Occasional need to access very very old/Wordpress-era versions of the site

[keeganland]

Alumni sometimes are looking for old articles, which sometimes experience code rot, or were even on old versions of the site running fundamentally different software.

---

Technical issue with accessing old articles and stopgap solution:

"I'm assuming [the article] was on the old wordpress which was SUPPOSED to be transferred entirely to the Django but things seem to fall through the cracks sometimes, and I don't think it's reasonable to let the old wordpress installation sit around complicating our google cloud setup, when it's not something that people actually need often. I do not think it can be accessed within a timeframe by us that anyone involved would find actually reasonable. It would derail any work that Web's currently doing only to get it to them later than they probably want it.

My best advice for them would be to talk to the AMS Archivist, who apparently tries to archive things that were on the site. I'm not sure exactly what this entails, though they do email us when we apparently accidentally break their workflow. But they do apparently do some archive work that's independent of our own archive work which is more likely to preserve old things because their archival work doesn't need to be balanced against serving the needs of the day-to-day readers."

Connected communication issues:

"Often I don't think it's fair to alumni to promise them to keep up their articles indefinitely. Code doesn't stay good forever, "code rot" is a standard idea in software engineering. Promising someone that an article they wrote in 2014 will be on the site in 2022 is sometimes as about as reasonable as promising a cake they baked in 2014 will last in the same condition in the office fridge until 2022: even if no human hands touch it, it can grow mold and decay.

Given that, we lose very very few articles.... but then I worry that the reasons that they can be lost are so poorly understood, alumni would be upset rather than understanding with us. I think it'd probably be best to make a point to tell people if they really want to save an article for future portfolio stuff, they can and should link it, but they should keep a copy of their own work too. Something that's sort of been on my personal wishlist is an easy way to download pages in a format that'd look nice in portfolios and stuff so a good version can live on its author's computer, and so they don't need to worry about its actual online counterpart. But that's a full blown feature and browsers can already do a rough approximation of that out of the box"

(Possible easy approach could be just to stick something about this on "About" or in an FAQ)