1. It needs to be "please burgle me". If you aren't at home, then you are being burgled, not robbed. A robbery is theft with violence or the threat of violence (at least in English law - Theft Act 1958 - it is). If nobody is at home, then nobody can be the victim of violence or the threat of violence. So your home is getting burgled - or, if you are an American, burglarized (what the hell kind of a word is that, right?).
2. PleaseRobMe seems to be built around the premise that one home = one person. If you know where I live, please be assured that I am currently not at my home. But other people live where I live. Families exist. Flat sharing exists. Communal living exists. (Yeah, go and raid the kibbutz - I'm sure it'll be empty!) This may be true for Web 2.0 valleyboys. It's not true of the rest of the planet.
That said, this kind of thing does show why most location-based services are stupidly designed. I have played around with a few of them, and the only one I'm a real big fan of is FireEagle. Sadly, it's been a bit neglected for business reasons - i.e. Yahoo! financial situation. What is great about FireEagle is you share you location with FireEagle, and they then share it with whatever services you want to share it with. So, I have the little iPhone app which updates FireEagle. FireEagle knows exactly where I am. Then there's a Facebook app which connects to FireEagle, but I don't necessarily have to let it broadcast my location if I don't want it to. Or I can only give a vague location - perhaps at a country or city level. I have it wired in to my SSH setup, so if I SSH in to my Linux box from certain places, it updates my location. Because it is a location broker, it can be updated in any way people think of, rather than having to use a specific application (say, for the iPhone) like FourSquare etc. do.
This is useful as I can build applications that sit on top of it. One I have been meaning to build is a "remind me when I'm at X" app. So I could basically dump a string (SMS/tweet length) into a database with a broad location in it. It could check against my location and when they match, I could be reminded of X. Remember to buy ice cream when I'm at the supermarket - well, when I'm at the supermarket, I should get a text message saying to buy ice cream.
Location-based services shouldn't be tied to devices but to people. This is what everyone gets wrong. They need really good granular privacy controls. They need a big "forget me" option. This is something Google Latitude doesn't have. There is no way I have found to tell Google Latitude "Hey, take me off the radar. I'm not anywhere anymore."
When I have some time to build it, I'd love to build something like FireEagle but running on my servers and just for me. Location is too important, useful and fun to trust Google or Yahoo! or some venture-backed Valley startup with. But if you are building location-based services, look at FireEagle and learn.
Sure, pseudocode is good for getting ideas across to other human beings and developing a rough idea of program flow... But it isn't going to compile. And it doesn't matter how much you argue that programming languages evolve over the years and get new features added and whatnot, your pseudocode still isn't going to compile.
Google "executable pseudocode". I think the Python community have worked pretty hard on producing a good pseudocode interpreter...
As you say, spoken English is different from written. Quoted and reported speech style is a difficult area - I'm pretty sure that newspapers would often lightly clean up quotes to make them match the style guide. It seems that may have fallen off in recent years. I'm quite ambivalent about it: as a reader, I would certainly prefer the quotes to be readable. As someone who has been subtly misquoted in a newspaper to the point where the meaning of what I said was changed (as well as my position being incorrectly attributed), I would rather like it if my words could be converted from audio to text with as little change as possible. If you wish to clean up what someone said, paraphrase it as reported speech rather than quotation.
Plenty of people do abuse spoken language too. One of my favourite examples is the "turned around" trope in British English. When people describe heated discussions and arguments they have, they use "turned around and said" instead of just "said". Someone will give a little monologue of the sort: "I was in the pub with my girlfriend, when all of a sudden she turned around to me and said [...] and I turned around and was like [...], and then she turned around and said [...]". When someone does this, all I can picture is a video I once saw of Whirling Dervishes spinning around and around, a spin around interrupting each line of the dialogue. It's kind of a strange nervous tic that people add to make what they say sound more dramatic when the words aren't quite enough. Som people seem to add it to every line which rather makes it lose any power.
Naah. PHP is old. Get with the times. Ruby, man. If only because you can use the following stack abbreviations:
LAMR = Linux, Apache, MySQL, Ruby
LARP = Linux, Apache, Ruby, PostgreSQL*
Don't be a LAMR (because everyone hates lamers) or a LAMPpost (dogs will pee on you) - be a LARPer instead! You get to fight people with polystyrene swords and wear a cape! And then come home and use a good RDBMS and a nice programming language!
* Justification for reversing the database and programming language: you can build an application quite easily that has no database backing. Flat-file or just in-memory storage or whatever. If you wrote an application that simply read an XML file from the filesystem, you would not feel the need to include the IO library or the XML parsing library in the stack diagram. So, therefore, the database and the programming language are really not depdendent on one another. The order of the letters in the abbreviation have little actual meaning, so the strategy of ordering them so that the abbreviation sounds cool and makes a nerdy reference is just a valid strategy as any other. Sort of.
Froze the account completely except for the £25 expedited transfer - profit trumping security in UK banking land, of course. It's a savings account (ISA), so there is no debit card. Still, when they were paying me 5.75% without tax pre-recession, these inconveniences could be handled.
Not to mention the huge amounts they've had in fees and fines - fines they've charged me due to the ineptitude of other banks (who have web security models designed for the aforementioned idiots).
Does that translate as "I can't remember my password and I keep getting locked out. Also, I'm so poor at managing my cashflow that I'm always overdrawn"?
You want to know what it actually is? A bank I used to use had an online security model which required you to answer with both a password and with a 'secret answer'. The 'secret question' was "What's the name of your first school?" - and it was the same for everybody. Ignoring the fact that now so many people are on Facebook and MySpace, that's hardly a useful security measure at all, there is a problem with punctuation. If you went to, oh, St. Paul's Roman Catholic Primary School, that could quite easily be Saint, St., or St, Pauls, Paul's, or Paul, Roman Catholic, Catholic, RC, R.C., Primary, Primary School or many other variants thereof. It turns out the bank was automatically stripping any characters that weren't [a-zA-Z\s] from the secret answer. Which is fine, except they don't tell you that. And you only have to mistype it twice or something and you are locked out. I didn't even mistype it - I have all my bank details in an encrypted text file, so I just copy and paste. If I write out the string, store it in a file, copy and paste it into the field to set it, and then come back a week later and enter it in the same field through copy and paste and it doesn't work, that's quite evidently not my fault.
Once they've locked you out from online banking because of their shitty software, it then takes them two weeks to unlock it. During which time, no transactions allowed. It took until the third time I'd been locked out of this process until they finally explained their byzantine string processing rules. And because it's internet banking, there is no other way of getting any money out. Except one. You can phone them up, and they won't do a BACS transfer for "security reasons" - but if you are willing to pay £25 they'll do a priority transfer. Think about that: you can't have money transferred slowly from your account for security reasons, but if you are willing to pay a fee, those security reasons disappear. I'm sure the fraudster intent on ripping off a savings account will pause in wait, wondering whether or not he wants to spend £25 of someone else's money in order to put a transfer through.
Sadly, stupid excuses about idiotic programmers and their inept handling of strings doesn't cut much ice with people one owes money to.
