afiler

"You're stupid." = name-calling, of course, but you consider "What you did there was stupid." to be name-calling? I don't think everyone would agree with that.

I wrote my own web-based terminal emulator, but shortly thereafter found shellinabox (http://code.google.com/p/shellinabox/) and I use that instead. It looks like GateOne might be a bit more configurable, and I don't think shellinabox does tiling or bookmarks. Can anyone else comment on other relevant differences?

"Well, actually, no, officer, I wasn't speeding. You see, while you clocked me at 90mph [c] between toll booths [atoms], once you factor in time at the booth, you'll see that I am actually driving much more slowly."

I have to say that your video is surprisingly beautiful given the subject matter, though the voiceover is done in a style that sounds like a commercial for a new wonder product that's both a floor wax AND a delicious dessert topping.

I meant it only in the superficial sense, that the call of len() is called as len(foo), not foo.len().

I guess this is another example of why using the term "object-oriented" at all can be unhelpful, as it may mean so many different things to different people depending on the particular situation.

I don't think you really mean "object orientation" so much as "a byzantine object hierarchy" (stereotypically associated with Java, of course). To me, Ruby's object-oriented string functions are a lot easier to keep track of than PHP's:

x.length vs strlen($x)

x.gsub('foo', 'bar') vs str_replace('foo', 'bar', $x)

x.strip vs trim($x)

x.upcase vs strtoupper($x)

All four of the PHP functions use a different form: strX, str_X, X, and strtoX. Some of this could be solved with consistent function names, like always using str_X, but then that makes me wonder why one wouldn't just want to have all string functions available as object-oriented methods on strings. Python takes a different tactic, and has a non-object-oriented len() method, but this isn't just used to get the length of a string, it will also get you the length of a list, tuple, or dictionary.

If they didn't put links in parentheses, they might break compatibility with teletypes! ...I'm not actually suggesting any newspaper still uses actual teletype machines, but I wouldn't be surprised if there are still plenty of workflows that evolved from teletype use.

The teletype and Baudot code do have a pretty amazing legacy. Baudot code was invented in 1870 as a 5-bit mode-shifting character set. This, of course, predates data processing with punch cards (1889), alphanumeric data processing (1929), and binary computers (1937-1941). Even after the introduction of ASCII, the smaller Baudot character set remained a common subset available on a wider range of machines. This is seen in C's "trigraphs", where ??/, for example, may be substituted for \ on machines that don't support that character. Even the Apple II+ had a teletype-style keyboard, supporting only the characters found in Baudot code (the Apple IIe keyboard was the first to support all of the 7-bit ASCII characters).

Seeing the title "Greece defaults" made me things had taken a turn for the worse compared to news I'd heard earlier in the day, but no, this is just the title of an opinion piece. Other news articles are more nuanced, and in particular, the Wall Street Journal asks "What Constitutes a Greek Default? And Who Decides?": http://online.wsj.com/article/SB1000142405311190346110457645...

The VOIP data is probably not delivered over the internet for local calls. The calls probably come out of the cable network and hit a softswitch in the same city. For local calls, they probably have a TDM interface (T1s/DS3s etc) to the incumbent carrier and possibly other CLECs too.

I see he says "(As in my previous post, I’m using an evolutionary explanation metaphorically — as a way to help explain a concrete phenomenon I’ve observed in my research and writing on this topic. Whether the evolutionary explanation for the phenomenon is strictly true is somewhat beside the point and beyond my expertise.)", which makes the evolutionary biology bit sound more like a vehicle for the argument than the argument itself.

(Yes, another reply to one of my own comments.)

After looking at your blog, I see that my example above was basically the geodata version of http://blog.buzzdata.com/post/7535032009/25-great-links-for-... . I think I'm in love.

If I were going to describe this to a friend, I'd probably say "like flickr, but for data"... not something you can put in marketing materials, I know.

I'd like to abuse this reply to ask that when you add tools for geodata, that you please please please do what you can to help people produce visualizations and rankings that are not horribly skewed by poor choices of geographic boundaries.

I know this is a hard problem, but for people who don't do this all the time, it's really easy to end up with a headline like "Manhattan Leads the City in Pedestrian Deaths Per Square Mile, Study Finds" (Manhattan is, of course, the densest borough in New York City, so this probably isn't terribly useful information). I've seen maps that might show, for example, the areas of the US with the highest number of coffee shops per square mile, but the map is done based on counties, so here in Seattle we end up not looking terribly dense since over half of gigantic King County is mountainous and unpopulated.

(I'm presupposing, of course, that "social network" sounds as overused of a term to everyone else as it does to me.)

