jaclaz

It's not you (alone) that are confused on the topic (a good half or maybe three quarters of the people have not entirely clear the matter, not their fault but that of confusing, partial or incomplete and contrasting information). If your EFI has a CSM (Compatibiity Support Mode, please read as BIOS emulation, most probably mislabeled by HP or by the Insyde guys as "legacy support mode") it would be better to use it (and the MBR scheme suggested). You are correct, if your device is 64 bit it will only work (in EFI) with a 64 bit OS and not with a 32 bit one. In a nutshell, there is NO reason whatsoever to have GPT (if not on larger than 2.2 Tb disks), and there is no reason to use EFI mode (i.e. it provides not any practical or noticeable advantage once the OS is booted). Anyway, you should consider these install attempts as "experiments", whether they work or not they may provide additional info on the behaviour of the device and OS, and hopefully eventually provide a good "final method" (that you may want to re-test restarting form scratch to verify it really works). Before I forget, a possible reason why the setup doesn't like that device could be because it sees it as "external" or "non-paging" :w00t:, so another experiment could be that of using Diskmod, see: http://reboot.pro/topic/9461-page-file-in-usb-hard-disk/ http://reboot.pro/topic/9461-page-file-in-usb-hard-disk/?p=86619 jaclaz

Just a guess, mind you, but right now the setup program doesn't *like* the device "as is". There may be several reasons (call them "precautions" if you like) why the good MS guys might have put some limitations to the setup program, making it detect a number of conditions that make a disk "eligible for install". But diskpart sees the device, so it is worth a try to (see this for some reference): http://www.jwgoerlich.us/blogengine/post/2009/11/05/Use-Diskpart-to-Create-and-Format-Partitions.aspx DISKPART> select volume 0 DISKPART> remove letter=C DISKPART> select disk (0) DISKPART> convert mbr DISKPART> create partition primary DISKPART> select part 1 DISKPART> active DISKPART> format fs=ntfs label=(name) quick DISKPART> assign letter=C DISKPART> list volume (right now the volume on the USB stick aka Disk 1 has got letter C, so you need to remove it in order to assign it to the internal disk) Double check again that the situation is as seen in the scrrenshots you posted, i.e. internal disk is disk 0 and Lexar USB device is disk 1 (and contains a volume with assigned letter C) before running the above commands, better be safe than sorry. The above commands are intended (I am not sure if this is the case) for a BIOS and "MBR style" disk, is that HP thingy BIOS or EFI/UEFI? If the latter you will probably need to make the disk GPT and create the FAT32 "boot" partition besides the "main" NTFS one. If the above set of command works, you can try again with the setup, maybe finding an active, formatted partition will allow the install Otherwise use the fujianaBC's method of applying the .wim (the WINNTSetup program mentioned earlier is based on that approach, but for the moment doing things manually gives more control/the possibility of having more feedback on the causes of the error): http://reboot.pro/topic/10126-nt-6x-fast-installer-install-win7-directly-to-usb-external-drive/ jaclaz

Well, not really (I was joking ). Once a system is switched off, it is off. Though there are differences between a "warm re-boot" (Ctrl+Alt+Del) and a "cold reboot" (switch off, then on again), a "cold reboot" is a "cold reboot". As you correctly stated one may want to wait a little bit to allow for the possibility that *something* is still powered (due to capacitors charge), something *like* like 5 or 10 seconds (not minutes) but anything more than that won't make any difference on relatively modern hardware. For PC's there is the issue of RAM "slow decay", the classic reference (JFYI) is the original research about the so called "cold boot attack" : https://citp.princeton.edu/research/memory/ but what the researcher did was to try and avoid the BIOS (when booting) to wipe/rewrite the memory as much as they can, and - on many systems - they needed to actually remove the sticks and read them "externally". The "wait 10" (or 15 or 30) seconds before powering on is (largely) a tradition, most probably "amplified" in suggested length of wait because of this or that. Anyway I don't doubt in the least that your exact procedure solved your problem , I was only making a note about it being seemingly more complex than what actually *strictly needed*. Practical example: Your car doesn't start in the morning. A common issue is that the battery is too low on voltage (because - say - you forgot your headlights on yesterday night) or has failed and another (less common but still happening) occurrence is that your starter motor is stuck (or as well gone bad). The normal troubleshooting path would begin with a test of the battery voltage and capacity and - in case - replace it with a working one. If the car still doesn't start test the starter motor, clean its solenoid contacts, verify and if needed replace brushes, or replace it with a spare one. Now, on a forum someone posts that since his car didn't start in the morning he replaced : battery starter engine block gearbox seats rear mirrors fuel tank and refilled with premium gasoline. Then, as soon as he tried again, after having waited 42 hours since the refill - the car started fine . No doubts about the procedure being effective, only a bit redundant and not providing any hint about what the actual problem was and which replaced item solved it. jaclaz

