glimcat

44 ms·

421.
by glimcat 15y ago
You don't really love computers unless you're working out every bit of the quantum mechanics necessary to design and model its integrated circuits on a chalkboard.
422.
by glimcat 15y ago
Complains about the camera's sensor, blames the lens. Wait, what? And what is objectively wrong with it? The big red text complains about "color reproduction" - does this guy know anything about photography? Neither camera is color-calibrat
423.
by glimcat 15y ago
A job post needs to tell me what problem you're trying to solve by hiring and why I should care about your problem. The overwhelming majority of job posts completely fail to outline a clear value proposition.
424.
by glimcat 15y ago
"C is faster" is a tired argument. The ability of this compiler or that to produce the most efficient possible code is almost never relevant outside of a few niche cases. Even UHF finance is such that the fractional return in computation ti
425.
by glimcat 15y ago
I think we're taking slightly different lessons from that post. iOS = $5914, Android and iOS = $10342.08 Between the question of statistical significance and the vagaries of mobile app promotion and discovery, it's hard to say that one clea
426.
by glimcat 15y ago
Python and Haskell for development. Weka and R for saving loads of time. Some competency in Java and C helps.
427.
by glimcat 15y ago
"They typically give a lecture about it to new PhD students since its very common in acadamia to feel like a fraud when starting out." They really do, you really will, and it will likely pop up again later as stress points emerge.
428.
by glimcat 15y ago
I know it's usually that, but it's at least nice to see them trying to figure out what motivates people. I'm still much more likely to bite if they actually say straight out what a few of those problems are instead of spewing a list of ev
429.
by glimcat 15y ago
Only if they're boring and badly written. Or if you're spammy about it. When in doubt, hop on IRC and see if someone else thinks it's worth submitting.
430.
by glimcat 15y ago
That pretty much matches my impressions. * The DB class is a solid introduction to the field. * The ML class is good as far as it goes, but it lacks the depth to be more than an overview. * The AI class has good content, but the attempts at
431.
by glimcat 15y ago
The surest road to "better" is often simply to optimize for a different segment of the market.
432.
by glimcat 15y ago
Some of them work well, but some of them clash or have poor contrast.
433.
by glimcat 15y ago
There's a lot of money in micropayments and there are loads of problems with existing solutions from the perspective of virtually everyone involved. The compromise solution tends to be purchasing service-restricted credits in larger increme
434.
by glimcat 15y ago
Not enough info before signup to motivate signup. Color choices need improvement.
435.
by glimcat 15y ago
Engineering is 99% knowing what to glue together.
436.
by glimcat 15y ago
1. Find people who know things you want to know. 2. Have coffee with them. 3. Repeat.
437.
by glimcat 15y ago
Salesforce? 37signals? Expanding sideways is pretty common among successful software and business services companies.
438.
by glimcat 15y ago
Collaboration. Augmented reality. Having a big monitor when you need it and a more apparently spacious work environment when you don't. What it's useful for really depends on the specs of the screen. It can be faked to some degree with a we
439.
by glimcat 15y ago
Crappy documentation.
440.
by glimcat 15y ago
Your graphs need work. What you've got is a series of linear ranges at different orders of magnitude. The peak in the middle of the x range is probably smaller than the Forrst peak, while the area under the HN peak is probably larger than e
441.
by glimcat 15y ago
Dozens of conflicting ways to do the same thing, none of which are adequately documented. Stuff left in despite the fact that it does absolutely nothing of use. Random feature bloat. Bugs in key features which go unaddressed indefinitely wh
442.
by glimcat 15y ago
Either's good experience, just different experience. I know a number of people who did both at once.
443.
by glimcat 15y ago
Yeah. There are plenty of risks ranging from the minimal (they waste a little bit of your time and you don't get paid) to rather extreme (they waste loads of your time for years and you lose your IP to legal wrangling). Book a competent law
444.
by glimcat 15y ago
Are there intractable IP concerns as a result of the old partner and investors? If not, your problem is "I don't have traction after a year." Without more information, it's hard to speak to that.
445.
by glimcat 15y ago
Quitting is a little bit better than not getting in to begin with. Getting an MS is getting an MS - loads of people "quit" grad school this way and it's rare that anyone outside sees it as more than you having an MS. Also, try talking to yo
446.
by glimcat 15y ago
Thousands to many millions. Your problem definition as it stands is overly broad. What are your goals for the project?
447.
by glimcat 15y ago
It takes a lot of time to become good at this. HOWEVER: I strongly feel that anyone who is going to be working with or managing programmers should spend at least some time learning to code. You will not be able to do their job, but you will
448.
by glimcat 15y ago
Buggy. It also pops up a nag dialog when you try to close it, which is a crucifixion offense in my books.
449.
by glimcat 15y ago
"leave AI to Popular Science magazine" Really? I guess I'll go tell those guys working on the video games and computational finance to pack it up. Can't have people developing expert systems, natural-language based search agents, or route-p
450.
by glimcat 15y ago
Sometimes API calls can be a meaningful expense, depending on what you're using and how. Weird resource usage can sometimes be a problem if you don't watch it. Infinite recursive operations, brute-force algorithms being called over and over

More ›