jaclaz

On which drive/partition you have NTLDR, NTDETECT.COM and BOOT.INI? Check this site: http://www.multibooters.co.uk/ http://www.multibooters.co.uk/multiboot.html jaclaz

NO. (meaning that it's not possible, not that noone can tell you this ) A "shell" program? A bootloader. There are several ones, typically: Microsoft own NTLDR Syslinux Grub4dos My personal suggestion is using grub4dos, as it is one of the most flexible solutions. Start from here: http://www.911cd.net/forums//index.php?sho...c=18846&hl= Use this if you are using a USB stick or you don't know how to prepare a HD: http://www.911cd.net/forums//index.php?sho...c=21702&hl= Read the guide: http://diddy.boot-land.net/grub4dos/Grub4dos.htm Browse the forum: http://www.boot-land.net/forums/index.php?showforum=66 jaclaz

A FOR /F should be all that is needed: http://www.robvanderwoude.com/ntfortokens.php Can you post the output of a "single" working "ping" and those with the "ping errors" that you want the alarm (redirection or whatever) to be triggered with? Another VBS script that might be useful: http://www.robvanderwoude.com/vbstech_network_ping.php jaclaz

Yep, right now the migrate.inf you produced with mkmigrateinf2.cmd is that of your Hard disk (normally C:\). I'll check the output files of MakeMigrateInf.exe, anyway. The only other (similar but not exactly the same) problem related with not working migrate.inf I can remember is on this thread: http://www.911cd.net/forums//index.php?showtopic=21682 Though I don't think you are in ANY of the situations that caused the problem, have a look at it, and see if you can spot any similarities with your setup/harware/whatever. jaclaz

These should be of help : http://www.laptoprepair101.com/laptop/2007...l-off-keyboard/ http://www.machinaelectronics.com/store/li...?f=key_type_k19 http://www.iraqigeek.com/2006/07/19/repair...aptop-keyboard/ jaclaz

CHKDSK with SP2 will be fine. NTFS didn't change. Though it is possible that just fixing the filesystem won't be enough. jaclaz

Maybe this helps: http://www.ss64.com/nt/dir.html http://www.ss64.com/nt/attrib.html jaclaz

