mxh

... but, aside from the higher tool cost, slower development time, and more expensive hosting, it's great?

I don't think performance is really a reason to pick the MSFT stack; MySpace does use it, proving that it can scale, but there are many more cases of web-scale systems built on open-source tech. So that's, at best, a push.

It is easier to find warm bodies for a MSFT shop than, say, a Python shop. But, as a startup, are you looking for 5 great guys, or 50 mediocre ones? And do you really think it's easier to find 5 great MSFT guys than 5 great Python (or Ruby, or Java, or even LISP/Haskell/Erlang) guys?

The last point seems debatable to me, the others, less so.

I agree with this sentiment, and would add a few (somewhat) related reasons.

a.) 'Planet Microsoft'. By this, I mean the gravitational pull all MSFT products seem to exert on one another, apparently by design and decree of the marketing department. With open-source stuff, it seems like I can mix-n-match components pretty freely. I can use Django on BSD, or Linux. With MySQL, or PostgreSQL. I can use its built-in templating, or not. With MSFT, it seems like I'd better be using Win2K3+SQL2K5+IIS6+ASP.NET or I'm taking my life in my hands. (And, when upgrade time comes, I better upgrade 'em all, or I'll have problems. And I better upgrade every system or they won't talk to each other.) Presumably MSFT could make its products more standalone, but prefers not to.

b.) Inevitable screwage. I believe MSFT's goal is to establish itself as an unavoidable presence in the IT business (ala the bad old days of DOS) and then use that position to demand money. For the moment, thanks to open standards, they can't do that. I'm not interested in helping them change that by using/endorsing their technology stack.

c.) MSFT MVPs. More generally, MSFT people. With all due respect to any Planet Microsoft hackers hereabouts, there is an enormous lack of perspective found in people who've only used MSFT tools; they tend to think 'One Microsoft Way' is a viable approach to building code, and not just an address. If you're reading this, and that offends you, I hereby acknowledge that there are exceptions, and you're probably one. I don't want to flame anyone, I just want to share an honest reason why I'd prefer to avoid the MSFT stack.

"With this amount of noise in the signal, it's hard to tell good hackers when you meet them. I can't tell, even now. You also can't tell from their resumes. It seems like the only way to judge a hacker is to work with him on something."

(From http://www.paulgraham.com/gh.html)

I wonder what the reddit guys think of aaronsw.

Balls to that. I'm armed with a bad attitude and worse math. By my reckoning:

On any given day, a given prisoner has a .99 probability of NOT visiting the room.

Over a period of N days, a given prisoner has a pow(.99, N) probability of having _NEVER_ visited the room. Therefore, the probability a given prisoner HAS visited the room at least once over a period of N days is (1-pow(.99, N)).

Therefore, the probability that X prisoners have each visited the room at least once over a period of N days is pow((1-pow(.99, N)), X)[1]. If I'm in the room, the question I need to answer is: How likely is it that each of the 99 other prisoners have visited the room at least once in the preceeding N days?

Let's visit Mr. Python:

   N    P(safe)  Cost[2]
 ----- -------- ----------
  100: 0.000000 (18250.00)
  200: 0.000001 (18249.99)
  300: 0.006887 (18124.31)
  400: 0.166419 (15212.86)
  500: 0.520678 ( 8747.63)
  600: 0.787901 ( 3870.80)
  700: 0.916504 ( 1523.81)
  800: 0.968598 (  573.08)
  900: 0.988391 (  211.87)
 1000: 0.995735 (   77.83)
 1100: 0.998437 (   28.53)
 1200: 0.999428 (   10.45)
 1300: 0.999790 (    3.82)
 1400: 0.999923 (    1.40)
 1500: 0.999972 (    0.51)
 1600: 0.999990 (    0.19)
 1700: 0.999996 (    0.07)
 1800: 0.999999 (    0.03)
 1900: 0.999999 (    0.01)
 2000: 1.000000 (    0.00)
 2100: 1.000000 (    0.00)
 2200: 1.000000 (    0.00)
 2300: 1.000000 (    0.00)
 2400: 1.000000 (    0.00)
 2500: 1.000000 (    0.00)
 2600: 1.000000 (    0.00)
 2700: 1.000000 (    0.00)
 2800: 1.000000 (    0.00)
 2900: 1.000000 (    0.00)
 3000: 1.000000 (    0.00)
 3100: 1.000000 (    0.00)
 3200: 1.000000 (    0.00)
 3300: 1.000000 (    0.00)
 3400: 1.000000 (    0.00)
 3500: 1.000000 (    0.00)

