xena

25 ms·

541.
by xena 2y ago
I'm really concerned by this as someone who has recently broken into DevRel. I understand that there will always be jobs for someone like me, but all of the weird skills I have kinda add up into DevRel. It's kind of a dream job fo
542.
by xena 2y ago
It's mostly software as of late, but there's bits like this: https://xeiaso.net/blog/2024/the-layoff/
543.
by xena 2y ago
Speaking as someone that has an "everything blog" ( https://xeiaso.net ), it's been fun. I've actually been considering splitting things out so that some of my joke series are on another blog (such as the one p
544.
by xena 2y ago
God yes, I made a personal API thing called mi that I've maintained over the last few years. It manages a few things, but the main thing it does now is repost my blogposts on a few other social channels when I publish new posts. I&#x27
545.
by xena 2y ago
Generally these things are useless for 3d artists because the wireframe is useless for them.
546.
by xena 2y ago
Oh, that's a satirical series referencing a famous onion article series: https://en.m.wikipedia.org/w/index.php?title=%27No_Way_to_Pr... I really do need to write that nuance post about why I think we should consi
547.
by xena 2y ago
I've never heard of this issue happening. Can you send me an email with more details?
548.
by xena 2y ago
Yeah, I try to compensate for this with a few rules that I undoubtedly fuck up: * Each character has very visually distinct designs (ideally each is as distinct as I can reasonably make them, to the point where I have different artists do e
549.
by xena 2y ago
You hit the nail on the head. This is exactly why I do these things, but the only difficult part is figuring out where to apply this experience and how to phrase things without sounding like a crazy person in the process. "Yeah I got b
550.
by xena 2y ago
This is something I'm aware of and need to fix when I have the time. Just need to figure out how to best write this migration. I know exactly what I need to do, just gotta write code to do it.
551.
by xena 2y ago
That way I don't have to rebuild the binary to update the website. This makes posting new articles take tens of seconds at most.
552.
by xena 2y ago
My website binary builds the static site on every commit to the repo. This putting things in a zipfile strategy is probably a bad idea, but it makes it easy for the preview site to handle: https://xeiaso.net/blog/2024&#
553.
MIME, RSS, and Existential Torment(xeiaso.net)
90 points by xena 2y ago|28 comments
554.
by xena 2y ago
It is a very hard thing to calibrate. I try to have my internal conflicts and the like get put onto the canvas. I imagine that part of the perceived distraction is because of the CSS being bad (I suck at CSS so much). I'll go dig throu
555.
by xena 2y ago
They're called kaomoji https://en.wikipedia.org/wiki/Kaomoji (face-character)
556.
Building a Constellation of Images with Earthly(xeiaso.net)
2 points by xena 2y ago|0 comments
557.
by xena 2y ago
Some of the Mastodon traffic is because of https://masto.host
558.
by xena 2y ago
The only exception is if the microcode responsible for this was written in C somehow. I highly doubt it though.
559.
Function calling in large language models(xeiaso.net)
2 points by xena 2y ago|0 comments
560.
by xena 2y ago
Really good weights that you don't have to run yourself
561.
by xena 2y ago
It's the number of tokens the model can output in one pass. There's subtle differences between running it multiple times to get a bigger output and running it once to get a bigger output. These are things that only really show up
562.
by xena 2y ago
Flatpak isn't really intended for command-line apps. Maybe this is where something like dockerc ( https://github.com/NilsIrl/dockerc ) can shine?
563.
by xena 2y ago
Everything's a tradeoff, but it seems that part of the tradeoffs include access to tools critical for your product to function correctly being taken away without a way to get it back. Maybe that can be an acceptable tradeoff, but I&#x2
564.
by xena 2y ago
I'd say it's less a price concern and more a consistency of output concern. It doesn't make much sense to continue incomplete JSON like that I don't think. I need to do some more research.
565.
by xena 2y ago
The main limit is that you can have 128k tokens of input, but only 4k tokens of output per run. gpt4-32k lets you have up to 32k tokens of output per run. Some applications need that much output. Especially for token dense things like code
566.
by xena 2y ago
One of the decent parts about open weights models is that you can't have this happen to you. You get to keep access to any model you want. This is essential for continuity of products.
567.
by xena 2y ago
I mean for their large language model training. They said they don't include low quality data and swearing. This means you can get out of it by swearing.
568.
by xena 2y ago
All you have to do is drop a token swear word into your content and they remove it from the dataset. Easy.
569.
The magic of fall-through caching with Tigris(xeiaso.net)
2 points by xena 2y ago|0 comments
570.
The H Programming Language(h.within.lgbt)
2 points by xena 2y ago|0 comments

More ›