Search options (Page 29 of 99)

Hi
this is the same driver C-Major requested.

I think we need standardised test reports more than ever.
We have had to hold back pretty cool ideas and the DriverPacks based on those ideas are promising.
Daydreams are not only about people you love. Daydreams can be about worthy ideas that are hard to implement.
Bâshrat the Sneaky has a future for us, and Overflow has cool ideas which are.. Well.. Really Cool.

In the meantime, we walk on trodden paths.
One of the cool ideas we started to implement was folder path shortening.
We recently learned that my handywork breaks KTD.

I did not know it, and released some of those optimized 'short-folder' packs.
I could, and maybe I should revert the edits what made shorter folders work in my tests.

On the other hand...
I could use automize and record the few commands I use to shorten path even further (to under 3900), and I could just as well record what I did to have the old paths (from approx 7000.) reduced to what we have released. (main packs in test have a path of approx 4500 in Method 1)

Mass storage DriverPacks.
I am not happy with what we have.
It is the best collection, it has a good load order (which can be changed quite easily) but it cannot have all of its contents working simultime.
(hence.. load order. If drivers was perfect and BIOSes were, then load order would not be an issue.)
Some drivers take the task when they should not take it, and that happens when we make them load later than the driver the machine should use. When you move the secction of driver it should use in the INI below the section of the "wrong one" and recompress.. they load in an other order.
It should not make a difference but it does.

We either educate and elucidate, or we drop some stuff what never worked together.
OR.. the brilliant coders make you choose.
We could then perhaps elucidate and tell a user that a01250 and Si3112r and megaIDE2 are mutually exclusive.
It is perhaps a daydream, but it may work. (after a hyllovvalotta work.)
> My personal opinion is that database driven driver selection with non filtered original drivers is never going to work because I saw what I saw.

we need an overseer or manager for REQuests in progress and have no systemised task sheduler.

Has this been handled yet?

TRIMMING TIPS
If you are going to TRIM  (aka prune) any of the DriverPacks, extract it in a separate folder because later on you have to repack its content along with its edited INI.

NEVER rename DriverPack_xxxxxxxxx_wnt5_x86-32.ini

You have to look at the content of the INI, because if you remove folders and do not edit the INI you will get a "file not found" error during slipstream.
Renaming a folder also results in "file not found" if you do not change the foldername in the INI.

An example:
In Mass storage, ALL folders have a corresponding section in DriverPack_MassStorage_wnt5_x86-32.ini.
The header [3-1] corresponds with D\M\3\1
Header [31] would correspond with D\M\31 and Header [3A-1] would correspond with D\M\3A\1

In Sound you could nowadays see D\SCR1 [SCR1] and in LAN you could see D\LNV3 [LNV3]
We reduced Method 1 path length about two thousand characters that way.
Some DriverPacks have NO INI.
These are, of course, the easiest to trim.

When you prune DriverPacks and remove folders from a pack's tree you must remove the corresponding sections in that pack's INI.

You can remove a section in the INI and leave the folder in the pack.
(you may wish to do that for troubleshooting)

This will not create an error during slipstream.

- - - -

You have to recompress the DriverPack.
DPs_base uses the driverpacks, not the extracted folders.
The name you give the repacked DriverPacks must follow the naming convention for a driverpack.
If you unpacked xxx8031, you could call it xxx8032 so as not to accidentally overwrite the original.
I figure you would want the original in case your modification was not succesful.
Its higher version number will be useful anyway.

An example:
I let 7zip unpack it to folder so it creates a new folder (which automatically gets the name of the driverpack..)
Then I rename that folder changing the version number to a higher version number.
An example, 803 (year/month) becomes 8031 (year/month/revision).
If it was 8031 to begin with it I change it to 8032.
(I think you don't want to call it 804.. We could publish a new one in april and that one would be 804.)

After that do what you want to do in that new folder.
(Examples:  remove sections because you deleted a folder, move sections to change load order, make it not skip a driver, disable a driver and keep it for PnP, and so on.).

When you are done you select the D folder and INI file so that both get highlited, and rightclick to use 7zip on selection.
It will automatically suggest the name of the folder, which you had already renamed to higher version..
Settings are: 7z/Ultra/LZMA/wordsize 256/compact.

NEVER rename the INI.

Let us know what you had to fix, and how you did it.
Help us help you.

funny that it fails on a win2000 file.
Let me guess?
Did you use DriverPacks on another OS (like 2000 or 2003) before you did XP?

if you did that, then you should clean the mass storage sub-folder in QSC.

IF you ran base to stream a new mass storage pack into a XP source you had already streamed an old mass into, you should start FRESH on a FRESH source that had NO mass storage streamed yet.
(the log actually shows you ran this over an older slipstream.. so I will advise to start clean slate, with a fresh source and mass storage deleted from QSC.).

We discovered this peculiar bug not too long ago.

Oh, you might as well read the tutorial you find in my signature.
The extra info mentions this peculiarity.

Do you want me to make us a 803B testpack? (which would restore Nvidia folders I had deleted from the pack.)

Hi
how long did your DriverPacks slipstream take?
I read you also tried method 1? how long did that take?

Hi Twig123..
Congratulations.

@ Jeff.
Smart move.

Hi, welcome again.
Can you link to the driver?

I found Biostar T780G ver 6.0 and this could be the wrong one.

Google translate..

Anyway, Welcome both.
Please try English?

If you was not aware or it, I will say it.
kernel 2008 is pure X64. It sits on the metal, or won't. Period.
(yep, period. ??? My understanding is that 2008 will not even TRY to install if the basic IO in hardware is not compliant.)

I will reword that...
The Kernel of server manager 2008 sits on X64 metal and the "driver repository" partition is at a higher level and can be populated with third party.
There is a X32 version of server 2008, but X32 won't have server manager.
X32 is more or less a client and can be remotely managed with the manager tool.
The manager is a "role" of server 2008 and that role cannot install if the machine is not fully 64bit compliant.