After 1000 days, I'm guessing everyone's visited. Screw you guys, I'm going home.[3]

[1]I know this can't be exactly right, because prisoner visits aren't independent events, but I figure it can't be that wrong, and I want to go home.

[2]Cost is calculated in expected days of life forgone, multiplying (1-P(safe)) by 50 years by 365. Arbitrary, but it gives some idea of how much expected life you gain by waiting longer to roll the dice.

[3]I know this is ducking the intention of the question, but I think the point that death isn't that bad a risk if the probability is low is a legitimate one.

Two possibilities:

a.) Physics. Always found it interesting, but not into academia as a way to pay the bills.

b.) Feeling poor compared to those bastards with 11 figure net worths.

Guns. Seriously. Marksmanship is a simple thing that takes a lot of focus. Concentrate on the front sight, mind your breathing, squeeze gently .... it's like meditation, only less sissy. (j/k!)

Interesting .... one corner-case I see has to do with mouseover. For instance, some sites (http://www.newegg.com) use JS to dynamically expand/contract menus based on mouse travel. I don't think a simple ImageMap would hack it in that case. Do you have a workaround in mind?

(Granted, sites are typically built to not _depend_ on this sort of fancy JS.)

Workforce experience reminds me of the old story about the shoe salesman who goes to look for a job:

 "I have 20 years experience", he says to his prospective boss.

 "No you don't.", the interviewer replies.  "You've had 6 months of experience 40 times over."

The corporate trenches are sort of like that. You do learn some stuff, but not much, and you learn it awful quickly. 90% of the interesting aspects of my "trade" that I've learned have been on my own.

Corporate work (for developers) seems ok (but not great) for gaining domain expertise. It is a great way to meet people with domain expertise. Other than that, it's not good for much. Large organizations seem considerably less than the sum of their parts.

"PLEASE SOMEONE TELL ME, HOW CAN I GET A JOB AT THE BIGGEST COMPANY ON EARTH?"

Just walk into any Wal-Mart[1]; I'm sure they've got applications lying around somewhere.

[1]http://money.cnn.com/magazines/fortune/fortune500/2007/perfo...

Talk to people. Tell them you're starting a company. Lay out an idea. It doesn't have to be good. It's been my experience that this scenario triggers an irresistible impulse in the other person to tell you _their_ ideas. Sooner or later, you'll hear something interesting. Explore it and refine it.

The biggest mistakes you can make, IMO, are (a.) thinking your idea is too stupid to start with, and (b.) keeping it a secret. The idea you start with almost certainly won't be what you finish with; the important thing is to get started. And if you don't talk about it, you'll never really explore, refine or (as you'll almost certainly need to) abandon it for something better.

Most people have something interesting to say. Find out what it is.

Yeah, '89 was something else. OTOH, it was centered off in the boonies (modulo Los Gatos), whereas this one looks to have been right outside of S.J., so a good ride even at less energy.

I'll still take quakes over tornadoes and hurricanes. They're usually harmless fun. Once in a while they're tremendously devastating, but what is life without adventure?

"...they get excited over revolutionary developments but grow weary at repetitive small tasks."

Blast! He's got me. I'm not claiming to be brilliant, but, boy oh boy, do I ever tire of repetitive small tasks. If only there were some way to automate them!

Perhaps, one day, some wondrous machine will be created that will relieve us of the burden of empty, mechanical processes. Mayhaps even a small industry might be built around the design and construction of such machines.

