jaclaz

Well, of course you will need to copy dsfo.exe from C:\MkPriLog\ to I:\ if you run the command from I:\ (which is fine BTW). From the FC results the sectors seem fine. So the issue seems the one about the volume bootsector being somehow "protected" and allowing not the write of the new bootsector. Try manually, running the Mountvol command. Check which volume has now (on the eSATA machine) the drive letter J:\. It will be something like: \\?\Volume{dcb73171-341c-11e3-b06c-001fc6bb76ce}\ copy and paste it *somewhere* (like a new .txt document). then run: mountvol J:\ /D this will remove the mounting point/drive letter from the volume. Then try writing manually the bootsector, that wii be: dsfi \\.\Volume{dcb73171-341c-11e3-b06c-001fc6bb76ce}\ 0 4096 as512NTFS.bss of course replace the Volume{dcb73171-341c-11e3-b06c-001fc6bb76ce}\ with the actual value you got. Please note how the question mark in \\?\ has been changed into a dot, i.e. \\.\ . Then run again mountvol, as: mountvol J:\ \\?\Volume{dcb73171-341c-11e3-b06c-001fc6bb76ce}\ to reassign the drive letter to the volume. Maybe that is the issue .... jaclaz

Are you sure you really meant "once" (and not "if") ? jaclaz

Maybe useful, maybe not, this should be a simpler batch: @ECHO OFFSETLOCAL ENABLEEXTENSIONSFOR /F "tokens=3,6 delims=: " %%A IN ('Handle ThemeSection') DO (ECHO handleid=%%BECHO Pid=%%AHandle -c %%B -p %%A -y)to the same effect which, provided that Command Extensions are enabled, could also be reduced to a oneliner: @FOR /F "tokens=3,6 delims=: " %%A IN ('Handle ThemeSection') DO Handle -c %%B -p %%A -yUNtested as I don't have a Windows 8/8.x or 10 install available. jaclaz

Well, with all due respect for edborg , in that thread he asked himself the "wrong" question and managed to get a "wrong" answer. What happened there was that he had an extended partition (LBA, type 0F) and that booting to grub4dos a ("wrong") menu.lst entry made that partition entry "hidden", changing it's type to 1F. All he had to do (in that case) would have been to change back the partition type to 0F. Right now I have no idea in which situation is your disk, why don't you run partinfo and provide a screenshot? http://www.msfn.org/board/topic/173016-triple-booting-windows-nt-4-98-and-2000/ http://www.msfn.org/board/topic/173016-triple-booting-windows-nt-4-98-and-2000/?p=1091850 (thread that might have BTW given some relevant info useful BEFORE you got into this ) Or run *any* tool showing the actual partition table layout? jaclaz

It's not that you have to explicit SETLOCAL ENABLEEXTENSIONS ? http://www.robvanderwoude.com/local.php http://www.robvanderwoude.com/ntfor.php jaclaz

Which "error message"? Does it come from the SFX engine or from the batch/command processor? jaclaz

Well, the first issue is then related to Mountvol needing to be run elevated, which we should be able to fix by going from to: But it is well possible that there is either an error in other parts of the batch or that the bootsector earlier created (for whatever reasons) has wrong values or has been deployed to a wrong address though this latter is improbable as the output of the batch is "as expected". What happens if you run again the switcher.cmd? In theory it could be run any number of times and prompt to switch only if "the other" bootsector is found. What happens if you put the whole disk offline and then bring it up online again? It is possible that this additional step is needed to "refresh" the view? Or maybe we are in this condition, where the bootsector is locked by the OS: http://reboot.pro/topic/8200-grubinstexe-write-failed-vista-ntfs-bootsector/ but it shouldn't happen for the bootsector. Or maybe it is better to revert the logic of the batch, right now it attempts to mount the volume assigning to it a driveletter and then deploys the bootsector, possibly it would be smarter/better to make sure that the volume is not mounted, deploy the bootsector and only then attempt to mount it. However, one thing at the time, let's first check if by any chance there is an issue with the bootsectors. Try running: fc /b as4kbNTFS.bss as512NTFS.bss and post the output. As well, get the "current" bootsector from the "J:" drive, that would be: dsfo \\.\J: 0 4096 mycurrent.bss and fc it against either of the two saved bootsectors. jaclaz

Good (I mean bad of course) . We will need to troubleshoot the switcher.cmd. On the USB machine you should be able to assign to it a drive letter manually, however. In the switcher.cmd, edit this: to The issue is likely to be in the previous code that attempts to use WMI and/or REG.EXE .... jaclaz

