Hey kids You might have noticed I haven’t…

Hey kids. You might have noticed I haven’t been as vocal as usual since we started 3.5, and I wanted to update you on why that is. I’m going to be taking a leave of absence from Automattic for a couple of months, and stepping back from core UI work in 3.5 at the same time. Though normally a leave would mean unplugging completely, I’m going to be continuing to work on bringing together the various contributor groups, and have an upper limit of 10 hours per week for this. (If I start edging into more time than that, my HR person will be severely disappointed in me!)

As core is also a contributor group (obviously), I’ll be available within that 10-hour limit to consult on ux/ui matters. What this means is that I won’t be doing wireframes, coming up with original designs, or project managing the dev group as I have in the past. What I will be doing is providing perspective and gut checks when asked on the designs and decisions that others come up with. I’ve talked with Koop about this, and I thought about making a flow chart for “Should we ping Jane?” but decided not to bother (even though it would have been funny).

Basically, if you want me to look and give an opinion (sometimes that opinion might come with a design suggestion, sometimes not) and/or a “Did you think about this?” check to be sure you haven’t overlooked any repercussions the new design may have, go ahead and ping me. I won’t be online as much as usual, so my response time likely won’t be immediate. Also, I will not be reading the firehose of Trac emails (gasp!) or checking the ux-feedback report or hanging out by default in the IRC channels while I’m on leave, so if there’s a ticket you want me to weigh in on, a ping would be good since I won’t catch it on my own.

Timing: the leave starts right after WCSF in early August, and ends October shortly before the community summit (of which the need to start planning is part of why I’m stepping back from core for a bit). So if you need to nab me during that time, email jane at wordpress.org or hit me on Skype if it says I’m online. Thanks, and have a great dev cycle!

#3-5, #availability

Agenda for July 18, 2012…

The agenda for today’s weekly project meeting:

  • Affirm deadlines and scope. Check in on the proposed feature scope for 3.5.
  • Teams. Form an initial set of feature development teams. This is about establishing responsibility for each feature, focusing ourselves on our goals, and making sure we have enough volunteers across the various projects. You don’t need to be at this meeting to “join” a team, nor do you need to be on a team to contribute.
  • APIs. Discuss potential platform and API improvements, particularly those that tie into the proposed features. (For example, ImageMagick support would be a great pitch, given changes to image handling.) XML-RPC and other perennial improvements should also be discussed.
  • Unit tests. What’s changed, and what our goals should be.

#3-5, #agenda, #dev-chat

Housekeeping

I’ll be posting a summary soon that covers Wednesday’s marathon meeting to scope out features for 3.5. In the meantime, some housekeeping:

make/core

This P2 blog has moved from wpdevel.wordpress.com to make.wordpress.org/core. The “make” network is still very young, but there are other P2 blogs already underway, including ui, accessibility, themes, and plugins. wppolyglots.wordpress.com also moved to make/polyglots.

Everything was migrated, including email and jabber subscriptions (using Jetpack). Being on the WordPress.org network opens up some possibilities, including custom features, better integration, and single sign-on.

New test framework

There will definitely be more to come on this, but in the last two weeks, the test suite was converted to a new test runner. You can read more about the effort on @maxcutler‘s blog. Tests are now easier and more straightforward to write, and the runner is also faster, leaner, and more stable. We’ve been working to increase our test coverage with every core commit, so this move was really important.

Unit tests and mailing lists

With the new test framework, we’re also looking to raise the visibility of our tests. We do plan to merge them into core’s Subversion repository in the future, but for now, we’ve merged some mailing lists. The wp-svn mailing list (every core commit, right in your inbox) now receives commits to the unit-tests repository as well. And wp-trac (every Trac ticket and comment — “the WordPress firehose”) now receives comments from the unit-tests Trac.

If you are mostly just interested in tests, the wp-unit-tests mailing list receives both commits and Trac notifications, as before. (Also, make yourself known!)

Daily Bug Scrubs

I’d like to continue having daily “office hours.” For now, we’ll continue them weekdays at 19:00 UTC (an hour before the dev chat usually is). A number of us are idling in IRC throughout the week anyway, but I think it has worked well to have a set time where you can stop by to help comb through Trac, bring up tickets for discussion, and pitch patches.

 

 

#3-5, #housekeeping, #unit-tests

Recognition, and news about the 3.5 cycle

Before we kick off 3.5 development tomorrow with the scope session, there are a few quick announcements!

Andrew Nacin’s prolific contributions, encyclopedic knowledge of the codebase, and his increasing development leadership have undoubtedly been known to you all. We’re now recognizing that leadership by promoting Nacin to a Lead Developer!

Daryl Koopersmith has had temporary commit for the past few release cycles, and should be known to you for his work on nifty features such as the theme customizer, distraction-free writing, and our fast-fast-fast linking dialog. We want him to keep bringing it for many releases to come, so we’re dropping the “temporary” bit and letting him keep ongoing commit access.

Next, Jon Cave (“duck_”) is having his temporary commit extended once again for the 3.5 cycle. Jon’s keen eye has been quite helpful, especially when it comes to improving WordPress security.

