bantic

Josh Waitzkin (the kid from Searching for Bobby Fischer) wrote a book called the Art of Learning and he talks a lot about the difference between these two modes of learning. He calls it Entity versus Incremental intelligence. (Really good cliff's notes on the book by Derek Siver are here: http://sivers.org/book/ArtOfLearning). He's even harsher on entity knowledge-ists -- they crumble under the pressure. It's a good theory to explain why some of the brightest kids I knew in high school have done nothing with their lives since: they are afraid to mess up their 'undefeated' academic records.

I wonder about the merits of this sort of thing as an interview screen, though. There's not necessarily a strong correlation between knowledge of language esoterica and other more important qualities like perseverance and raw intelligence.

I've always wondered what I might be revealing (demographics-wise) through my cookies while browsing, but I never though I might be getting shafted by them!

Wonder if Bing is revenue-sharing w/ their vendors this way somehow.

Still, $5M to local charities is all for the good, right? Is your argument that they're too stingy or that trying to buy goodwill by making donations is wrong?

I always thought this sort of systematic obfuscation ( /@/at/s, /\./dot/s, etc) was as machine-readable as an actual email address. I've just gone on the assumption that there are spam harvesters out there using regexes that catch "bob at domain dot com" as well as bob@domain.com.

Richard Feynman. I majored in physics because I was inspired by his intellectual curiosity. "Surely you're joking, Mr. Feynman" is a great collection of his memoirs: http://tenniselbow.org/scott/feyn_surely.pdf

the title should be != More Capacity

Yeah I agree. I was expecting a bunch of utf8 ligatures (or maybe emoji!) used to condense the text.

erlang. because somehow my install of ejabberd failed and I don't know the first thing about how to troubleshoot it. plus I want to learn more about languages that are good with parallelization.

Yes, that's true. In my experience doing batch sms-emailing (with relatively small batches -- several hundred at a time only), we would take use two strategies: 1 -- queue the messages and send them slowly, maybe between 5-25 every minute. and 2 -- distribute across carriers. Group the users by carrier and then for each queued batch take from as many of the different carriers as possible, thus limiting the number of times you sent an email over any specific carrier's gateway in a given queued clump of messages.

This seemed to work but it was hard to guarantee message delivery.

There's a new peepcode pdf about this that looks interesting. https://peepcode.com/products/mms2r-pdf

I'm sure the carrier negotiations in order to use shortcodes are different in countries outside the US. I have heard that docomo in Japan actually makes this sort of stuff really easy for developers (and that Japan has a burgeoning diy market of cool phone-based services as a result).

What country are you in?

I've done work using the email2sms gateways for both sms and mms. For sms it's kind of a pain because the gateways can't always be inferred just from carrier and number (T-mobile for instance, sometimes uses a nickname that you set up rather than your phone number, so your phone's sms email address ends up something like coolguy@tmomail.net instead of 2125551234@tmomail.net). Receiving MMS is a HUGE pain because the carriers all include the attachments differently. Some carriers include multiple attachments -- some of them are spacer gifs and things like that -- and you have to figure out the correct one. Other carriers, like Sprint, don't even include an attachment, they have a link in the MMS email that takes you to a page, and that page has another link to the picture. I had to use come complicated regex's and multiple wget calls to download the picture. The whole thing was very fragile as a result. But, to Sprint's credit, they are so bloated and slow to change that in 9 months they never modified the way those messages were structured so nothing broke.

I haven't worked on this stuff in over a year, so it may be out of date, but given the glacial rate of change in telco's (esp in the US), I'd be surprised if the situation has changed much.

Again, though, could be a very different story depending on where you're actually located.