Of course, once it had cost me my second or third £12 bank fine, I took my business elsewhere. And, no, I generally manage my money pretty well - my student loan is about a third of the size of a lot of my friends, I clear my credit card every month, my laptop is pretty much my only luxury - bla bla bla.
My point was very simple: if you go a pound overdrawn and the bank then wham you with a £12 charge for it, the least they can do in return is to process some damn cheques without whining about how much it costs.
That's exactly the right point. The Payments Council don't seem to have really thought out person-to-person transfers or small business transfers. If someone wants to pay me money using a cheque, all they need to know is my name (or the registered trading name of a business etc.). With a bank transfer, you need to know the bank account details of the person - and nobody is quite sure as to how public bank account numbers ought to be. Cheque still rules for getting paid expenses too. I've done the PayPal thing, which is cool - except PayPal take a slice of the transaction. I've done the online banking stuff and it's painful - crappy websites, no notification (how about an e-mail from my bank every time a transaction goes through my account? GPG exists, goddamnit - use it!), security designed for the sort of people who set their passwords to "password1" and tell all their idiotic friends their MySpace password and then wonder why they get "hacked". If you think writing a cheque is inconvenient, how about carrying a laptop and 3G dongle around with you in order to do bank transfers on the website instead?
Yeah, in shops, use a damn debit card. But for person-to-person transactions, cheques are pretty convenient. The security sucks, admittedly - a shared, public key that anyone can copy is not really security at all (a bit like credit cards: credit card fraud, at least until a few years ago, consisted of writing down the number and details that was given out to the merchant upon every purchase - that's real secure). But with cheques, you can post them - slip 'em inside birthday cards. You can give them to third parties (children, employees etc.) who can hand them to their eventual target. You can use them even where there isn't an Internet connection - and, well, outside of the big cities, there's plenty of rural countryside with no 3G service. You can post-date them, and the recipient can return them - my old school used to do this with their after-hours activity programmes - you'd give them half in a current-dated cheque and the other half in a post-dated cheque, so if you don't decide to finish the activity, they just return the second half of the payment.
The only problem with cheques is that I have to walk to the bank to pay them in and it costs the banks money to process them. The walk is quite good exercise, and since the banks got £300 billion of taxpayers money last year under the Special Liquidity Scheme, and they pay themselves HUGE FUCKING BONUSES, I figure the odd 25p here or there to process my damn cheques is pretty reasonable. Not to mention the huge amounts they've had in fees and fines - fines they've charged me due to the ineptitude of other banks (who have web security models designed for the aforementioned idiots). Not to mention interest they make off the money we keep in them. They seem to want to have it both ways: they argue that cheques cost a lot to process AND nobody is using them. But if nobody is using them, surely the number of people you need to employ processing cheques is pretty minimal.
While I'm ranting about banks, here's another thing: it's only in the last year that UK banks have actually got their shit together to be able to move money between accounts and it take less than four fucking days. A family member moves the housekeeping money from one account to another every month. He does this by going to one bank, drawing out however much it is (a few hundred pounds), walking down the street to the other bank and paying it in. Every time, the helpful bank assistant seems to suggest that he could do this electronically. The difference is, if you pay in cash, it is immediately available - while a BACS transer takes 3-5 days. They've only recently changed this so that it takes at most a few hours. But Christ-on-a-fucking-wheel, why did it take until 2009 to be able to move money instantly from one account to another? And it's still only certain banks that do the same-day transfers. These guys are absolutely retarded. With the billions they make every second, you'd think they'd be able to install a few broadband lines between their offices and make it so money can get transferred quickly. Three days - seriously?!
'Too big to fail' presumes the banks aren't the epitome of fail to start with.
Here's my solution: get a laptop and two netbooks, one with a spare battery. Use laptop until it runs out of battery. Then switch to netbook #1. Use that until it runs out of battery. Then switch to netbook #2. Use that, switch battery, continue.
That's how I roll. It's totally insane (and probably hernia-inducing), and I so don't want to get mugged, but it lasted for 6+ hours while on a long train ride recently. It's a complete accident - I went to a conference where they gave everyone a netbook about a week after I bought a netbook from a friend.
But, yes, a netbook with 15 hour battery life would be awesome. Part of the problem is recharge: yes, when I'm working, I can plug my computer in to recharge. But it takes a long time. Most of the day, in fact. I think we are really missing out on super high-speed battery charging - both for computers and for electric cars. If it takes 2 hours to discharge a laptop battery, it should take ~20 minutes to charge it up again. We've gone wireless for networking. I can't wait until we have batteries that'll last long enough so we can go wireless for computing too...
I don't think we nerds are bad at naming things. From all I can tell, 'cloud computing' is the invention of consultants, VCs, vendors and other notorious buzzword merchants. Nerds come up with names like scrotwm.
As for cloud computing? It's a really dumb buzzword. People have had applications running on the Internet for a long time. Some of these are somehow 'cloud' and others aren't. Apparently, Gmail's IMAP servers are in the cloud, while my university's IMAP servers are just on the dull old Internet. If I store stuff on Dropbox or whatever, I've got stuff in the cloud. If I store them on an FTP server, just dull old Internet. If cloud computing is accessing applications over the Internet, then we've been doing cloud computing since the 80s with telnet. If cloud computing is virtualization, then we've been doing that for some time too. But it's neither: it's just a buzzword. See Larry Ellison's brilliant rant.
UbuntuOne was the last straw for me - I saw that and am moving all my Ubuntu machines over to Debian Sid. I trust the Debian community not to get sucked into stupid marketing hype.
Well, ideally, they'd be able to get excited about doing it differently, maybe doing it better. We have to reconcile ourselves to the fact that non-programmers may make decisions about software for less-than-optimal reasons (A: "We're using XML"; B: "Why?"; A: "Well, anything that begins with 'X' has to be cool and the industry press has talked it to death."; B: "But, but..."; A: "I pay your salary."; B: "Okay then. I'll just take our CSV data and wrap it in a 'CSVData' element."), but that we can at least try to enjoy the ride.
What you said about software "both unbelievably ephemeral, yet also incredibly long-lived" is true of lots of things, but it reminds me a lot of jazz. That's certainly how I try and think of programming: whatever else changes, try and have fun with it, and try to work on interesting projects. If the implementation stays there for 50 years, great. If it doesn't, hopefully you still had fun implementing it, it still helps someone do something useful, and you have become a better programmer.
Damn right. I walk in London and use buses/trains to get around. Cyclists are the biggest bunch of inconsiderate arseholes you can find. Many of them think red lights only apply to cars, don't stop at zebra crossings, don't signal, go the wrong way up one-way streets, scoot out in front of traffic. They clutter up the pavements - outside the Lyceum Theatre on Wellington Street, there's a special cyclists crossing across the corner of Strand/Aldwych to get onto Waterloo Bridge. But it goes all across the pavement, and if a crew of cyclists are waiting at the crossing, they make it so that it's actually impossible for pedestrians to use the pavement.
