branchandroot | *hairy eyeball*

Dear AO3 decision makers,

How can you be lying so much? Your noses should be growing. Shame!

No, tag filtering is not down because FF.N had a brief witchhunt for too-porny fic and all-caps summaries, and now everyone is coming to AO3. You might wish that, but no, and a good thing, too, considering the actual reason.

Tag filtering is down because that filtering menu was written so astonishingly thoughtlessly that that menu alone took up almost two thirds of the server load. Not new accounts, not new fic being posted, not too many notification emails being sent. No. That one menu was sucking down so much horsepower that even a modest increase in traffic, as for example from a new movie fandom coming to read on AO3, flatlined the servers even before FF.N had it's latest little round of "no hot porn, we really mean it".

The reason, the real reason, is not an increased load. It's the incredibly poor programming choices initially made for that menu. You put up graphs demonstrating this fact on your own blog, for pity's sake. So, please, stop lying through your teeth about how it's all because of how beautiful and popular you are, and incidentally how evil the competition is. That's really embarrassing to watch.

Also, if you want to convince people of how hard you're working to fix it, try not making experienced programmers jump through one-way-mirror paperwork and "training" hoops to help. Then you might manage to not burn out the few programmers who are still locked in, and possibly even avoid setting yourselves up for yet another clusterfuck down the road. Concept!

Yours in deep exasperation,
Branch

P.S. Repeat after me: "Blacklight", "Solr", "not reinventing the wheel". Honestly.

Flat | Top-Level Comments Only

I think it would help make development a much healthier process. Without that gate in place, it couldn't be as black-box as it is now, and it's that black-box that's letting one or two people drive the whole thing. And smothering any open discussion about how this might all be made to work.

There's also the fact that a lot of things are said to be not on the agenda because other things are more important. It's one thing to say that to someone who just happens to be asking, but when you're telling someone who wants to help code, "No, you can't come work on that because other stuff is more important," it's not like they're necessarily going to just fall in line and code what you tell them! If someone has a choice between "code this small thing that I can do and learn from" or "eat some lasagna while watching Community," and they're told, "Well, you can't code that one small thing, but you can work on this monstrous project that looks intimidating and out of your reach," then pretty much the only people who benefit are lasagna ingredient manufacturers. Certainly the archive isn't benefitting by turning would-be programmers away before they even get to the door, and neither are archive users.

Blah blah blah hey choir how ya doin'. :)

*sings the "amen"*

Exactly that, yes. The heart of an open source project is "here's all the bugs, have at 'em!". And, yeah, some things have to be done before others, but attempting to "prioritize" coding into some kind of utterly linear progression that you then require all coders to follow or they can't code is just... *throws up hands* So wrongheaded I have no words for it. Except possibly "psycho control freak".

And this is precisely the argument to be made for an OSS project having paid employees, even part-time contractors, over and above OSS contributors, and why most large-ish OSS projects have same. It's harder to project-manage (and there is the potential for a weird hierarchy thing that many, many projects have struggled with over the years) but WHEN YOU PAY PEOPLE, EVEN A LITTLE BIT, EVEN AS CONTRACTORS, YOU CAN TELL THEM WHAT TO WORK ON. (And when you run out of paid contractors to order around, that's when you start promising people hand-knit socks. Not that I'm frantically trying to finish this pair before OSCON or anything.)

Just ... the complete ignorance of *every single modern open source project that went before* is staggering. I don't expect you to know the history of modern open source all the way back to the Linux kernel release and the development of the first GPL in order to start a new OSS project and I'm certainly not saying you have to emulate them in all things -- I'm the first to admit that contemporary OSS projects are kind of dysfunctional at times -- but for gods' sake, when you are a more dysfunctional OSS project than aforementioned Linux kernel, HOUSTON WE HAVE A PROBLEM.

BEST PRACTICES, PEOPLE. BEST PRACTICES. TWO DECADES OF WORK HAVE GONE INTO THEM. IF YOU ARE IGNORANT OF THEM, YOU WILL REINVENT THE THINGS THAT FAILED AND RESULTED IN THEM. BADLY.

*breathes deeply*

*hands you a cup of soothing tea*

I mean, so much of a successful open source project is... like throwing a bag of catnip mice on the floor in a roomful of cats. You hope they'll like them, and play with them, but you know some of them will get bored, and some will just shred their mice, and others will think this is their cue to go get some dead mice to add to the pile. So you hire a few people who are more like trained service animals, to keep the whole thing from devolving into a mess. But thinking you can get those cats to act like the service dogs? Not happening. You need to hire those. And also make the pile of mice really accessible. Hell, just hacking around the edges of freaking /Wordpress/, I know this.

*considers her metaphor* Possibly that glass of wine went faster than I thought it had.

AND YET IT MAKES SO MUCH SENSE.

And, I mean, there are ways to make the mice more appealing! and there are always volunteers who want the project to succeed and will ask "okay what should I work on next?" but there needs to be someone with the AUTHORITY TO ANSWER THOSE QUESTIONS.

the lack of a product manager shows more and more and more ...

The whole situation makes my stomach hurt. It does not have to be this hard! And there are reasons why people-managing and project-managing are discrete skills. They could buy a book or hire someone or even just listen to advice from people who have done things! ARRRRGH.

...maybe I need more tea. No, even tea will not help.

And I know there are people who have been trying to change things, Jenny in particular has had a bunch of conversations with me about the things that could be improved, but nothing ever gets fixed even once the flaws are identified. Which, you know, means there's something majorly, majorly wrong in there.

Now I understand better why it looked like you were biting the hell out of your tongue during certain parts of that con.txt panel.

So, hey, storytime.

Once upon a time I planned to start a rare fandom exchange that would take place in summer as a more convenient time for many people for such a thing.

I contacted shalott to see if she would be willing to share her matching script and the modifications she had made to Automated Archive to host Yuletide. I said it was fine if they were a technical mess, I am pretty savvy and had a husband to volunteer into helping me get everything up and running.

She declined, saying she was too busy to package it up. In fact, she indicated she was busy coding the replacement archive! And really the fastest way to solve the problem would just be tothe send my helpful tech savvy coder over to help code on the OTWarchive software. One presumes precisely to her specifications of the road map.

I did not pass on that offer to my husband, no.

hairy eyeball

no subject

no subject

no subject

no subject

no subject

no subject

no subject

no subject

no subject

no subject