Drew DeVault

I write code

Natural language proficiency:
★★★★★ 🇺🇸
★★★☆☆ 🇯🇵
★★☆☆☆ 🇳🇱
★★☆☆☆ 🤟
★☆☆☆☆ 🇫🇷

Does not want to talk to you about cryptocurrency or AI

Website: https://drewdevault.com
Code: https://sr.ht/~sircmpwn
Location: Netherlands
Pronouns: he/him

Public Key

npub1d88002d7uxl3cg9pwtgsnpgxs7qrmrgetqara0p4kw7d3yy4r50qtz57ex

NIP-05 Address

drewdevault_at_fosstodon.org@momostr.pink

Profile Code

nprofile1qqsxnnhh4xlwr0cuyzsh95gfs5rg0qpa35v4sw37hs6m80xcjz2368spz3mhxue69uhhyetvv9ujuerpd46hxtnfduqs6amnwvaz7tmwdaejumr0dsuhuxka

Publishing to

relay.damus.io

nos.lol

Author Public Key

npub1d88002d7uxl3cg9pwtgsnpgxs7qrmrgetqara0p4kw7d3yy4r50qtz57ex

Show more details

Last Notes

2024-11-23 14:25:37 UTC

- reply

the system requirements I've seen for a relay are perfectly accessible for small organizations and enthusiasts. Apache could easily afford to run one, or your university

2024-11-23 12:55:12 UTC

- reply

yes, I agree... but if you can run a relay and it interconnects then that's still federation, even if it's costly. And I don't take costly as axiomatically bad

2024-11-23 12:39:56 UTC

- reply

fediverse people tend to think that any design other than ActivityPub doesn't count as federation and I'm kind of tired of it

2024-11-23 09:36:23 UTC

- reply

And for the record I think a BlueSky relay would cost a lot less to operate than many people seem to believe, though I don't think it's going to scale well so that won't be true soon. In short anyone who plugs the minimum requirements into the AWS price estimator or whatever is doing it wrong. At this scale it's much more economical to buy a server and colocate.

2024-11-23 09:34:37 UTC

- reply

If the minimum node size raises the barrier of entry to that of, say, someone with a budget of about $50k annually to spend on federation, I don't necessarily feel that this condemns the model. In fact there are advantages to this approach, in particular it makes moderation easier.
If the price floor to run a node is more like $1M then I'm not so comfortable with that.

2024-11-23 09:32:15 UTC

One thing which frames my thinking on federation is "what is the minimum node size". On Mastodon the sweet spot is about neighborhood scale, i.e. 100-500 users. The sweet spot is different from the hard limits; you can run a single user Mastodon instance but the trade-offs make less sense.
Some people feel axiomatically that a minimum node size of one, i.e. fully distributed/self-hosted, is or ought to be the goal; I disagree.

2024-11-23 09:03:15 UTC

- reply

They are mostly open source and promise to become moreso and they have kept their promises so far. They have a standard and their model is flawed but based on a genuinely open, federated protocol. I have some qualms about the privacy afforded by their model but that's neither here nor there in terms of open source strategy.
Is it strategic for us, the open source community, to say "this is good enough" to hop onto this platform and gain access to this growing audience of, frankly, normies?

2024-11-23 09:01:34 UTC

So a lot of breath has been used up on how federated or distributed or open BlueSky is or is not (the answer is, to the best of my understanding, "somewhat"). But I wonder: how does BlueSky fare if we evaluate it in terms of being willing to compromise to meet people where they are?
That's not to say being willing to *capitulate* to meet people where they are, I'm not about to hop on Discord. But *compromise*? Maybe.

2024-11-21 10:02:49 UTC

- reply

1. Appeal to a kernel maintainer to approve your changes in their subsystem, they say no
2. Complain about them to a general public which has no stake in the matter or domain knowledge about it
3. The gatekeeper is still in charge of the gate and now they don't like you
4. ????

2024-11-20 09:57:02 UTC

- reply

as I understand it, it is possible but onerous to run 100% of the federation stuff yourself. The relays are expensive to operate, but definitely within reach for an enthusiast. I think that's a pretty acceptable approach to federation.
Right now I'm just running a PDS and I'm running the web app locally rather than using bsky.app.