Yep , these three were the three in three years time I was referring to, but I seem to remember that there were also some slight differences for Windows 2003 drivers, possibly introduced with SP1 or SP2 (but cannot really say for sure), while I am pretty sure that for some reasons (but cannot really remember the specifics) I did keep some Windows 2000 Server installs because there were not available drivers for the 2003 and the 2k one's did not work (and thus went "forward compatibility" of the WDM's). To be fair I believe that - just like UEFI/GPT and their artificial or senseless incompatibilities - the main culprit in this case is more Intel than MS (but of course I may well be wrong). jaclaz

Good. I am still thinking about the *need* of a suitable way to have the mechanism work, now that the issue has been found, in a more "consistent" way or if it is ok, as it seems, to leave the bootsector with the 4Kb value at the end of mkprilog.cmd and convert it to 512 byte value at the start of mkdualdisk.cmd. I could of course make sure about which value (the 512 or the 4 Kb) to write to the MBR by getting it from either bootsectors written in the image, but in order to do so I would need to access (and thus make a temporary copy) of the one or the other, while till now, we make only a temporary copy of the MBR for the checks which is "more elegant" (provided that a half-@§§ed batch can be defined elegant). In theory I could also "unify" the mkprilog.cmd and the mkdualdisk.cmd into a single batch, but I would need to "insert" between the two a set of diskpart commands, to create automagically the second NTFS partition, something that may (or may be not) a good idea, as it would greatly limit the freedom of the user, which may well decide to make instead of the huge NTFS partition spanning the whole disk - say - two more NTFS partitions, one only readable/accessible when on 512 b/s device and one only readable/accessible when on 4096 b/s device. Anyway for the moment let's see if the "end result", as is, works fine. jaclaz

OT , but as often happens not that much , I previously posted my overall not-too-bad impression of use of a no-name, el-cheapo Android based tablet, that a friend of mine bought for around 90 €: http://www.msfn.org/board/topic/155290-windows-8-deeper-impressions/page-235#entry1084400 I found out how similar devices: http://www.forensicfocus.com/Forums/viewtopic/t=12579/ can now be found with Windows 8.x installed for *anything* between 70 and 100 £, i.e. something like 100 to 130 €, and possibly with much better underlying hardware... Since basically the Android OS is free, the cost of hardware should be roughly comparable and there must be some margin of profit both for the seller and for the manufacturer, I wonder how much does MS get for one of these "OEM" licenses. jaclaz

Wasn't the main "switch" in Vista ? http://en.wikipedia.org/wiki/Windows_Driver_Model#Criticism jaclaz

Typo (actually missing operands ) in the proposed edit, this (part of the lines added to mkdualdisk.cmd): Should really be: jaclaz

Which is sad , as it shows how old I am , having lost it some 15 years ago : http://www.911cd.net/forums//index.php?s=&showtopic=11383&view=findpost&p=70389 jaclaz

The BreadBox Ensemble (just like it's predecessor NewDeal Office) is/was last evolution of GEOS/GeoWorks: http://en.wikipedia.org/wiki/GEOS_(16-bit_operating_system) http://toastytech.com/guis/indexgeos.html Of course the original was more Windows 3.x-like, while the New Deal thingy was pretty much 95ish: http://toastytech.com/guis/nd32.html The thingy worked very well AFAICR, had very little hardware requirements (being essentially a DOS shell). If you put it in perspective, or if you ever loaded the Windows 95 first edition through floppies you can understand how at the time it wasn't a bad idea to come out with something providing a good enough GUI on low power machines, remember how when the Windows 95 came out it needed rather "beefy" machines, a "standard" Windows 3.x would run with 4 Mb and a "good, fast system" had double that, 8 Mb typically. Just for the record, I would like to highlight a comment on the given Toastytech site: http://toastytech.com/guis/nd32.html jaclaz

Which is good, as jaclaz could swear that he replied to that reply , most probably a glitch in the matrix (there have been a few issues with the board in the last few days): http://www.msfn.org/board/topic/161919-why-was-msfn-down-yesterday-and-why-is-it-still-slow/?p=1092824 Till now we don't know if your disk is in BUSY state or not. You should try to check with mhdd or hdat1 or victoria to make sure that it is (or it is not). IF the disk is in BUSY state, it won't communicate properly through the serial (hence the need to "force" a "loop exit" by generating a hardware error, i.e. insulating the PCB contacts, the head ones, or the motor ones or both and/or additionally shorting the Read Channel, depending on models and situation). Imagine that the disk is running DOS and an endless loop batch was started. You need to press CTRL+C to stop the execution and get to command line, BUT IF BREAK=OFF was specified: http://www.computerhope.com/breakhlp.htm you cannot anymore, but if you generate a hardware error execution may well be stopped.... There is however a difference between the insulating of one or the other (or both) the contacts and the Read Channel shorting, the first is to access the interface, the second is a way to avoid that a specific command enters another loop of it's own. jaclaz