Finally, we’re trying something a little bit different for the 3.5 cycle. We’re going to have a specific release leader — someone who has both primary authority and primary responsibility for the release. They’ll lead the scope discussion, they’ll coördinate the development process, they’ll crack the whip when people aren’t delivering on things they promised, and they’ll make the hard decisions about whether features stick or get punted. If this works, we can make this a rotating gig.

For the 3.5 cycle, Andrew Nacin will take the lead, and Daryl Koopersmith will be his second-in-command. These are your point people for this release. I, for one, am excited to see what they (and you) can accomplish!

#3-5, #release-lead

Version 3.5 Scope Session on Wednesday, July 11

Per our previous dev meeting, the 3.5 feature scope session will be taking place at 20:00 UTC on July 11 (tomorrow). (#wordpress-dev on irc.freenode.net — this is our regular place and time; see the sidebar.)

These meetings usually go long (1.5-2 hours), so this one will focus primarily on determining deadlines and planning out features. API/platform development (you know, the “Under the Hood” stuff) happens organically throughout the cycle anyway, so we may push those conversations to next week. (We’ve already been focusing on quite a few enhancements and bugs over the last few weeks thanks to near-daily bug scrubs, which I want to continue.)

Come armed with ideas and an open mind — see you tomorrow!

#3-5, #agenda, #dev-chat, #scope

As Twenty Twelve is punted to 3.5 it…

As Twenty Twelve is punted to 3.5, it is being removed from core. It can be found here for now, and will be brought back in for 3.5.

#3-5, #bundled-theme, #twentytwelve

GSoC/Summer/3.5

Normally we don’t start talking about the next release until the current one is out the door, or at least in beta/RC, so this post jumps the gun a bit, but for a good reason: the GSoC deadline. There are two approaches we could take toward our participation in GSoC this year, and one of them is tied closely to 3.5.

Historically

  • Good GSoC mentoring takes time. Time is hard to come by at the best of times, even harder for many during the summer.
  • Many of our previous GSoC mentors have held the position for several years and could use a break from trying to mentor while simultaneously working on features for a regular release.
  • Almost none of our GSoC projects have actually made it into core. A few because they were plugins, but most because once GSoC is over there hasn’t been a concerted effort to follow up on these projects.
  • We often run late on dev cycles.

Since 3.4

  • We have ramped up several core contributors to more responsible/trusted roles as a result of the 3.4 process experiment (teams, cycles, updates, etc). This could mean more mentors.
  • We are running late in our dev cycle, and with SXSW about to eat a week, I’m thinking we’re about to get even more behind. My guess is we’re looking at a May launch, not April.
  • The stated intention of having all feature dev for the cycle tied to a central goal of making it easier to customize your site didn’t really happen. There were at least 3 teams working on features that had nothing to do with this, and another couple that were related, but not smack in the middle of it. Good features all, but we failed in sticking to that goal as a unifying concept.

Proposal

What if for 3.5, instead of it being a “regular” cycle, we made it a mentoring cycle tied to the GSoC schedule (shorter than normal)? If we assume 3.4 will launch sometime at the end of April or early May (and if it does happen earlier, awesome), that would put us in a position to start working on 3.5 right when the GSoC accepted students are announced.

If we chose a “release concept” (like customizing your theme, but something different) and outlined every feature/enhancement/bug that’s related, we could make those things be the potential GSoC projects. We could work in teams like in 3.4, but in this case each team would have a student or two working on things with them closely. Since these would be the only features being worked on (additional bug-fixing always ongoing, obviously),

  • Students would be guaranteed mentor attention and working with core
  • We would be more likely to do the work necessary to get student work to commit-worthy status
  • We would target a launch for late August to coincide with the end of GSoC (so we could do one more small release before end of year)
  • We could do additional outreach to include new contributors who do not qualify for GSoC (too young, too old, not in college, etc), improve our mentoring skills and processes
  • At the end of this mentorship-focused summer, we would not only have the features developed by mentees, but we would have an ideal pool of people to help us create documentation to help new contributors.

I’m thinking that what might make sense would be for there to be a team or two that doesn’t mentor or work on a feature for 3.5, but begins working on one of the more complex things we keep putting off, so that it could be the first thing into 3.6 (like gallery management or something similar).

Deciding on a release concept that could be done in a 2.5-3 month cycle would be important. I’m thinking maybe it could be the feedback loop — improving comments and communication with readers via html emails, forms, etc on the front end and a UI facelift of the comments/related screens on the back end, putting something cool into Twenty Twelve around this (or just support for something in core related to same), etc. There are a number of projects around this that have been done in the past that could be looked to for inspiration and/or what not to do, it’s needed attention for some time, and it’s not as complicated as something like media or multisite.

Thoughts? Specifically, thoughts on:

  • Doing a mentorship-focused release timed to GSoC
  • Potential areas of focus for 3.5 if we were to do this
  • Mentoring in teams like 3.4
  • Wanting to mentor in this case
  • How many students you think we could take on if we used teams like in 3.4

Comment here today, and tomorrow I’l round up the core team to see what people think based on the conversation so we can make a decision and I can update our application before the application deadline if needed. If we don’t do something like this, then I’m planning on reducing our GSoC student allotment to 5-6 students (we’ve asked for up to 15 in the past) to ensure enough mentors and adequate attention/follow-up on projects.

Thanks for your input!

#3-4, #3-5, #gsoc, #mentorship