Forgot your password?
typodupeerror

Comment: who cares that 3.2's "features" aren't backported? (Score 0) 164

by portscan (#35272632) Attached to: Python 3.2 Released

let's take a look at the "major" changes in 3.2:
- stable ABI, .pyc, .so stuff: only really relevant to developers of python itself, not programmers who program in python
- argparse: a nice update to optparse, but can easily be ported since it's just a module written in pure python. definitely not compelling enough of a reason to upgrade
- dictionary-based configuration for logging: again, not compelling at all when weighed against the syntax changes in py3k. also, just a module, can easily be back-ported.
- concurrency tools. there might be something interesting there
- wsgi, email library modules: looks like they are fixing things that were fouled up by the ascii->unicode switch. these are problems you would not have if you had stayed with python2

etc. etc. etc. Python3.{0,1,2} includes many things that would be nice additions to the language, but are totally not worth the headache. Think if all of those library authors had actually been enhancing their own libraries instead of rewriting the same features to fit into the py3k language changes. Take a look at PEP3000 and PEP3100. There really is no killer feature, nor do all the little upgrades collectively constitute a killer feature. It's just bad PR to cause so much work for the Python community for no discernible benefit.

Sure there are several things that "didn't quite make sense with python" but a lot of them were in libraries that could be deprecated and then removed. Or just deprecated. Successful languages don't change. The "purity" of the language is a bunch of ivory tower computer scientist BS that may very well tank the popularity of the language just as it was coming into its own. I hope they find a way to fix this that does not require everyone to just rewrite their python code to be py3k compatible. I know it's really not that much work, but again, there is no tangible reward for this effort.

When someone can point me to a real benefit that py3k brings that *could not* be achieved without the backward-compatibility break, then I will re-consider my position and even consider upgrading myself. Until then, I will be frustrated that the core dev team has caused busy work for everyone.

Comment: Re:DIY Credit Union (Score 1) 158

by portscan (#32233066) Attached to: Developer-Friendly Banks?

Well options are very useful for all the situations listed when the quantities are very uncertain. Credit Default Swaps are useful for investors to hedge against default of bonds (or to specifically assume just the credit risk component of a bond). For all the harm they caused, CDOs (really more of a structured product than a derivative) help to keep mortgage (and all other type of loan) rates low by enabling more capital to be deployed into loans. A pension fund, for example, will not make loans directly to homeowners, but it is happy to by mortgage-backed bonds. The bank, knowing that it can repackage and sell mortgages, is able to offer a lower interest rate to borrowers. When not taken to the absurd excess of ~2004-2007, everybody wins: borrowers get better rates, investors get asset class diversification and securities with reasonable yields and the banks' capital remains free for other activities.

The purpose of a derivative is to transfer risk from someone who wants it to someone who doesn't. In my example of currency swaps above, the natural user of such a product is a multinational corporation. A bank enters a swap with such a corporation, but usually does not want the other side of that exposure. Banks typically do not want to be speculating on FX rates. There are many investors, however, that do. Hedge funds often believe themselves to have superior prediction power relative to the rest of the market, so they may want to go long some currency or another. They can take the other side of a trade, and the bank makes a small bid-ask spread. Some institutional investors just want to invest in currencies as an asset class. They are another natural counterparty. None of these is really "covered" shorts, but the ability to transfer risk from hedgers to speculators is an important value provided by derivatives.

Simple put options on stocks can be valuable as well, by giving investors a way to bet against a company. It is important to have such an ability because it is investors with short exposure that are incentivized to call the BS so many corporations are spouting (enron, worldcom, lehman brothers, etc. etc. etc.). short-sellers stop those companies from continuing to con and defraud investors.

Comment: Re:DIY Credit Union (Score 1) 158

by portscan (#32224964) Attached to: Developer-Friendly Banks?

How about foreign exchange swaps/forwards? Any company that does substantial international business, but reports earnings in the currency where it is domiciled is exposed to significant foreign exchange (currency) risk. The strength of a company is making and selling its production, not speculating in currency markets. Using derivatives, a company can "lock in" a rate of exchange so that its success or failure depends on its business performance, not on fluctuating exchange rates. In fact, any international enterprise that does not use currency derivatives to hedge its currency exposure is recklessly speculating in currency markets!

Establishing a currency swap or forward with a bank is of vital importance to every international business in the world.

I've got plenty of others if you want.

Iphone

+ - Microsoft, employees embarassed about iPhone use-> 2

Submitted by portscan
portscan (140282) writes "There is an entertaining and telling article in the Wall Street Journal about iPhone use by Microsoft employees. Apparently, despite it being frowned upon by senior management, iPhone use is rampant among the Redmond rank an file. The head of Microsoft's mobile division tried to explain it away as employees wanting "to better understand the competition," although few believe this. Nowhere does the article mention attempts by the company to understand why the iPhone is more attractive to much of Microsoft's tech-savvy workforce than the company's own products."
Link to Original Source

Comment: darpa (Score 4, Funny) 259

by portscan (#31046616) Attached to: Internet Nominated For 2010 Nobel Peace Prize

So, the award would really have to be given to DARPA -- the US Defense Advanced Research Projects Agency, which is where the Internet was invented. After all the attention last year given to Obama receiving the prize while waging two wars, I think giving it to the US Military directly would really drive some people over the edge.

Comment: That's an investment bank (Score 1) 165

by portscan (#30628520) Attached to: Finding Someone To Manage Selling a Software Company?

