jhgg

8 ms·

61.
by jhgg 3y ago
Some would argue that it allows for cutting edge research that could potentially massively benefit humanity. Whether or not that pans out is to be determined, but that is the stated goal. The point then, as advocates for this technology bel
62.
by jhgg 3y ago
Not replying to something isn't pushing back. Chances are this one feedback post item got lost in the sea of them. If it's any consolation I'll reply to this feedback post saying it's planned and in progress. As for the
63.
by jhgg 3y ago
Yubikey support is coming soon to users and is being internally tested on staff right now. Also, I'm not sure what's given you the impression that there has been any push-back on the feature?
64.
by jhgg 3y ago
What is a bit unclear which isn't addressed in the original blog post is how they actually were able to successfully scan over all the data and copy it to the new cluster, in spite of corrupt sstables. CDC would not handle back filling
65.
by jhgg 3y ago
Yes because the flag is on your account.
66.
by jhgg 3y ago
They're called servers because we wanted to pick words that were familiar to what we were originally competing against, which was team-speak/ventrilo/mumble servers. Picking words people are familiar with in order to convey a
67.
by jhgg 3y ago
I don't think they took offense to the "rita" part.
68.
by jhgg 3y ago
Yeah it's a pretty cool design. Discord's real time system operates on a similar (although much more complicated system) on top of hash rings as well.
69.
by jhgg 3y ago
Discord uses websockets on all clients/platforms.
70.
by jhgg 3y ago
If it feels problematic, it's probably because you are missing the fact that there is a control and treatment group of this experiment to control for what you are saying.
71.
by jhgg 3y ago
These are all reasonable things to do at compile time, however for historical reasons since go did not have generics till recently, this is probably the best they can do without compiler specific specialization (eg what was done prior for m
72.
by jhgg 3y ago
It would be really awesome if code search could one day consume LSIF for precise results in its index similar to source graph. The symbol search is good now, but approximate. Having more precise code search by allowing devs to upload LSIF d
73.
by jhgg 4y ago
cloud sql proxy makes connecting to cloud managed postgres instances pretty simple/easy. honestly, we dont have a "great" reason other than, we set it up a long time ago, and it's continued to work haha. so pgbouncer con
74.
by jhgg 4y ago
nah we run a pool of common pgbouncers that route to the various postgres instances we run. they also run google's cloud sql proxy as well. it's actually haproxy -> pgbouncer -> cloud sql proxy -> postgres. everything but
75.
by jhgg 4y ago
at work we run pgbouncer on 4 vCPU (GCP c2-standard-4) machines, at 60% load, and they do ~15k TPS at peak, however we only peak at ~45k connections per node. however, i think that you may have a lot left on the table in terms of optimizati
76.
by jhgg 4y ago
I am curious if GPT-4 would have performed better.
77.
Structural Alignment of LLMs with ControLogits(github.com)
1 points by jhgg 4y ago|0 comments
78.
by jhgg 4y ago
I was mulling over the idea of building keyword image search (say, with CLIP based embeddings). However I'm not really sure the cost, or whether or not this is the best solution. Do you have any case studies about large deployments of
79.
by jhgg 4y ago
Does Zilliz do BYOC?
80.
by jhgg 4y ago
These are optional features. You can decide whether or not you want to use them. No one is forcing you to do these checks at compile time. My point is, the capability is useful to some people, and there are many other ways that doing arbitr
81.
by jhgg 4y ago
Then avoid crates that do such things. Other people however are able to make use of compile time code execution to do some pretty awesome things. For example, a database library sqlx can check all the SQL in your code as being syntactically
82.
by jhgg 4y ago
The spark migrator was needlessly complicated in pursuit of being general. It is also much slower. Ours is quite simple! We take the token range (int64 min -> int64 max), and sub-divide that into N equally sized ranges within the range t
83.
by jhgg 4y ago
Do you know of any service that is running at scale out there that stores their data as 5mb plain text files?
84.
by jhgg 4y ago
Because it would be obscenely expensive (in terms of computational resources) to search history via grep. Hence why we use an inverted index.
85.
by jhgg 4y ago
Discord originally hosted on Linode in 2015. Unfortunately we had to move off, as we knew Linode would be the limiting factor in being able to continue to grow. So we chose to go with GCP. Unfortunately, I suspect this is something that hap
86.
by jhgg 4y ago
And this is why we should all adopt webuathn, and get rid of totp based 2fa. This attack vector is significantly harder to pull off if a hardware authenticator will assert that the user is logging into the correct domain.
87.
by jhgg 4y ago
I open sourced a very minimal proof of concept for this here: https://github.com/jhgg/hyrule which shows the syntax.
88.
by jhgg 4y ago
We used to use Smyte at my place of work prior to the twitter acquisition. Very cool stuff. Ended up rewriting it in house. We ended up with a SQRL variant that is a restricted type-checked subset of python. It is not python however, just
89.
by jhgg 4y ago
Correct.
90.
by jhgg 4y ago
Hey all, I work at discord and specifically have been tracking this issue. We recently released support for AV1 hardware encoding for go live streams. However, the process of probing the GPU for what codecs it supports has triggered a laten

More ›