r7479 - branches/udev_update/BOOK

Matthew Burgess matthew at linuxfromscratch.org
Thu Mar 30 10:11:49 PST 2006


Ken Moffat wrote:
> On Thu, 30 Mar 2006, Matthew Burgess wrote:
> 
>>Hmmm, what's with this apparent convention of changing the date in its own
>>revision?
> 
>  Atomicity.  (i.) if you want to revert a Real Change, it should be 
> cleaner. 

But if only one change was made to the book on a particular day then 
reverting the date change would be the right thing to do, IMO. 
Obviously we can all come up with cases where changing the date 
separately or with a Real Change make sense.

> (ii.) now that I've got svn merge actually working, no 
> conflicts (the dates are accompanied by branch titles, so trying to 
> merge adate change in general.ent produces a lot of mess).

Yes, maybe I'm lazy with this but I've got used to the conflicts.  I 
simply go into general.ent, change the date entities to what they should 
be, remove the conflict markers then 'svn resolved' to remove the 
.merge* files that svn produces.  It takes just a few seconds although 
it sounds like a lot of work.  It's probably about the same timewise as 
issuing two different 'svn commit' commands.

Anyway, I'm not so bothered about this, just interested in the 
motivations behind it which you've now provided.

Cheers,

Matt.



More information about the lfs-book mailing list