This is exactly the business of investment banks. They evaluate your company's worth and use their existing network of clients to find a buyer. If your company is actually worth something, call them. Yes, I'm talking about Goldman Sachs, Morgan Stanley, etc. If you are a smaller company (which seems very likely, given that you don't know much about business), then try the smaller firms and possibly those with a technology focus: http://www.google.com/finance?catid=us-66043603

Or try VCs, local rich dudes you might know, etc. Put together a small pitch-book about your company and send it around. If it's worth anything, someone will come knocking.

Comment: Re:Fix SAS (Score 1) 318

by portscan (#29483951) Attached to: Who Wants To Be a Billionaire Coder?

SAS maintains backward compatibility across releases so a certain amount of oddness in the syntax must remain, unless they want to write a whole new language on top of the same stats engine, which is most likely a waste of developer resources. Plus, programming SAS on non-mainframe platforms really isn't so bad. It is quite a flexible language for data analysis, actually. Ok, so it's not python, but it is totally useable once you've gotten used to it, which takes very little time if you use it with any frequency.

Furthermore, I'm not sure what you consider "large scale" but nothing comes close to SAS in that arena. R, while a very nice tool would choke on the sort of data SAS handles with ease.

Comment: seems like a very easy hack (Score 1) 121

by portscan (#28725167) Attached to: Integrating Wikipedia With a Local Intranet Wiki

unless i am completely misunderstanding you, this seems like a pretty easy hack on any wiki engine. just query the page's title at other wikis and append the content to the bottom. for example: you have a page called Server Farm -- detailing your companies server farm. whenever that page is loaded in a browser, the dynamic content generator in the website downloads the page with the same name from wikipedia, strips out their formatting, and sticks it at the bottom of your page. your users can only edit your local content. this should probably take only a couple hours to fully implement and test.

if you wanted, you could also parse the "see also" section, and fetch those links and add them. or if you want to be a little more clever, you can allow people to embed keywords for relevant wikipedia articles (or other URLs) in your internal wiki's articles and then fetch those, too. in the previous example, say your server farm consisted entirely of apple xserves. you could fetch that wikipedia article and maybe the spec sheet from apple, etc.

yawn.

Comment: Re:Input page (Score 1) 218

by portscan (#27975891) Attached to: Wolfram Alpha Launches Tonight, On Camera

the site is overloaded. i was using it for about an hour or so. periodically, it would say that the site was too busy (it gave a picture of HAL and an "i'm sorry, dave" error message).

now it seems that the test is over or almost over. it will still let me search for some stuff occasionally, but usually i just get the behavior you described.

from the few dozen queries i executed, i'd say the site is cool, but not amazing. it definitely has a lot of potential.

Comment: Re:here are some (Score 1) 451

by portscan (#27889265) Attached to: Classic Books of Science?

i'm pretty sure griffiths's quantum mechanics has a good chapter on it. that's what we used in (undergrad) QM class.

wolfgang pauli's book wave mechanics looks pretty good. there is a preview on google books. hydrogen atom starts on page 88.

linus pauling's book introduction to quantum mechanics also looks good. preview on google books. hydrogen atom is chapter 5, starting on page 112.

disclaimer: i have not read the last two in full, just skimmed them online. they look quite good, although of course they require a decent understanding of techniques for solving PDEs. as dover editions, they are about $10 each, which seems well worth it to me.

Comment: my thoughts (Score 1) 412

by portscan (#27889153) Attached to: What To Do When a Megacorp Wants To Buy You?

it sounds like you have already decided not to sell, but here is my advice:

sell, you moron. pretty much all new businesses fail. most existing business consider being bought out by a larger company a huge success.

if you are worried that the money you are getting is not enough, sell 51% of the company to them and retain 49% for yourselves, so you basically split the money down the middle, but they have the control that they want.

if you are worried about the autonomy, then sell, make sure there is an employment contract with satisfactory conditions attached, and try that for a while. then quit at the end of your tenure if you are unhappy. ride the gravy train to the top if you still enjoy it.

a big business with "excellent" employment conditions is probably too good to be true, but anything close to that makes this a no-brainer.

make sure the offer is cash, not stock!!!! if it's stock, make sure you can sell it right away. the last thing you want is to be holding all of your wealth in the stock of one company. once you have $50 bn, it's ok if $49 bn is in MSFT shares, b/c you still have $1bn left even if it completely evaporates. (and believe me, gates has lots of other investments). but for the type of money we are talking about, you want cash.

get a good lawyer to straighten this out for you. the $750/hr now will be a pittance compared to the millions it can save you down the road.

Comment: Re:True, but... (Score 1) 412

by portscan (#27889037) Attached to: What To Do When a Megacorp Wants To Buy You?

actually, it has been widely shown that the correlation between wealth and happiness is very low and if anything, it is negative.

i suppose most people who are "rich" got that way by working 80+ hrs per week, ignoring their kids, families, and friends. either that or your parents were like that and you haven't had to work a day in your life and instead have bounced around the offices of various psychiatrists and rehab centers.

defining success by earnings and what it can buy you is and endless game of self-inflicted misery. you will not be the next bill gates or warren buffett, so there will always be someone with more money than you, nicer toys than you, etc. so you bought the 60" plasma, but now you want a whole movie theater room in your house. your ferrari is pretty nice, but the new lamborghini just came out. a 70' yacht is nice, but wouldn't 110' be so much nicer... and so on.

mo money, mo problems, so to speak.

still, he should sell. most new businesses fail miserably and a bird in the hand is worth two in the bush.

"Look! There! Evil!.. pure and simple, total evil from the Eighth Dimension!" -- Buckaroo Banzai

Working...