forrest-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Ross Gardler <rgard...@apache.org>
Subject Re: use of whiteboard in forrest
Date Mon, 29 Aug 2005 10:16:31 GMT
Thorsten Scherler wrote:
> On Sat, 2005-08-27 at 14:51 +0100, Ross Gardler wrote:
> 
>>David Crossley wrote:
>>
>>>Thorsten Scherler wrote:
>>>
>>>
>>>>c) refactoring of "old" code into a new version which would break the
>>>>usability of the old code while refactoring.
>>>
>>>
>>>I think that would be better done in a branch.
>>
>>
>>+1
> 
> 
> Actually I was referring to your example of refactoring the Daisy
> plugin. I have chosen different wording I have to admit.

The "refactoring" of the Daisy plugin is a good example of code that 
should be done in a branch. I broke some functionality of the 0.1 plugin 
during this refactoring whilst I worked out how it should be done. In 
this case it was done in the locationmap branch because it used the 
locationmap. If I'd done it in core then it would not have always been 
releasable.

If it had been a straight refactoring (tidying of code) I would agree 
with you though.

NOTE: If we agree to move all plugins out into a separate repo module we 
will be able to branch individual plugins.

> IMO branches should only be used if necessary. Refactoring code not
> always have to be done in branches. If they are components that can be
> capsuled then they should be refactored in "incubation". That is more
> efficient (I do not have to check out a whole branch).

Plugins cannot have a version in whiteboard/incubation and one in the 
core - their names would clash. This is another good reason to move 
plugins into their own repo, we can then branch just that plugin code 
and not the whole tree.

> Core code normally do not allow that, another good reason to try to make
> the core as small as possible. 

+1

Ross

Mime
View raw message