Though I am now out of the game, I guess that the preliminary work and maybe some of the conclusions I got to might be of use to someone willing to take part in it. This post will be extremely long, and in some parts will be more similar to a RANT than to a valid technical post, please try forgiving my verbosity and my critical attitude. Let me state beforehand a few facts and personal opinions: I find HP has been for yesrs the best firm in the world for pocket calculators (in my life I bought several tens of them) I find HP has been for years the best firm in world for Laser printers (as well in my life I bought several tens of them) I find HP has been for years the best firm in the world for plotters (as well in my life I bought several tens of them) I find Compaq has been for years the best firm in the world for plotters (as well in my life I bought several tens of them - in the years 1992 up to 1995) I find current HP policies with software and particularly with OS simply appalling . HP was facing to be condemned for similar practices in a Class action and opted for a settlement: http://web.archive.org/web/20060402204440/...settlement.com/ It seems like it is a common practice for HP to close with a settlement cases where they were accused to provide defective products and/or failed to give Customers satisfacting repairs/fixes: http://barrettgrider-v-hpcompaq.com/index.html In Italy there are currently several cases opened against HP following its refusal to reimbourse customers that do not want Vista on their PC's (Italian): http://punto-informatico.it/1354479/PI/New...unciata-hp.aspx http://punto-informatico.it/2096808/PI/New...rso-dovuto.aspx http://punto-informatico.it/2341415/PI/New...-tribunale.aspx and an appeal to the Antitrust European Committee has been filed. (but the above bears no connection to the issue at hand) Let's get back to topic. From what I have seen and understood from the reports, the "Softthinks" MBRINST.EXE ( in some cases renamed to MBR.DLL) is a MBR installing utility that works roughly this way: writes a "special" MBR analyzes the partitions on the hard disk if a partition with certain (unknown at the moment) settings/data is found, THEN: adds to the partition bootsector the word "RECOVERY" changes some data in the MBR accordingly The above procedure is NOT repeatable, i.e. once the "RECOVERY" partition is made, the MBRINST.EXE does not recognize it as "RECOVERY" anymore. Additionally, a bit seems like being set in the MBR if, even once, the MBR is booted and the "Recovery" partition is not found. (to be more exact byte 1B4h changes from 01 to 00) Of course by disassembling or reverse engineering the MBRINST.EXE it should be possible to find out which checks are made, and thus being able to re-create the conditions where MBRINST.EXE recognizes the partition, but this procedure is out of the scope of the thread/board. (at least as I see it) On the other hand, no Law prohibits trying to find how the thing works. Some notes: There are several versions of the MBRINST.EXE file. I was able to find/examine three of them: 147,456 bytes in size dated 30-09-2005 159,744 bytes in size dated 12-03-2008 165,248 bytes in size dated 16-10-2008 They are all the same thing, but with "added options". #1 has 4 options for "While counting display": nothing Digit Dots Qplay #2 has 7 options for "While counting display": nothing Digit Dots Qplay HP Qplay GW HP F11 NTFS HP F11 BITL #3 has 8 options for "While counting display": nothing Digit Dots Qplay HP Qplay GW HP F11 NTFS HP F11 BITL HP F11 BL;QP Now again my personal opinion: the tool was initially programmed "properly" (in a version before the first one listed) then it was "customized" and options were added to it (in the "wrong place", it simply makes no sense to have options in "While counting display" in my opinion this is the result of lazy programmers that do quick and dirty hacks instead of doing properly something The app contains a number of different MBR's and also a few bootsectors. For the curious ones, I wrote a small batch that, using gsar and dumphex, can extract these sectors from the app and "divide" them in three categories: MBR's BS's UNKNOWN's (most probably parts of a multisector MBR) Here it is. You will need in the same directory of the batch, besides MBRINST.EXE, gsar and dumphex: gsar: http://home.online.no/~tjaberg/ dumphex: http://rbach.priv.at/DumpHex/ @ECHO OFF SETLOCAL ENABLEEXTENSIONS SETLOCAL ENABLEDELAYEDEXPANSION CD "%~dp0" IF "%~nx1"=="" GOTO :Nofile ECHO All files in current directory with extension .sec and .dmp will be deleted. ECHO to abort press Ctrl+C NOW! PAUSE Del *.sec Del *.dmp PAUSE Set Source="%~dpnx1" Set /A Counter=0 FOR /F "tokens=3 delims=:" %%A IN ('gsar -b -s:x00:x00:x55:xAA %Source% ^|FIND /V "found"') DO ( SET /A Counter=!Counter!+1 SET /A FoundOffset=%%A SET /A Offset=!FoundOffset!-508 IF !Counter! lss 10 (SET Number=0!Counter!) ELSE (SET Number=!Counter!) dsfo %Source% !Offset! 512 !Number!_!FoundOffset!.sec ) Del *.dmp FOR /F %%? IN ('DIR /B *.sec') DO ( ECHO %%? CALL :Dorename %%? CALL :Checktype !NewName!.sec ) GOTO :EOF :Dorename FOR /F "tokens=2,3 delims=: " %%B IN ('dumphex -s1B6 /l2 /nc %1 2^>nul') DO ( SET NewName=%~n1_%%B%%C REN %1 !NewName!.sec ) GOTO :EOF :CheckType FOR /F "tokens=2,3 delims=: " %%D IN ('dumphex /l1 /nc %1 2^>nul') DO ( SET Type=UNKNOWN IF /I %%D==33 SET Type=MBR IF /I %%D==E9 SET Type=BS SET NewerName=%~n1_!Type! REN %1 !NewerName!.sec dumphex /o!NewerName!.dmp !NewerName!.sec ) GOTO :EOF :Nofile ECHO you must supply a target file! ECHO example %~nx0 MBRINST.EXE PAUSE GOTO :EOF Result will be a number of "couples" of files, named in the form: (progressive number)_(Offset within_MBRINST.EXE)_(value of bytes 1B6h&1B7h)_(Type of sector).sec (progressive number)_(Offset within_MBRINST.EXE)_(value of bytes 1B6h&1B7h)_(Type of sector).dmp The .sec files are the binary sectors, the .dmp files are dumps, that you can open in Notepad or any tect editor. Bytes at offset 1B6h&1B7h appear to be a way to "categorize the different MBR's. Another interesting thing to do with the file is running BinText against it: http://www.foundstone.com/us/resources/proddesc/bintext.htm An interesting number of things can be found from reading the output, most noticeably the text description of the various MBR's (which is often truncated in the text box when running the tool): And some info about WHO actually wrote the tool: Finding of other interesting strings is left to the reader. I guess that people suffering from the problem may want to try contacting the author, Ulf Loesche through his site http://www.xss.com/ to either: ask him for help in solving the matter show him their appreciation for the program A last bit of text in the .exe, that as I see describes the whole stuff very well : jaclaz

