[darcs-users] More darcs merging fun.

Ganesh Sittampalam ganesh at earth.li
Tue Apr 20 20:55:06 UTC 2010


On Tue, 20 Apr 2010, Axel Simon wrote:

> I'm still trying to work out how to get the Gtk2Hs repo in sync with what I 
> have locally.

For what it's worth, you can just copy an entire repo around using normal 
filesystem operations, if you want to get it into exact sync. That would 
also clobber things like repo settings and posthooks though.

> I've obliterated a few patches in code.haskell.org:/srv/gtk2hs such that 
> nothing occurred when saying darcs pull in my local repo. I've added all the 
> patches I needed and then did this:
>
> ~/source/gtk2hs $ darcs push code.haskell.org:/srv/code/gtk2hs
> Thu Nov 12 22:39:16 CET 2009  Axel.Simon at ens.fr
> * Add the TreeModelFilter module and a demo.
> Shall I push this patch? (1/38)  [ynWvplxdaqjk], or ? for help: a
> Reading inventory of repository /srv/code/gtk2hs inventory
> Examining patches for conflicts 2/38
> darcs: Couldn't find patch matching "hash 
> 20100409164412-c227b-45619374ae95d31dd4b6df2f1d5b11e46e1d0e68.gz"
> Xtract: Ran out of input (EOF)
> Finished applying...
> Push successful.

The Xtract: Ran out of input message comes from your posthook, and is 
presumably triggered by the previous error from darcs. My best guess from 
the output and from reading the posthook script on code.haskell.org is 
that the error from darcs also comes from that script, so the script 
itself is buggy/out of date in some way, but the actual initial darcs 
operation was fine as indicated by "Push successful".

> The resulting repository is identical to my local one (darcs pull/darcs push 
> doesn't do anything). But I have this feeling that I should worry. Should I 
> worry? Has the error really occurred at the 2nd patch or might this be an 
> artifact?

Do a 'darcs check' on the code.haskell repo just to be sure, but I think 
it's fine.

Cheers,

Ganesh


More information about the darcs-users mailing list