What were Larry Wall's 3 programmer virtues? Impatience, laziness, and hubris?

I live in Silicon Valley. My spending is actually down slightly from 2003 - that mostly appears due to minor lifestyle changes (eating out less, turned off cable) but is certainly incompatible with rampant inflation. Cost-of-living looks basically flat. I shop at the same grocery store (Draegers .... Mmmmmm) that I did in 2003.

On the other hand, I have noticed a spike in the office vending machine snack prices (to $.80/item!) and, of course, gas looks to have roughly doubled. But neither of those are big expenses for me.

At the risk of engaging of contradiction, that 30% number doesn't jibe with my data whatsoever.

Have to disagree that current mortgage rates have little to do with current inflation rates. Additionally, inflation is such a tricky thing to measure (because the prices of different things fluctuate with respect to one another, and through time, all the time - I can give you a (bad) argument that we're experiencing massive deflation by looking only at the prices of 100GB HDs) that it seems entirely fair to me to point out that someone's inflation statistic is incompatible with the behaviour of many, many people who are paid to get this sort of thing right.

Those numbers don't pass the smell test; Interest rates (e.g. mortgage rates) are nowhere near 12%, so if those figures are accurate, all sorts of lenders are paying you to take their money. Additionally, and FWIW, I've been obsessively tracking my spending for the past 6 years, and have seen little if any signs of inflation. My grocery bills from 2003 are indistinguishable from those from 2007. (Yes, I'm a geek.)

I might be willing to buy a 12% number for inflation under some conditions, since there are lots of ways to calculate that number, and an emphasis on Gold/Oil prices and/or exchange rates might cough up a 12% figure. I'd still consider it unrealistic unless you were heavily involved in such things. However, the 40% for food seems fantastic, meaningless, or both.

"...and a massive stock valuation. They can make acquisitions of key technology and talent without spending much."

I don't agree with this. There's no liquidity in the stock, and just because MSFT thinks the company should be valued at $15B, I don't see why I should, or should expect anyone else to.

Wouldn't a $15B valuation make their (newly-issued) stock options _less_ attractive to new employees, since they would only be in-the-money if Facebook IPO'd with a (kinda high) $15B valuation by the public markets?

It's a bit of a different question for acquisitions, in which they can trade stock outright. Still, to do a deal, it seems they'll have to buy at very high valuations, giving their own highly-valued stock out in return. So it should still cost them (about) the same %age of FB to acquire a company post- and pre-investment.

It's certainly not _bad_ to raise a quarter-billion dollars in exchange for 1.6% of the company, but I don't see how this makes their stock much more valuable to an independent observer; without liquidity, it's still anyone's guess. Granted, people will probably guess higher now, but I still don't see that valuation as a free ride to low expenses.

I have never modded you down; until today, I'd never run across one of your comments mentioning kratom. However, here's why I think people might be modding you down:

Frankly, and with respect, you sound like a crank and, on this subject, you are a bore. I'm bored of the subject based only on a comment of yours I read today, and the few posts I could stand to read on this thread.

I have no time to evaluate your claims, which are indistinguishable from the claims of 100s of bullshit products. I might be missing out on some wonder-substance, but I know I'll waste my life if I look into every thing that someone promotes as a wonder-substance. So, I'd rather spend my time thinking about things that interest me (code, money, women) and I suspect others, similarly inclined, are responsible for the down-mods.

You seem like a bright guy, but I think this topic is unwelcome, as evidenced by the down-modding. I suggest you drop it.

To the extent that thought experiments suggest real experiments, they are indeed valuable. I disagree that they can 'reveal ... results', however. For instance, Einstein's thought experiment was designed to 'prove' that QM was wrong. Obviously, that result was incorrect.

Does the Copenhagen Interpretation strike anyone else as eerily like lazy evaluation?

My other thought is that once you're doing 'thought experiments', you've become rather unmoored from science as I understand it: Observation, Hypothesis, Prediction, Experiment, and so forth. Hopefully, the 'thought experiments' will lead to something testable.

