Jamey Sharp

@davidgerard@circumstances.run I'm so disappointed about this because my profile here has said "an advocate of nuance and historical context" since years before AI boosters co-opted the word and now I guess I have to find a different way to say what I meant

@djsundog@fedi.reclaim.technology good news! we've created the library from Borges' classic novel, The Library of Babel I attended a talk once where the speaker proposed a compression algorithm for storing multiple files: sort them and store only the differences between successive pairs. He noted that the more files you compress together, the less space each one needs. In the limit, you can store all possible files in no space at all, but at that point identifying which entry you want in the set requires as much data as the original file. He then proposed asking copyright holders to identify where in the compressed set their content was located. It was a memorable talk, comparable to his talk involving music synthesis via Jello. (edit: my mistake, it was arithmetic performed by playing music through Jello. https://web.archive.org/web/20160505053421/http://opensourcebridge.org/wiki/2010/When_Everything_Looks_Like_A_Nail and https://web.archive.org/web/20191020043219/http://opensourcebridge.org/sessions/453)

@skade@hachyderm.io I keep being sad when I see that there's a Rust consulting firm named Corrode. I don't know, maybe they're doing good work, but I can't get past their choice of name to find out.

@jojo@lgbtqia.space I lean toward "from the moment you get there", or preferably "during the first interview", but I think it actually depends on how much you need the job. (Congrats, in any case!)

@roknrol @mayintoronto For sharing files between devices, have you evaluated https://syncthing.net/ Is there some reason it doesn't meet your needs? My current way to keep notes is in a directory of text files, synced between my laptop and phone with SyncThing. On my phone I use https://quillpad.github.io to edit them; it has UX resembling Google Keep but has nice editing for Markdown files. I'm only using this approach for small files of notes, grocery lists, etc, but I think it might be usable for larger projects too.

@xgranade My personal version of this is walking into doctors' offices and seeing signs all over the place saying "NO SHARPS"

@jyn I had assumed you were joking. the single-core time just to enumerate every f64 should be hundreds of years, right?

@pho4cexa all the usual hazards of putting your heroes on a pedestal apply; everyone has blind spots and foibles, and it seems to me that anybody who'd spend decades on the thankless job of defending copyleft would necessarily have an inclination toward stubbornness and a strong belief in the rightness of their positions, for both good and bad. still, I too hope they listen

@pho4cexa yeah there definitely are not many. fsfe was the only other one I was thinking of, but I'm something of an optimist and hoped I'd merely forgotten about somebody

@micahflee Oh, I see this is from all of DHS, not just ICE, and also that it includes contracts which ended years ago. That's good to be aware of. I actually worked on part of this one contract, over a decade ago, and it was purely about improving cybersecurity for people's smartphones; it had nothing to do with immigration. https://micahflee.github.io/ice-contracts/?state=OR&modalType=contract&modalAwardId=937

@beandreams For a while after we moved in here, our quite elderly neighbor kept coming over and edging our lawn, until we finally worked up the nerve to ask him to stop. He did a great job, sure, but it was deeply uncomfortable, and we felt that he was doing it because he didn't like looking at our less than perfectly neat yard.

@idlestate @pho4cexa I get where you're coming from, but I'm not convinced. for one thing, your characterization of Conservancy leadership doesn't match my personal experience of them. they are just as hard-line about their own avoidance of non-free software even when that forces other people to work around them; just ask them about banking and taking donations without running proprietary JavaScript, for example. I think they would agree that this is roughly their position, although I'm probably missing nuance. I can even respect them for that stance. but it does make this blog post more strange, in a way, for the reason you point out. anyway, even given a very generous reading in which we assume they were fundamentally only trying to make a "love the sinner" argument, that leaves too many crucial issues unaddressed which really can't be left out of a discussion of generative AI's impact on free software in 2026. I think the thread starting at https://social.coop/@cwebber/116411429436945287 says everything I'd want to about this. I'm very short on sleep right now so I'm not sure how coherent you'll find the above, and there's only so much I'm willing to say about this anyway, but I wanted to acknowledge your perspective even though I don't find it convincing.

