sqrt17

48 ms·

391.
by sqrt17 16y ago
Optimizing for one screen resolution is a bad idea: people do have different screen resolutions, and will want the application to fit their needs, whether they live on a 1024x768 screen and pull the application window to 860 width because
392.
by sqrt17 16y ago
Seeing statically typed variants of Python such as ShedSkin or PyPy's RPython, I think that one major annoyance is gone. You still need polymorphism for some things, and having the possibility to have "implicit interfaces" (you just write x
393.
by sqrt17 16y ago
Knowing what you can do in Python, I guess they'd like something where you can't do the fancy metaclass things that you can do with Python, because they are rather tedious to implement - ask the Jython and IronPython people. None of their a
394.
by sqrt17 16y ago
Now, where are the people complaining that they can't install MacOSX properly on their cheapo laptop? Or that [important part from a nameless company in Taiwan] stopped being supported in MacOSX Hyena? I mean, if Apple were not heavy-handed
395.
by sqrt17 16y ago
He's not. Any program code written today that solves an NP-hard problem without exploiting special properties of the input data will take exponential time. If we discover a smart way to do NP-hard problems in polynomial time, we'll have a b
396.
by sqrt17 16y ago
Once the model is loaded, you can get around the ~200ms roundtrip latency that it takes to connect the server. Obviously, there's lots of ways to hide these 200ms from the user.
397.
by sqrt17 16y ago
I don't have a JS version, but here's a functioning HMM tagger in under 300 lines of Python: http://gist.github.com/503784 (model loading doesn't work yet for some reason, but you see what it's doing in principle). This uses a smoothed tr
398.
by sqrt17 16y ago
London is a rather expensive place to be bootstrapping a startup. What would make it more appealing than Birmingham (nearby, pretty big), Edinburgh (medium-size city, with one of the best UK universities) or any other of the big cities with
399.
by sqrt17 16y ago
Statistical methods do quite a good job of pushing that ambiguity back under the rug (see link to Berkeley parser further up, or the work on statistical disambiguation in Alpino). As to the time consumption and complexity, that's a known pr
400.
by sqrt17 16y ago
Darn, now I'm tempted to write a pure javascript POS tagger just to show that you don't really need anything server-side (or maybe just little bits here and there so the web page doesn't need to load a 20MB model right away - the computatio
401.
by sqrt17 16y ago
Pretty much. If you replace it with unambiguous words, go/VB lemon/NN yourself/PRP doesn't sound good, whereas go/VB shave/VB yourself/PRP is fine. IMO, go should also be a VBP, not a VB.
402.
by sqrt17 16y ago
Nope. Parts of speech of words are ambiguous and you need some statistical method of disambiguation. Plus, you need something to guess the possible tags for unknown words (e.g. speakers of English are able to guess that 'supercalifragilisti
403.
by sqrt17 16y ago
There's several central ideas in CS that most people know. (Not everyone knows every single of them, but most people know many of them). One of them is the idea of finite-state automata. If you have a finite set of strings, or a regular lan
404.
by sqrt17 16y ago
Actually, if you want to see whether the edit distance (Levenshtein) is within a fixed bound, say k, it's no longer O(mn), but around O(kn) (since you're only ever filling a narrow band of width ~k along the diagonal instead of the whole re
405.
by sqrt17 16y ago
The real reason why you want ->> can be seen in the code example for rule #7: instead of mapping the result of partitioning, it (wrongly) reuses a (i.e., the original list). If bugs result in irregularity, then making regularity exp
406.
by sqrt17 16y ago
One thing that always bugged me about air travel in the US is the time you spend on the ground, standing in queues, pulling your shoes off, putting them back on again. You'll say that this is for better security. It's not an effective way t
407.
by sqrt17 16y ago
Can someone who has more of a network theory background say why this would be interesting? From an NLP angle, both what they're doing (text classification) and how they're doing it (constructing a co-occurrence matrix) don't sound particula
408.
by sqrt17 16y ago
Actually, you can code very nice lightweight apps using an embedded servlet container like Jetty, a JSP library, and an ORM (MyBatis or Hibernate). Which is kind of like the idea of a microframework. I've used it for a demo I've made (some
409.
by sqrt17 16y ago
In my understanding, they could not, just as they could not force any other software pirate to do what they please. It's just tht opening GPL-violating code is probably a cheaper alternative than paying damages (and as far as I know, there'
410.
by sqrt17 16y ago
Karma is really a "black swan" thing: sometimes you get one or two points, sometimes you lose a point because people don't agree with you and downvote you, and once in a while you post something totally innocuous (but potentially well-infor
411.
by sqrt17 16y ago
Network latency is not that much of an issue when you have your application in the cloud as well. And I'll only believe claims about backups if the contract contains an enforcable clause that I'm entitled to a pound of the CEO's flesh if/wh
412.
by sqrt17 16y ago
Where do the magic numbers (p, b1-b5) come from? Are these empirically found using curve fitting? Also, wouldn't it be better to write things like (+ (- (* (exp (* (- r) t)) strike) s) call))) in a more schemer-friendly way as (
413.
by sqrt17 16y ago
Lesbians can certainly bear children by themselves, just as any woman could. By that point, it does not matter (for the biological notion of conception) if that woman is lesbian, heterosexual, asexual, or anything else. As opposed to that,
414.
by sqrt17 16y ago
As common wisdom goes, children are created through an act by one male and one female human, and subsequently carry about 50% of each partner's core DNA and 100% of the female partner's mitochondrial DNA. As a result, neither egocentrical m
415.
by sqrt17 16y ago
ouch. If you encapsulate all the data about a customer in a "CustomerData" object, do you really need a separate "Customer" object? Or do you mean, not the whole customer but larger bundles of information like addresses etc. that don't ne
416.
by sqrt17 16y ago
The idea is that, if you think the reader will just accept something as a fact, you don't need to hedge it. If your reader will see it as an arguable point (instead of an uncontroversial fact), then you should make (or at least sketch) the
417.
by sqrt17 16y ago
Agreed. Where I went to college (Germany), they also had waterless urinals. The biggest problem seemed to be that people would throw cigarettes and stuff into them.
418.
by sqrt17 16y ago
I for one like the concept of storing JSON-like data structures and using javascript-based indices on top of that, like you can do with CouchDB or MongoDB. For the greater part, having something where you can store big amorphous blobs is mu
419.
by sqrt17 16y ago
... as in, those people who use memcachedb but would like to have something bigger?
420.
by sqrt17 16y ago
GTK/Glib/whatever is more than just UI just as Cocoa/Foundation/whatever is more than just UI. Oh, and GTK runs on far more platforms than Cocoa (Cocoa runs on Macs and iPhones/iPads, whereas you'd be able to use GTK on Mac, Windows, deskto

More ›