And then they whine endlessly about how there aren't enough cycle lanes. And so the government spend all this money building all these stupid cycle lanes on roads nobody uses in order to please the cycling lobby. In the university area of London, on Malet Street WC1, they've built all these cycle tracks. All very important, of course, since there's maybe two cars an hour go down the damn road (and the cyclists have a bunch of cut-throughs across the campus). But the cyclists just won't survive unless they have some special lanes to go down. Don't worry, though, they pay for it through their cycle tax, and cycle insurance and, oh, hang on...
I even found this: Stop At Red, a campaign by cyclists to pledge to obey the Highway Code and other road laws. That something like this gets created says a lot.
Don't get me wrong: I love cycling and if I could snap my fingers and make half the cars in the country disappear and be replaced with cycles, I'd do it in an instant. But as a pedestrian in London, cyclists can be really fucking inconsiderate.
12/12A - for videos, nobody under age 12 is allowed to purchase it. for movies shown in a cinema, under twelves can watch it if they are accompanied by an adult.
15 - suitable for 15+
18 - suitable for 18+
R18 - Restricted 18 - basically porn. Can only be shown in specially licensed venues and sold only in licensed sex shops.
'E' is on some videos. It's not actually a rating, but it's just a symbol put on by video producers to specify that the film is exempt from rating. Things like videos of sports matches, musical performances, educational videos don't get rated. Most imported videos
The BBFC also now produce some text that accompanies the rating symbol which broadly gives the reason why the film is rated that way. For instance, it might say "Contains frequent strong bloody violence and very strong language" next to an 18 certificate.
They introduced R18 a few years ago when they realised that hardcore porn was mostly being circulated through the black market and by people distributing copied tapes. Of course, now, we have the new rules on "violent and extreme" pornography which actually makes it a crime to possess pornography that depicts violent scenarios - rape fantasies, that kind of stuff. With one hand, the government make porn a bit more legal, and with the other hand, they've created a new black market that the Internet supplies.
The BBFC is generally, imo, pretty fair - I mean, as fair as a bunch of censoring, free-speech-restricting thugs can be. Perhaps it's just bias living here, but BBFC seem to get it right a lot more often than the MPAA ratings do, and they are a hell of a lot less squeamish about depiction of sex and nudity - they make a distinction for 'natural nudity' where it's non-sexual, so we don't have idiotic philistines sticking big blocky pixels over Dürer woodcuts, Titian paintings and Michaelangelo's David (etc.) because they've got HUR HUR HUR DONGS LOL. There's an interesting set of articles by people who have worked at BBFC, describing exactly what it's like censoring movies and video games for a living.
There are some things I think location-based services, now including Twitter, need to learn from previous experience of other location-based services like FireEagle:
1. Don't require me to use any specific technology to update it. I don't have a smartphone, and don't really want one either. I have an iPod touch and I have a mobile phone. I quite like them being separate (the fact that my phone lasts a few days on standby and isn't tied to any specific provider is pretty nice). That said, I do use location based technologies: I just use some interesting ways to update them. The primary way I update my location is using the/var/log/auth.log data on my Linux machine. Whenever I log in over SSH, my computer checks to see the domain I'm logging in from. I have a file on the computer that maps the IP addresses to real street addresses, and then notifies FireEagle, Yahoo's location brokerage service, which then appears in Facebook (and also lets me go to sites like pubs.iamnear.net and find out pubs that I am near). Decouple data collection from data use. My phone doesn't know where I am - it's a pretty dumb phone. But that doesn't mean none of the gadgets or technology I use don't.
2. History isn't that much of a problem, but live data is. I don't care that the whole world knows I was wandering around central London a few days ago. I don't mind my friends knowing where I am now (so long as I have some kind of prominent kill switch that cleanses the Internet of my current location). The problem is when everyone knows where I am right now. If someone knows that I'm some particular place, they know I'm not at home - which means they can burgle my home.
3. And, well, I don't want even my friends - or, rather Facebook/Twitter type friends - to know where I live unless I tell them. What location services that hook up to social networks need is a way to say "when I am near this address, don't tell anyone".
The privacy stuff is important, and I hope Twitter take it really slowly and listen to user and developer feedback. There is a good side to location services which too much focus on privacy ignores: they let you meet up with your friends. I posted on Twitter a while back that I was working in the British Library. Within minutes, a friend responded that so was he. So we had lunch. That's pretty cool. Allowing people who know each other online to serendipitously meet up IRL is a great thing, and we need to maximize it without enabling creepy stalkers.
I have had a Wave Sandbox account for a few weeks. I don't know why. Something to do with going along to two Google developer events in London.
I don't buy that it's going to replace e-mail. The great thing about e-mail is that it is already massively distributed and established. There are e-mail clients already existing for every platform. Pretty much the first application that gets written for any operating system these days is an e-mail client or a web browser. E-mails are sent in an extremely easy-to-understand, widely implemented format - plain text. There are already hundreds of e-mail servers, and if you are using a mature Linux distro like Debian or whatever, e-mail isn't that complex to set up.
Wave requires much more technology: you've got to host XMPP, you've got to host whatever interface magic you are doing to serve up the Wave experience to the end user (which means another HTTP service, at least until Wave clients start being available for Windows, Mac and $YOUR_PLATFORM), you've got to host all the media that gets dumped into the Wave. Beneath all of that, you've got to understand a fair bit about XMPP, and you've got to understand a fair amount of the complexities of XML, HTML, JS and so on. And, well, is the extra complexity and extra cost (in resources) of Wave really worth it? I'm not sure it is.
As a technology, Wave is awesome. Drag and drop files into your browser and watch them appear magically on another computer. The tech demo is really impressive. I'm impressed by Wave as a demo of just what is possible using the browser and a stack of open standards and open source technologies. I'm just not convinced that it's really been thought through. It won't replace e-mail or IM: it's too complex compared to them. As for collaboration, compare it with wikis or EtherPad - both are also a lot simpler. The Wave UI is mind-bogglingly complicated. There's no way my parents are going to be riding the Wave.
Wave seems to be predicated on there being something wrong with e-mail. I don't think there is. The only problems with e-mail are social, not technological. The primary problem with e-mail is that people don't follow RFC 1855 - people sending stupid rumours and chain letters, spam and spammy notifications, HTML crapola, and not being able to figure out how to quote correctly. Oh, and paranoid corporations who feel the need to shove 20K of legal crap at the bottom of the e-mail telling me that it's confidential (I never signed anything), that if I got it in error I should delete it (no, fuck off), not to print it out (I'll do what I want with my e-mail and my laser printer) and that it's been through their virus scanner (okay, how does that affect me?).
Summary: Wave is a cool tech demo, but I don't buy that it's a replacement for e-mail.
I first read it in 'Getting Started with FreeBSD' - a shortened, printed version of the FreeBSD handbook, if I recall correctly. I had to go to a local free software distributor to buy the CD-ROMs and a printed manual. I'm guessing this was 1997 or 98, and I desperately wanted an operating system that wasn't Windows 95. I've still got the disk package floating around somewhere - version 2.2.6.
I'm not sure if I'm "the most diligent of programmers", but I'm pretty sure HelloWorldPrinter.java doesn't quite do enough to crash the OS:
class HelloWorldPrinter { public static void main(String[] args) {
System.out.println("You say 'Hello, world!'");
System.out.println("The world sighs back.");
} }
Perhaps I'm not as imaginative as the Cardinal, but I cannot figure out how to crash the OS with just this code.