I find improbable that the stick is the problem. Try reading these: http://www.msfn.org/board/index.php?showtopic=125945 http://www.msfn.org/board/profiles-directo...me-t125819.html and use the MakeMigrateInf.exe: http://www.msfn.org/board/index.php?act=at...st&id=23940 Run it TWO times, one with the stick inserted and mounted: makemigrateinf.exe>with_stick.txt and one without it: makemigrateinf.exe>no_stick.txt Attach the two resulting files, together with the output of mkmigrateinf.cmd. jaclaz

I see it's absolutely of no use participating to this game. It has been fun while it lasted. So long and thanks for all the sectors.... jaclaz

Make it what? The posted workaround? <-this is the thing that needs to be tested! Or running MBRINST.EXE or MBR.DLL renamed? <- this is the thing for which there is NO solution (yet ) jaclaz

Hopefully a workaround (NOT the solution) by using mbldr. Get mbldr here: http://mbldr.sourceforge.net/ http://mbldr.sourceforge.net/#Download http://prdownloads.sourceforge.net/mbldr/m...46.zip?download Unzip to a directory, then run mbldrgui.exe. Then: Select the drive. Add> first the RECOVERY partition and specify the "RECOVERY" label for it - it will be the smallest of the two and probably listed as second on the left window Add> second the "Normal" partition (and specify the "Normal" label for it) - press "Mark current" button Uncheck "Hide other primary partitions at boot" Change the drop-down box on the left "Progress bar" from "Decreasing digits" to "Dots" Change the drop-down box on the right "Keis choosing what to boot" from "1,2,3" to "Custom scan code" A popup will show, READ it's contents and click OK then replace the "59" with "87" Now edit the text: delete line "mbldr v1.46" - a popup will show, READ it's contents and click OK delete line "RECOVERY" delete line "*Normal" Enter a line "Press F11 for Emergency Recovery" Now backup and save: Press the "Backup MBR", this means that you are saving to a file the current MBR of the chosen drive. Press the "Save MBR", this means that you are saving to a file the NEW MBR that you just built with mbldr. Now write the NEW MBR: Change the small drop-down box on bottom left corner from "file" to "sector" A popup will appear, READ it's contents and click OK. Press the "Save MBR", this means that you are saving to a sector the NEW MBR that you just built with mbldr, a popup will appear with already the preset "0" sector, leave it as is and press OK. Another popup will show for confirmation, click OK. Try rebooting, you should be able by pressing F11 to boot to the "RECOVERY" partition, and by waiting or pressing [ENTER] to boot normally. DON'T EVEN THINK of messing with the MBR if you have not a tested alternative way of booting the PC (see a few posts before for an easy one using a USB stick). Report if this workaround works. jaclaz

Try re-reading this: jaclaz

NO. Don't do ANYTHING you may later regret. Before doing ANY attempt, and definitely NOT with "sophisticated" tools like Partition Magic, Partition Doctor, Acronis or paragon's Partition Manager, make sure by testing and double testing that you have a way back, as I showed ahmad2080 before. Are you willing to take part in game #1? Let see if I can put together the essentials: from factory there is a "special" MBR on HP/Compaq laptops that allows when booting to press F11 to access 2nd partition (Recovery) and from it start ..... (please fill the gap, what is supposed to do when F11 is pressed? Should the option to Recover the drive or create the disk start? this capacity is lost by replacing the MBR with a "normal" one (or however a different one) the MBR you posted appears to be the original "special" one, but with a couple of "suspect", as I see it, bytes how was the MBR you posted made? Did you already run MBRINST.EXE or MBR.DLL on that PC? Or otherways when did the problem happen? jaclaz

Niru, there are three games played on the board: 1) is the one I am playing, that at the moment is still going on, but with no progresses as my partner abandoned.... 2) is the one ahmad2080 played and won by cheating (and later abandoned) 3) is the one that Tripredacus played (and won ) using MBR.EXE from Terabyte: http://www.msfn.org/board/index.php?showtopic=130609 Your best bet right now is #3. jaclaz

@snack The bootsector you posted is perfectly unuseful, a NTFS bootsector is 16 sectors. jaclaz