2024-11-20 09:46:13 UTC

I am evaluating bluesky, find me at drew.ddevault.org

2024-11-19 11:42:13 UTC

- reply

Minecraft physics is dodgy as fuck but if you're curious about it here's my implementation:
https://git.sr.ht/~sircmpwn/betamine/tree/master/item/engine/physics.ha
It's based on NSDex's 2016 reverse engineering write-up for TrueCraft:
https://github.com/ddevault/TrueCraft/wiki/Entity-Movement-And-Physics

2024-11-19 11:34:53 UTC

- reply

✅ Gravity and jumping
Not sure if I want to do water next or just move on from physics for now... could be a good opportunity to add walking sound effects
https://cdn.fosstodon.org/media_attachments/files/113/509/404/168/017/260/original/71ebf89ad4b1f67c.mp4

2024-11-19 10:30:18 UTC

- reply

Basic physics with terrain collisions. Will add gravity, jumping, and swimming physics shortly.
The physics simulation is designed to apply to any entity so it can be reused for mobs and falling sand and arrows and such once those are added.
https://cdn.fosstodon.org/media_attachments/files/113/509/141/346/379/912/original/a2156ed7eca45e29.mp4

2024-11-18 16:32:45 UTC

- reply

you'd be surprised how little 1,000 lines of code is haha

2024-11-18 16:30:50 UTC

- reply

would be nice to develop a nice creative commons asset pack, though!

2024-11-18 16:27:35 UTC

- reply

haha no they don't allow reditribution, hence the script for end-users to fetch it themselves

2024-11-18 16:26:05 UTC

- reply

these are official Minecraft assets
https://git.sr.ht/~sircmpwn/betamine/tree/master/item/scripts/fetch-mojang-assets

2024-11-18 16:19:31 UTC

- reply

About 13 total days of work, some days more than others, and currently ~8500 lines of Hare code

2024-11-18 16:10:12 UTC

- reply

https://cdn.fosstodon.org/media_attachments/files/113/504/826/153/910/653/original/8cb44de210e97ea7.png

2024-11-18 15:09:16 UTC

- reply

https://cdn.fosstodon.org/media_attachments/files/113/504/585/493/627/686/original/377bc5955a0c8bfc.png

2024-11-18 14:29:36 UTC

- reply

https://cdn.fosstodon.org/media_attachments/files/113/504/429/823/588/825/original/aa086fb7cdc40db4.png

2024-11-16 14:59:19 UTC

- reply

Maybe a fun thing to do in the foreseeable future would be audio. Probably should wait until some basic physics are there since the physics engine will be an obvious place to ask "what sort of block is at the player's feet"

2024-10-28 12:39:15 UTC

- reply

that's a LOT more than just "generics"

2024-09-13 14:01:03 UTC

- reply

Things appear to be coming back up. The issue is largely out of our control but the optimistic view is that it's back for good.

2024-09-13 13:57:27 UTC

We're aware of users experiencing issues accessing SourceHut. Our upstream network provider has reported an outage that affects us and is working on a solution. Ideally, a fix is expected within 15 minutes.

2024-09-10 17:10:29 UTC

- reply

per the standard true does not accept any arguments and that is important

2024-09-10 17:05:43 UTC

- reply

the entire purpose of true(1) is to unconditionally exit with status code zero

2024-09-10 16:50:04 UTC

https://cdn.fosstodon.org/media_attachments/files/113/114/282/958/974/392/original/1d504598a663272c.png

2024-09-05 18:24:16 UTC

Send some ehugs if you can spare them

2024-09-04 17:15:44 UTC

Property is an illusion, intellectual property doubly so

2024-09-04 10:29:16 UTC

- reply

yes, it doesn't work well

2024-09-04 10:22:02 UTC

- reply

only one that consistently works for me is xpdf and I hate xpdf

2024-09-04 10:16:24 UTC

I have had it up to HERE with PDF readers that cannot handle the ARM manual

2024-09-03 09:35:34 UTC

- reply

sourcehut's robots.txt might be a good place to start:
https://git.sr.ht/robots.txt
It has comments on why each entry is there

2024-08-31 10:15:04 UTC

- reply

I very much dislike the assumption that any serious greenfield osdev work is a lost cause. It's not.

2024-08-31 10:14:50 UTC

- reply

And honestly the idea that we *have* to do it in Linux to get anything useful done at a practical scale for kernels is a really fatalistic assumption that spits in the face of just about any serious osdev project and condemns us to live with the design issues of Linux forever -- and believe me Linux has a lot of design issues. Even making a new kernel based on its general design gives us an opportunity to make a ton of sorely needed improvements.