Unless, of course, you hand the code to someone else, alias their Java compiler or runtime to echo "The Java compiler needs root privileges due to a weird Debian [or your distro here] permissions problem." and then sudo rm -rf /
That requires you to think like a bastard though.
(Written in Java, because if I had written it in Smalltalk, Python, Ruby or Lisp or whatever, six lines of code would be enough to do something significantly cleverer than printing two lines.)
I've been watching the farce that is the new ICANN process. They seem to think that the only test for a new top-level domain suffix is that it will make money for the corporate owner.
I think it's bullshit. The new domains they have added so far -.aero,.coop,.museum,.biz,.info,.pro,.jobs,.tel - all suck. Okay, I quite like to be able to type http://british.museum/ - but.biz and.info are pretty much used exclusively by make-money-fast scammers, phishers, scammers and other assorted lowlife who like it because it costs a third of the price of a real domain name. I can think of only a small handful of.info domain names that are actually legit - http://rdfa.info/ comes to mind..aero? What the fuck? Are Boeing and Delta Airlines really having difficulty with domain squatting?.tel and.jobs are complete bullshit. I mean, if you suddenly decide that you want to know the telephone number of your local Starbucks, are you more likely to (a) go to Google and type in "Starbucks Manhattan" or (b) type in manhattan.starbucks.tel. Same for jobs. If you decide you want to get a job working for BigCompany, you go to their website and you click on 'jobs' or 'careers' or whatever. Or you Google. You don't think "Oh, I better go to BigCompany.jobs."
We've also had.mobi, which is completely stupid. The current generation of mobile devices - iPhone, Android G1, Pre, Blackberry - are becoming increasingly capable of showing normal websites. And rather than registering.mobi, most sites seem to think m.whatever.com (for generic mobile: some are also doing i.whatever.com for iPhone-specific sites, and back in the day people had pda.whatever.com for PDA specific sites, wap.whatever.com for stuff written in WML and served as WAP content - I'm sure in Japan, there's probably i-mode specific subdomains or pathnames).
But now ICANN seem to want to make it so that anyone can have a TLD if they pay. There have already been people suggesting.nyc,.london,.paris and other city names (hint:.us,.uk,.fr). Big companies are thinking about putting in bids for their trademarks: I've seriously seen it suggested that someone like Michael Jordan register.jordan, then can have clothes.jordan, shoes.jordan, basketball.jordan and so on.
As for.eco: the only justification for it seems to be that the sale of.eco domains could be used for some useful social end - sell domains, plant trees, reverse global warming or whatever. Okay. Great. That's not nearly enough justification. Hosts that are going to want a.eco domain are either commercial entities (use.com), non-commercial entities (use.org), country-specific domains (use the country code TLD), government (use.gov or your local government national TLD -.gov.uk or whatever), or maybe academic (use.edu or.ac.uk or your country's national TLD).
ICANN are corporate whores and have proven they are completely untrustworthy. The domain name system needs to have a Revenge of the Nerds moment. We need to take it back and put it under the careful guidance of an IETF-like body who will decide TLD decisions on the basis of the needs of the Internet community, in an open and democratic process, rather than a process where you basically turn up with a wheelbarrow full of cash and a dumb idea and then get proclaimed owner of a new TLD. TLDs need to be distributed by IETF-like wise beardy Internet veterans rather than assholes who think the
If you need to do LaTeX, just use LyX. It's ugly, but powerful. Since switching to LyX, I've had no reason to use Microsoft Office or OOo. I very occasionally use Apple's iWork suite when I'm using OS X. My only substantial word processing need is academic, so I just use LyX. Everything else is just plain text, so Vim does the job. One day I'll learn LaTeX properly, and just edit it in Vim.
Word processors are for secretaries and office drones.
SubEthaEdit is different - what the grand-parent poster is pointing to is remote sessions - where you basically take your text editor (or LISP runtime or whatever Emacs is these days) and can log into it remotely and execute commands. A bit like GNU Screen or a bit like VNC.
What makes SubEthaEdit cool is the collaborative editing environment: that lets you have one document with multiple people editing it at the same time all using their own installations of SubEthaEdit. Well, there actually is a way of doing that with Emacs too. Gobby is an X11/GTK+ app that does much of what SubEthaEdit does (but without the Mac OS X prettiness). IIRC, there's a plugin for Emacs available which lets it join Gobby/Sobby sessions. (Sadly, nothing for us Vim users.)
There's also EtherPad which is all written in JavaScript and runs in the browser. I use it for collaboratively scribing meetings. Sadly, much as I'd like Gobby to be cross-platform, it's not quite there: it works great on Linux. I've heard nothing bad about it on Windows. But on Mac OS X, it's a pain in the arse to get working.
If you were a vi/Vim user, then Vimperator. It's Firefox with the totally rational key-bindings Vim users like me expect.
Emacs users can use Conkeror instead. This blog post is a pretty neat intro to Conkeror.
Looks like Conkeror has what Vimperator calls hinting - in Vimperator you hit f or F (for new tab hinting) and it labels all the links on the screen with a number. You type in the number you want and it opens it. M-l does the same in Conkeror.
One of the nice things I like in Vimperator is that you can write (in JavaScript, thankfully, not VimL) really simple plugins and remaps. There's a plugin you can get which makes it so you can use the:shorten command and it takes the URL of the current page, shortens it through $URL_SHORTENER_OF_YOUR_CHOICE and then yanks (that is, copies to the system pasteboard) the result. Or if you want to yank just the current URL, you can tap "y".
Firefox extensions that make Firefox like editors almost makes the GUI usable. Vimperator is certainly very cool on a small netbook as it really helps getting all the GUI clutter out of the way.
Twitter is actually pretty handy: if I've got a tech problem that I can't resolve by Reading The Fucking Manual or Googling, I post it on Twitter and usually get really good answers and links, often within a few minutes. But then I don't follow Oprah Winfrey and Ashton Kutcher, I follow a bunch of knowledgable geeks.
I'm not sure what the problem is with the #amazonfail thing. People make mistakes. It wasn't a lynch mob. It was a bunch of people discussing something on Twitter with a hashtag. Big fucking deal. Go on to Twitter Search and type in the name of a game or a programming language. Are all those people part of an angry mob?
I remember the amazonfail. Most of the posts were pretty sceptical. They were like: this is a bit weird - Amazon have deranked all the gay-themed books. If this is legit and not a hack, Amazon aren't getting any more business. (Note the conditional word "if".) There was also some discussion about how much of our lives are stored on Amazon servers - pointing out that if #amazonfail turned out to be true, it might actually be quite a bit of work to untangle all the Amazon Web Services stuff (S3, EC2 etc.). Thankfully, it turned out to have a perfectly reasonable explanation. I think the level of belief on Twitter was pretty proportional to the evidence.
Slashdot Mirror
Theft Act 1968, not '58. Damn computer should output what I think not what I type.
