I do not know how to take a picture that conveys how brutal this hill was. still had a good time!!!
Remote
Adrian Sampson
@adrian@discuss.systems
Carly Rae Jepsen fan account.
1364
Followers
459
Following
45
Posts
Joined November 08, 2022
Posts
the trail was just packed with people talking about the #mosstodon hashtag
tentacle debt
I’m still looking for where the Carly Rae Jepsen community is hiding on the Fediverse, so in the meantime I must subject y’all to an excellent new song https://youtu.be/_Xqexqiqkdk
just in time for the solstice, I seem to have sorted out my summertime proofing routine #breadposting
I have been working on a fake Unix shell, which is 50% practical joke and 50% earnest attempt to make computational pangenomics go fast. https://www.cs.cornell.edu/~asampson/blog/flash.html
I’m still pretty bad at pizza, as indicated by this gradient from charcoal to edible #breadposting
It’s finally the time of year when it’s 70° somewhere in my house, so things can actually rise properly #breadposting
practicing for the impending pizza season #breadposting
I didn’t do the egg wash that the recipe called far because (a) I am lazy and (b) I don’t have a little silicone brush, but they’re still pretty good! #breadposting
I think it sucks that compute capacity is apparently measured in watts now! at least report the FLOPS, you weirdos
Replying to
@adrian@discuss.systems
like hiding a pill inside a treat for a dog
My toxic trait:
“Hey Adrian, what do you think about this technical problem?”
“I dunno, seems complicated; we had better be careful about the effort/reward trade-offs”
“What if we framed it as ✨designing and implementing a programming language✨?”
“I’m in”
I just had a lil $20k REU supplement proposal rejected! this used to be a rubber-stamp kinda thing! it really feels like there is no research money anywhere! and all the CFPs are just for appearances!
Replying to
@adrian@discuss.systems
fixed a bug where Analog Library would error out on mononymous authors. RIP Arvind https://al.radbox.org/doi/10.1145/69558.69562
Replying to
@peter_sewell_@types.pl
@peter_sewell_@types.pl I'm currently generating the BibTeX myself, using data from the Crossref API. https://codeberg.org/samps/analog-library/src/commit/53da27b01e14f2aaf56f944b396fd1d65f1aff43/src/bib.rs#L40-L67 DBLP has a nice API but I haven't found anything I need from it yet
Replying to
@krono@toot.berlin
@krono@toot.berlin good point. maybe we can talk about this in an issue? https://codeberg.org/samps/analog-library/issues seems to require a bit of thought on how to do it right
Replying to
@GUIpsp@types.pl
@GUIpsp@types.pl sounds interesting! can I have an example to start? I can look into it
Replying to
@adrian@discuss.systems
Analog Library now links to other versions of a DOI, so you can easily get to the PLDI version from the “SIGPLAN Notices” version of a paper you probably don’t want
Replying to
@mattpd@mastodon.social
@mattpd@mastodon.social this should be in place! seems to work for some papers but not all https://codeberg.org/samps/analog-library/pulls/25
Replying to
@addison@nothing-ever.works
@addison@nothing-ever.works do you mean in the BibTeX, or just as separate fields on the page? (fwiw I have tried to keep the BibTeX minimal because that's how some people want it, but we could explore multiple options for verbosity)
Replying to
@mattpd@mastodon.social
@mattpd@mastodon.social ah, thanks, yes, I had noticed something wrong with ASPLOS 2026 papers in particular: https://codeberg.org/samps/analog-library/issues/22
shouldn’t be too hard to fix, thankfully
Replying to
@adrian@discuss.systems
if you’re a Kagi user, you might enjoy this redirect rule when you search for papers:
^https?://dl.acm.org/doi/(.*)|https://al.radbox.org/doi/$1
install it in your settings here: https://kagi.com/settings/redirects
Replying to
@markusde@mathstodon.xyz
@markusde@mathstodon.xyz yay, glad it might be useful at this point!
Replying to
@adrian@discuss.systems
Analog Library now has a bookmarklet (or a userscript) you can use to bypass Digital Library crud
Replying to
@adrian@discuss.systems
24 is kind of a lot of papers: it’s the largest LATTE since it started in 2020. This has so far been a high-acceptance-rate, low-key, discussion-first workshop, but maybe we need to do something different next year.
☕️ LATTE, our little workshop on hardware design languages/compilers/etc., has 24 (!) rad-looking position papers this year. It’s on Monday, and you can attend on Zoom or in Pittsburgh: https://capra.cs.cornell.edu/latte26/
what we have here is gluten inflation #breadposting #breadposting
Replying to
@notypes@discuss.systems
@notypes@discuss.systems all I know is that all my chairs are fully indistinguishable
every year, I get an invitation to a Cornell “Distinguished Chair Holders Luncheon,” which I have never attended, but this is what I imagine it looks like
Replying to
@adrian@discuss.systems
Analog Library is running on a $2/month VPS, and I guess I just think it’s kinda rad that there exists a non-useless $2 VPS in 2026
Replying to
@adrian@discuss.systems
By popular demand* Analog Library now generates minimum viable BibTeX.
*one person asked
Replying to
@nemo@camp.crates.im
@nemo@camp.crates.im good call; it turned out to be pretty easy to flip on LetsEncrypt on my crappy shared host. (and yes, that is my friend Todd's proudest achievement)
Because I have poor self control, I made a thing to avoid looking at those increasingly terrible ACM Digital Library pages. Introducing Analog Library: https://al.radbox.org
My most difficult teaching moment, you ask? Last week, I live-coded an assembly program that eventually added 42 and 27 in front of 300 students, and when the simulator showed register a0 containing 69, I only had a split second to decide whether to say “nice”
Delighted to announce that both of our PLDI submissions have been rejected! ✨
Replying to
@drmorr@hachyderm.io
@drmorr@hachyderm.io oh nooooo! I hope you make another one and it's even more fluffy and delicious
it has been a busy week #breadposting #breadposting
Thank *you*, weird web form. I agree.
☕️ For the sixth time (!), we’re running our little ASPLOS workshop on hardware design languages/compilers/etc. Papers are just two pages! So easy! Please submit! #latte26 https://capra.cs.cornell.edu/latte26/
super sorry to the guy with the username @reset on GitHub, whom we keep accidentally sending email to when we talk about setting registers to zero https://github.com/cucapra/calyx/issues/1640#issuecomment-1652270808