11 Comments

  • Yep. Yep, yep all the way. Our biggest download success by a massive margin, the search/replace thing, isn’t in the repo. Meanwhile, the amount of effort we’ve put into getting code just so for the repo hasn’t always felt like it’s been paid back. We exposed the amount of information the API received about an install many many years ago but it went largely unnoticed. Nobody cared, really.

    I’m not sure what the answer is, but having your plugin on the repository amongst the white noise of everything else that’s on there is barely a benefit any more and we’re certainly having similar conversations internally.

    • Our biggest download success by a massive margin, the search/replace thing, isn’t in the repo.

      I hadn’t had people come to me because of my plugins. But I had people use my repo plugins, because of my reputation elsewhere. Arguably I had been promoting repo more than it had been promoting me. :)

      We exposed the amount of information the API received about an install many many years ago but it went largely unnoticed. Nobody cared, really.

      People do care. Alas not the people who could overturn these practices and not many enough to make them care.

      I’m not sure what the answer is

      I am not sure what the answer is, but by know I am sure enough to pick an alternate direction for myself and spend my attention differently.

  • Damn skippy. Couldn’t agree more. I want to mouthpunch somebody — anybody — every time I need to use SVN to update the repo. And the support forums must be killed with fire — I have a single pinned post in mine, saying “Go to GitHub for support, you WILL be ignored here.”

    Your post also mentions a tonne of other things I hadn’t even considered (privacy implications; the license thing which *is* utterly batshit crazy). Very good points all around.

    The sad thing is, it could be much better than this. I’m a Git-vetted contributor on Drupal.org (meaning I can create new projects in the Drupal-equivalent of the WordPress Plugin Directory), and getting to that point was an enormously long process — first I had to write my module, then I had to lint it against community standards. That much is similar to WordPress. But then I needed to post it to a code review, wherein both established community members and other new developers analysed what I’d written for consistency, security, standards, best practices, etc.

    In order to get an established member to review my code in a reasonable amount of time, I had to review three other new projects, in turn improving my knowledge of the Drupal API and community standards. Once my code failed the review by the established member, I then needed to review three *more* new projects before getting a re-review of my own code. Then, once approved by the established community member, it was then reviewed by a member of *the freaking security team* — after which I was granted Git access.

    Long and arduous? Yes. Frustrating? Often at times. Do I feel more confident about the quality of my Drupal contributions than I do of my WordPress contributions? Absolutely, no question whatsoever.

    While I’m not sure I advocate such a community-intensive process for WordPress (my guess as to why WordPress has way more themes than Drupal is this process — I kind of doubt most designer-y, front-endian folks could be arsed to endure something as technical and long as the Drupal.org Git-vetting process), *some* better sense of community around 3rd party development would be most welcome.

    As for me, I’ll probably keep using the directory, but only as the most user-land representation of my code — everything substantive will be on GitHub.

    • WordPress always took pride in having a lot of extensions, same way it takes pride in having a lot of market share. It’s very flat and flattering presentation of situation with either.

      In case of extensions the low entry bar to repo have always been beneficial to quantity. Which resulted in abysmal average quality.

      I often feel unfair with critique of review and approval process for repositories, since in the end it needs people resources. However I feel it’s not too upfront of WordPress project to take such credit solely on quantity, swiping other issues under the rug (almost literally — in hiding those 2+ years plugins).

  • > The only plugins that do receive promotion are those with highest downloads count and short “featured” list.

    When you create a detailed readme.txt and you properly tag your plugin you are benefiting on several places at the plugin repository. E.g. Popular tags, highest rated, recently updated and so on. The secret is to periodically update your plugin and you get a lot of new downloads and user.

    About SVN:
    Personally i use for my plugin “Mashshare” the WP SVN repository as a pure download repository and use git for developing.

    Best,
    René

    • E.g. Popular tags, highest rated, recently updated and so on.

      You might have point about tags, I hadn’t thought much about that facet.

      Most popular / highest rated from quick look have respective first pages covered with plugins with upwards of 500K downloads. It just makes big plugin bigger.

      The secret is to periodically update your plugin and you get a lot of new downloads and user.

      Except that too frequent updates are considered to be gaming the system and explicitly forbidden by rules. ;) And I want to spend my time improving my code, not spamming to “stay ahead”. As for me having to do this to bump visibility is a hack and downside, not benefit.

  • It’s a hard decision to take, and I can’t deny your logic. But I’ve a different view. An icon like you, if raise voice somewhere, would take effect. WordPress has some mirrors in GitHub. I don’t know the mechanism, but it seems we can simply shift from SVN to Dist. VC. If you leave because the system is in wrong shape, it won’t be fixed anymore. I want to see it from a different perspective – if anything is wrong, why not I first start fighting against it, by NOT leaving, but being IN it with my VOICE, and CHANGES. Rarst, it’d be a drought if we miss you in WP Plugin repo. Truly. :(

    • An icon like you, if raise voice somewhere, would take effect.

      You are overestimating my status in community. While I have some visibility, I do not think I have much (if any) practical influence in mainline WordPress project matters. I had about one thing I count as success in that regard — getting licensing rules in line with de–facto situation (allowing Apache License v2, etc) while back.

      WordPress has some mirrors in GitHub. I don’t know the mechanism, but it seems we can simply shift from SVN to Dist.

      Just this weekend (after this post was written and already scheduled) it was announced that WP will trial accepting contributions to core via GitHub pull requests. I consider that is amazing shift, but it doesn’t impact repositories for now and it’s not enough to overturn literally years of issues, covered in the post, overnight.

      I want to see it from a different perspective – if anything is wrong, why not I first start fighting against it, by NOT leaving, but being IN it with my VOICE, and CHANGES.

      I am one person with only so much capacity to do things. Sometimes this means walking away from things towards things that matter more for me.

      I don’t feel like fighting with repo.

      Rarst, it’d be a drought if we miss you in WP Plugin repo. Truly. :(

      I very much doubt absence of my plugins would matter much. I had never been a popular plugin developer and that makes it much easier for me to leave like this.

      And the point I am making here is that there are different ways to build and manage our infrastructure. Sometimes it’s playing to the central repository, sometimes it’s not.

  • Hi. Can’t wait to see your next shares on github :)

    But here I’m all at sea !

    Considering the fact you do not want to waste your time with support for free plugins you should not releasing your stuffs on wordpress.org and github is the perfect place.

    Moreover you CAN build your own repository. A lot of premium WP Businesses have their own repo with updates and stuffs like this.

    Indeed if you’re not satisfied with wordpress.org as hosting for your plugins then leave, we know where to find you :) again Github is the perfect place.

    Codes you share hardly regard average users. To me wordpress.org IS for average users. So yeah this is probably not the best place for you.

    Nevertheless I think WordPress needs good programmers and developers so it could be highly improved in the next years. Maybe it’s not time for leaving…

    • Considering the fact you do not want to waste your time with support for free plugins

      Hadn’t said that. I am perfectly fine (and do) provide reasonable support for my plugins. But dot org forums are not my happy place and I want better workflows and control for it than that.

      Moreover you CAN build your own repository.

      There is a lot of nuance between technical possibility and getting to favorable results.

      Composer infrastructure was shaped from the start with complete open source and even two levels of it in Packagist/Satis for different needs. Everyone uses same tools, everyone improves same tools.

      WordPress infrastructure is that closed source API you need to bend backwards to hack around. And when you do — likely no one will care but you. We have literally dozens implementations, I wouldn’t even call competing — they are just floating around.

      Codes you share hardly regard average users.

      Yep. I solve my own problems primarily.

Comments are closed.