1. It needs to be "please burgle me". If you aren't at home, then you are being burgled, not robbed. A robbery is theft with violence or the threat of violence (at least in English law - Theft Act 1958 - it is). If nobody is at home, then nobody can be the victim of violence or the threat of violence. So your home is getting burgled - or, if you are an American, burglarized (what the hell kind of a word is that, right?).
2. PleaseRobMe seems to be built around the premise that one home = one person. If you know where I live, please be assured that I am currently not at my home. But other people live where I live. Families exist. Flat sharing exists. Communal living exists. (Yeah, go and raid the kibbutz - I'm sure it'll be empty!) This may be true for Web 2.0 valleyboys. It's not true of the rest of the planet.
That said, this kind of thing does show why most location-based services are stupidly designed. I have played around with a few of them, and the only one I'm a real big fan of is FireEagle. Sadly, it's been a bit neglected for business reasons - i.e. Yahoo! financial situation. What is great about FireEagle is you share you location with FireEagle, and they then share it with whatever services you want to share it with. So, I have the little iPhone app which updates FireEagle. FireEagle knows exactly where I am. Then there's a Facebook app which connects to FireEagle, but I don't necessarily have to let it broadcast my location if I don't want it to. Or I can only give a vague location - perhaps at a country or city level. I have it wired in to my SSH setup, so if I SSH in to my Linux box from certain places, it updates my location. Because it is a location broker, it can be updated in any way people think of, rather than having to use a specific application (say, for the iPhone) like FourSquare etc. do.
This is useful as I can build applications that sit on top of it. One I have been meaning to build is a "remind me when I'm at X" app. So I could basically dump a string (SMS/tweet length) into a database with a broad location in it. It could check against my location and when they match, I could be reminded of X. Remember to buy ice cream when I'm at the supermarket - well, when I'm at the supermarket, I should get a text message saying to buy ice cream.
Location-based services shouldn't be tied to devices but to people. This is what everyone gets wrong. They need really good granular privacy controls. They need a big "forget me" option. This is something Google Latitude doesn't have. There is no way I have found to tell Google Latitude "Hey, take me off the radar. I'm not anywhere anymore."
When I have some time to build it, I'd love to build something like FireEagle but running on my servers and just for me. Location is too important, useful and fun to trust Google or Yahoo! or some venture-backed Valley startup with. But if you are building location-based services, look at FireEagle and learn.
Sure, pseudocode is good for getting ideas across to other human beings and developing a rough idea of program flow... But it isn't going to compile. And it doesn't matter how much you argue that programming languages evolve over the years and get new features added and whatnot, your pseudocode still isn't going to compile.
Google "executable pseudocode". I think the Python community have worked pretty hard on producing a good pseudocode interpreter...
As you say, spoken English is different from written. Quoted and reported speech style is a difficult area - I'm pretty sure that newspapers would often lightly clean up quotes to make them match the style guide. It seems that may have fallen off in recent years. I'm quite ambivalent about it: as a reader, I would certainly prefer the quotes to be readable. As someone who has been subtly misquoted in a newspaper to the point where the meaning of what I said was changed (as well as my position being incorrectly attributed), I would rather like it if my words could be converted from audio to text with as little change as possible. If you wish to clean up what someone said, paraphrase it as reported speech rather than quotation.
Plenty of people do abuse spoken language too. One of my favourite examples is the "turned around" trope in British English. When people describe heated discussions and arguments they have, they use "turned around and said" instead of just "said". Someone will give a little monologue of the sort: "I was in the pub with my girlfriend, when all of a sudden she turned around to me and said [...] and I turned around and was like [...], and then she turned around and said [...]". When someone does this, all I can picture is a video I once saw of Whirling Dervishes spinning around and around, a spin around interrupting each line of the dialogue. It's kind of a strange nervous tic that people add to make what they say sound more dramatic when the words aren't quite enough. Som people seem to add it to every line which rather makes it lose any power.
LAMP = Linux, Apache, MySQL, PHP
Naah. PHP is old. Get with the times. Ruby, man. If only because you can use the following stack abbreviations:
LAMR = Linux, Apache, MySQL, Ruby
LARP = Linux, Apache, Ruby, PostgreSQL*
Don't be a LAMR (because everyone hates lamers) or a LAMPpost (dogs will pee on you) - be a LARPer instead! You get to fight people with polystyrene swords and wear a cape! And then come home and use a good RDBMS and a nice programming language!
* Justification for reversing the database and programming language: you can build an application quite easily that has no database backing. Flat-file or just in-memory storage or whatever. If you wrote an application that simply read an XML file from the filesystem, you would not feel the need to include the IO library or the XML parsing library in the stack diagram. So, therefore, the database and the programming language are really not depdendent on one another. The order of the letters in the abbreviation have little actual meaning, so the strategy of ordering them so that the abbreviation sounds cool and makes a nerdy reference is just a valid strategy as any other. Sort of.
Froze the account completely except for the £25 expedited transfer - profit trumping security in UK banking land, of course. It's a savings account (ISA), so there is no debit card. Still, when they were paying me 5.75% without tax pre-recession, these inconveniences could be handled.
Does that translate as "I can't remember my password and I keep getting locked out. Also, I'm so poor at managing my cashflow that I'm always overdrawn"?
You want to know what it actually is? A bank I used to use had an online security model which required you to answer with both a password and with a 'secret answer'. The 'secret question' was "What's the name of your first school?" - and it was the same for everybody. Ignoring the fact that now so many people are on Facebook and MySpace, that's hardly a useful security measure at all, there is a problem with punctuation. If you went to, oh, St. Paul's Roman Catholic Primary School, that could quite easily be Saint, St., or St, Pauls, Paul's, or Paul, Roman Catholic, Catholic, RC, R.C., Primary, Primary School or many other variants thereof. It turns out the bank was automatically stripping any characters that weren't [a-zA-Z\s] from the secret answer. Which is fine, except they don't tell you that. And you only have to mistype it twice or something and you are locked out. I didn't even mistype it - I have all my bank details in an encrypted text file, so I just copy and paste. If I write out the string, store it in a file, copy and paste it into the field to set it, and then come back a week later and enter it in the same field through copy and paste and it doesn't work, that's quite evidently not my fault.
Once they've locked you out from online banking because of their shitty software, it then takes them two weeks to unlock it. During which time, no transactions allowed. It took until the third time I'd been locked out of this process until they finally explained their byzantine string processing rules. And because it's internet banking, there is no other way of getting any money out. Except one. You can phone them up, and they won't do a BACS transfer for "security reasons" - but if you are willing to pay £25 they'll do a priority transfer. Think about that: you can't have money transferred slowly from your account for security reasons, but if you are willing to pay a fee, those security reasons disappear. I'm sure the fraudster intent on ripping off a savings account will pause in wait, wondering whether or not he wants to spend £25 of someone else's money in order to put a transfer through.
Sadly, stupid excuses about idiotic programmers and their inept handling of strings doesn't cut much ice with people one owes money to.
Of course, once it had cost me my second or third £12 bank fine, I took my business elsewhere. And, no, I generally manage my money pretty well - my student loan is about a third of the size of a lot of my friends, I clear my credit card every month, my laptop is pretty much my only luxury - bla bla bla.
