I have flagged this on Github.
There’s a CORS error occuring with some TADS 3 games via Parchment across browsers. It doesn’t seem to be all of them (for example, According to Cain is fine). I’m not sure what the pattern is.
Has anyone else had this problem?
I don’t know nuthin’ 'bout Parchment, but the errors here aren’t CORS errors, that’s just the error the Parchment code defaults to.
This file is missing: https://unbox.ifarchive.org/1363p9wc5y/I-Am-Prey-Beta-Patch-6.t3
And these two files aren’t served by ifarchive with the &open param. If you remove that, the files are served.
Bad: http://www.ifarchive.org/if-archive/games/tads/return-to-ditch-day.zip&open=ditch3.t3
Ok: http://www.ifarchive.org/if-archive/games/tads/return-to-ditch-day.zip
Bad: http://www.ifarchive.org/if-archive/games/tads/LydiasHeart.zip&open=LydiasHeart.t3
Ok: http://www.ifarchive.org/if-archive/games/tads/LydiasHeart.zip
What you’re describing is consistent with the behaviour you see when you’re in the UK, which is still geoblocked by the IF Archive. Is it that case that “play online” links to the Archive unboxing service are the ones that fail and links to story files hosted elsewhere work? Is it possible that your internet connection is being identified as originating from the UK for some reason?
Wut
EDIT: Sorry, with less Gen-Z in my veins this time: Was my game removed from the archive, or…? I thought I would have been emailed about that, if it happened.
EDIT 2: Oh, I guess a geoblock would also look like file removal, huh?
I think it was just a bad URL.
From Aster’s post: https://unbox.ifarchive.org/1363p9wc5y/I-Am-Prey-Beta-Patch-6.t3
But the game is available here: https://ifarchive.org/if-archive/games/tads/I-Am-Prey.zip
unbox.ifarchive.org is currently experiencing downtime. Some URLs still work, because they’re cached in Cloudflare, but the underlying service is down. @Dannii
to be clear, these are the direct links i got from ifdb
no, it was across a few devices, neither with a vpn on
so it’s for every game not just tads 3? huh. i only tried a few and couldn’t discern a pattern.
It’s back up now. The pattern was just “what was already cached in Cloudflare.”
Someone from Britannia can test if the Realm’s lapse into late Queen Victoria era’s proverbial prude has to do with that parchment issue ?
Best regards from Italy,
dott. Piergiorgio.
No it was just the recurring full hard drive problem I haven’t been able to diagnose yet. Probably a race condition.
full hard drive problem related to parchment or IF Archive ? in the latter case, is worrisome, because everyone known that by early september will be uploaded a very sizeable .zip file which WILL be accessed by scores, if not hundreds of online players…
Worried regards from Italy,
dott. Piergiorgio.
As Dan said, it’s specifically the Unboxer. The IF Archive is unaffected.
More specifically ? The unboxer is at the core of IF online playing, and the prospect of scores of online IFComp judges requesting unboxing from the big IFComp .zip file is my main worry in light of this issue…
Best regards from Italy,
dott. Piergiorgio.
unbox.ifarchive.org and ifarchive.org are different machines(*) running different software.
Since, as you say, the IFComp Big Zip is generally the largest zip file of the year, we set a flag to make sure that unbox doesn’t touch it. If you look at Index: if-archive/games/competition2025 , you’ll see that IFComp2025.zip has no “view contents” link; only the individual game zips do.
(* I should say, virtual machines hosted by Linode. They’re not physical boxes, but they’re configured as distinct Linux servers and each one has its own RAM and disk quota.)
The Unboxer is actually perfectly happy to unbox the big ifcomp zips, though it does need to evict a lot of other zips out of its cache to do so. I’m pretty sure there’s some kind of race condition which means that files aren’t being deleted from the cache when they’re meant to. I’m thinking of just running a script on a timer to check if the cache is bigger than it should be because even though it wouldn’t be fixing the underlying problem it would still stop the server from dying.
Thanks for your explanation, Zarf ! I’m now more at ease now.
Dannii, if you have now a diagnosis (the race condition being in the cache handling) I think is a substantial step toward closing the issue 
Best regards from Italy,
dott. Piergiorgio.
FYI, I’m doing some server maintenance on the Unboxer server right now.
Edit: back online now.
