Dave Reid

Note: this is a serious post, but I had to have a little fun and wanted to parody Barack Obama's 2009 State of the Union speech.

I've come here today not only to address the distinguished men and women in this great community, but to speak frankly and directly about everyone's favorite Drupal module, XML sitemap. I know that for many Drupalers reading right now, the state of the module is a concern. And rightly so. If you haven't been personally affected by this module, you probably know someone who has--a friend; a neighbor; a co-worker. But while the module's reputation may be weakened and our confidence shaken, tonight I want every Drupaler to know this: We will rebuild, we will recover, and the XML sitemap module will emerge stronger than before.

So how did things get this bad?

For those of you who haven't had the pleasure of using XML sitemap yet, I think Angie Byron (webchick) and Jeff Eaton said it best:

webchick: "Dear XMLSiteMap module: Please die. And not in that cute, friendly way. I mean I literally hope you get hit by a bus. Twice."
eaton: "Uhoh. Brainstorming a #drupal site assessment drinking game with @quicksketch ... "XMLSitemap? Finish your drink."

The last stable/official release of the module was the 5.x-1.6 version from almost a year ago. There is a 5.x-2.x version that was supposed to fix problems in 5.x-1.6. For over a year, the 6.x-1.x-dev version has been a work-in progress port of the unfinished 5.x-2.x version. Despite there not being any stable 6.x release, there are over 10,000 people using development builds which are buggy, frequently failed on updates, and could not scale to large sites. This of course, leads to several issues in the queue with over 200 replies with recurring themes of frustration and impatience.

Much like the current US/world recession, I can't assign blame to any one person or problem. The current maintainer, Kiam was very active and would keep hacking away at the module, fixing problems, but sometimes introducing new ones. For the 25th most popular Drupal module, there was a lack of regulation (peer code review) that we usually see in other popular or large modules. There was no way to say that a change won't work, or something is a bad idea until after the code had been changed. Some developers that needed a working module for their clients would take one look at the code and issue queue and would promptly run away, seeing no other short-term solution besides writing their own custom implementation. The people who's input and time could help the most did not need or want to help contribute back. Those who did make honest efforts were frustrated with language barriers or misunderstandings. There was no test coverage for the module, which I also can't blame since writing SimpleTests for an existing, largish module which you didn't originally write is really hard. But new features would be added and bugs would be fixed without ensuring that they would work correctly.

Yikes! So were are we now?

I got just as frustrated and upset as every one else. And like some of the developers before me, I went a rewrote the module how I thought it should be done. I took careful planning in a lot of the major design decisions so that the module could avoid the architectural and scaling problems. I debated releasing the module as a duplicate project or keeping it just for my personal use. But I decided to put the code into my personal CVS sandbox and then shared it with a few people via e-mail and on IRC with what I thought were the major problems that needed to be fixed, how to fix them, and wondering where do we go from there. I need to say major thanks to Angie, Greg Dunlap (heyrocker), and caktux who helped test, inspect, and review my sandbox code as well as made sure I could explain and justify why this code should move forward.

So now at this point we've got some code with potential, but I was still unsure on how to proceed. Do we forcefully replace the maintainer of a module? That could lead to messy results and hurt users. Do we fork the module? That could result in confused end-users and two similar projects. Angie, the awesome and fantastic person that she is, stepped up and helped open communication between the module's original maintainer, Darren Oh, Kiam, and myself. After lots of discussion, we have all agreed on the following plan:

1. Kiam, assisted by Earnie, will continue to work on getting a stable, bug-free 6.x-1.x version released. They have both done a great job so far and made a lot of progress in the last week. A release candidate could possibly be out by the end of the week.
2. I have been adopted as the official maintainer of the 6.x-2.x branch, which is a complete rewrite based on the sandboxed code. So that problems can be fixed efficiently and correctly, there is currently no upgrade path to this version.
3. In order to concentrate on a bug-free release, the 6.x-1.x branch is no longer accepting feature requests. As soon as the 6.x-2.x branch becomes stable and has an upgrade path, the 6.x-1.x branch will be declared unsupported.

Ok, so where are we headed?

• Regulation! With three active maintainers (not including Darren Oh), we can help each other out with reviewing code, ideas, bus tests, etc.
• Cutbacks! In an effort to focus on the 80% of features that most everyone uses, some advanced features will probably be cut or removed in favor of integration with other modules that could do a better job.
• Safety! The 6.x-2.x branch will aim to have 100% test coverage. No new features added or bugs fixed without a test. Sound familiar? :)
• Leadership! Since I'm active in keeping up with Drupal core development, I will have XML sitemap ported to Drupal 7 and stable before Drupal 7.0 is officially released. Hopefully other major modules will follow suit and not languish back like the mess we had porting from Drupal 5 to 6.

Awesome! So how can I help?

I'm so glad you asked! There are four ways you can help:

1. Financially: The 6.x-2.x rewrite has already taken a much larger chunk of my time than I planned. And there is a lot more work to go. If you or your company would like to sponsor development of the rewrite or in general, please contact me. If you'd like to help donate to the rewrite efforts, I have opened a ChipIn at http://davereid.chipin.com/ where you can show your support. I will personally thank and attribute every single person that donates. And no, I won't stop working if no one donates, but if people show support, it is additional incentive to work a little harder, longer or faster.
2. Testing: We need people verifying bugs and testing patches in the issue queue for the 6.x-1.x version, and the 6.x-2.x version if you're feeling brave. Read http://drupal.org/node/314199 if you need help getting started. If you're testing the 6.x-2.x version, please remember that there is no upgrade path and that you must run it as a clean install. It's also incomplete, with only the common and menu modules "converted", but as a sign of my confidence or ignorance, I'm using it here on my own blog.
3. Documentation: If you like to edit or write, we'd love your help improving the module's documentation. Here's a list of the current documentation tasks, including the big wall-of-text on the project page, and out of date handbook pages. We're getting too many support requests for known issues that could be better documented. I should see if I can just steal add1sun away for a while.
4. Translation: I'd like to aim to have official releases of the module translated at least 10 major languages. Hopefully we can add a translation server in the drupal.org redesign to help ease this process of contributing.

In summary, we are not quitters. If we come together and lift this module from the depths of this crisis; if we confront without fear the challenges of our time and summon that enduring spirit of a community that does not quit, then someday years from now our children can tell their children that this was the time when we performed something worthy to be remembered. Thank you and long live Drupal!