My point was very simple: if you go a pound overdrawn and the bank then wham you with a £12 charge for it, the least they can do in return is to process some damn cheques without whining about how much it costs.
That's exactly the right point. The Payments Council don't seem to have really thought out person-to-person transfers or small business transfers. If someone wants to pay me money using a cheque, all they need to know is my name (or the registered trading name of a business etc.). With a bank transfer, you need to know the bank account details of the person - and nobody is quite sure as to how public bank account numbers ought to be. Cheque still rules for getting paid expenses too. I've done the PayPal thing, which is cool - except PayPal take a slice of the transaction. I've done the online banking stuff and it's painful - crappy websites, no notification (how about an e-mail from my bank every time a transaction goes through my account? GPG exists, goddamnit - use it!), security designed for the sort of people who set their passwords to "password1" and tell all their idiotic friends their MySpace password and then wonder why they get "hacked". If you think writing a cheque is inconvenient, how about carrying a laptop and 3G dongle around with you in order to do bank transfers on the website instead?
Yeah, in shops, use a damn debit card. But for person-to-person transactions, cheques are pretty convenient. The security sucks, admittedly - a shared, public key that anyone can copy is not really security at all (a bit like credit cards: credit card fraud, at least until a few years ago, consisted of writing down the number and details that was given out to the merchant upon every purchase - that's real secure). But with cheques, you can post them - slip 'em inside birthday cards. You can give them to third parties (children, employees etc.) who can hand them to their eventual target. You can use them even where there isn't an Internet connection - and, well, outside of the big cities, there's plenty of rural countryside with no 3G service. You can post-date them, and the recipient can return them - my old school used to do this with their after-hours activity programmes - you'd give them half in a current-dated cheque and the other half in a post-dated cheque, so if you don't decide to finish the activity, they just return the second half of the payment.
The only problem with cheques is that I have to walk to the bank to pay them in and it costs the banks money to process them. The walk is quite good exercise, and since the banks got £300 billion of taxpayers money last year under the Special Liquidity Scheme, and they pay themselves HUGE FUCKING BONUSES, I figure the odd 25p here or there to process my damn cheques is pretty reasonable. Not to mention the huge amounts they've had in fees and fines - fines they've charged me due to the ineptitude of other banks (who have web security models designed for the aforementioned idiots). Not to mention interest they make off the money we keep in them. They seem to want to have it both ways: they argue that cheques cost a lot to process AND nobody is using them. But if nobody is using them, surely the number of people you need to employ processing cheques is pretty minimal.
While I'm ranting about banks, here's another thing: it's only in the last year that UK banks have actually got their shit together to be able to move money between accounts and it take less than four fucking days. A family member moves the housekeeping money from one account to another every month. He does this by going to one bank, drawing out however much it is (a few hundred pounds), walking down the street to the other bank and paying it in. Every time, the helpful bank assistant seems to suggest that he could do this electronically. The difference is, if you pay in cash, it is immediately available - while a BACS transer takes 3-5 days. They've only recently changed this so that it takes at most a few hours. But Christ-on-a-fucking-wheel, why did it take until 2009 to be able to move money instantly from one account to another? And it's still only certain banks that do the same-day transfers. These guys are absolutely retarded. With the billions they make every second, you'd think they'd be able to install a few broadband lines between their offices and make it so money can get transferred quickly. Three days - seriously?!
'Too big to fail' presumes the banks aren't the epitome of fail to start with.
Here's my solution: get a laptop and two netbooks, one with a spare battery. Use laptop until it runs out of battery. Then switch to netbook #1. Use that until it runs out of battery. Then switch to netbook #2. Use that, switch battery, continue.
That's how I roll. It's totally insane (and probably hernia-inducing), and I so don't want to get mugged, but it lasted for 6+ hours while on a long train ride recently. It's a complete accident - I went to a conference where they gave everyone a netbook about a week after I bought a netbook from a friend.
But, yes, a netbook with 15 hour battery life would be awesome. Part of the problem is recharge: yes, when I'm working, I can plug my computer in to recharge. But it takes a long time. Most of the day, in fact. I think we are really missing out on super high-speed battery charging - both for computers and for electric cars. If it takes 2 hours to discharge a laptop battery, it should take ~20 minutes to charge it up again. We've gone wireless for networking. I can't wait until we have batteries that'll last long enough so we can go wireless for computing too...
I don't think we nerds are bad at naming things. From all I can tell, 'cloud computing' is the invention of consultants, VCs, vendors and other notorious buzzword merchants. Nerds come up with names like scrotwm.
As for cloud computing? It's a really dumb buzzword. People have had applications running on the Internet for a long time. Some of these are somehow 'cloud' and others aren't. Apparently, Gmail's IMAP servers are in the cloud, while my university's IMAP servers are just on the dull old Internet. If I store stuff on Dropbox or whatever, I've got stuff in the cloud. If I store them on an FTP server, just dull old Internet. If cloud computing is accessing applications over the Internet, then we've been doing cloud computing since the 80s with telnet. If cloud computing is virtualization, then we've been doing that for some time too. But it's neither: it's just a buzzword. See Larry Ellison's brilliant rant.
UbuntuOne was the last straw for me - I saw that and am moving all my Ubuntu machines over to Debian Sid. I trust the Debian community not to get sucked into stupid marketing hype.
Well, ideally, they'd be able to get excited about doing it differently, maybe doing it better. We have to reconcile ourselves to the fact that non-programmers may make decisions about software for less-than-optimal reasons (A: "We're using XML"; B: "Why?"; A: "Well, anything that begins with 'X' has to be cool and the industry press has talked it to death."; B: "But, but..."; A: "I pay your salary."; B: "Okay then. I'll just take our CSV data and wrap it in a 'CSVData' element."), but that we can at least try to enjoy the ride.
What you said about software "both unbelievably ephemeral, yet also incredibly long-lived" is true of lots of things, but it reminds me a lot of jazz. That's certainly how I try and think of programming: whatever else changes, try and have fun with it, and try to work on interesting projects. If the implementation stays there for 50 years, great. If it doesn't, hopefully you still had fun implementing it, it still helps someone do something useful, and you have become a better programmer.
Damn right. I walk in London and use buses/trains to get around. Cyclists are the biggest bunch of inconsiderate arseholes you can find. Many of them think red lights only apply to cars, don't stop at zebra crossings, don't signal, go the wrong way up one-way streets, scoot out in front of traffic. They clutter up the pavements - outside the Lyceum Theatre on Wellington Street, there's a special cyclists crossing across the corner of Strand/Aldwych to get onto Waterloo Bridge. But it goes all across the pavement, and if a crew of cyclists are waiting at the crossing, they make it so that it's actually impossible for pedestrians to use the pavement.
And then they whine endlessly about how there aren't enough cycle lanes. And so the government spend all this money building all these stupid cycle lanes on roads nobody uses in order to please the cycling lobby. In the university area of London, on Malet Street WC1, they've built all these cycle tracks. All very important, of course, since there's maybe two cars an hour go down the damn road (and the cyclists have a bunch of cut-throughs across the campus). But the cyclists just won't survive unless they have some special lanes to go down. Don't worry, though, they pay for it through their cycle tax, and cycle insurance and, oh, hang on...