Yep , that's the idea of note [1] in my post. I know nothing or very little about drivers, mind you, but I have seen too many failed attempts to write proper drivers and too many subsequent revisions/versions of what should normally be a "plain enough" piece of code, and I suspect that - besides the complexity of the NT structure - there is a large amount of misssing or even plainly wrong documentation. jaclaz

Yep , though I did not specified it I had expected something more along the lines of: Condition #1 is OK: ::first partition must be type 0F, start at offset 63 and be 1 sector in length Condition #2 is OK: ::second partition must be type 01 and start at offset 64 Condition #3 AND #4 Fail while either #3 or #4 should have succeeded: ::... and it's start+length should ::be 8 times the start of third partition (if formatted MB aligned) or 8 times the :: start of the third partition - 63 ::if formatted head/cylinder aligned Condition #5 is OK: ::third partition must be type 07 Condition #6 is OK: ::fourth partition must be type 00 Let's see what is the issue in #3/#4, the one that applies in XP is #4 and is tested OK here (but due to the different/virtual environment in which I experimented it may well need to be corrected as well, on further tests, so for the moment we need to check just #3: The partition you created starts Entry478_Start=65536 Immediately after the FAT12 partition Entry462_Start=64 Entry462_Length=65472 sum462=65536 and not, as expected , at 65536/8=8192 The issue lies actually in the "other" batch, mkprilog, due to the limitations of the virtual disk driver I am using to test, I need to create the second NTFS partition with the disk mounted as 512 b/s and run it with the /testonly switch, while you are (obviously) doing everything in 4Kb/s. In a nutshell, we need to find a way to set temporarily the Entry462_Length to 1/8th of what it is now to allow you to create the NTFS partition in Disk Manager at the "right" address. Temporarily, you need to edit mkprilog.cmd, these lines: should become: and add a line to mkdualdisk.cmd here: should become: These two edits should allow you to go further in the creation of the dual mode disk. In the meantime I will think about a way to pass the value wanted needed from mkprilog.cmd to mkdualdisk.cmd in a more "proper" or "sensible" way. jaclaz

Basically you are confirming that the MS choice of removing PAE from XP was - though a perverted one - valid , what happened at the time, more or less was: Since the hardware, but mainly the Third Party's software and particularly drivers suck (and suck big)[1] we remove the feature, limiting it to Server 2003(and ONLY on the Enterprise/Datacenter editions and - but senselessly limited to 8 Gb - on the lesser known "Windows Storage Server 2003") where both hardware and drivers are better checked, i.e. there is more money per unit involved... ... since we are MS we will put the blame on the Third Parties and start senselessly pushing 64 bit computing, creating to both the end users and to the Third Parties involved, even those that were involved in the writing of those sucking drivers, every kind of new problem/issue. This has the not-so trifling advantage that when we will starting pushing for a new (artificially made incompatible) disk partitioning structure and a new (artificially made incompatible) firmware, people won't be shocked by it, as they will by that time be used to all kind of arrogant, senseless or plainly stupid move we will make. And of course, as soon as the issues caused by these two unneeded changes will start to fade away because people will either get used to them or find workarounds, we will make a couple new Operating Systems that will be made artificially incompatible even visually with anything else.... jaclaz [1] ... and maybe the fact that we changed three times in three years the understructure and the driver models, each time failing to provide complete documentation or to provide real-world working sampled more complex than hello-world-type simplified and largely malfunctioning sample drivers is marginally involved in this ...

