harmony-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Geir Magnusson Jr." <ge...@apache.org>
Subject Re: Where do we put it? (Was Re: [jchevm] Porting JCHEVM to OSX/PPC)
Date Wed, 07 Dec 2005 05:35:19 GMT

On Dec 5, 2005, at 11:48 AM, Archie Cobbs wrote:

> Geir Magnusson Jr. wrote:
>> On Dec 5, 2005, at 10:17 AM, Archie Cobbs wrote:
>>> Geir Magnusson Jr. wrote:
>>>>>>> I stripped out lots of ELF specific stuff, so my version  
>>>>>>> could be
>>>>>>> interesting for a port to windows too. I just don't want to 
>>>>>>> commit it
>>>>>>> to "trunk" yet because it's completely untested.
>>>>>> You could create a branch to play with in Subversion. If you do,
>>>>>> I recommend also using svnmerge... http://dellroad.org/svnmerge
>>>>> Good Idea, but where would I place that branch?
>>>>> "harmony/enhanced/branches/sandbox/contribs/jchevm/osx_port"?
>>>>> I'll look at the rest of the issues later...
>>>> Is there a way to safely get this code in the main trunk?  I   
>>>> don't  mind (what about others?) if you need to work in the  
>>>> corner  of the  sandbox, but this seems like enhancement to the  
>>>> trunk  rather than a  revolutionary experiment...
>>> That's OK with me, but since it's still "completely untested" it  
>>> seems
>>> like a branch may be better... but in any case, if there is some   
>>> uncertainy,
>>> why not create a branch? Branches are essentially zero cost with   
>>> Subversion,
>>> even if it's only for a short while to get things "in order", etc.
>> How about testing it? :)
> Sure.. but for that to happen, the patches have to be committed  
> somewhere
> first (let's try to avoid emailing patches around and use SVN  
> instead).

The developer should test locally first.  If it's a matter of the  
developer not having the platform to test on, that's another story.

>> I believe that the changes shouldn't break jcheVM on platforms  
>> other  than OSX/PPC, but if they don't work yet on OSX/PPC, that's  
>> fine.    If that can't be done - if David can't do that, then  
>> maybe a sandbox  is right, but at some point, it *has* to come  
>> back to the tree, which  means that it doesn't break the existing  
>> code, and I think that if we  can do that now in the trunk that's  
>> better.
> OK, that's fine with me.
>> We'd like to have a culture of "check in stuff that doesn't break   
>> others", but "commit early and often so others can see and work  
>> with  what you are doing"
> I couldn't agree more. This is exactly what branches allow you to  
> do :-)

Yes, but then you aren't working together on the codebase :)


Geir Magnusson Jr                                  +1-203-665-6437

View raw message