On Sat, Oct 10, 2009 at 22:19:12 -0700, thomas hartman wrote:
> quoth Max Batcher on the darcs users mailing list: "Darcs send checks
> "upstream" in order to determine which patches need to be sent, just
> like darcs push, and just like darcs push it defaults this check to
> _darcs/prefs/defaultrepo. Darcs doesn't copy the _darcs/prefs/email,
> it simply expects to be provided a remote repo."
>
> So, it seems creating the _darcs/prefs/email won't do anything useful.
I'm sure you saw Trent's explanation on the darcs-users mailing list,
but just in case: darcs uses the *remote* _darcs/prefs/email to
determine where to send the email to, so it would be very useful indeed.
I realise that our use of _darcs/prefs is very confusing because we
use the same directory for three very different things:
remote prefs:
- motd <-- this could also be useful for patch-taggers
- email
local prefs:
- author
- defaultrepo
- defaults
- sources
- boring
- binaries
internal metadata:
- prefs <-- used by setpref
Darcs users/hackers : perhaps a good thing would be for future darcs to
have a _darcs/localprefs to avoid this confusion. I don't think there
is much we can do about _darcs/prefs/prefs, on the other hand.
Regards,
--
Eric Kow <http://www.nltg.brighton.ac.uk/home/Eric.Kow>
PGP Key ID: 08AC04F9
http://blog.patch-tag.com/2009/10/14/patch-tag-is-now-darcs-send-friendly/
Let me know if you have any idea for additional improvements.
:)
thomas.
On Sun, Oct 11, 2009 at 7:27 AM, Eric Kow <ko...@darcs.net> wrote:
> On Sun, Oct 11, 2009 at 04:54:38 -0400, Max Battcher wrote:
>> On 10/11/2009 4:02, Eric Kow wrote:
>> >Also, I should correct one of the errors in my earlier message:
>> >_darcs/prefs/sources should be treated as internal Darcs metadata,
>> >not a local pref
>>
>> I would say that's a fine line distinction right there-- I've had to
>> edit _darcs/prefs/sources for various reasons and I could see
>> someone manually pruning it as appropriate...
>
> I we need to think a bit more carefully about this, especially why we
> make these distinctions. I seem to be confusing myself at least!
>
> The reason why I first said it was local information was because as you
> said above, you sometimes edit it as appropriate.
>
> The reason I later backtracked and said it was internal metadata was
> because Darcs manipulates it and consults it from remote repos.
> Notably, darcs copies the _darcs/prefs/sources from the remote repo and
> augments it with local information. Editing _darcs/prefs/prefs should
> be probably discouraged; editing the _darcs/prefs/sources on the other
> hand... is maybe kosher. Confusing the issue is the fact that Darcs
> does not currently handle the _darcs/prefs/sources very intelligently.
>
> See
> - http://bugs.darcs.net/issue1599 (authomatically expire unused caches)
> - http://bugs.darcs.net/issue1159 (unionCache does not adjust cache
> paths wrt remote repository location)
>
> Perhaps if Darcs were smarter about caches, users wouldn't need to
> intervene manually and we really can treat it as purely internal
> data.
>
> --
> Eric Kow <http://www.nltg.brighton.ac.uk/home/Eric.Kow>
> PGP Key ID: 08AC04F9
>
> _______________________________________________
> darcs-users mailing list
> darcs...@darcs.net
> http://lists.osuosl.org/mailman/listinfo/darcs-users
>
>
--
Need somewhere to put your code? http://patch-tag.com