So, disk 0 is the emmc and disk 1 is the Lexar USB. Everything seems "cool". The Windows Setup won't list the USB as possible "target" drive for the install (possibly being "external", "removable"), but only the ("internal", "fixed") disk 0. Now it is time to understand why Windows believes that disk cannot be used for the install. I would try to attempt initializing and partitioning and formatting it in diskpart then see if you can apply to it a an install.wim. jaclaz

I tried waiting only 137 seconds (i.e. a little more than 2 minutes) and it worked fine as well . It must be because I have a remarkably fast machine, particularly when it is off . Or maybe it was because I kept my finger crossed all the time. Seriously, maybe you went through more work than really-really needed, it is rare that a single error has so many causes (System.ini AND VMM32.VXD AND SETUPX.DLL AND CONFIGMG.VXD AND WINSPOOL.DRV AND SYSTEM.DAT AND re-install drivers), it is more likely that only one or two of these items needed checking/replacing. About the VMM32.VXD, you may find of interest this article: http://www.helpwithwindows.com/techfiles/vmm32.html jaclaz

I don't think that the display/video part(s) of the Advanced Server are in *any way* dfferent from "Normal" Professional or Server in 2K. Drivers is another thing, from the symptoms it sounds like an issue with RAM. Can you try reproducing with Task Manager loaded checking how much memory the various programs use? Also some hardware specs? jaclaz

Yep, but this ASUS implementation is much worse than DELL's one (not tha the DELL's one is actually "good"), JFYI: https://duo.com/blog/out-of-box-exploitation-a-security-analysis-of-oem-updaters jaclaz

OT , but not much what could actually be worse than windows auto-updating to Windows 10? Rather obviously, unauthenticated automatic updates from the hardware manufacturer, JFYI: http://teletext.zaibatsutel.net/post/145370716258/deadupdate-or-how-i-learned-to-stop-worrying-and The really worrying part is not the (flawed) design or the (senselessly irresponsible) implementation, it is the fact that the ASUS guys didn't respond to the disclosure notice, ad not even to CERT/CC ... jaclaz

The Nusb is a "more complete" and/or "more complicated" package, offering "more" than a simple USB driver, when it works, it works very nicely of course , simpler packages, like Rloew's one, the one Aiko Chan mentioned and this one, etc.: are simpler, which is what the OP asked, but as said each and everything available needs to be tested and verified working (or not working) on the specific setup/hardware. jaclaz

A little hyped, as often happens, but it raises a couple of points about pushing upgrades automatically at the wrong time/without user intervention, etc.: http://www.theregister.co.uk/2016/06/03/windows_10_upgrade_satellite_link/ jaclaz

Nlite should be able to Though depending on size/features needed, you might also want to look here (JFYI): Historical: http://reboot.pro/3717/ Current: http://reboot.pro/topic/16765-minixp/ Actual site is down , but files and documentation can be got via Wayback Machine: https://web.archive.org/web/20151127225813/http://minixp.reboot.pro/ Also, a slightly bigger build (with MUCH more functionality) can be the Wimb's project of small "universal" XP in VHD: http://reboot.pro/topic/9830-universal-hdd-image-files-for-xp-and-windows-7/ jaclaz

Huge amount of RAM/SATA/crazy large hard disk sizes Check Rloew's patches: http://rloew.x10host.com/ He is a member here, you might want to contact him about (possibly already tested) compatibility of the hard disk related patches with your motherboard (the huge RAM one should work OK, independently from the specific motherboard). His (freeware) USB driver pack (based on a Lexar released driver) is simpler than MD NUSB, which attempts to integrate more to the system and has had (on some motherboards/systems at least) some installation issues. Be prepared anyway to several iterations of install/re-install, having a functional Windows 98 on modern hardware usually requires quite a lot of tweaking/adjustments. If you could "choose" one among the systems listed here: it might be easier. jaclaz

You made it sound a lot like: http://www.imdb.com/title/tt0118571/quotes?item=qt0412274 jaclaz

... maybe you should add "(related to computers)" or something to the same effect jaclaz

Yep, that would be very fine, it would be an "on request" or "on demand", much better than "no, you can't have it anymore" . As a side note, and JFYI, I personally find that your ethical concern is however not something to worry much, it is what happens everyday with *any* product. VERY early adopters/buyers of *any* product are likely to have some promotional discount (but may well receive an unfinished, buggy, or unpolished product) those coming after will receive the (mature) product at full price and those coming even after will get the same product at a discount (end of production, clearing of warehouse, rebates to remain competitive with competitors, reduced commerciability due to new versions/models etc.). The key here is "time" or "when" and/or "amount of need". If someone *really needs* to reduce a Windows 7 and needs it "now", he/she pays 10 bucks for the little nifty program, if he can postpone his/her *need* some weeks/months/years, likely the same program will be available at a discount, or at a fraction of the original selling price (or even eventually for free), it is something that happens and how it works more or less for most products, particularly in matters related to computing, think of RAM or USB sticks. I don't think that your customers could be offended by this, in your case maybe it is just a matter of the very short timeframe. jaclaz