I'm glad to see something coming in this space. I had been very sad to see DabbleDB go away, and Google Fusion Tables, Google Refine, and Google Docs Spreadsheets seem promising yet somewhat neglected, unfocused, and disjoint.

Surely, though, there must be a more descriptive term for this than a "social network"? Even though you could consider it (to some degree) to be a social network, Github doesn't describe itself as one.

The problem of blasting out non-relevant messages to friends/followers/people who have you in a circle also occurs with Twitter and Facebook. Twitterers often solve this by opening more accounts and separating "my cat coughed up a hairball" and "I just wrote some awesome new app"-type posts into different accounts. Is there a way this is solved on Facebook? It seems like the closest practical solution is to start a group (that ends up primarily being dominated by its creator).

As Google+ is more focused on having a "real name" like Facebook is, will throwaway accounts work as well? The suggestion in this article is interesting, though I'd worry about excessive UI complexity. Then again, if Google were to let you create multiple accounts that ended up as sub-accounts (or just call them "roles" as in the article), one could elect to either follow an entire person or just one of their particular roles (if multiple roles exist).

I had to look this up after you mentioned it. Looks like it was run by a company called US Videotel: http://www.digibarn.com/collections/systems/minitel-usa/inde...

I know the US also had some trials of teletext (text pages over broadcast TV) too, but I think that was even less successful than videotel.

I'm sure my neighbor, who's blind, would appreciate piezo buttons. I had to help her deduce why her screen reader would "die" on her Dell laptop periodically. It turns out that it has capacitive buttons for Mute/Volume, and the only feedback from them is an LED. She would occasionally brush her fingers over the buttons and end up inadvertently muting the laptop.

Adding the -S option to python (to skip 'import site') as seen in another post here brings python's time down to 1.229sec.

To my great astonishment, Ruby's performance was much closer to bash than Python in my own test: 0.572sec for 101 runs in Ruby, versus 4.276sec with Python and 0.318sec with bash. I now feel slightly less bad for all those shell scripts I've written in Ruby instead of bash.

OS X/Cocoa includes some Emacs-style key bindings by default (see http://www.hcs.harvard.edu/~jrus/Site/system-bindings.html), so I don't think leaving other key shortcuts out makes the keyboard "interfere with a novice user" as you say. Dialog boxes don't support keyboard accelerators other than Esc and Enter, and the Finder doesn't support things like cut-and-paste or opening folders with the Enter key, and I find these things incredibly frustrating having used Windows in the past. Try doing something simple like moving a file from a directory to its parent in Finder versus Explorer.

More on the "Cocoa Text System": http://www.hcs.harvard.edu/~jrus/Site/Cocoa%20Text%20System....

Prompted by downloading a .doc file from Qwest only to find out that inside was a monospaced text file, I set up a small, nearly UI-free site for doing document conversions. http://doc.mar.cx/<url>; gives an HTML or other sensible rendering of an url (e.g. http://doc.mar.cx/http://www.itu.int/dms_pub/itu-t/oth/02/02... ) and http://doc.mar.cx/<extension>/<url>; attempts to convert the url into the format with the given extension (e.g. http://doc.mar.cx/txt/http://www.itu.int/dms_pub/itu-t/oth/0... ).

I use wvHtml for doc->html, wvPDF for doc->pdf, but antiword for doc->txt. To convert .docx, .xls, .xlsx, and WordPerfect files to HTML, I use OpenOffice, by way of jodconverter. For ODF files, I use OdfConverter. Conversion of Excel files to .csv files uses xls2csv. For PowerPoint files, I use ppthtml to convert to html, and catppt to convert to text. For Lotus 1-2-3 files (I added this after downloading some historical telecom data from the FCC!), I use ssconvert.

Any conversion that results in an HTML file (e.g. doc or pdf to html) I bundle all the images into a single file using the data: url scheme. To do this, I wrote a utility called pagecan: http://afiler.com/pagecan/

Spakman, I'm very excited about this. And I say this as someone who's writing his own ruby shell/shell-complement. Mine is mostly intended to be invoked from Bash the way you'd invoke grep, though it can also be run interactively, where it behaves like irb. The one clever thing I did is allow you to avoid escaping by putting your ruby command after a # on the bash command-line. Bash then ignores everything after the #, and crush (my shell) grabs the last line in the bash history and evaluates the part after the #.

I will definitely be trying out your shell. If you're curious about mine (feel free to steal code, no promises that everything works as committed): https://github.com/afiler/crush

I figured the concept for the display devices was based on microfilm/microfiche readers rather than CRTs. Microfilm was "high-tech" back then, and Vannevar Bush's Memex (of a couple decades earlier) was microfilm-based.