2006/07/30

Migrating S3-Objc to Google Code Hosting (sort of)

I created a project on Google Code Hosting to eventually migrate all of my OS X / S3 code there. Well, that was the excuse. What I really wanted was to test their code hosting service.

The administration interface is straightforward, and gives an good impression of avoiding limitations (unlimited links/groups/blogs) while remaining clean and consistent. The issue tracker is ok. I'm not 100% fan of the let's-edit-predefined-values-in-a-textfield-with-a-basic-syntax. It's working for simple fields (open/close issues status values) but is a bit awkward in the predefined issue labels section than can be restricted in arity based on the label prefix (confusing ? told you so).

Using the issue tracker itself is ok, with some strange GUI decisions (it's not 100% obvious that just to change status or labels you must use the "add comment" section). The "typed", comma-separated labels (for instance Type-Enhancement, Priority-Medium, Priority-High) are not optimal for readability.

Even if it's a tiny project with a very small number of revision, I think it's not acceptable to loose revision history. So the source code is not yet imported, and will be only when Google implements svn dump import. It's also extremely negative that "At this time, there aren't any import or export features. But look for them as the service matures.". This is locking, period.

I sincerely hope this is just a start, because it feels really primitive. I understand the big plus is the huge scalability that Google is able to offer, but from a project "management" point of view, Google Code Hosting is missing almost all of the common tools really useful in other tools: Wiki, Web/SVN browser, RSS timelines, commit/ticket integration (a la trac), file releases, ...


Comments: Post a Comment

<< Home

This page is powered by Blogger. Isn't yours?