@pho4cexa I'm not certain SFC is the only good organization that defends copyleft but otherwise I'm there with you

@mirth @regehr I do have some experience with combinatorial search, mostly when I was a student a couple decades ago, and I love it! In a way I think it's kind of disappointing that SMT solvers are so good, because they make it not worth learning search techniques for most problems unless you get extra deep into a problem, and that takes away some of the joy of studying search.

@pheonix I'm confused by this response, and I think this is an interesting exercise, so I hope you don't mind this reply. Either the ad server has enough information to validate the bids, in which case why ask the client to collect them at all? Or it doesn't, and the client could lie, like @rndeon suggested. What am I missing? In particular, if the client queries several advertisers and forwards all the responses to the ad server, I'd expect a graceful fallback for some of the requests failing. If the ad load fails when any request fails, then every additional bidder multiplies the failure rate. So the client should be able to look at the responses and decide to drop some of them. While signing might help validate bids, encryption might keep the client from making decisions on bid contents, and a signed timestamp might help prevent replaying old bids, even cryptography can't keep the client from pretending it never received a response. And the client could drop responses at random or based on metadata, instead of on actual bid contents, and still screw with the results. That said, and like you point out, this is much more complicated than just blocking the entire ad loading process, which is safer and more efficient for the client, so I'm not sure it's worth trying to manipulate the auction outcomes. Sending decompression bombs or other invalid data to ad servers to waste their CPU cycles is still tempting though…

@h4kor@chaos.social Yeah, that can be super useful! There's also a HashMap::get_disjoint_mut, which doesn't have an equivalent alternative like split_at_mut.

@bal4e I'm not an expert on the nuances of Rust macros, but after reading your blog post, it sounds to me like the matching part describes only regular languages. (I think the token-tree rules are what really make this possible, perhaps?) If that's true then, in principle, matching can always be done in time linear in the length of the input. Although the connection with Earley parsing makes sense, the algorithm you describe sounds a little like a Pike-style VM for regular expression matching, and a lot like the implementation from "A Play on Regular Expressions". Do you think regular expression matching algorithms apply here? If that's too restrictive, I'd want to go re-read the "Parsing with Derivatives" papers (by Matt Might, IIRC).

@nein nice! I built something similar years ago for the same reason: extracting the filesystem contents of a bunch of .deb packages and merging them all into a single image is way easier than the alternatives and goes a long way. I'm not sure I have the code for my implementation lying around anywhere, but I remember I wrote it before Rust so I used C and libarchive. anyway I think it's a great approach and deserves to be more widely known, so, nice work!

@mrencyclopedia@retro.pizza I do want to be clear that I don't believe there's any more than a hint of truth to this joke, though one part of why it's funny to me is how widely believed it is

@idlestate @pho4cexa I certainly over-simplified in talking about "hard-line" approaches and I appreciate your restraint from diving into that. especially after reading this reply, I think our positions are similar overall. the four hours of sleep I've gotten since last posting is not enough for more depth on the topic so I just want to thank both of you for your thoughtfulness

@ansate I guess my dad and I both got established with primary care there quite a few years ago so it's totally plausible they don't have availability now, but based on my dad's experience I can at least attest that if you can get in they will take Medicare 😅

@ansate OHSU takes Medicare (and Medicaid/OHP as well), and I think that goes for all doctors there, but I could easily be wrong. You might see if the Adventist hospital near 205 (which merged into OHSU a while back) has primary care doctors with availability. I'm only familiar with some of their specialists, but if they also have primary care I think that might be a good choice. I really like OHSU's Richmond clinic at 39th and Division but that might be less conveniently located for your family, if that's who you're asking for

@mirth @regehr I meant a different thing. If you have a hypothetical tool that tells you whether your source code meets your requirements, run a fuzzer on that tool: it may eventually find an input program that satisfies your requirements. This is not that different from using an LLM in a feedback loop with such a tool, except a coverage-guided fuzzer can see inside the tool to make potentially better guesses about what result you want. I claim random fuzzing should be "smarter" than an LLM for that purpose.

