varunnangia.com

Wired has a good article up about why the future of digital music is still the crusty old MP3 format from the late 1980s. Probably the most important reason, though Wired mentions it only in passing is compatibility. Everything – and I really, truly mean everything – supports MP3. Looking around the house, I’ve yet to find a device or gadget that plays music in a format other than MP3. Sure the iPhone supports AAC, and the N800 supports Vorbis, but both devices support MP3 and that’s why I think MP3 will be around a long, long, long time yet. And judging by popular reaction, I’m not alone in this frame of mind.

In fact, the best way to understand MP3′s wide compatibility is to look at a market where there is not a standard like MP3. For example, I would very much like to backup a number of my Indian movies, which are already suffering from disc rot. However, there is no one single file format that I can save my movies too. If I save it in DivX, the most widely used video file format, I cannot play them back on the television, since none of the devices connected fully support the playback of DivX. Neither can I watch them on my iPhone, since DivX is an alien concept to Apple. On the other hand, saving the file as a H.264 file ensures I can play back on the Xbox, but the file will stutter on the iPhone. If I choose to setup a file that doesn’t stutter on the iPhone, I artifacts that make the file unwatchable on the Xbox. And heaven forbid I should try to watch this on my N800, or on the computer. I find myself in the unenviable task of having to making five different versions in order to ensure universal playback, a situation that should never arise in the first place, if there was a standard that just worked, like MP3 does for audio.

So, if anyone has a good solution to my problem, I’d love to hear it – what’s the video version of MP3?

Until I started reading Orson Scott Card’s Ender’s Game, I had not understood the expression “insists upon itself”. However, reading it for the first time in my quest to go through all the major works of sci-fi, I finally grasped the concept for the first time.

I’ve never really liked a story that exists for a purpose other than telling a good yarn. A good example of this is language use. For example, I personally found the language of Paolo Coelho and Arundhati Roy to be very similar in their flowery nature. However, Roy’s book exists to win a prize, to be dissected for meaning, for those to whom language takes precedence over story. Coelho, for all his literary techniques, tells a great story, made greater still by the fact that the language is fascinating. In the former, there is insistence upon itself as a great work of literature; in the latter, quietly a great book is written.

The story itself was something that I had anticipated. Or rather, the insistence upon itself gave away. It was pretty clear from the get-go how Ender Wiggins’ unusual talents would take him. Indeed, I found the story so monotonously plodding on, that I was irritated more than anything else by Card’s attempts to put a third (or really, even a second) dimension on Ender’s character. Unlike a great coming of age story (like, say, Harry Potter), or a great reviled character story (Frank Chalmers in Red Mars comes to mind), Card’s anti-hero plods on, with predictable and scheduled moments of self-doubt. And perhaps I’m too cynical, but I expected the ending. Rather I predicted one of two things would happen – the buggers turning out to be fake was one of the two alternatives (they’re very much real), and the actual ending was my other guess.

At the end of the day, I basically found myself thinking, “basic pulp sci-fi, not particularly interesting”. Unlike Isaac Asimov’s Foundation series, Ender’s Game does not have a compelling mythos. Unlike Arthur Clarke’s Rama Tetraology or Space Odyssey series, Ender’s Game does not cause genuine reflection into what it means to be human. Unlike Alastair Reynolds’ Revelation Space universe, Ender’s Game does not have a truly menacing threat. And finally, unlike Kim Stanley Robinson’s Mars Trilogy or Dan Simmons’ Hyperion Cantos, Ender’s Game does not have a really interesting inhabited world (or worlds). I’m unsure what to make of it. I’m glad I read it if for no reason other than to get references on the Internet, but if that is the most compelling reason to read it, then I would say skip it. I will at least attempt to read the next book in the series, but I strongly suspect that it will be another book that expects to be a great novel, and will fall flat attempting to do so.

Or to paraphrase Star Trek: First Contact, “don’t write a great book, just write a book – and let readers make their own judgement”.

One of the weirder things I’ve noticed after building out a proper network, and building a storage server, is that I start wondering “what more can I build?” For example, I’ve wondered about building another storage server (no, expand the existing one), a webserver (no, I already have one), an Asterisk server (no, I don’t need one) and a DNS server (NO!!)

For a brief while, I was also contemplating building a router, which fortunately did not pan out, otherwise, I’d likely be sitting here quite upset when connectivity dropped, for any reason. But beware the urge to build stuff: once you build one thing, the urge to build something – anything! – else is very, very, very strong.

Over the past few days, ever since a hard drive came magically back to life, I’ve been attempting to consolidate all the data I have into a more coherent organizational system. Some of these are causing quite a lot of think and rethink, others are simply on the lines of merging all the folders called “software” together. In the process, I’ve stumbled upon a weird Vista Explorer copying glitch, which just makes me even less happy about the state that Vista is in, but… that’s another story for another day.

First, robocopy simply does not work. It worked just fine under XP, but for reasons which are unclear to me, it fails while trying to copy anything to anything. Not sure why. Even a basic copy of one large file with no options just dies. So I’ve been forced to resort to using the GUI to do basic file system stuff, which, as you can imagine, is quite a challenge for a keyboard guy.

