[mythtv] [mythtv-commits] mythtv commit: r12794 by jarod

Jarod Wilson lists at wilsonet.com
Wed Feb 14 21:26:31 UTC 2007

On Feb 14, 2007, at 16:07, Chris Pinkham wrote:

> * On Wed Feb 14, 2007 at 08:51:57PM +0000, mythtv at cvs.mythtv.org  
> wrote:
>>       Author: jarod
>>         Date: 2007-02-14 20:51:56 +0000 (Wed, 14 Feb 2007)
>> New Revision: 12794
>>    Changeset: http://cvs.mythtv.org/trac/changeset/12794
>> Modified:
>>    branches/release-0-20-fixes/mythtv/libs/libmyth/mythcontext.h
>> Log:
>> Update version date string, lots has changed since August of 2006...
> We should really do this a different way.  This change makes it look
> like the 0.20 release is a newer binary version that SVN HEAD from  
> trunk.
> Trunk has this:
> 	#define MYTH_BINARY_VERSION "0.20.20070124-1"
> while -fixes now has the following whic makes -fixes look newer:
> 	#define MYTH_BINARY_VERSION "0.20.20070214-3"
> Normally we only change the binary version when the libmyth binary
> compatibility changes.  If we want to differentiate this in the -fixes
> branch, perhaps we should change the number at the end to -fixes.
> Usually this number is used just in case we change the binary  
> version more
> than once in one day.

Yeah, I realized about 2 seconds after committing that it was a  
stupid idea. Snow-Man was after someone to commit something to make  
sure some changes to the commits list hadn't broken things, and I  
stupidly figured "why not change this!" without reading the comments  
above it.

However, I think it'd be good if this were somehow populated with a  
date from the most recent svn commit in the branch, in conjunction  
with the binary version in trunk nudged up to 0.21.x-y. I remember  
firing up a patched build the other day and noting the August date  
and thinking that it really should be updated (which led to the commit).

We can certainly revert that change though.

Bad me, no biscuit...

Jarod Wilson
jarod at wilsonet.com

More information about the mythtv-dev mailing list