Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
qhoxie
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
23 ms
·
181.
▲
by
qhoxie
16y ago
This is something we are quite proud of. You will get to read more about how it all came together when some of our engineers like matthiaskramm publish their blog posts.
182.
▲
by
qhoxie
16y ago
We have some ideas about hooking the slider in with thumbnails. What we think the slider affords us is a nice way to give more context to moving around a document. Currently, it gives page numbers as you move and snaps to pages. These ar
183.
▲
by
qhoxie
16y ago
I literally just changed this off of that view mode! We have a number of documents that we are launching this with, and we wanted to showcase a number of different view modes.
184.
▲
by
qhoxie
16y ago
Really glad you are enjoying it. Are you talking about the toolbar auto-hiding? That was a decision we went back and forth on, so I would imagine it is not set in stone. Would you prefer it expanded at all times?
185.
▲
by
qhoxie
16y ago
There are some oddities with selections because of layering, but this is something I often notice in native PDF viewers. Have you seen any particular patterns which cause this?
186.
▲
by
qhoxie
16y ago
We have discussed some really cool things to do with page thumbnails. It is one of the features I am particularly excited about, because I feel that navigation of online viewers has a long way to go. How do you like the slider/scrubber in
187.
▲
by
qhoxie
16y ago
I might be missing your point, but Scribd does not do any drawing. Any non-text elements are represented as images.
188.
▲
by
qhoxie
16y ago
IE6+. In less-compatible browsers there are workarounds to implement the same features, so the UX is basically the same.
189.
▲
by
qhoxie
16y ago
We work with a significant amount of data and traffic, so our problems stay very interesting. If anyone has questions for a Scribd engineer, you are more than welcome to contact me. Email in profile.
190.
▲
by
qhoxie
17y ago
Yes, that is basically the point and the reason they call it "instant" - they don't want anything to be a click away. The opt-out prompt is in the form of a blue bar at the top of the page. Once you say "No thanks" they are required to rem
191.
▲
by
qhoxie
17y ago
Many (all?) of the distributed features (including the sharding) are part of gizzard, which it sits on top of. http://github.com/twitter/gizzard
192.
▲
FlockDB: Twitter's distributed, fault-tolerant graph database
(github.com)
79 points
by
qhoxie
17y ago
|
23 comments
193.
▲
by
qhoxie
17y ago
Nick gave a nice response in the comments. http://romanroe.blogspot.com/2010/02/why-i-hate-everything-y...
194.
▲
by
qhoxie
17y ago
They didn't expand much on the improvements to actors and support for continuations, both of which are quite interesting. Actors: http://www.scala-lang.org/node/2041 Continuations: http://www.scala-lang.org/node/2096
195.
▲
by
qhoxie
17y ago
In the post, he links to the draft of a book he is writing about libevent. Looks to be worth following: http://www.wangafu.net/~nickm/libevent-book/
196.
▲
by
qhoxie
17y ago
Ruby and Shoes are a great pair to introduce people to code. Quick and elegant results.
197.
▲
by
qhoxie
17y ago
Sounds like an interesting idea for some scenarios. You should definitely give it a try. Assuming you do and it's open, drop a link to it on here.
198.
▲
by
qhoxie
17y ago
I share many of the author's feelings with one notable exception. He praises testing libraries in his pros section. While they show promise, I find these libraries to be in need of maturing. This is obviously true for much of the Scala e
199.
▲
by
qhoxie
17y ago
If you are not too familiar with functional paradigms, many of the code examples seems obscurely clever . The nice thing about Scala is there is always a procedural way of doing the same operation that is more digestible for most people.
200.
▲
by
qhoxie
17y ago
That's not the point. I don't think anyone is arguing against what he says about worst case for a well-implemented hash table. The problem is the straw-man: all of the 'bad implementations' he has seen, which is all he really attacks.
201.
▲
by
qhoxie
17y ago
It seems unproductive to respond to this discussing how difficult it would be to fix/implement. I think Stroustrup probably knows that better than most anyone else, given, as he noted, his unique perspective. What is productive is gaining
202.
▲
by
qhoxie
17y ago
You probably mean http://www.metalabdesign.com They are quite good.
203.
▲
by
qhoxie
17y ago
These are surely aspects that differ based on the coder, but here is my take on it. I have a strong Ruby background and the last few months I have been using Scala alongside Ruby. 1. In short, yes. I think the key for me is using a mix o
204.
▲
by
qhoxie
17y ago
I agree for the most part, but I would not say EC2 is consistent. Maybe if you are light enough that you don't push its limits, but if you do, you see spikes. This and the fact that nodes frequently (at the scale I spin up) don't fully bo
205.
▲
by
qhoxie
17y ago
Anecdotally, that is not true. I work with a reasonably sized team on a large website. We are collaborating with a well-respected design firm and our teams are both using Sass/Compass. Thus far, I would say it has been a great success. V
206.
▲
by
qhoxie
17y ago
The most basic area where the evented model performs poorly is the most obvious one given its architecture - slow transactions. Depending on the set up, the event loop could be completely blocked by a slow transaction, thus nullifying many
207.
▲
by
qhoxie
17y ago
For well-polished basics, I don't think you can do any better than posterous.
208.
▲
by
qhoxie
17y ago
We have 2 offices. Our main office is here in Portland, but we have another in Lahore, Pakistan. The dev team here is a really solid group of 6 people. We wear many hats and likewise have many responsibilities.
209.
▲
by
qhoxie
17y ago
AboutUs is hiring full-time developers in Portland, OR. Most of our work is in Ruby and Scala. We routinely spin up 100+ instances in EC2 to do long-running tasks and are always investigating new technologies to help solve our problems. If
210.
▲
by
qhoxie
17y ago
This really is true and under-appreciated. Carsten's (raster, rasterman.com) early work on EFL was fantastic, and even though the delays of E17 became a bit of a joke in some circles, the progress was always impressive.
More ›