Coordinate with IFDB on creating game listings

[dfabulich]

Every year, I run my automated tool to import IFComp listings into IFDB, and every year, it creates problems, because there's no way to ensure that we don't create duplicates.

There are at least two ways duplicates can happen.

• The very second that the competition opens, a horde of insatiable volunteers inevitably begins creating IFDB entries by hand. This year, the competition went live around 7am Pacific time. I ran my script around 11am, which was plenty of time to create lots of duplicates for me to clean up.

• Before the comp even opens, some people create IFDB entries for their unreleased entries. This is encouraged in the Author's Handbook. https://ifcomp.org/about/how_to_enter#ifdb

  Unlike with the IF Archive, IFComp does not take care of creating or updating records in The Interative Fiction Database about your IFComp entries. Some kind soul might put your game on IFDB, but there is no guarantee.

  While we don’t require it, we do recommend that IFComp authors create or update IFDB pages about their own entries. (And yes, you are allowed to update your game’s IFDB page during the competition.)

  If you do want to ensure that your game appears on IFDB appropriately, it’s a good idea to swing by a few days after the competition ends to ensure that the IFDB points to your game’s permanent location on the IF Archive. And, of course, if do any post-comp releases, you’ll also want to update your IFDB page to reflect that.

The best possible thing in the future would be for the IFComp team run a tool themselves, so nobody could have a chance to manually do work that conflicts with the tool, or, barring that, some kinda coordinated release, so that I can have my finger ready to push the button the second the games drop.

[dfabulich]

I've filed #425 on this, which has something to do with the process, but may or may not be what we want.

Perhaps the IFComp team would send me games.html before the competition opens? But then, IFDB would "break the news" of all the IFComp games before ifcomp.org does. That feels like a bummer.

If we wait until release day, well, I might as well just read the game data off of the /ballot after all. And, if the data doesn't get conveyed to me right at the very moment the competition starts, that still leaves plenty of time for volunteers to create a bunch of duplicates. (The duplicates can even be created before the competition starts, e.g. by authors.)

So, I don't know what we want to do here. Our best bet might be to just coordinate release timing with me, so next time I'd know that the games are going live at 7am Pacific. I could then set my alarm and run my importer ASAP after the competition opens, minimizing the risk of duplicates.

Or, alternately, I could just deal with the cleanup. I could try to write code to make it easier to deal with, e.g. using a better duplicate detector, improved merges, etc. etc.

I'm afraid that the real answer here is, indeed, to just deal with it…