I thought the whole point was to find the way to run the original app to re-install a working MBR as in factory. Using the way Tripredacus suggested before (using the MBR.EXE from Terabyte) or using grub4dos, you could have done the same. So, I am happy that your initial problem was solved, but still we haven't found the solution . Once you have done the recovery disks, try applying the attached MBR and post results. Also, I seem to be missing two things: the bootsector form your Recovery partition (16 sectors) the bootsector from the recovery partition of your friend (the first one) also 16 sectors. I need them, when you have time, in order to try and understand which mods (if any) are made to them. jaclaz MBR_from_j.zip

Though seemingly unrelated, I guess some info may come from this "Recovery Console" thread: http://www.boot-land.net/forums/index.php?showtopic=5316 jaclaz

Basically, missing to read the present thread. Quick resume: the thread by Markimoo is rather old there are new versions of grub4dos there are new methods to install grldr (or actually NOT installing it) there is a dedicated Forum with tens of threads delving deeper on the usage of grub4dos there is a new Guide for grub4dos In other words: jaclaz

As said, now that you have alternate ways to boot, you can try everything you wish to (within limits ). In my experience, however, the infinite time and monkeys approach: http://en.wikipedia.org/wiki/Infinite_monkey_theorem has rarely produced a good result in an acceptable time, and often has also prevented the "right" solution to be later implemented. For example installing the "upgrade" you may well have messed up things in such a way that there is NO way back anymore, one thing is to replace a MBR (for which you have a sure way back, and another thing is to apply an "upgrade" dated BEFORE the files you have, I like to define this behaviour "asking for troubles" ) Of course it is perfectly possible that the install of the "upgrade" carried no consequences at all.... Rest assured the MBR is ALWAYS 512 bytes long, the SPACE (number of bytes) available in it for "Master Boot Record Strings" can change depending on the type of MBR (i.e. if the MBR code is longer, there remains LESS available space for the text connected to messages). Try counting the characters in the messages in MBR.INI: 12345678901234567890123456789012345678901 Press F11 for Emergency Recovery Err1 Err2 Err3 And subtract from it the "fixed part"? 12345678901 Press Fxx for That would make 30. 1234567890123456789012345678901234567890123456789012345678 Press F11 for Emergency Recovery MBR Error OS Missing Press a key. That would make 47. It is possible that without a MBR.INI the "embedded" values lead to 53.... There are a couple of bytes in the MBR you re-created following my instructions that differ from those on your friend's (working) MBR, that may be the key, but I need some time to do some tests....., you need some patience. jaclaz

Why not, NOW you have way backs (at least up to this day trial). Worst that can happen, it won't boot, and you can restore the last working MBR. jaclaz

Try sending me the new resulting MBR. Even if you won't believe me a small step in the right (supposedly) direction has been made. You should now be able to see "Press F11 for recovery partition" when booting. I also need the bootsector (16 sectors) taken with HDhacker of both your and your friends "Recovery partition" (choose logical drive and the appropriate drive letter). jaclaz

Well, I guess you can write your own batch to do that. Here is an example: http://www.msfn.org/board/Hand-BOOTINI-t66101.html http://www.msfn.org/board/Hand-BOOTINI-t66101.html&st=15 Just replace the argument of FIND /V with the exact text of the entry you want to remove. jaclaz

Do you have a French keyboard? http://www.boot-land.net/forums/index.php?...ic=7653&hl= To be accurate, the file you made with HDhacker on 06-04-2009 14:30 differs from the one taken by MBRWIZ on 08-04-2009 14:19. Since the "new" MBR captured by MBRWIZ appears to be like one made by MBRINST.EXE, but WITHOUT the proper settings, I do suspect that in the meantime MBRINST.EXE has been run on the machine. I would try to re-capture from Vista and with HDhacker the MBR, then, immediately after, re-capture it from FreeDOS and MBRWIZ booted from the USB stick, and compare these last two files. It's a known thing that little green men sneakily enter households at night making socks odd pairs and runnning MBRINST.EXE..... jaclaz

Read is one thing and Read/Write is another. Ext4 Fs is - up to a certain point - "forward compatible" with Ext2 and Ext3: http://en.wikipedia.org/wiki/Ext4 and as such, ext2fsd should be able to Read it with no problems: http://ext2fsd.sourceforge.net/ As well, "trivial" write operation such as updating a small text file should be possible, as well as copying one at the time files, I wouldn't set a paging file on a ext4 partition, nor use multithreaded apps to write to the filesystem. Basically you will access the ext4 as it was an ext2 FS: http://www.linuxjournal.com/article/9449 jaclaz