[PATCH] add explicit extension disabling syntax to hgrc.5

Adrian Buehlmann adrian at cadifra.com
Wed Feb 27 03:29:40 CST 2008


On 27.02.2008 09:03, Thomas Arendsen Hein wrote:
> * Adrian Buehlmann <adrian at cadifra.com> [20080227 00:42]:
>> On 26.02.2008 22:14, Patrick Mézard wrote:
>>> Steve Borho a écrit :
>>>> # HG changeset patch
>>>> # User Steve Borho <steve at borho.org>
>>>> # Date 1204028711 21600
>>>> # Node ID 96ef88ccbcb0e790b25f5c4eecbe2bd29033287e
>>>> # Parent  434139080ed4007d99e843b203c7da0850238744
>>>> add explicit extension disabling syntax to hgrc.5
>>> Slightly edited and pushed in crew as 7f9f3233a2c6. Thanks !
>> So I wonder why you didn't make two changesets of this.
>> One with Steve's original patch, followed by your edit.
>>
>> Just an observation. Possibly a very bad example here since your edit
>> is certainly good. But for me it looks a bit like someone signs
>> something using the signature of somebody else. A bit like using
>> the wrong sender address in an email. Something I really wouldn't like
>> to do myself.
> 
> Usually if it is more than fixing a typo it will be a separate
> patch. In this particular case Steve said on IRC that we should feel
> free to extend it a little bit to make path vs. no path clearer.
> 
> So this is just a shortcut for "Steve, please change this for that
> and make a new patch".

Sorry for having nitpicked about this. Patrick certainly did the right 
thing here.

My point is that one of the reasons why I switched to Mercurial is to
have full transparency and an exact, complete immutable recording of 
who did what based on what.

But that is of course a matter of repository policy and probably does
not apply to crew repo. But I do feel that this is a little bit of a 
missed chance here since newbies like myself naturally look at how 
successful real world Mercurial repos are used to learn some best 
practices.

And honestly, not wanting to clutter up history is normally a rather 
weak argument, IMHO. This leads to the well known habit of big submits 
(which -- to be clear -- absolutely does not apply to Patrick's edit 
here).

>> On another note, it seems that repository pullers will never know who
>> pushed the change into crew by looking into the crew repo alone.
>> An information that would be nice to have as well.
> 
> Sometimes I miss that, too, and already thought about adding this
> information to the notifications :) This wouldn't be part of the
> changeset history of course.

Couldn't this information be added into the user field as well? I 
mean, Mercurial doesn't parse what's stored in the user field?

Maybe: "Steve Borho <email>, imported by Patrick Mézard <email>"

As a side note, monotone's "mtn cert" use case might be interesting to 
look at: http://monotone.ca/docs/Certificate.html#Certificate



More information about the Mercurial-devel mailing list