HFSLIP - Test releases

[Camarade_Tux]

Ok. In fact, this is a problem with hfslip (and I think I already had it last month, CPU is stable and this happened on more than 5 isos)

-> Filemon.

Edited August 9, 2006 by Camarade_Tux

[Kiki Burgh]

hi Camarade Tux! try not to recompress ASMS ...

[Super-Magician]

hi Camarade Tux! try not to recompress ASMS ...

I second that! I wouldn't do too much experimentation with HFSLIP + nLite if I were you .

[Camarade_Tux]

I tried without nlite and I still got the error and I think hfslip cabs the files itself.

I'm slowly debugging that. Extracting files one by one. Hopefully, I don't have to go through the whole textmode part of setup.

edit : I had to extract the content of 4 folder to get this back.

Now, I wonder what the error is. I suspect it has something to do with compression method.

PS: T-39 is later on and filemon is now running : 260.000th entry in the log has been reached. :woot:

Err.... 300.000 entry I mean. No, 310.0000...

Edited August 9, 2006 by Camarade_Tux

[Camarade_Tux]

WOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTT !

I finally got it, ... thanks to Tomcat.

The worst thing is he gave me the solution immediately : *.MAN files have to remain uncompressed. That's all.

No, I have to check if hflsip cabs this file on its own. :?

[Camarade_Tux]

I can confirm the error is in hfslip.

Or at least in a hotfix : my man files are CABbed.

[Tomcat76]

KB917008 for Win2K is not OK. I'll revert to HFSLIP_60808a once I get my computer running again (spilled coke on the keyboard... again!). I'll then try to figure out how to have it slipstreamed properly.

@Camarade: I'll check it out.

[Tomcat76]

@Camarade: Try 60809a...

[Camarade_Tux]

Workign, but why not compress everything except .man files ?

Have a look at this page for ready-to-use implementation (thanks hp38quser).

Edited August 10, 2006 by Camarade_Tux

[Yzöwl]

Workign, but why not compress everything except .man files ?

Have a look at this page for ready-to-use implementation (thanks hp38quser).

Correction for your link

this page

Correction for the code given

FOR /F "DELIMS=" %%I IN (

'DIR/B/A-D %LOCATION%^|FINDSTR/EVIL ".CAB .CMD .INI .MAN _"') DO (

MAKECAB /D CompressionMemory=etc.

[Camarade_Tux]

Thanks for your corrections.

I wonder how I ended up with "http://page" in my link. :?

And you're incredibly fluent with BATCH, much more than I will ever be.

And why "/EVIL" ? Is it just something like foobar ?

[Tomcat76]

E = end of line

V = exclude

I = case insensitive

L = literal search

L is probably used so you can do without a backslash in front of the dots in these search strings.

I was thinking of having HFSLIP compress new ASMS files anyway but just exclude *.man files. But I don't know how nLite copes with a source of which some ASMS files are already cabbed and some not so I thought it safer to just leave all files uncompressed.

[Tomcat76]

On second thought, I think it isn't such a good idea to compress the new files (except .man). Either you compress both old and new ASMS binaries, or neither. There's no point in just compressing the new ones as people who'd care would prefer to do them all so they'll use a different program for it anyway.

[vci]

just wanted to say:

latest test release works fine, everything is beeing slipstreamed perfectly!

[Camarade_Tux]

On second thought, I think it isn't such a good idea to compress the new files (except .man). Either you compress both old and new ASMS binaries, or neither. There's no point in just compressing the new ones as people who'd care would prefer to do them all so they'll use a different program for it anyway.

In nlite, there's an option for that and the corresponding tooltip says it may create bugs. I thought it was an old comment and asked if it could be changed. Nuhi told me some OSes would still bug if one of the file was compressed.

Better safe than sorry, in the end, it is probably better not to cab everything.

And now, I can say this test releases is working perfectly. B)

Thanks again.