@pho4cexa I've found very little country music that I enjoy, but every once in a while there's an oddball… like https://soundcloud.com/songsbyjason/my-girlfriends-boyfriends-truck

@xhr@infosec.exchange To be fair, less introduced the & command for filtering lines in 2008, so everyone has only been missing it for 17 years rather than 25 (I say, as someone who also only learned about this feature recently despite over 25 years of using less)

@mstange@mastodon.social @brendangregg@aus.social We produce general-purpose DWARF (which register is this variable in, etc), which I believe happens to include enough for unwinding. I'm not super familiar with the details though, I could have some of this wrong.

@brendangregg@aus.social As another data point: In the Wasmtime JIT for WebAssembly we've had frame pointers turned on unconditionally on x86-64 for several years, and started doing that for all architectures a couple years ago because we need to quickly capture a stack trace if the guest traps. That said I'm confused by your suggestion that JITs can't produce DWARF, since Wasmtime does. That code is a regular source of bugs, and I'm not certain whether profilers can see the generated DWARF, but when it works, I believe at least gdb supports it. That doesn't change your point that DWARF isn't a great choice for profiling, and for the same reasons we don't want to use DWARF for stack-walking. But we've discussed that we may have to stop relying on frame pointers in the future to support WebAssembly exceptions proposals. We also don't currently do any inlining, so frame pointers can potentially be a significant hit if we have a ton of tiny leaf functions. So we still have open questions about what's the best thing for us to do.

@lain I loved how the author stated this "problem" with such certainty, then immediately after seemed to realize that readers don't actually want it solved, and tried to justify it by claiming there's no other feasible solution besides making advertisers happy. As far as I can tell, no matter how happy you make advertisers, they won't pay enough to keep publishers afloat, so who cares what they want?

@FeloniousPunk I keep hoping

@SnoopJ I don't have an answer but after your edit I've realized I'm also confused about the question. By "professional organizations", do you mean along the same lines as, say, ACM or IEEE, but specifically for queer/trans devs? Is the goal to find peer support, to find work, to get certifications, or something else? Does location matter? Like, in my area there's a meetup called T4Tech of trans people who like tech and just hang out together every couple of weeks, and that meets some of those goals very well and others not at all.

@ansate It's a nice clinic with a pharmacy and phlebotomy, and there's a giant trans-pride flag hanging behind the front desk which I'm always happy to see. I've been pretty satisfied with all the providers I've seen there over the years, though my dad has not always meshed well with every provider he's seen. I rely on that clinic for almost all my medical care, and I'd happily recommend it. There's also an associated immediate-care clinic two blocks east which I've needed a few times too.

@FeloniousPunk Oh, I've participated in rocket launches real close to there, but never made it to the observatory. What led you to it?

@Elizafox I think I was one of the people who voted to remove that language from the Oregon constitution. That would have been the first election in which I was eligible to vote and I don't remember for sure if I was on top of things enough to register to vote in time

@brendangregg@aus.social Yeah, as I said, you're right that DWARF isn't a good choice for profiling, for several reasons. It just isn't literally impossible, that's all. 😁

@agnes Not me, but I think it's really neat that they offer this: I bet showing up with a buddy helps motivate both people to keep doing it, so I love that they priced it to encourage that. I hope you find someone you want to train with!

@patrick I'm happy to see another SNI middlebox! I like the concept; I wrote about the advantages a few years ago: https://jamey.thesharps.us/2020/04/24/most-general-reverse-proxy/ How are you using eBPF to make the client address visible to the backend? I had looked into that but it seemed like the best I could do was passing SKBs from one socket to another, which I thought wouldn't carry the remote address info along. I also considered using Unix socket file descriptor-passing, but then needed some way to peek at the initial part of the stream without consuming it, and also would have needed to modify backends to get incoming sockets that way instead of by calling accept. So I'd love to hear how you apparently got the best of all worlds!