference with NTFS/$MFT errors is to be seen. jaclaz

Personally I think we should actually get rid of the "likes" (as opposed to the addition of "dislikes"). Likes are senseless metrics, they are about consensus (by the masses), not value in absolute terms. More specifically, I think you guys should be taking some long walks outside instead of insisting on these themes, this stuff is getting old. jaclaz

I don't like that CHKDSK report. Besides and before the actual errors, WHAT (the heck) is using the volume? Personally I would still have a first partition within the first 2.2 TB (and use ONLY that one in XP), the 32-bit limit is inherent in the "storing" of the partition data in the MBR style of partitioning but we don't know if any "limited to 32 bit LBA addressing" math operation is in any of the drivers or tools. The experiment to do is anyway: 1) repeat the backup/whatever operation that triggered the behaviour (in Windows 10) 2) unmount/disconnect and re-mount/re-connect the disk under Windows 10 3) run a CHKDSK manually under WIndows 10 (without any parameter) 4) if there are NO issues, reboot to Windows XP 5) if the CHKDSK is not automatically started, run it manually (again without parameters) jaclaz

That adapter is entirely passive, there is no difference between sata and esata, only the shape of the connector. No idea how that card (and its driver) works, but at least in theory it is perfectly possible that it can distinguish between the sata and esata ports, if they are all treated the same, then of course there is no reason to prefer the esata to the sata ports. About the Blu-Ray, and the hard drive caddy, it seems to me that your plan to swap the connections is fine. jaclaz

Happy to know that it (overall) works. I don't think that it may make a difference , but have you tried connecting the device(s) temporarily to the actual esata ports on the card? It is possible that that checkbox only affects the esata ports but I doubt it . Adapters from esata to sata actually exist, though they cost an awful lot of money (compared to the riser): https://www.testsieger.de/sata-kabel/2CAZ10CRHK9T4-inline-27502-esata-stecker-an-sata-buchse-adapter.html If you are into testing and wiling to do crazy experiments, you could try one of the "removable to fixed disk" filter drivers (cfadisk.sys, dummydisk.sys or diskmod.sys, this latter is the most recent and more featured one). Edit: Also, maybe: https://fightingforalostcause.net/1044 (if you want to only remove the DVD/Blue-Ray), you can also remove the whole stuff (since you are running Hot-Swap): https://msfn.org/board/topic/34987-safely-remove-hardware-iconthat-everyone-hate/ https://msfn.org/board/topic/176084-trouble-with-stobjectdll-maybe-newdevdll/ jaclaz

From what I understand it needs to be connected to power. There are three types, as seen: 1) flat cable (that has all wires) 2) flat cable (that has all wires) with additional power connector (for power hungry devices) 3) Usb 3.0 extender (the one you have) that need external power as the USB cables has only nine wires, according to the already posted article, PCI uses 14 pins for data (but 6 of them are ground), so 14-6=8+1=9 there are no wires available for power.

Someone attempted to register on MSFN but had issues, he reported the problem on reboot.pro: http://reboot.pro/index.php?showtopic=22827 Maybe something can be done to solve the problem? Tripredacus has an account on reboot.pro, if any other Mod/Admin needs to contact the user I can try to serve as "communication officer". jaclaz

@JFX There is an user of your nice WINNTSETUP that for *some reasons* (I will also post in site and forum issues) cannot register here on MSFN. He posted on reboot.pro, here: http://reboot.pro/index.php?showtopic=22827 what he asks for seems to me reasonable, maybe you can suggest a way (or officially exclude there is one) to directly install on a VHD. jaclaz

That link goes to the whole directory of available Bluetooth dongles (880 results), WHICH one are you interested in? jaclaz

The slack time in the GANNT chart for that project has been exceeded, we will have to either shorten the time dedicated to testing or delay the completion date. You (the British) often fail to appreciate the usefulness of Latin time. Project estimated duration: 5 days Translated to Latin time: 10 weeks (double the amount and pass to the next larger unit) We are all well within the deadline. jaclaz

There is no script? If a formatting attribute is *somehow* kept, and the hypothetical script is a batch script, the CMD interpreter will remove the attribute. Create a (say) new MS Word document, copy and paste in it: ECHO boot.ini here the BOOT is italics bold, the dot is red, the INI is underlined red. When you paste it in a "rich" text editor these characteristics will be kept. Now open a command prompt and paste again in it, and press ENTER. The CMD strips the attributes (actually it is the clipboard that pastes only the text, I believe) The script, on the other hand needs to be a plain text file, with extension either .BAT (wrong) or .CMD (right) other formats (i.e. those that can keep the formatting) won't be parsed correctly an throw an error jaclaz

Sure, that would be: boot[ color = #000000].[ / color]ini

What jumper posted is actually something like (without the spaces) this: b o o t [ i ] . [ / i ] i n i the part inside the square brackets is what tells the board software that the dot between the two sets of square brackets is to be rendered as italic. Just like you do not see the square brackets, when you copy and paste you do not have them. In any case that is just a string and you cannot execute a string, the file with that name is a plain text configuration file and anyway it is never executed, the lines inside that file are parsed and then IF valid, used as parameters for the use of NTLDR (and the parser is quite strict). A file name is NEVER executed, and even IF it was a file, a .ini file is NOT executed. In DOS/WINDOWS the extension (what is after the dot in file names) is a way to identify a file type and take actions depending on it. jaclaz