Is this image for real? :00t: It reminds me somehow of the Psion 3 interface: http://www.guidebookgallery.org/screenshots/sibo3a or of the NewDeal/Breadbox Ensemble "Preferences" one: http://toastytech.com/guis/bbe.html (which the difference that the Psion 3 icons were actually nice and that the Breadbox ones had colours ). So the teacher has deprived the 5 year old kids designing the interface of their colour pastels and they were left with just a dark blue felt tip pen? jaclaz

Good (which means bad in this case) . The "recognition" of the disk happens through a "five point voting system": One or more of the five checks fails, from the output you posted they all seem to verify , try (using Notepad ) a line after each "SET /A PriLog+=1" *like: ECHO Condition #1, Prilog=%Prilog%And, just before the "IF %Prilog% geq 5 (" a couple lines to see the values involved in the check: SET Entry4SET sumSET byjaclaz

You mean you ran the mkPriLog.cmd by double ckicking on it? I guess I'll have to add a final PAUSE to it, since the idea that batches, particularly when experimenting, should ALWAYS, ALWAYS, ALWAYS run from an open command prompt in which you navigate to the directory where the batch is and then you type it's name and hit the [ENTER] key is something that simply doesn't go through as a message, not even getting to an experienced member like you are. And before you ask, the switcher.cmd is an exception to the above general rule and CAN be run by double clicking on it, as it was designed for it. I guess I will have to review the other two batches making sure that they can be run also by double clicking on them having all their functionalities. jaclaz

Yep, I must have changed names while developing the batches , the correct file name is "mkdualdisk.cmd" of course, I'll edit/correct the readme.txt. When you run mkprilog.cmd, at the end it should display: Now , check the drive letter assigned in Disk Manager to the 256 MB FAT (or assign to it one), you will see how that that same drive letter in "My Computer" is assigned to the 31.6 MB FAT partition (the size of the FAT partition comes from the choice you made in mkprilog.cmd, where you are asked to provide a size for the partition between 1 and 32 Mb). The batch create however an image with only one partition, the first one, primary (on the left in Disk Manager) which is a (smallish) FAT(12) one and the second one (on the right in Disk Manager) which is a (largish) NTFS partition that you create manually in DIsk Manager, which is later trasformed by the riunning of mkdualdisk.cmd into a Logical Volume inside Extended so guess to which partition I refer with: I know that the choice of using dark blue for primary partition and of light blue for logical volumes and a tiny light green border around the Extended partition is not the best choice in the world to distinguish something visually, but you should sue MS for this . I cannot refer to the partition as having a given size, as it depends on the choice you make when running mkprilog, and the size of the "other" partition of course depends from the size of the disk and/or from the size of the partition you created manually in Disk Manager, I thought, that order (first), filesystem used (FAT12) and nature of partition (primary) was clear enough, but I am open to any suggestion to make this more clear. The .bss files should have been generated in the same folder where mkdualmode.cmd was run, i.e. C:\MkPriLog\ in your case, and where should be also the dsfi.exe, elevate.exe and Switcher.cmd, maybe I could call this "tool folder" to better make it clear? The file as4kbNTFS.bss and as512NTFS.bss are created by "mkdualdisk.cmd" when it displays: I may add a couple more lines f text so that it becomes: jaclaz

Well, there are hundred of thousands of pages on the MS website, and possibly tens related to or containing guides to build a WinPE form any among Windows Vista, 7 or 8/8.1, if you followed one for making a BIOS bootable WinPE, what you have in your hands may be a BIOS bootable USB stick, which won't normally boot from UEFI. UEFI (stupidly) needs a FAT (FAT12, FAT16 or FAT32) partition/volume and needs it's own EFI bootloader. You may have followed "latest" instructions (related to Windows 8/8.1) and in this case you should have created a USB stick with a single FAT32 partition that should be bootable on BOTH BIOS and UEFI: https://technet.microsoft.com/en-us/library/hh825109.aspx BUT you may need to access the UEFI controls on the Surface and select the .EFI file manually and/or change a few settings, like disabling TPM and secure boot, see: http://www.microsoft.com/surface/en-us/support/warranty-service-and-recovery/how-to-use-the-bios-uefi you can also use this as a reference: https://winfe.files.wordpress.com/2014/10/image-a-surface-pro-using-bootable-uefi-winfe1.pdf which is appropriate, since most probably what you really want to try doing is to use the WinPE (or WinFE) to make an image of the internal disk (if your Surface has a single USB port you will need a powered USB hub, besides a suitable USB hard disk). This is going to be a bit "technical" though. jaclaz

Yes/no Meaning that yes , there is actually an "extra dot after the com" in the posted link , which is a (recent) copy/paste from the generic DeBoyne Pollard's FGA's page: http://homepage.ntlworld.com/jonathan.deboynepollard/FGA/ but no , the link works fine here with or without the extra dot (... like we were talking here of special characters....) I wonder HOW I managed to insert it jaclaz