Live data from Hacker News

Slack LaTeX

blog.c0nrad.io

11–14 of 14 posts

Re: Slack LaTeX

#12
post #10
post #8

Earlier quoted context omitted.

> Interesting article. How much requests can you support for 20/month before losing money with this stack? Ha, I haven't profiled it yet. I'm not too sure. But, I think a decent amount. If a slack team with 10000+ physicists signs up, I may need to talk to them offline. I do plan on doing some stress testing in the future for fun, I'll try to remember to reply here. > partner frontend Yep, to repeat what steve_adams_…

Teams with this number of physicists exists only in CERN which uses self-hosted mattermost. Don't worry from that for now.

Thanks for the insight!

Re: Slack LaTeX

#13
post #11

On my Safari 14.1.2 the hero image (demo) on the landing page does not show properly. The URL is https://slacklatex.com/static/media/demo_img.5ddca5f3.webp and I assume the format is a bit too esoteric.

Ah thanks for the heads up. We'll wrap it in a html picture block. (Google's https://web.dev/measure/ recommends webp, but I should have done more research)

Re: Slack LaTeX

#14
this is a great idea! i'm all for making math typesetting more accessible and useful!

i am wondering why they use a headless to render from KaTeX though. Is there an advantage to doing that compared to, say, shelling out to xetex? You'd get better compatibility with potentially useful libraries like tikz that way (and potentially higher throughput/lower costs), at the cost of investing some thought in sandboxing. Curious about the considerations behind that decision.

Post reply on HN