Re: Proposal: Switch back to savannah using GIT

[Richard Frith-Macdonald]

On 29 May 2015, at 14:31, David Chisnall <address@hidden> wrote:
> 
> On 29 May 2015, at 13:18, Richard Frith-Macdonald <address@hidden> wrote:
>> 
>>> Oh, and this doesn’t actually do what I do (and what is best practice to 
>>> do) to ensure a clean release tarball, which is tar up the result of svn 
>>> export, to ensure that the tarball exactly matches the contents of svn.  
>>> Instead, it deletes some files, hoping that it gets everything.
>> 
>> Actually this is FUD (untrue/mistaken); gnustep-make *does* do an svn export 
>> to produce a clean tar ball (as far as I know it always ... since we moved 
>> to svn ... has).
> 
> No, it is true.  I tested it before saying it, because I’d not used make dist 
> before.  I did a clean checkout, ran configure, modified a random file, and 
> did gmake dist.  The tarball I got contained my random modification.

Sorry, I missed that you were shifting the discussion to something different ...
Riccardo wasn't particularly talking about 'make dist' (he mentioned it as a 
'quickie', but then make it clear he was primarily talking about the commands 
to make a formal distribution).

> And this completely misses the point of the rest of my email which is *as 
> someone who packages GNUstep and GNUstep-using apps*, being able to 
> automatically get a tarball for any revision is invaluable.  make dist does 
> not do this unless I want to check out that revision and construct (and host) 
> the tarball myself.

That's like complaining that 'ls foo' doesn't remove  the file 'foo' 
Of course it doesn't because it's not the command to do that!
If you use the command to make a distribution/package of your current local 
work, then of course that's what you get.

  make svn-dist
gives the latest tagged release

  make svn-snapshot
gives the latest committed version

  make diet
gives your uncommitted version ... useful for moving code around for testing on 
other systems before committing