Yes, I spent about 3.5yrs in it. I can't really agree with the underpaid/overworked thing; the pay was fair, and the hours not unreasonable.

I left the industry because the idea of spending 2+ years on something that may-or-may-not pan out, with no way to tell before release, was beginning to get me down.

I realize startups are sort of the same thing, but I think they actually offer earlier feedback and greater opportunities for recovery than games.

Certain things in this article indicate that it was not written by a serious man. The confusion between the words 'effected' and 'affected'. The reference to 'capitalists' as an identifiable class of people. The assumption that America lacks "the ability to produce products that the rest of the world wants to buy", apparently irrespective of their cost. Gratuitous references to 'obese' Americans.

I want to make two general points in response.

1.) Many things in Economics have two sides to them, and exchange rates are one such thing. Yes, a weaker dollar means those who hold dollars are, effectively, poorer. OTOH, a weaker dollar makes our exports more attractive to the rest of the world, and will redress the trade imbalance that the professional doom-and-gloomers are always on about. If you're in business, selling to the rest of the world, a weaker dollar is _good_. See, for instance, all the whining about 'artificially' low pegs for the Yuan and Yen.

2.) The _current_ (floating) exchange rates represent the aggregate wisdom of the market on the question of _future_ exchange rates. If the dollar was "obviously" going to depreciate further, it would depreciate now, as traders looked to short it. Predicting deviations from the current market rates is simple speculation. If that's your bag, knock yourself out, but don't act like the future is transcendentally obvious. That sort of talk is simple kookery.

A bit o/t, but I hereby nominate 'serial', as applied to 'entrepreneur' and/or 'founder', as the most pointless, irritating, and obnoxious word of 2007.

You don't describe someone who works a bunch of jobs as a 'serial employee'. You don't describe an athlete who competes in successive Olympics as a 'serial Olympian'. Etc.

This 'serial' business is nothing more than noise and puffery. 'Entrepreneur' tells me something, though not much. 'Successful entrepreneur' tells me a tiny, tiny bit more. 'Serial entrepreneur' tells me someone is a self-promoting git.

I even saw this term applied to PG in an article some weeks ago. Considering that PG has repeatedly written that a startup is kinda stressful, and something he'd be very reluctant to do again, I think that marked the time the term slipped into complete meaninglessness for me. (Note that PG didn't refer to himself that way, 'twas the idiot writer/reporter.)

Let's drop the adjective, and shun those who use it.

/rant

Frankly, even if the improvement is worth the work, I think this is the wrong kind of work. Writing a custom toolbar (that doesn't, according to the article, work the same as all the other toolbars the user is used to seeing) seems like the wrong call. Better to alter the design to be (a.) spiffy and (b.) compatible with the technical limitations of the system.

I think it's often better to work within the constraints of a system than to fight against them. Sometimes, of course, it's a Big Win to fight them, and sometimes it's fun, but you gotta pick those battles. (As I'm sure everyone agrees. Sounds like Hamming's talk, in a different context, actually.)

"Programs developed using this [sic] techniques are significantly easier to write, debug, and maintain than programs written with mutexes, CRITICAL_SECTIONs, or semaphores."

No, they damn well aren't. Or, at least, not always. Like most things, it's horses for courses.

This article got under my skin because I had the distinct "pleasure" of maintaining/debugging/enhancing a system designed according to these principles. The problem that crops up over and over and over is that one wants to compute f(X, Y), where X and Y are members of different objects, assigned to different threads. So, one must send messages to retrieve the values and trigger a computation .... and then the result of the computation has to make its way back to the caller.

It. Fucking. Sucks.

I'm sure this isn't a problem if the data is laid out just so, and the system never diverges from its original design, etc. But there are cases where a mutex would be just fine, thank you very much.

As an aside: Can performance considerations _really_ play a role in arguing for messaging as a interthread communication solution? Even if you can construct an environment specific argument that messaging has lower overhead than mutexes (and I find the linked article unconvincing) isn't the latency going to go through the roof? Doesn't that matter?

