On Thu, Jul 14, 2011 at 6:18 AM, Ryan Schmidt <[email protected]> wrote:
>
> On Jul 13, 2011, at 23:48, [email protected] wrote:
>
>> Revision: 80511
>>          http://trac.macports.org/changeset/80511
>> Author:   [email protected]
>> Date:     2011-07-13 21:48:08 -0700 (Wed, 13 Jul 2011)
>> Log Message:
>> -----------
>> Initial commit of statistics collection Ruby on Rails app.
>
> This is a lot of files. I would have expected to see many more smaller 
> commits instead of one huge one. Commit early and commit often, so that 
> individual logical changes appear as individual commits and can be reviewed 
> on their own merits, and errors can be spotted sooner. I'll guess nobody is 
> going to review this large commit, so errors might be overlooked.
>

Thanks for the advice, I'll keep my next commits small and focussed. I
think the large commit came about as a result of me wanting to have
the basics of the app in place which ultimately led to many files
being created and edited. A better approach would have been to commit
an empty Rails project and then make smaller commits as I implement
features.

>
>> - Seed database with all ports and categories (seeds.rb) - data comes from 
>> mpwa script
>
> This file is enormous. The commit mail was over 7MB, mostly due to this file, 
> which as you say contains a database of all ports and categories. Naturally, 
> such a database will quickly be out of date. Why does it need to be committed 
> at all? Isn't it dynamically generated somehow?
>

That's a very good point, the file is dynamically generated and should
never have been included in the commit. My apologies for the large
email.

Derek
_______________________________________________
macports-dev mailing list
[email protected]
http://lists.macosforge.org/mailman/listinfo.cgi/macports-dev

Reply via email to