Le 08/04/12 14:44, Daniel Friesen wrote :
We're using gerrit now. And as we use it we all keep running into issues stemming fundamentally from how gerrit is designed.
<snip>
Hello Daniel,
We already had the experience of writing a code review tool from scratch and although it filled a need at that time, I am not going to regret it.
I do not think the foundation should fund the project of writing a new app from scratch which probably nobody else will use anyway and based on a niche Framework which is MediaWiki. Instead we should stick to reuses existing tools and have them enhanced to fit our needs, writing a whole new app should be a solution of last resort.
Before starting to write code in a rush, your really first step should be to find out what is *functionally* wrong with Gerrit and find out what we need. I have yet to find out anyone describing what is wrong with Gerrit although we all have a sentiment that the UI is confusing and some obvious technical issues. They do not justify a rewrite in my opinion.
Once you have a list of requirements, though, we could look at the possibility of fixing Gerrit to fit our needs. If that is not possible, we could use that list as a basis to find out another tool. You could coordinate with John Du Hart who has been working on Phabricator, a suite of tools from Facebook which include a code review tool as I understand it. Other examples that comes to mind are GitHub and Gitorious.
Meanwhile, dont waste your time coding something we are most certainly never going to use.
cheers,