A guide to getting involved with SQLAlchemy.
A key to participation is being tuned in to the project's current status. The current released version of SQLAlchemy is always at the top of the homepage on the right hand side. Usually there are two branches released at a time, such as "0.6" and "0.7". The higher number, i.e. "0.7", is the "default" branch, and "0.6" is the "maintenance" branch. The movement to a new pair of major branches spans around 12-18 months. Releases within each branch are between one and two months apart. At some point there will be a 1.0 release. But our approach of moving slowly and steadily has so far been very beneficial to the project, allowing great strides in architecture and usage before a larger pool of new users is taken on.
The overall status of development can be gleaned by viewing the Roadmap in Trac. We try to assign all bugs and new features to a specific version. A specific number, like "0.6.8", means we'd like to get the ticket completed as of that version (though this is not always guaranteed). An "open ended" number like "0.7.xx" means the ticket is "in the queue", but is not determined to be part of any specific version. A lot of these tickets are of lower priority, some very involved and tedious as well, and are pushed along major releases, in some cases several times. You can help out with some of them!
Major new developmental initiatives are also discussed on the development list. This list has historically been a little dead but we hope to revive it. New releases are always announced on the main mailing list, and more recently on the SQLAlchemy Blog, recently upgraded to allow a higher volume of posts. Occasional more in-depth posts can also be found at techspot.zzzeek.org.
SQLAlchemy uses Trac for bug reporting and issue tracking, Wiki pages, and source browsing. Trac now uses a registration system, so create an account via the "registration" link to log in and participate in tickets.
Bugs are reported using Trac at the new ticket link (note you have to be logged in to use it).
Bugs are generally reported only against the two most recent release branches, which comprise the default branch and the maintenance branch. Branches prior to these are rarely modified. Bug reports regarding library behavior as well as documentation issues are reported in nearly equal measure.
When reporting a behavioral bug we ask that you:
- Ensure that the issue is still present on the most recent released version of SQLAlchemy within either the default branch or the maintance branch, and preferably against the latest development tip of the series in use. SQLAlchemy provides easy download links of the most recent development versions on the download page.
- Check that the issue doesn't already exist in Trac. This can be a little tricky as Trac's search interface isn't the best, but we ask that you give it a try.
- Only report issues that you're pretty sure is a bug. It is perfectly OK to report a usage issue on the mailing list that may or may not be a bug - we'll escalate bugs to the bug tracker if it really is a bug.
- Create a succinct test case which reproduces the issue. This needs to be a script that we can actually run - so it should not require any imports that the SQLAlchemy developers don't have access to, and in the vast majority of cases should not have any imports outside of SQLAlchemy itself. It needs to include whatever table definitions and data are required to reproduce the issue. While we have access to most database backends, SQLite is preferred unless the issue is specific to a certain backend.
- Tell us exactly what version of SQLAlchemy the issue is being observed with, as well as details about the database in use. Often, an issue has already been fixed in the upcoming series and could not be backported.
- As you'll create an account to post a bug, updates on the ticket will be sent to the email address you register with. Feel free to add other addresses as CCs to the issue.
- Monitor the issue. Lots of times the developers have immediate followup questions, and might even have patches for you to try in case the issue is hard to reproduce. Please keep the conversation going until we can get the bug fixed !
Ready? Let's report a bug!
SQLAlchemy always needs people to help answer questions, particularly from new users.
The Mailing list is active every day, and can sometimes receive a dozen new users in one week. Help with fielding new users and questions on the list is always appreciated ! Regular users can also volunteer to help with approval of new posters.
The IRC channel
#sqlalchemy on Freenode is similarly busy most days.
There's a core set of "regulars" who are pretty friendly, and more helpers are always welcome.
SQLAlchemy places great emphasis on polite, thoughtful, and constructive communication between users and developers. Rudeness, personal insults, or overly brusque answers are never appropriate, even for users with unreasonable requests. We also try to ensure that no message on the mailing list goes unanswered, even if the answer is simply to politely direct the user towards the appropriate section of documentation. The core SQLAlchemy developers would like to encourage all users to help with this task - if you see a very basic question sitting on the list for a few days, that's us hoping you'll respond to it ! You have our permission :).
We love for people to spread the word about SQLAlchemy. Some common venues:
- Twitter - we get a lot of comments good and bad via Twitter - if the tweet contains "sqlalchemy" or a link to the site, it will show up under searches for the term. Expect SQLAlchemy developers to respond to provocative tweets!
- The "Organizations Using SQLAlchemy" page lists prominent and not-so-prominent organizations who make use of SQLAlchemy. Please contact us to have your organization added here.
- Blog Posts - Users often write short tutorials or blog posts describing how to achieve something with SQLAlchemy. It all becomes part of the larger collection of knowledge so share freely!
- User group presentations - lots of users present SQLAlchemy, or an application that SQLAlchemy was instrumental towards, to their local Python group or other software development group. Put your slides up on the web after the presentation!