Grants talk:Project/Timeless/Post-deployment support/Midpoint

Report accepted

edit

Dear Isarra,

Thank you for submitting this Midpoint Report. I am accepting it now with the following comments:

  • "The enormous backlog of bugs, feedback, and issues to consider across various projects has been generally consolidated into a single location: the Timeless phabricator workboard, with more feedback constantly incoming."
    • Congratulations on getting everything organized and centralized. I imagine this was a whole lot of work!
  • "Improvements and compatibility fixes from Timeless and a few other skins have been consolidated into the Example skin so that all new skins made from it should now have much better out-of-the-box support for expected features."
    • This is great. Thanks for thinking beyond Timeless.
  • "Development has improved the internal architecture of the Timeless skin to just generally break things less. It's still pretty bad."
    • Would you say this is in line with what you expected to be true at this point in the project? Or have things gone differently than anticipated?
  • Since you wrote this report, has there been any movement with the RfC? How likely is this to move forward, from what you can tell? If it does not go forward, how will it impact your project?
  • "getting back into the mind to compile and report on things after the fact has been difficult"
    • For what it is worth, this feeling about report-writing is shared by many (most?), not just developers. And for waht it's worth, you did a lovely job with this report.
  • "Much of the code that I am working with is very bad. Some of it is my fault, some of it is just ancient, and some is the result of various conflicting viewpoints on what is proper being put down to the repositories. As much as I knew about this going in, actually working with it is a whole other matter. This work has to be done on some level, however, so the only solution is simply to continue to move forward and address what is presently addressable, and make note of what is not. Resolving the underlying issues will be a matter for a later project."
    • This sounds like the heart of the matter, in terms of the challenges of this project. When we talk, let's explore what this means for your work on Timeless.
  • "I have encountered various delays with the project due to personal health issues, which were particularly problematic as I am just one person: first illness, and later suffering a concussion. I would rather prefer to avoid repeating any of this, because getting started again after two months away from any project is difficult enough even without the side effects of a head injury, no matter how understanding the affected communities might be."
    • Isarra, I'm so sorry to hear what you have been going through (so much!) with your physical well-being. Of course, this must have been very challenging. Again, let's check in about this a little more when we talk.
  • A very wise learning pattern that I think all of us could stand to remind ourselves about now and again.
  • "I've been very impressed by how understanding and downright friendly the editing communities have been in light of all the delays, bugs, and random setbacks that have occurred so far in this project"
    • This is so nice to hear--music to my ears, I would even say. :-)

Looking forward to speaking with you more about this project. Thank you for all of the work you've done so far.

Kind regards,

Marti (WMF) (talk) 01:34, 16 March 2019 (UTC)Reply

@Mjohnson (WMF): Thanks for the feedback and questions, and thanks for your patience on this, between this cold and even more travel (conferences! Yay?). Anyway, trying to figure out how best to reply, I'll just highly specific things like you did:
  • "Thanks for thinking beyond Timeless."
    Well, this was always about that - a specific example from which to figure out where we stand and where, perhaps, we should be going. I still worry that even with this I'm remaining too close to where we're currently stuck, as opposed to where we should be going with it. But this relates to the following, too:
  • "Has there been any movement with the RfC? How likely is this to move forward, from what you can tell? If it does not go forward, how will it impact your project?"
    That RfC, no - it seems to be stuck, and I'm really not sure how to unstick it. While this is annoying, and means I'll lack the dependency to test out one of the larger proposed features here (although I could implement it anyway, deploy it to Uncyclopedia, and then just tell everyone there to try it out instead, which could at least get some interesting feedback), it's not really a blocker, particularly. More just... no cool toys for Wikimedians for the time being, so much for the community wishlist requests.
    BUT: As I mentioned in the call, there is another RfC, and this one is much more ambitious and perhaps really what we should have been going for in the first place: an inheritance system for skinning templating that will basically replace the entire current stack. Skizzerz is handling the bulk of this one, where Jack Phoenix and I are providing most of the actual specifics when it comes to what we need from skins, and we just discussed it with techcom today, and it... might actually go somewhere. Continuing with this project will also help us to determine what we will need to be supporting with the initial prototypes to implement the RfC properly for moving forward, so this, all in all, seems to be pretty decent timing.
  • About the architecture: "Would you say this is in line with what you expected to be true at this point in the project? Or have things gone differently than anticipated?"
    Well... yes, I kind of did expect to be further along. And I didn't particularly expect most of the fixes to consist simply of removing things. But really, that's good - it means we're coming together between various skins to be more consistent, because the less any of them have to do, the better, and the less that differs between them, the fewer compatibility problems they all have.
  • Bad code and such: "This sounds like the heart of the matter, in terms of the challenges of this project."
    Well, yes and no. I don't know how much I got into this, but the code is bad... but it's bad at a fairly ordinary level. It's soup, and it's disorganised, but it's actually a bit like what we found with WikiProjects. Because where they're just articles written in wikitext and thus only as useful as the people writing that wikitext make them, skins are html soup that's often only as good as anyone can organise it into in any given skin, but ultimately that isn't very complicated. It's a mess, and it's ugly, but we're trying to figure out how to make it less ugly, and that's where it actually gets complicated. We're trying to give something structure based on the somewhat arbitrary structures that grew organically over years.
    I'm not entirely sure what I'm saying here - it's a problem, just... we have the potential to make it into a serious problem, if we screw up this RfC?
    And in the meantime it sometimes feels like I'm spending as much time arguing with people as actually doing anything. Fortunately I managed to avoid most of that in the first half, but that could also get interesting in the next few months as well, especially with these RfCs...
We have an amazing community, though. I kind of love it, when I can even keep up with what people are saying. I just hope we don't break too much for them. -— Isarra 01:00, 4 April 2019 (UTC)Reply
Return to "Project/Timeless/Post-deployment support/Midpoint" page.