Bah. Another example of taking something hard and simple, trying to make it easy and complicated, and leaving it instead hard and complicated.

Multi-core's most immediate application won't be to algorithmic parallelization, which is a ball-breaker of a problem even w/o awkward theoretical results. It will be to multitasking, VMs, etc. - scenarios where the utility of extra cores is pretty easy to see.

Those scenarios make the most immediate sense inside the data center. I think part of the push towards SaaS and web-based applications will come from the fact that Moore's law will begin to apply disproportionately to hardware used to serve those solutions. Soon, the same CPU might be able to run either one desktop or 50 virtual servers, making servers cheap relative to desktops.

This assumes all the cars are charged at night. The grid is heavily taxed during the day.

Check out http://www.caiso.com/outlook/SystemStatus.html on a hot day (A/C load has a huge impact on system demand), and this article might look a little optimistic.

Ironically, SCE is in the CAISO, not NEPOOL (i.e. Maine, where the reporter lives.), control area.

Silverlight is EEE for the web.

First, MSFT deploys Silverlight clients broadly. In this phase, they will do just about anything to see it deployed on any platform available, specifically including Linux and MacOS. Continue until essentially all new computers include a Silverlight client.

Secondly, persuade developers that Silverlight is a viable platform. This requires (1) - mass developer adoption of a sparsely-deployed or Windows-only web platform won't happen, not even with MSFT's marketing muscle. Continue until Silverlight achieves Flash-like ubiquity.

Thirdly, either through patents, DMCA, obscurity, or all three, kill off any non-MSFT Silverlight server-side solutions. Ensure that Silverlight app quality is superior to web apps by, if necessary, sabotaging web standards efforts and keeping IE buggy.

At that point, if you want to deploy a 'serious' app over the Internet, you'll (hopefully, from MSFT's perspective) feel strongly compelled to pay the MSFT tax and buy Silverlight servers. (For extra fun, in a few years, maybe MSFT chokes off non-Windows Silverlight client implementations. That's extra-speculative, however. My core speculation is that Silverlight is best understood as MSFT's play for the datacenter.)

A lot of money has been made by companies who successfully set up 'toll booths' between buyers and sellers. MSFT's OS monopoly was based upon providing the means for application developers to deliver their wares to consumers. Game console maker's profits are based upon their ability to restrict access to the gamers who own their consoles to companies who give them money. (I'm not saying this is unethical, just how the money is made.) The web undermines this model; the standards are open and (pretty) simple, clients and servers are freely available, and therefore there's no way to erect a toll booth between app developers/content publishers and consumers.

It looks to me that MSFT would like to change that. (So would Adobe, but they're a lot less able to pull it off.)

Anyway, a little paranoia for you there. Enjoy!

Twisted is an interesting framework, though a little lower-level than others. I think you could build a self-contained, in-memory solution around it, as DBs aren't particularly central to its operation.

Oooh - Brian Harvey. Brings up happy memories of CS60B; C, MIPS ASM, and the SPIM simulator. This guy is one of the best lecturers/speakers I know of.

I don't care about FB either. I probably should, because it's a huge phenomenon, but, no matter how hard I try, I just can't.

In the first place, all these pure social networking sites (i.e. those that don't do much useful if you're the only user) rub me slightly the wrong way. I feel I'm being asked to invest my effort into creating value for someone else, and that I'm being locked into a single-vendor implementation. It's not a good feeling.

In the second place, I just don't see what FB _does_ for me that I can't do more simply some other way.

Although I can't remember name of the fellow who said it, or the exact wording, I share the sentiment that "I already belong to a social network. It's called the Internet." I've got an e-mail address, and a (few!) websites. I don't see why I should have to play in somebody's walled garden to connect with other people who've got the same.

The link is to Kathy Sierra's "people on the web are mean" (google can doubtless explain further) post, which is followed by a lot of her favorite pictures/graphs/bullet points about design/management/stuff. The latter are, presumably, the things every startup team should read. The story headline probably ought to have explained that, since the lead-in part is long, and its meaning obscure.