I even found this: Stop At Red, a campaign by cyclists to pledge to obey the Highway Code and other road laws. That something like this gets created says a lot.
Don't get me wrong: I love cycling and if I could snap my fingers and make half the cars in the country disappear and be replaced with cycles, I'd do it in an instant. But as a pedestrian in London, cyclists can be really fucking inconsiderate.
British movie/game ratings are pretty simple.
There is:
U - Universal - suitable for all
PG - Parental Guidance
12/12A - for videos, nobody under age 12 is allowed to purchase it. for movies shown in a cinema, under twelves can watch it if they are accompanied by an adult.
15 - suitable for 15+
18 - suitable for 18+
R18 - Restricted 18 - basically porn. Can only be shown in specially licensed venues and sold only in licensed sex shops.
'E' is on some videos. It's not actually a rating, but it's just a symbol put on by video producers to specify that the film is exempt from rating. Things like videos of sports matches, musical performances, educational videos don't get rated. Most imported videos
The BBFC also now produce some text that accompanies the rating symbol which broadly gives the reason why the film is rated that way. For instance, it might say "Contains frequent strong bloody violence and very strong language" next to an 18 certificate.
They introduced R18 a few years ago when they realised that hardcore porn was mostly being circulated through the black market and by people distributing copied tapes. Of course, now, we have the new rules on "violent and extreme" pornography which actually makes it a crime to possess pornography that depicts violent scenarios - rape fantasies, that kind of stuff. With one hand, the government make porn a bit more legal, and with the other hand, they've created a new black market that the Internet supplies.
The BBFC is generally, imo, pretty fair - I mean, as fair as a bunch of censoring, free-speech-restricting thugs can be. Perhaps it's just bias living here, but BBFC seem to get it right a lot more often than the MPAA ratings do, and they are a hell of a lot less squeamish about depiction of sex and nudity - they make a distinction for 'natural nudity' where it's non-sexual, so we don't have idiotic philistines sticking big blocky pixels over Dürer woodcuts, Titian paintings and Michaelangelo's David (etc.) because they've got HUR HUR HUR DONGS LOL. There's an interesting set of articles by people who have worked at BBFC, describing exactly what it's like censoring movies and video games for a living.
There are some things I think location-based services, now including Twitter, need to learn from previous experience of other location-based services like FireEagle:
1. Don't require me to use any specific technology to update it. I don't have a smartphone, and don't really want one either. I have an iPod touch and I have a mobile phone. I quite like them being separate (the fact that my phone lasts a few days on standby and isn't tied to any specific provider is pretty nice). That said, I do use location based technologies: I just use some interesting ways to update them. The primary way I update my location is using the /var/log/auth.log data on my Linux machine. Whenever I log in over SSH, my computer checks to see the domain I'm logging in from. I have a file on the computer that maps the IP addresses to real street addresses, and then notifies FireEagle, Yahoo's location brokerage service, which then appears in Facebook (and also lets me go to sites like pubs.iamnear.net and find out pubs that I am near). Decouple data collection from data use. My phone doesn't know where I am - it's a pretty dumb phone. But that doesn't mean none of the gadgets or technology I use don't.
2. History isn't that much of a problem, but live data is. I don't care that the whole world knows I was wandering around central London a few days ago. I don't mind my friends knowing where I am now (so long as I have some kind of prominent kill switch that cleanses the Internet of my current location). The problem is when everyone knows where I am right now. If someone knows that I'm some particular place, they know I'm not at home - which means they can burgle my home.
3. And, well, I don't want even my friends - or, rather Facebook/Twitter type friends - to know where I live unless I tell them. What location services that hook up to social networks need is a way to say "when I am near this address, don't tell anyone".
The privacy stuff is important, and I hope Twitter take it really slowly and listen to user and developer feedback. There is a good side to location services which too much focus on privacy ignores: they let you meet up with your friends. I posted on Twitter a while back that I was working in the British Library. Within minutes, a friend responded that so was he. So we had lunch. That's pretty cool. Allowing people who know each other online to serendipitously meet up IRL is a great thing, and we need to maximize it without enabling creepy stalkers.
I have had a Wave Sandbox account for a few weeks. I don't know why. Something to do with going along to two Google developer events in London.
I don't buy that it's going to replace e-mail. The great thing about e-mail is that it is already massively distributed and established. There are e-mail clients already existing for every platform. Pretty much the first application that gets written for any operating system these days is an e-mail client or a web browser. E-mails are sent in an extremely easy-to-understand, widely implemented format - plain text. There are already hundreds of e-mail servers, and if you are using a mature Linux distro like Debian or whatever, e-mail isn't that complex to set up.
Wave requires much more technology: you've got to host XMPP, you've got to host whatever interface magic you are doing to serve up the Wave experience to the end user (which means another HTTP service, at least until Wave clients start being available for Windows, Mac and $YOUR_PLATFORM), you've got to host all the media that gets dumped into the Wave. Beneath all of that, you've got to understand a fair bit about XMPP, and you've got to understand a fair amount of the complexities of XML, HTML, JS and so on. And, well, is the extra complexity and extra cost (in resources) of Wave really worth it? I'm not sure it is.
As a technology, Wave is awesome. Drag and drop files into your browser and watch them appear magically on another computer. The tech demo is really impressive. I'm impressed by Wave as a demo of just what is possible using the browser and a stack of open standards and open source technologies. I'm just not convinced that it's really been thought through. It won't replace e-mail or IM: it's too complex compared to them. As for collaboration, compare it with wikis or EtherPad - both are also a lot simpler. The Wave UI is mind-bogglingly complicated. There's no way my parents are going to be riding the Wave.
Wave seems to be predicated on there being something wrong with e-mail. I don't think there is. The only problems with e-mail are social, not technological. The primary problem with e-mail is that people don't follow RFC 1855 - people sending stupid rumours and chain letters, spam and spammy notifications, HTML crapola, and not being able to figure out how to quote correctly. Oh, and paranoid corporations who feel the need to shove 20K of legal crap at the bottom of the e-mail telling me that it's confidential (I never signed anything), that if I got it in error I should delete it (no, fuck off), not to print it out (I'll do what I want with my e-mail and my laser printer) and that it's been through their virus scanner (okay, how does that affect me?).
Summary: Wave is a cool tech demo, but I don't buy that it's a replacement for e-mail.
A huge chunk of Yahoo! runs on PHP and a huge chunk of Google runs on Python.
I first read it in 'Getting Started with FreeBSD' - a shortened, printed version of the FreeBSD handbook, if I recall correctly. I had to go to a local free software distributor to buy the CD-ROMs and a printed manual. I'm guessing this was 1997 or 98, and I desperately wanted an operating system that wasn't Windows 95. I've still got the disk package floating around somewhere - version 2.2.6.
I'm not sure if I'm "the most diligent of programmers", but I'm pretty sure HelloWorldPrinter.java doesn't quite do enough to crash the OS:
class HelloWorldPrinter {
public static void main(String[] args) {
System.out.println("You say 'Hello, world!'");
System.out.println("The world sighs back.");
}
}
Perhaps I'm not as imaginative as the Cardinal, but I cannot figure out how to crash the OS with just this code.