2024-08-31 10:13:52 UTC

- reply

Yes, the road I propose is not necessarily the fastest way to getting secure kernels in the hands of users, and it would be faster and easier if we could get Rust in Linux without having to play the political game. But given that the political game isn't going away, we have to consider the other road as an *option*.

2024-08-31 10:11:40 UTC

- reply

And it's entirely true that while the political environment of Linux is, on the whole, pretty toxic, that there are legitimate political challenges to overcome with respect to Rust-for-Linux and even absent the toxicity it's going to call for a lot of work which isn't material to the goals of getting Rust in kernel space.

2024-08-31 10:10:55 UTC

- reply

Yes, a better solution than leaving the Linux community to make your own way in the face of toxicity would be to address that toxicity head-on. And believe me I condemn in no uncertain terms the toxic environment of the Linux kernel. But a complete answer to this problem would be distracting from the goals of Rust-for-Linux, which is to say getting Rust in Linux, and if people are already burnt out I'd rather care for their needs first than put them aside to address the causes right away.

2024-08-31 10:08:54 UTC

Finding the kernels of truth among the bullshit and doing my best to respond in good faith: regarding the scope of a Linux-compatible kernel written from scratch... yes, it's large. But there are exactly zero real-world programs that use that utilize the entire ABI surface of the Linux kernel. You implement the parts that you need to run the software you want and it's useful for that purpose right away. e.g. a kernel that can run nginx is useful even if it doesn't have GPU drivers, for instance.

2024-08-31 09:45:22 UTC

- reply

even if the community's party line is an obstinate disavowal of the Rust Evangelism Strike Force, it's still an attitude that seeps into the discourse unchecked.

2024-08-31 09:43:15 UTC

- reply

the positioning of Rust and its memory safety guarantees as a moral imperative for software engineering has been an incredibly toxic and destructive framing for the Rust community to operate under

2024-08-31 09:34:41 UTC

- reply

"Drew is a passionate Rust hater, nothing he says about it should be taken seriously. He's also the author of a low-level C-like language (without safety assurances of Rust), so it's likely whatever he's saying is motivated by pure spite and jealousy."
"some of [Drew]'s opinions are truly insufferable"
"Drew DeVault's name has managed to stick in my head as someone I just can't stand"

2024-08-31 09:32:22 UTC

- reply

Some of the Rust community's responses to my blog post:
"Sometimes people [Drew] would benefit from thinking a bit longer than they did.", followed by moving the goalposts several miles
"I am a Drew Devault hater, so buckle up. [...] I never have the sense Drew has ever thought critically about his opinions at all"

2024-08-31 09:27:58 UTC

Jesus, the eagerness with which the Rust community jumps to answer my discourse not with any kind of good faith, but with personal attacks and put-downs and horrible ad-hominem nonsense is just... insane. It's almost as bad as when I post about "woke" stuff and the far-right descends on me with harassment and death threats.
Your community has a big fucking problem, Rust, and cosplaying as inclusive and healthy and welcoming is disgraceful in the face of reality.

2024-08-30 11:43:11 UTC

- reply

I'd like everyone to shut the hell up about driver compatibility. Seriously. It was incredibly fucking annoying when Linux didn't have a competitive driver suite, too, and somehow we still managed to get to the point where Linux does have great driver support.
Fuck's sake. You write the drivers you need and then use the hardware that's supported, and if you want to support new hardware you write new drivers, you do not need to implement every driver overnight.

2024-08-30 11:17:19 UTC

- reply

Anyone coming in to @ me with bad-faith insults targeting Rust developers is cruisin' for a blockin'

2024-08-30 10:57:41 UTC

- reply

Rust would be a really good fit for a big Linux-style monolithic kernel and they could realize a lot of gains very, very quickly IMO.

2024-08-30 10:57:19 UTC

- reply

I think there needs to be some time to heal but the Rust-in-Linux folks should really rally around a new Linux-compatible kernel project from scratch. It'd be wild. It would probably actually work, whereas Rust-in-Linux is still a huge gamble and a burnout machine.

2024-08-30 10:50:06 UTC

- reply

Making a novel OS design is hard (Redox is trying) but if you are aiming to match the design of an established system and maximize compatibility then it's much, much easier.

2024-08-30 10:49:15 UTC

A motivated group of Rust developers could build a production ready mostly Linux-compatible kernel from scratch within 5 years without doing any politics on LKML
Passionate 3-4 person hobby OSes have achieved impressive levels of Linux compatibility very quickly. Pour the level of enthusiasm that Rust-in-Linux has into a greenfield project to build a Linux-compatible kernel and you would have astonishing gains very quickly.

