[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: retrieving corrupted binary revisions!
From: |
Nils9 |
Subject: |
Re: retrieving corrupted binary revisions! |
Date: |
Wed, 5 Sep 2001 23:41:23 +0300 |
Exactly! Without the -p option it works perfect. Hm.. now, how
to make the WinCVS to not use the -p ?!
Looks like I have to download the WinCVS source, track down
the option and recompile without it?
-Nils Jakobson
SWHTechnology
----- Original Message -----
From: "Eric Siegerman" <address@hidden>
To: <address@hidden>
Sent: Wednesday, September 05, 2001 8:55 PM
Subject: Re: retrieving corrupted binary revisions!
> On Wed, Sep 05, 2001 at 03:00:11PM +0300, Nils9 wrote:
> >
> > Environment: RadHat running repository, pserver. WinCVS 1.2
> > clients on NT4/W2K.
> >
> > In short: I commit word.doc as a binary (-kb)
> > [...]
> > Now I select the 1.1
> > in the graph and from menu choose Graph -> Retrieve revision (the
> > respective command: cvs -z9 update -p -r 1.1 word.doc), it
> > creates the non-cvs file (word.doc.#.1.1) with the 1.1 revision, but it
> > is corrupted file - its half the length it should be!
>
> That "-p" is suspicious. I'm just guessing, but it would stand
> to reason that CVS's stdout is always in text mode, causing the
> file's -kb setting to be effectively ignored. To avoid this, CVS
> would have to close and reopen stdout to switch its mode back and
> forth -- and that would be a *bad* idea.
>
> To test this theory, try typing that command yourself, without
> the "-p", at the WinCVS command prompt. If that fixes it, the
> next challenge will be to convince WinCVS not to use "-p"...
>
> --
>
> | | /\
> |-_|/ > Eric Siegerman, Toronto, Ont. address@hidden
> | | /
> With sufficient thrust, pigs fly just fine. However, this is not
> necessarily a good idea.
> - RFC 1925 (quoting an unnamed source)
>
> _______________________________________________
> Info-cvs mailing list
> address@hidden
> http://mail.gnu.org/mailman/listinfo/info-cvs
>