(Almost) Nightlies
After doing a statistics article last year, there was some information I wanted to see that was either impossible to assemble, or at the very least a pain in the butt in terms of working with various dates. So halfway through January 2022 I set up a script to run nightly and record a handful of values at the time of running. How big was the queue? How many files were being hosted? That sort of thing. This plan worked well. Almost!
Alas, towards the end of August I did some database work and forgot about this random little script, never updating it to reflect schema changes. By the time I noticed, a little over a month of data points had vanished. So unfortunately, every single one of these charts is going to have an unnatural looking slope as the data stops on August 20th, and resumes on September 27th. The good news is that most of these stats do boil down to "how much is the number going up" so that loss isn't actually anything dramatic.
Upload Queue Size
The 2021 queue had a huge swelling of uploads in the spring when asie was actively streaming, and I was actively working retail, getting so absurdly out of hand that the queue was spilling well into a third page! Publication Packs today are a lot more consistent even if they remain unscheduled. These days, hitting the double digits is usually enough to get me thinking it's time to publish, and the idea of a second page is scary prospect. Unless somebody decides to upload a dozen of their historical files at once rather than bundling them, I don't think there should ever be a second page again.
The chart shows the philosophy in action. After clearing things out for the start of the year, whenever things creep above ten they drop by a good amount, though not always back to zero. For packs with newer worlds that shouldn't sit around too long, smaller packs get produced. There were only 27 recorded dates in which the queue was entirely empty. This stems from the nature of the Publication Pack article that accompanies publishing these days as well as how the front page of the Museum displays published files.
Putting together a publication pack is easy enough. A tool on the site lets me select from the queued material, provide an identifier to prefix the included screenshots with, and then assembles a file for me to populate with short descriptions of everything. Where it becomes more painful is in the cross-posting. I've never actually assembled a proper script to post everything off to Twitter and so instead, I have a copy/paste view that speeds up the process, but still requires well, copying and pasting a bunch. For 2023 as I try to get WoZZT established in places beyond just Twitter, the need for a proper tool is a lot more important, but needless to say, it does get to be a slog when posting 15+ items. So it's in my own interests when the queue does creep up in size to not necessarily empty it all at once.
There are other reasons the queue doesn't go for broke on publication. The front page divides recently published files into two categories: "New Releases", which are newly created worlds, and "New Finds", which are defined as anything published outside of the current year. It probably won't be on the front page by the time this article goes public, but currently "Zee!", is listed on the front page as a New Find because it was a 2021 world not published until 2022.
The goal of this split is to allow newly created ZZT games to not be buried by AOL-era work. It's important for people to see ZZT still thriving today and for authors to not have to compete with titles from decades ago, so these new releases get their own half of the list. Screen resolution non-withstanding, the latest twelve uploads in each category are listed, but only the top four are visible without any scrolling.
A lot of what to publish and when comes down to not wanting to end up burying new releases with more new releases! As we saw with Cyber Purge last year, being on the front page for a prolonged period of time meant more eyes on the game. I actively consider the order a set of files are published because of this. Same goes for the ordering within Publication Packs themselves when they contain a variety of particularly interesting content. Twitter threads like to show the beginning and end more than the middle, so if at all possible the good stuff bookends things. (Fact check: This isn't a given rule. Sometimes I just want to get the pack out the door ASAP.)
Larger, more impactful (subjective) releases tend to get the top spots, so single board experiments get pushed behind full games. For new finds, the yellow bordered scrawlings get to be below the surprise hits.
In instances where there are too many new releases, things get tougher. This is why the queue sometimes doesn't actually empty. I don't want anything new being stuck in the queue for too long, while also wanting to give other releases time to breathe on the front page. It's tough!! Oktrollberfest and the end of the year are the two times in a year when a large number of newly dated uploads can be expected, so I try to work the queue in such a way that there's room for it all. As long as the private unpreserved queue used for streaming has games left, it's easy to strike a balance between old and new. That queue is slowly dwindling though, and afterwards I suspect the size of Publication Packs will end up shrinking, perhaps even becoming a misnomer if a new release shows up with no others to be published alongside. A pack of one...
Total Files
Do you like lines that go up? This chart is for you.
In rare instances items are actually removed, usually because there's a duplicate that wasn't noticed, or somebody didn't use the edit functionality and uploaded a second revision of an unpublished game. Rarest of all are emails by authors asking to remove something they made ages ago. We're talking maybe one of those a year if that. This is a number that goes up.
Up by more than 300 for the year in fact! It's absurd in every way. Nearly a hundred new releases, more than 200 worlds that weren't readily available now being easy to find on the Museum. I wouldn't have expected there to still be so many ZZT worlds to share.
Total Authors and Companies
I'm also always curious about just how many people have made ZZT worlds. These charts count the unique number of authors and companies. They also have weird dips around the data loss period as the database changes that broke things were going from being stored as strings of text for every file to have dedicated tables that would be associated with files. There was plenty of merging duplicate and mis-entered data.
The adjustments make it hard to compare the start and end points for the year, but hey, well over fifty authors that were previously unknown made their first appearance this year, whether by having something from 1995 or 2022 published for the first time.
The companies count also went up a solid amount. ZZT company names are such wonderful things, that even if the vast majority of them are utterly meaningless, having had no other members, no online presence, and no community involvement, you can't help but smile when you see worlds credited to Awesome Software Incorporated, Diamond Point Entertainment, ButtSoft(tm), or the Quest Arts ZZT Role-Playing Division. It's so serious, and yet not at all. Perhaps that's the nature of ZZT, too.
The Grand Finale
Click To View Full Data Table
Year Previous
Count2022 Change 2022 Total Unknown 137 76 213 1991 24 0 24 1992 46 1 47 1993 88 1 89 1994 101 12 113 1995 159 23 182 1996 223 17 240 1997 233 37 270 1998 193 11 204 1999 320 13 333 2000 332 9 341 2001 404 4 408 2002 221 1 222 2003 155 3 158 2004 82 5 87 2005 51 0 51 2006 33 3 36 2007 86 -2 84 2008 27 0 27 2009 15 1 16 2010 9 1 10 2011 0 0 0 2012 18 0 18 2013 8 0 8 2014 7 1 8 2015 7 0 7 2016 7 0 7 2017 3 0 3 2018 12 0 12 2019 20 0 20 2020 50 -1 49 2021 59 1 60 2022 0 79 79 SZZT 54 8 62 ZIG 18 0 18 Utilities 133 1 134 ZZM 26 0 26 Featured 126 5 131
To cap things off, let's end with seeing the distribution of new files for the year. The previous count values come from last year's statistics which used the file counts seen on the Mass Downloads page (hence the extra categories at the end), so I've continued to use those numbers. In short, there were 137 ZZT worlds with unknown release dates in 2021. In 2022, that number was 213, so 76 worlds with unknown dates were added for the year.
It's worth noting that dates on the Museum do change, and those changes will be considered 2022 additions. This also explains the few negative values. I was very curious what the heck LibreOffice would do when I said there were a negative number of additions. Apparently it puts them on the bottom. That sounds about as good as it could be. I think we all learned something today.
The new additions don't provide any revelations! Some finale. Years which are known to have a lot of releases have a lot of newly discovered releases. Years that are lacking, well, remain lacking. The 2000s are just a tough nut to crack no matter how you look at it.
Anyway, I hope you enjoyed taking a look at some numbers for the year. I hope next year we can continue to have good numbers.
Still holding out hope on a complete game dated 2011...