2024-08-28 09:06:24 UTC

- reply

radical is a value-neutral term IMHO

2024-08-28 08:18:23 UTC

- reply

trans-exclusionary radical fuckwits

2024-08-28 08:06:56 UTC

- reply

"Hare, a programming language created by Drew DeVault..."
Wrong, Hare is a programming language built by a community of over a hundred people, who collectively own its intellectual property, who are governed by a social contract that can be revised at any time by any person, including you, assuming you can build sufficient enough consensus to achieve your goals, consensus which can be built from within or without.
This is true of every FOSS project even if upstream believes otherwise.

2024-08-28 08:03:48 UTC

Daily reminder for FOSS participants to reject auteurs and embrace communal ownership
Especially if the auteur in question is a business

2024-08-28 07:53:16 UTC

"Microsoft donates the Mono Project to the Wine team"
What does that even mean? FOSS projects don't belong to anyone. Was there a trademark that was handed over? Do they mean that Microsoft isn't going to maintain it anymore and Wine volunteered to? Or Microsoft is pulling their sponsorship of Mono?

2024-08-25 21:19:23 UTC

- reply

when I was served a Scottish breakfast there was haggis on it so

2024-08-25 21:17:05 UTC

- reply

I will hand it to the Brits regarding the full English breakfast though, well done

2024-08-25 21:12:37 UTC

- reply

Also, chicken biscuit is easily one of the best breakfasts imaginable

2024-08-25 21:07:13 UTC

- reply

Any offended Europeans can come have a biscuit at my place, I'll make you some and you'll understand

2024-08-25 21:04:43 UTC

- reply

"You're supposed to put jam on it or something", listen here you little shit

2024-08-25 21:02:54 UTC

- reply

Why would I want to choke to death on a lame ass scone when I could have a buttery flakey soft moist biscuit (with gravy please)

2024-08-25 20:59:20 UTC

Scones suck, biscuits are where it's at, get it together Brits

2024-08-20 18:50:35 UTC

- reply

they don't deal with bad actors on their service and enable spam/abuse

2024-08-20 18:44:10 UTC

What's the least bad domain registrair? Asking for a friend, not happy enough with what I use now to recommend it

2024-08-13 08:50:37 UTC

- reply

me: does nothing
firefox: lemme pin this CPU
me: switches tabs
firefox: that'll be ready in 30 seconds
me: opens xkcd.com
firefox: aight imma crash

2024-08-13 08:48:35 UTC

Firefox is being a piece of shit today

2024-07-31 09:25:54 UTC

- reply

we'd use a managed thing like masto.host, I don't want to run an instance -- but yeah, moderation is a concern if we allow sourcehut users on board
That said we already do a lot of moderating, but this would definitely up the workload

2024-07-31 09:17:16 UTC

- reply

mastodon instance

2024-07-31 09:12:10 UTC

Thoughts on a potential soc.sr.ht fedi thing?

2024-07-24 15:52:24 UTC

- reply

Since there is some confusion about this, I never supposed that Hyprland was some kind of obscure "dark web" project. Instead I compared the discourse around Hyprland (and connections to it via Brodie, Lunduke, etc) to the broader social conception of the "Intellectual dark web", which I cited in a follow up to clarify:
https://en.wikipedia.org/wiki/Intellectual_dark_web
Happy to debate the applicability of this comparison but not really to debate something I didn't say re: "dark web"

2024-07-24 15:46:33 UTC

- reply

I agree that we should not misgender anyone, including those we disagree with; please desist and/or edit replies which did so.

2024-07-24 15:00:41 UTC

"manipulates facts to fit his argument"
"calls a project more popular than his, a dark web underground who-even-knows-this"
Fact check
https://pkgstats.archlinux.de/packages?query=hyprland
https://pkgstats.archlinux.de/packages?query=sway
I last committed anything meaningful to to sway or wlroots in 2020 and I officially handed off maintenance, so it's hardly "my project" either.
https://cdn.fosstodon.org/media_attachments/files/112/842/058/125/438/023/original/7265ea68da008bb6.png
https://cdn.fosstodon.org/media_attachments/files/112/842/058/126/552/521/original/cd425c6097148145.png

2024-07-03 09:35:57 UTC

- reply

obviously AWS would be a lot better if the platform were FOSS, but that applies to basically all nonfree software. However, AWS has invested a *ton* of labor and money into free software projects, much moreso than, say, Elastic

2024-07-02 15:35:20 UTC

- reply

5. The Free Software Foundation. #hottake
The one thing all of these issues have in common is that the FSF has no answers for them. The EUPL and MPL are doing a better job of facilitating copyleft than the GPL family. Savannah isn't a match for Codeberg, let alone GitHub. Most egregiously, the FSF has utterly failed to address diversity and social issues, especially sexism, and with RMS's position restored and maintained at the helm that is never going to change.

