quilt-dev
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Quilt-dev] merging with upstream and adding new files


From: Andreas Gruenbacher
Subject: Re: [Quilt-dev] merging with upstream and adding new files
Date: Mon, 30 May 2005 16:14:48 +0200
User-agent: KMail/1.8

On Saturday 28 May 2005 11:13, jerome lacoste wrote:
> Now some notes. I just started to use quilt this week. I am using it
> to manage local patches
>
> - removing patches
>
> One of the patches I had applied locally has been merged upstream.
> I made a cvs up on the code with my patches applied and this created
> an issue. I think I've also refreshed my patch later on, I thought
> that this would make the patch 'empty'. Except that this created a
> reverse patch!
> I guess I shouldn't have poped all patches first, do a cvs up and push
> them afterwards, removing the obsolete ones.

The right thing to do would be:
 - pop all patches
 - cvs up
 - push all patches again, merge rejects, and remove patches which have
   meanwhile been integrated.

> Am I correct? If so, maybe the documentation (PDF) should contain some
> notes about merging upstream patches.

Yes, I agree. Do you want to write something up and send a patch?

> - Furthermore It took me time to understand that in order to get a new
> file into a patch, I had to do something like
>
> touch myfile
> quilt add myfile
> cp myfile/from/somewhere myfile
> quilt refresh
>
> Without creating the empty file first and adding it, quilt was not
> doing anything.

Wrong order of operations. Do it like this:
 - quilt add myfile
 - cp myfile/from/somewhere myfile .
 - quilt refresh

> [...]
> I understand why it's like that now, but I wonder if the doc could be
> more explicit about it

Hmm, the documentation says:

> [...] Files must be added to a patch with quilt add before they are 
modified. Note that this is slightly different from the CVS style of 
interaction: with CVS, files are in the repository, and adding them before 
committing (but after modifying them) is enough. [...]

How do you want this to be explained instead?

Thanks,
Andreas.




reply via email to

[Prev in Thread] Current Thread [Next in Thread]