Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
andres
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
22 ms
·
211.
▲
by
andres
15y ago
I'm getting a hit for "SML on Stilts" from 5 months ago: http://www.hnsearch.com/search#request/all&q=SML+on+Stil... Did you get different results? Were you expecting another submission?
212.
▲
by
andres
15y ago
The thing that excites us the most is the flexible schema because that can cut down on development time dramatically. The REST API is also optimized for developer happiness. We're working on some examples and should have them ready soon. Be
213.
▲
by
andres
15y ago
We use Python server-side and the Python implementation of Google protocol buffers is extremely slow.
214.
▲
by
andres
15y ago
I actually stumbled into thrudb a few months ago after we had already written v1. We talked to some people at Disrupt that are doing similar things with search and Thrift as well.
215.
▲
by
andres
15y ago
Would love to hear more about hummingbird db. Email me at andres@octopart.com. We chose Thrift because the schema was flexible and that was the biggest problem for us at Octopart.
216.
▲
by
andres
15y ago
Good point. We decided to make it available as a (free) cloud service just so we could get hacker feedback as quickly as possible. If we had waited until we had a version that was easy to install, it would have taken us a lot more time. If
217.
▲
by
andres
15y ago
ThriftDB uses Facebook's Thrift serialization internally so the schema is completely flexible.
218.
▲
by
andres
15y ago
In grad school I worked on IceCube which is a neutrino telescope at the South Pole. The detector is an array of photomultiplier tubes embedded in the ice underneath the Pole. At the time I was mostly working on electronic devices for calib
219.
▲
by
andres
15y ago
Thanks so much for the comment. It's great to hear that hardware hackers find Octopart useful. We have a lot of ideas on how to improve Octopart and now we can finally implement them.
220.
▲
by
andres
16y ago
In terms of the product itself, "Kayak of electronic parts" might be a better description. In terms of technology, the first thing we did was actually take a Google-like approach to the part search problem. For example, before we started wo
221.
▲
by
andres
16y ago
Our backend is already equipped to handle this and now we're starting to focus on getting manufacturer data feeds. If you have a chance, please send me an email to let me know which manufacturers you think we should prioritize (andres@octop
222.
▲
History for Dollars
(nytimes.com)
4 points
by
andres
16y ago
|
1 comments
223.
▲
Our Fix-It Faith and the Oil Spill
(nytimes.com)
5 points
by
andres
16y ago
|
0 comments
224.
▲
by
andres
17y ago
Originally I wanted to write it in Cappuccino and I actually started learning Cappuccino for that purpose. However, I had used JavascriptMVC for a previous project and in the interest of speed I decided to go with that. Also, I was developi
225.
▲
by
andres
17y ago
What do you mean by filter icon?
226.
▲
by
andres
17y ago
Thanks for the feedback. If you click on the magnifying glass in each column header you'll get a filter for that column.
227.
▲
by
andres
17y ago
I was planning to upgrade to JavascriptMVC 3.0 before open sourcing it but if you want to see the source as-is let me know.
228.
▲
LHC produces first collisions
(nytimes.com)
32 points
by
andres
17y ago
|
21 comments
229.
▲
Nuclear Bomb Detector Stopped by Shortage of Helium 3
(nytimes.com)
19 points
by
andres
17y ago
|
0 comments
230.
▲
by
andres
17y ago
haha! Thanks! We recently switched from a hacked-version of sphinx to solr. We're using some custom solr features (including some that we're contributing back) but most of the speed comes from solr faceting. It wasn't trivial to setup thoug
231.
▲
by
andres
17y ago
We have 9M+ parts in our database so we use algorithmic methods to parse and normalize technical information. In general, the data comes from many different sources including increasingly from the manufacturers themselves. We're not parsing
232.
▲
by
andres
17y ago
Fixed. Thanks.
233.
▲
by
andres
17y ago
We'd love to step aside to let developers build on top of the API. I'd even like to see someone revamp the Octopart UI! Here are some other ideas: - a really good project manager (with collaborative features and publishing tools) - br
234.
▲
by
andres
17y ago
Some background - We think that hackers with hardware experience are the best people to build engineering design tools. Unfortunately, it's quite difficult to get startup capital for design tools so there hasn't been much innovation in the
235.
▲
Octopart releases free API for developers
(octopart.com)
73 points
by
andres
17y ago
|
15 comments
236.
▲
by
andres
17y ago
thanks for asking mitERS for their input. please let everyone know to email me with suggestions on how we can help them search for parts (andres@octopart.com).
237.
▲
by
andres
17y ago
Google's documentation claims that protocol buffers are designed to be fast so I tested them out and I also found that their Python implementation was too slow to use. I posted a question to the Discussion Forum and found out that it's a kn
238.
▲
Financial Times Feels Vindicated by Web Strategy
(nytimes.com)
1 points
by
andres
17y ago
|
0 comments
239.
▲
Bay Area Science: Techs and the City
(economist.com)
1 points
by
andres
17y ago
|
0 comments
240.
▲
by
andres
17y ago
It's been a while since I learned something new from a nytimes tech article.
More ›