Unless, of course, you hand the code to someone else, alias their Java compiler or runtime to echo "The Java compiler needs root privileges due to a weird Debian [or your distro here] permissions problem." and then sudo rm -rf /
That requires you to think like a bastard though.
(Written in Java, because if I had written it in Smalltalk, Python, Ruby or Lisp or whatever, six lines of code would be enough to do something significantly cleverer than printing two lines.)
Yep, and they also miss out the BSD Daemon, and Hexley the Platypus - which beat the corporate Windows and OS X logos any day. And lacking BSD, they miss the story of the two Texans reacting to the BSD daemon T-shirt, one of the best stories in BSD history.
I've been watching the farce that is the new ICANN process. They seem to think that the only test for a new top-level domain suffix is that it will make money for the corporate owner.
I think it's bullshit. The new domains they have added so far - .aero, .coop, .museum, .biz, .info, .pro, .jobs, .tel - all suck. Okay, I quite like to be able to type http://british.museum/ - but .biz and .info are pretty much used exclusively by make-money-fast scammers, phishers, scammers and other assorted lowlife who like it because it costs a third of the price of a real domain name. I can think of only a small handful of .info domain names that are actually legit - http://rdfa.info/ comes to mind. .aero? What the fuck? Are Boeing and Delta Airlines really having difficulty with domain squatting? .tel and .jobs are complete bullshit. I mean, if you suddenly decide that you want to know the telephone number of your local Starbucks, are you more likely to (a) go to Google and type in "Starbucks Manhattan" or (b) type in manhattan.starbucks.tel. Same for jobs. If you decide you want to get a job working for BigCompany, you go to their website and you click on 'jobs' or 'careers' or whatever. Or you Google. You don't think "Oh, I better go to BigCompany.jobs."
We've also had .mobi, which is completely stupid. The current generation of mobile devices - iPhone, Android G1, Pre, Blackberry - are becoming increasingly capable of showing normal websites. And rather than registering .mobi, most sites seem to think m.whatever.com (for generic mobile: some are also doing i.whatever.com for iPhone-specific sites, and back in the day people had pda.whatever.com for PDA specific sites, wap.whatever.com for stuff written in WML and served as WAP content - I'm sure in Japan, there's probably i-mode specific subdomains or pathnames).
But now ICANN seem to want to make it so that anyone can have a TLD if they pay. There have already been people suggesting .nyc, .london, .paris and other city names (hint: .us, .uk, .fr). Big companies are thinking about putting in bids for their trademarks: I've seriously seen it suggested that someone like Michael Jordan register .jordan, then can have clothes.jordan, shoes.jordan, basketball.jordan and so on.
As for .eco: the only justification for it seems to be that the sale of .eco domains could be used for some useful social end - sell domains, plant trees, reverse global warming or whatever. Okay. Great. That's not nearly enough justification. Hosts that are going to want a .eco domain are either commercial entities (use .com), non-commercial entities (use .org), country-specific domains (use the country code TLD), government (use .gov or your local government national TLD - .gov.uk or whatever), or maybe academic (use .edu or .ac.uk or your country's national TLD).
ICANN are corporate whores and have proven they are completely untrustworthy. The domain name system needs to have a Revenge of the Nerds moment. We need to take it back and put it under the careful guidance of an IETF-like body who will decide TLD decisions on the basis of the needs of the Internet community, in an open and democratic process, rather than a process where you basically turn up with a wheelbarrow full of cash and a dumb idea and then get proclaimed owner of a new TLD. TLDs need to be distributed by IETF-like wise beardy Internet veterans rather than assholes who think the
If you need to do LaTeX, just use LyX. It's ugly, but powerful. Since switching to LyX, I've had no reason to use Microsoft Office or OOo. I very occasionally use Apple's iWork suite when I'm using OS X. My only substantial word processing need is academic, so I just use LyX. Everything else is just plain text, so Vim does the job. One day I'll learn LaTeX properly, and just edit it in Vim.
Word processors are for secretaries and office drones.
SubEthaEdit is different - what the grand-parent poster is pointing to is remote sessions - where you basically take your text editor (or LISP runtime or whatever Emacs is these days) and can log into it remotely and execute commands. A bit like GNU Screen or a bit like VNC.
What makes SubEthaEdit cool is the collaborative editing environment: that lets you have one document with multiple people editing it at the same time all using their own installations of SubEthaEdit. Well, there actually is a way of doing that with Emacs too. Gobby is an X11/GTK+ app that does much of what SubEthaEdit does (but without the Mac OS X prettiness). IIRC, there's a plugin for Emacs available which lets it join Gobby/Sobby sessions. (Sadly, nothing for us Vim users.)
There's also EtherPad which is all written in JavaScript and runs in the browser. I use it for collaboratively scribing meetings. Sadly, much as I'd like Gobby to be cross-platform, it's not quite there: it works great on Linux. I've heard nothing bad about it on Windows. But on Mac OS X, it's a pain in the arse to get working.
If you were a vi/Vim user, then Vimperator. It's Firefox with the totally rational key-bindings Vim users like me expect.
Emacs users can use Conkeror instead. This blog post is a pretty neat intro to Conkeror.
Looks like Conkeror has what Vimperator calls hinting - in Vimperator you hit f or F (for new tab hinting) and it labels all the links on the screen with a number. You type in the number you want and it opens it. M-l does the same in Conkeror.
One of the nice things I like in Vimperator is that you can write (in JavaScript, thankfully, not VimL) really simple plugins and remaps. There's a plugin you can get which makes it so you can use the :shorten command and it takes the URL of the current page, shortens it through $URL_SHORTENER_OF_YOUR_CHOICE and then yanks (that is, copies to the system pasteboard) the result. Or if you want to yank just the current URL, you can tap "y".
Firefox extensions that make Firefox like editors almost makes the GUI usable. Vimperator is certainly very cool on a small netbook as it really helps getting all the GUI clutter out of the way.
Twitter is actually pretty handy: if I've got a tech problem that I can't resolve by Reading The Fucking Manual or Googling, I post it on Twitter and usually get really good answers and links, often within a few minutes. But then I don't follow Oprah Winfrey and Ashton Kutcher, I follow a bunch of knowledgable geeks.
I'm not sure what the problem is with the #amazonfail thing. People make mistakes. It wasn't a lynch mob. It was a bunch of people discussing something on Twitter with a hashtag. Big fucking deal. Go on to Twitter Search and type in the name of a game or a programming language. Are all those people part of an angry mob?
I remember the amazonfail. Most of the posts were pretty sceptical. They were like: this is a bit weird - Amazon have deranked all the gay-themed books. If this is legit and not a hack, Amazon aren't getting any more business. (Note the conditional word "if".) There was also some discussion about how much of our lives are stored on Amazon servers - pointing out that if #amazonfail turned out to be true, it might actually be quite a bit of work to untangle all the Amazon Web Services stuff (S3, EC2 etc.). Thankfully, it turned out to have a perfectly reasonable explanation. I think the level of belief on Twitter was pretty proportional to the evidence.