Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
sdan
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
43 ms
·
451.
▲
Frances Arnold Joins Alphabet Board of Directors
(twitter.com)
1 points
by
sdan
7y ago
|
0 comments
452.
▲
by
sdan
7y ago
This is why I think Keybase is so awesome. I've gotten some of my friends on it and so far the encryption/exploding messages and all is working great.
453.
▲
by
sdan
7y ago
Fairly sure he'd change the last names or something to conceal who they are. Otherwise, they'd eventually be recognized.
454.
▲
by
sdan
7y ago
They're using Python and webpack. Not sure what they were using before and unsure if Python is really dragging them down, but I like their design (although watching videos on mobile and going through posts on desktop is extremely gimmi
455.
▲
by
sdan
7y ago
Added a gradient to spice what I wanted to be a plain website: https://sdan.io/surya
456.
▲
by
sdan
7y ago
I like that email response time graph... thinking of doing that as well when I have time.
457.
▲
by
sdan
7y ago
Or you can build your own CDN with Cloudflare Workers: https://developers.cloudflare.com/workers/tutorials/configur... Did it with https://sdan.io/cdn (I host a ton of files so my other websites c
458.
▲
by
sdan
7y ago
I really like https://distill.pub 's design. Which is why I remastered it into a Ghost theme and put it on my blog: https://sdan.io/blog
459.
▲
by
sdan
7y ago
So does it replicate what Mujoco and Pybullet do? What exactly do you mean by simulation?
460.
▲
by
sdan
7y ago
"Fortunately, Uber AI’s Plug and Play Language Model allows researchers to make use of the few pretrained models out there: rather than requiring everyone to train their own wooly mammoth, PPLM lets users combine small attribute models
461.
▲
by
sdan
7y ago
Yeah I'm lacking a ton of stuff. No GUI, nothing in fact. But the tracker I've implemented (not into prod yet) does get under 90ms globally. Your service, which I've seen is serverless, is actually pretty fast compared to oth
462.
▲
by
sdan
7y ago
Actually addressed that on the website. Currently I'm getting an average of 1 second globally which is super slow. I'm going to migrate to the faster branch soon, which wouldn't be much of a problem, but am going to release a
463.
▲
Plug and Play Language Model: Steer topic and attributes of GPT-2 models
(github.com)
44 points
by
sdan
7y ago
|
3 comments
464.
▲
by
sdan
7y ago
I use the Developer Program to create Personal Access Token pretty frequently, what will happen to that?
465.
▲
by
sdan
7y ago
As a student (and a developer) I see paying $14 a month a huge expense. To really promote the privacy and accessibility for all, I'm building this to make it free and will absorb the compute costs if I have to for the sake of value for
466.
▲
by
sdan
7y ago
That's a lot of great points. Actually was thinking of doing something like this and now I see there's even more potential to help people host stuff quickly and efficiently.
467.
▲
by
sdan
7y ago
I can't either. As someone who just wants simple, private, and cheap analytics, I wanted a product that could do so (all current products are pretty expensive). That's why, I'm building https://sdan.io/pingpon
468.
▲
by
sdan
7y ago
It definitely is. As someone who just made some simple HTML sites on Github Pages, I just want some simple analytics without having to pay much. That's why, I'm building https://sdan.io/pingpong , which should be f
469.
▲
by
sdan
7y ago
In my experience Matomo is pretty bulky and hard to use. At the same time, Simple Analytics and Fathom are relatively expensive (but at a business standpoint, their pricing is good). Looking at these as a developer, I'm building https
470.
▲
by
sdan
7y ago
It definitely is. I hacked it together a couple nights ago and left the mess there. Working on the backend to make it faster and reliable and I'll fix up the frontend.
471.
▲
by
sdan
7y ago
Given that I'm targeting the developer community, I understand that paying for stuff like this is a huge turn off (at least that's one of the main reasons I'm building this). In the long term I think I'll try making this
472.
▲
by
sdan
7y ago
Check out Logflare. In either case, I'm building a cheaper and faster version of Fathom ( http://sdan.io/pingpong ). Currently I'm onboarding some customers. I'll try to see if I can incorporate web access logs
473.
▲
by
sdan
7y ago
I'm building one that's both cheaper (mostly free but at most $2 a month) and faster than Fathom. I'm onboarding some customers but I'm planning to launch in January! http://sdan.io/pingpong If you want
474.
▲
by
sdan
7y ago
Looks pretty interesting (automatically deleting duplicate files). Bookmarking to see if I can utilize this!
475.
▲
Wayve – Learned Urban Driving
(wayve.ai)
2 points
by
sdan
7y ago
|
0 comments
476.
▲
Facebook built a chatbot to help employees deflect criticism over the holidays
(theverge.com)
4 points
by
sdan
7y ago
|
1 comments
477.
▲
by
sdan
7y ago
Can you point me towards getting an AWS/GCP instance for around $20/year? Is it AWS or GCP? Also what type of instance? What are the specs? At the moment I just use EC2 and GCP instances (and a couple of GCS buckets)
478.
▲
by
sdan
7y ago
Agreed. I was somewhat confused on why they had more focus on their Podcast than telling me it was an AWS for startups/people. Maybe sometime in the future when I have time I can go through these long-form types of content, but they sa
479.
▲
by
sdan
7y ago
This is a bad landing page. 1. Bad HN title (doesn't say anything about what they did about the computer they made) 2. Landing page image has some generic background and as far as I saw, no tangible product, other than some shelves of
480.
▲
by
sdan
7y ago
Yeah, exorbitant. Didn't like paying $300 a year just to run some simple websites/databases
More ›