lmeyerov

81 ms·

961.
by lmeyerov 5y ago
Oh wow, I read the post and thought congrats Moxie, job indeed done. We are still not there, so a lot of empathy for the feelings behind that :) ... Except if it is donations and esp. Brian Acton's, or even say Firefox with most of th
962.
by lmeyerov 5y ago
Yep, and apparently forced bc they give the $ on day 1. That's a lot of %, and before most founders understand what is happening, esp. in their target demographic....
963.
by lmeyerov 5y ago
It took me a second to work through. Basically, they just said they will always do some of their effective pro rata. So more weight on the cap table. Hopefully it isn't required, bc sometimes there isn't a lot of room, and their v
964.
by lmeyerov 5y ago
I don't think academia should be one thing, that's kind of the point of tenure You're promoting both a super traditional view yet also one that seems divorced from the last ~100 years of sociology studies wrt how society prog
965.
by lmeyerov 5y ago
A lot of systems research is socially guided where community-of-practice approaches help. If you haven't, I recommend reading something like Rogers' surveys of papers in 'Diffusion of Innovation' for why, especially befo
966.
by lmeyerov 5y ago
The student picks the PI, and as part of that, the field of methods they want to learn, which may include scrappy prototyping. If they don't think that kind of R&D is their thing, they should pick another PI and, potentially, field
967.
by lmeyerov 5y ago
My response is to the Reviewer #2 aspect of, instead of say championing a proposal on merit and, for their allocation issue, advocating for say increased budget for extra staff... rejecting it because of belief in Not My Job for something a
968.
by lmeyerov 5y ago
I've been on the other end of such reviewers. Building community and engaging in it takes a whole ton of seemingly non-core work, but community-of-practice research is totally a thing, and for folks who empirically study innovation, o
969.
by lmeyerov 5y ago
Agreed in part, I'd indeed want careful motivation and change management of such a change from my language or DB vendor -- JS/Python takes years for tinier semantic changes :) The average OSS project is just ~1 fulltime maintainer
970.
by lmeyerov 5y ago
I agree - we probably err too long on avoiding breaking legacy code: we are about to do our first deprecation in years as the old protocols are getting too unconscionably insecure + costly to maintain relative to our new ones. But that&#x27
971.
by lmeyerov 5y ago
It's the users who need to audit, the social contract of OSS is just not to be sneaky/sloppy about it (when posing as a serious project). Semvar is beautiful bc it lets you be explicit. Likewise, end users can judge "wow majo
972.
by lmeyerov 5y ago
Breaking changes are things, no project is ever done -- ye olde django is still doing big moves like adding async colors, and that's great! But... CHANGELOG.md with BREAKING CHANGE sections and a major version bump, maybe a DEPRECATED
973.
by lmeyerov 5y ago
A project being late involving a junior dev due to architectural and scoping issues is probably not because of the junior dev but the tech lead and product/project manager. The junior dev is still at risk in practice because the team i
974.
by lmeyerov 5y ago
Graphistry | Visualization engineer - multiple | Part-time + Full-time | SF / US / Global / Remote (contract / contract-to-hire) | https://www.graphistry.com/careers Graphistry is the first GPU visual gr
975.
by lmeyerov 5y ago
Startups are hard: even for ultimately winning ideas, likely to see poor win/loss rate, unknowns, few resources, no 40 hour work week, etc. with 5-10 years to exit or more likely, failure. Even when things go super well, they just go t
976.
by lmeyerov 5y ago
Most are: The mantra 3, 3, 2, 2 (annual) adds up to 9x revenue growth in first 2 years, and presumably, higher growth in upstream metrics to feed that . They will invest at 20-100x (and even 1000x+) multiples of revenue if they believe it.
977.
by lmeyerov 5y ago
Cool to see this stuff getting out into the wild, long time coming! What is the current state of these -- do simd wasm ops run in ffox/chrome/edge/safari, and does the data have to be slowly bulk copied back-and-forth from a
978.
by lmeyerov 5y ago
I think the intuition on processes vs threads was right: 1. The Arc use is a case (I believe! Am not a Rust'er!) of true sharing of references. The cacheline will be contention across cores. Doing processes would have eliminated true s
979.
by lmeyerov 5y ago
Hopefully, it should be a leading indicator for negative churn -- per-account and new account growth / virality -- and for more complicated products, a way to slice that across different personas and features. ('bakers at high-en
980.
by lmeyerov 5y ago
See the above example notebook If node or edge attribute information is useful for clustering, such as not just account IDs but say entity or event sizes & time, you may want to see sub-cluster separation based on that. A naive initial
981.
by lmeyerov 5y ago
This! We do it all the time in fraud, genomics, social media, security, etc data tables, it is one of our favorite techniques. We do one more step for UMAP and other dimensionality reducers in practice: connect the nearest neighbors to turn
982.
by lmeyerov 5y ago
Graphistry has a free tier, GPU stuff included :)
983.
by lmeyerov 5y ago
He is the editor (see line 2). For his direct research here, he wrote many papers on computational geometry / graphs - his dblp is quite long!
984.
by lmeyerov 5y ago
Roberto is fun! I didn't appreciate his work until way later =/ Funny enough, if someone is into modern versions of these graph techniques, we have been building out an end-to-end GPU visual graph intelligence platform used by all
985.
by lmeyerov 5y ago
Sort of, in a depressing way: - It's pretty doable to get someone's hours for this, especially within the contractor or one of their DC buddies: there's already a customer and a contractor, and presumably, some sort of purcha
986.
by lmeyerov 5y ago
Great, suspected something like that :) Interesting wrt workers -- how does TCP vs sockets start looking in a multiple worker + multicore scenario, esp. wrt peak QPS? That's more confusing for schedulers :) Also, FWIW, the GIL thing mi
987.
by lmeyerov 5y ago
I didn't look too closely, but I'm wondering if this is Python's GIL. So instead of nginx -> multiple ~independent Python processes, each async handler is fighting for the lock, even if running on a different core. So read
988.
by lmeyerov 5y ago
Yep. Professors are not only able to get an answer, but able to define the problem. It varies by region and domain for when that is meant to be true, so you need to understand the system a bit better. Ex: US CS PhDs from the Top 30 are gene
989.
by lmeyerov 5y ago
Conversely, I was just visiting one of the biggest fintechs (so not California culture), and after Omicron hit, they nixed their return to work plans. We were almost the only people on the entire floor.
990.
by lmeyerov 5y ago
consider $bigco making $10B year that's $20K/minute chances are, $shiny_acquisition isn't making $20K/minute, and woe is the helpful IT responder who kills the golden goose. it'd be crazy if anyone can do that, so m

More ›