2024-07-02 15:32:55 UTC

- reply

yeah, or install a package if it's available for your distro. New release soon

2024-07-02 15:28:57 UTC

- reply

it depends

2024-07-02 15:26:48 UTC

- reply

Moreover, we need more people than just programmers. You know why we don't have an answer to Discord? A big part of it is that we don't have people doing visual design, translations, marketing, etc. Learn about people who differ from you and your expectations: their struggles, their victories, and their needs and what's important to them. Become a student of the other, and take your skills to your leadership and moderation roles, to make safer and more inviting spaces.

2024-07-02 15:26:14 UTC

- reply

4. Lack of diversity in FOSS. Yes, I mean the culture war sort, but also others besides. Most of the people reading this post still fit the mold: the relatively affluent white male cishet computer science major. We live in a time of struggle and if we don't foster solidarity with other political movements it's going to break down our doors sooner or later. We need different perspectives and backgrounds among our *peers* and *leaders*, and so far they're mostly relegated to users if even that.

2024-07-02 15:13:52 UTC

- reply

It's also exclusionary, and even unsafe, for many other marginalized participants. Many Discord servers are infested with far right recruitment and campaigns of hate and harassment. Many harassment campaigns are launched from Discord and racist and queerphobic sentiments go unpoliced. It's an issue with the platform, which retains ultimate control and refuses to use it to moderate this behavior, and with the culture, who develop small monarchies where these problems go unanswered.

2024-07-02 15:11:47 UTC

- reply

3. Discord. This is another proprietary platform and many of the same arguments regarding GitHub apply here. It's an incredibly popular platform, and it's no surprise to find it landing on our shores. However, it is especially toxic for FOSS in many respects: it seeks legal action to prevent anyone from attempting to build software on top of it. It's a walled garden: Discord is an incredibly exclusionary platform which has countless accessibility issues for the disabled, poor, and many others.

2024-07-02 15:06:33 UTC

- reply

Oh, and on this subject: AWS is not really a threat to FOSS, if anything it's mostly been a boon for us. Any time someone cites AWS while taking any of the four freedoms away from you, you should start asking some pointed questions.

2024-07-02 15:05:40 UTC

- reply

Moverover, we need to re-discover reciprocal licenses like GPL and EUPL and hold copyrights in common between all stakeholders. Which is why you should never sign a CLA!

2024-07-02 15:05:25 UTC

- reply

To address this we need to better educate people on how money and free software can co-exist in a way which does not threaten FOSS, and for people to learn and *grok* the social and economic dynamics of free software. The commercialization of FOSS is not a bad thing -- so long as the *critical* provisions that all users are equally entitled to share that wealth and capitalize on it on equal terms is upheld.

2024-07-02 15:04:41 UTC

- reply

2. Open washing, which is to say the proliferation of licenses that look FOSS if you squint but don't work if you look closer, and practices related to these licenses. Here we have big players like Elastic, Redis, MongoDB, and numerous smaller cases as well. The practice of building off of the lavish advantages of being in the FOSS ecosystem, then pulling the rug and seeking exclusive commercial monopolization of the end result.

2024-07-02 14:59:44 UTC

- reply

Consumption of proprietary platforms in our infrastructure erodes support for the very real participatory projects that exist and need your support and legitimacy to succeed. The solution is to eschew consumption of proprietary software, and build your infrastructure on free software platforms. This is a problem of critical mass, and it won't be reached unless enough projects are bold enough to take the loss of access to these proprietary platforms and their reach on principled grounds.

2024-07-02 14:59:37 UTC

- reply

1. GitHub, and to a lesser extent GitLab. These platforms commodify the FOSS community and minimize the social aspects that matter -- people over stars. They are proprietary (or open core) and consequently encourage a consumption culture with respect to FOSS: users consume GitHub, rather than participating in it; and this culture fans out to the broader culture associated with projects and people that work with it.

2024-07-02 14:58:19 UTC

🧵 Biggest threats to FOSS and some proposed solutions