MASON-INTEREST-L Archives

June 2015

MASON-INTEREST-L@LISTSERV.GMU.EDU

Options: Use Monospaced Font
Show HTML Part by Default
Show All Mail Headers

Message: [<< First] [< Prev] [Next >] [Last >>]
Topic: [<< First] [< Prev] [Next >] [Last >>]
Author: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Subject:
From:
Luís de Sousa <[log in to unmask]>
Reply To:
MASON Multiagent Simulation Toolkit <[log in to unmask]>
Date:
Thu, 4 Jun 2015 22:07:22 +0200
Content-Type:
text/plain
Parts/Attachments:
text/plain (32 lines)
Hi Sean, hi all,

First of all let me just express my joy for this decision. I believe
it can create an interesting dynamic around MASON.

Regarding your question, I do not have the experience myself, but it
seems the folk at GitHub have taken some care to guarantee things go
smoothly. They enforce a particular repository structure clearly
defining where branches and tags  reside:

https://github.com/blog/966-improved-subversion-client-support

Thus all svn actions should unequivocally map to a git action (or set
of actions), avoiding unexpected repository states.

Naturally, svn lacks loads of interesting stuff existing in git, such
as stashing, sub-modules, etc. If one of these modern things is pushed
to GitHub, then whomever accesses the repository with svn might start
missing some of the fun.

Regards,

Luís


On 4 June 2015 at 15:21, Sean Luke <[log in to unmask]> wrote:
> So I'm thinking we may migrate ecj and mason to GitHub, which has dual git/SVN access.  I'll probably still maintain things in SVN, but if you're into git you can access the repositories that way.
>
> I imagine read/write from SVN and *read* from git works fine.  But does anyone with experience in this model know how well doing read/write from both git and svn works in reality?
>
> Sean

ATOM RSS1 RSS2