It didn't, but it was an OS for Professional use, not (also) for the entertainment, the numbers were small because very few people actually needed (or used) a PC at work, the numbers came with XP (that was released only one year later), a typical office in those years ran several copies of Windows 95/98 (everyone but the following), very few NT 4.00 (Servers) and some more NT 4.00 (Engineers workstations and similar), when Windows 2000 came out, the Windows 95/98's were not upgraded, only the Servers and Engineer workstations. Until Server 2003 (actually Server 2003 SP1) came out all servers would be still running NT 4.00 and 2000 server, and as well all workstations continued using 2K, the much bigger impact of XP was because it replaced all the win 9x'x (and a very small number of Me's) machines, and of course a couple years had passed and *everyone* would have a PC. Once upon a time (JFYI) computers were made to work (as opposed to watch videos on youtube, twit and chat ) at the time Internet access was *luxury*, not something *standard* on every device. jaclaz

Citing myself for no apparent reason : http://www.theregister.co.uk/2016/06/01/teamviewer_mass_breach_report/ jaclaz

And with wrong theories , XP is ALREADY bloated, 2K is the "real thing" , and yes, I still run a couple NT 4.0 machines. jaclaz

Dude? Contents of a .msu CAN be extracted with 7-zip, contents of (MSDELTA compressed) .cab(s) inside .msu is another thing. Really, besides calling people "dude", you could take some time reading what was posted before replying. jaclaz

Any reason not to use the Convenience rollup instead? Or just the list of its contents. This way a number of superseded/replaced/cross*linked/whatever updates will be out of the equation. jaclaz

NO, actually it makes NO sense whatsoever, as .msu files do not exist . Have you actually EVER examined a .msu file in a hex editor or have done *any* research on it? EVER wondered WHY EXACTLY its contents can be extracted using EXPAND.EXE (or 7-zip)? How long would have been needed to understand how they are nothing but .cab files with a changed extension? This info is hidden inside this Wikipedia entry, BTW: https://en.wikipedia.org/wiki/Cabinet_(file_format) Now, if you are talking of making the "...properties.txt", and the .xml configuration files, and creating MSDELTA or IPD (Intra-Package Delta) .cab's that might be another thing and possibly an interesting question (as a side note and JFYI): Though I believe that the use of MSDELTA/IPD is just a choice by the good MS guys to "compact" the file and more traditional cab compression would work anyway. jaclaz

Well, if I may, personally I don't like the approach at all. It is perfectly fine to me that you would like to get a few bucks for your hard work. It is true that until you produced that thingy no similar tool existed. But what is the sense of putting it in such a way that it will "cease to exist" on the 30th June 2016? If we agree that by design you won't get anyway any more money (little as it might be) after June 30, 2016, why not re-release the thingy as unsupported freeware (or similar) keeping the Copyright and intellectual property? It won't cost you a single cent more than what you will have already spent (in time, resources, hosting, support, whatever) until that date, and a few people which might actually *need* or *want* to use it might still be able to find it. jaclaz

Traitor! Or maybe you wanted some revenge for that time he broke the lawnmower you lent him . jaclaz

You are welcome , though you will notice that the processor will get roughly 50% usage just by running the batch . jaclaz

In a nutshell, yes. The setting in BIOS changes the device (controller) VID and PID, or if you prefer the controller has two PID's, one connected to IDE mode and one connected to SATA mode. To the OS driver the controller is either a IDE controller (actually Parallel ATA) or a SATA (AHCI) one. Since ATA (Parallel ATA) tops at 133 Mb/s in the latest standards, i.e. ATA/ATAPI 7, you can attach to that bus *whatever* "fast" device you like to, but it won't ever exceed the 133 Mb/s. Your board (and OS driver) believe to be using an IDE controller and an ATA drive. You have to understand how a good half (or maybe three quarters) of the names we commonly use in the computing field are the result of a total failure to communicate (by the good engineeers) and total failure to be accurate (by the marketing folks), and as well quite a bit of the "perceived information" is either false or wrong. You will need to go through these: https://en.wikipedia.org/wiki/Parallel_ATA https://en.wikipedia.org/wiki/Serial_ATA and possibly also these: https://www.phildev.net/ata-modes.html https://expertester.wordpress.com/2008/07/24/ahci-vs-ide-–-benchmark-advantage/ As you have just noticed experimentally a good "real world" hard disk speed pre-SATA III is between (say) 50 and 140 Mb/s and (rarely) a "top range" SATA III exceeds 200 Mb/s, see as an example: http://www.tomshardware.com/charts/hdd-charts-2013/-02-Read-Throughput-Maximum-h2benchw-3.16,Marque_fbrandx42,2900.html Your 75 Mb/s sounds "just right" considering that the hard disks are older than the above. So - basically - there are NOT any difference between IDE/ATA 133 Mb/s and SATA I (150 Mb/s) there might be - maybe - some marginal differences with SATA II devices in AHCI mode if NCQ (Native Command Queing) is working, the "real thing" with SATA II and SATA III controllers can be appreciated with SSD's, which are waaay faster than any rotating hard disk. jaclaz