Colin Gordon

@[email protected]
498 Followers
1.5K Following
2.4K Posts

Programming languages professor, kernel hacker, aspiring linguist (syntax & compositional semantics).

Currently figuring out how to combine all of my interests by mechanically translating English into formal specifications of a formally verified OS kernel for RISC-V.

      

pronounshe/him, er/ihn
languagesenglish (native), deutsch (~B2, Ich spreche ein bisschen Deutsch, aber nicht genug für alles zu benutzen), français (<A1, je parle seulement un peut le français), linguae latinae (relearning bit-rotted ~B1)
alts@csgordon and @csgordon
homepagehttps://csgordon.github.io/
Colin GordonMay 17

So following up on
https://discuss.systems/@csgordon/116389175834499609

I finally got around to setting up a Dell Wyse 5070 thin client as a desktop. As in, stuck an M.2 SATA SSD in there (doesn't support NVMe) and installed FreeBSD locally. It's running with the stock max 8GB of RAM (there are reports of it working with up to 32GB, but that's like $200 now, and I paid $50 for this machine, plus disk). And it's... fine?

As-is, it's perfectly snappy. I generally do about 2 full days' worth of work on it each week (otherwise I'm on campus), and possibly double that working in the evenings the last couple weeks for deadlines. I'm running i3 as the window manager, but usually something like XFCE isn't that much more memory. It does occasionally kill the Zotero desktop client if I open too many Elsevier tabs in LibreWolf (Firefox). And things were a little touchy when I was building Servo from source, so I closed the browser for a bit and kept working. I may upgrade to 16GB at some point, or add swap. So not a flawless experience, but honestly I have more (different) problems with my M4 MacBook Pro.

Max power draw is 15W. Comparable to a high-output LED light bulb. (Not including the monitor of course.)

Colin GordonApr 10
I have so many questions:
- Why does a seminary need a Program Director of AI Business Strategy?
- Why is this role combined with a *theological seminary* faculty role?
- WTF did I click on for this to be something they email me about? (I live kinda near Eastern University, and I do work in higher ed, but...?)
Colin GordonMar 12
Looks like the underpricing of these generative "AI" tools to get people hooked might be under strain... Normally verified students get access to all kinds of expensive goodies because companies are hoping to get the students to drive future business by setting their expectations for professional tools.
Show threadColin GordonFeb 20
It sped up, so I could do this before I left for the weekend #CheriBSD #cheri
Show threadColin GordonFeb 20
Unfortunately I think I'm going to have to leave for the weekend before all of the packages finish downloading from cheribsd.org
Colin GordonFeb 20
It's here! #cheri #CheriBSD
Show threadColin GordonDec 18, 2025

Continuing on with the absurdity of the new ACM DL changes... beyond the broken AI summaries replacing human-written (peer-reviewed!) abstracts, and the ridiculous decision that an author *can't view their own profile*, here's another change I just noticed:

If you're logged in with premium access, you get the normal links to download a PDF or (weird) ePub directly. If you're not logged in, you get a "PDF/eReader" link which directs you to an Elsevier-style PDF reader. In fact, it looks *exactly* like the PDF reader Elsevier sends users to, which spies on readers while they use the reader: https://golem.ph.utexas.edu/category/2021/12/surveillance_publishing.html#c061007

I might be more pissed off about this than about the idiotic AI summaries.

Update: I've been poking around in developer tools looking for signs of it actually sending telemetry when I'm moving around the PDF, but have not observed anything. But I'm also a kernel / compiler developer, not a web developer, so I'd appreciate if someone more experienced would give this a try.

Colin GordonDec 17, 2025

So apparently ACM has not only prioritized AI summaries of articles, including slapping them on articles that are licensed to the ACM in a way that forbids derivative work (see link), apparently they're now paywalling the second page of ACM author profiles (see picture)
https://discuss.systems/@csgordon/115732091730446890

Update: I actually can't view the second page of my own profile even when logged in as myself.

Colin GordonDec 16, 2025
wtf is this
Colin GordonDec 9, 2025

I know I semi-regularly complain about Web of Science, which my University unfortunately takes seriously, but .... I still can't believe this crap. Not only does it auto-add publications which I'm then not allowed to merge when it finds multiple sources, but check out these entries in my publication history. They're blank. I can't edit them. I can't even look at them to see if there's some identifiable metadata. WTF is this?

BTW they have contact addresses you're supposed to email to fix this sort of thing, but I've tried a few times and they never responded or fixed anything, so it's not clear why they're there.