Which brings me to the glitch: Explorer “forgets” that it has copied the first x dozen files and directories, and attempts to recopy them at the end of the copy. So for example if I have FolderA, FolderB, FolderC all being copied into Folder1 from Folder2, Vista will copy FolderA, FolderB, FolderC, and then doubles back to copy FolderA again. Then, obviously, it brings up the confusing “Merge” dialog box, even though the copy happened perfectly the first time. The problem is even more clear when moving multiple items – theoretically, Vista is supposed to read, copy, delete in that order; in reality, for the first x dozen files and directories, it reads and copies, and then doesn’t delete… and comes back to try to merge the folders again.

I’ve no idea why it happens, but in case anyone has any idea why Vista tries to recopy or double backs on certain files and folders, please let me know how to fix it. It’s driving me crazy to initiate a copy and come back to find that after a night, it’s still got 18GB left, because it stopped copying to ask me about a merge.

Also, it’s slow as molasses doing any file system transactions compared to XP. But I think we all know that.

Should you have a home network and find yourself in the unenviable position of administrator, here are a couple of tips to make your life easier:

1. Make a map. I don’t mean use nmap and scan for ports and all that – though that can also be helpful. I mean, draw the thing out, either on a piece of paper using a writing instrument or on an application that allows you to draw. It really helps figure out what might be malfunctioning between two different points.
2. If building with the cheapest possible components from a (supposedly) reputable vendor, do not, at any cost, change any settings away from the default. I’ve just spent 90 minutes trying to figure out why my NAS decided to drop off the map for one computer, but not the other. And not just one computer, really – one share on one computer. Tracked it down, at long last, to a change made to the MTU – several weeks ago. Why changing the MTU should make a computer barf on one share, but not the other, I’m still uncertain. But my head hurts, my eyes are bleary, I’m going to call it a night – and don’t you touch that MTU setting.

Some months ago, I wrote about the untimely demise of two of my hard drives.

A quick search on the internet on how to recover hard drives that were clicking elicited this odd suggestion: freeze the drive. Literally.

I jokingly tossed senali, my 1.8″ 60GB IDE hard drive into the freezer.

And good lord – it started right back up!

I’m coming up on nearly four hours of use.

So the next time your drive clicks, try tossing it in the freezer. It might just work…

I must be getting old. Back in the day, I would install and try out scores, if not hundreds, of applications. My last Windows XP installation had, at the time I formatted it, just over 430 installed applications, not including service packs or updates. This computer? “Just” 130. And I’m cutting that list pretty quick, too!

In fact, I’m trying my damndest to reduce the amount of clutter on my computer and in my life. Part of the revelation was finding blogs like Lifehacker and Unclutterer. A second was a realization that if I ever had to move – particularly in a hurry – I would never be able to take everything, because I had no accounting of it. That was more than a little freaky a realization. So de-clutterization is the new mantra, and goal for this summer.

There are a few things I can’t throw or give away – like books – but the rest, I’m trying pretty hard. I have discovered that if I can’t bear the thought of throwing something out, if I stick it in a “potentially tossable” pile, and come back a few days or weeks later and find my emotional attachment has abated and I can toss the stuff.

Oddly, I’ve never been this successful before in clearing up. Inevitably, I would “clean up” by basically taking one pile of stuff, and putting it in another, supposedly more logical, pile. Last time I was home, I looked through the stuff in the drawers of my desk and couldn’t figure out for the life of me what I’d kept random sheets of paper for. Time constraints limited my ability to do such a mass clean-up there, unfortunately, but I think I’d be able to do a lot of the clearing up now.

One useful trick if you’re considering engaging in some en masse cleaning up: pick one spot of the room or data store at a time. For years, I never got around to sorting my bookshelf, until one day, I started with one shelf. And the next day, another shelf. It might have been slow, but by the end of the week, my entire bookshelf was more orderly than it has been in decades. Or I picked the letter M and cleared up all the songs in my iTunes library where the artist’s name began with the letter M – proper MP3 tags, high-quality CD art, sorted into play lists. Or, I decided to clean up the Cthulhu‘s child of cables behind my computer by choosing a certain type of cable – all power cables and adapters on one day, all data connections on the next. Breaking the task up into manageable segments has really helped in getting stuff done. I know this may be a “duh” moment for a lot of you, but it’s been a recent revelation for me outside of work or school, and has allowed me to get a lot more done with a lot fewer hours – though I still pine for a 28 hour day.

Happy de-cluttering, though.

I really wanted to like Vernor Vinge’s Hugo-winning 2006 novel, Rainbows End.

And sadly, I knew from about the first 50 pages that I would be beginning my review of the novel with those words.

Rainbows End came out about the time I was busy graduating from college, so there wasn’t much time to any of the reading then. I was intrigued by it, because I had just read Vinge’s work on the coming technological singularity, and had been reading Charlie Stross‘ Accelerando on and off as it came out in parts. Anyway, some two years after I put it on my must read list, I finally got around to it. Kind of shows you how long this list is, eh?

After reading and loving Alastair Reynolds‘ Revelation Space universe, I felt that I was ready for another “serious scientist” turned author. Never having read A Fire Upon the Deep or A Deepness in the Sky, I was unable to compare the writing style. I’m told however that Rainbows End is probably the most readable of Vinge’s novel, which is slightly scary, since I found it a pretty difficult novel to read, and I didn’t particularly end up liking it.

What I mean by difficult is perhaps slightly different from what is normally meant by that word. For me, difficult speaks to the ease with which I could get into the story line. I found Dan Simmons‘ Hyperion similarly difficult to get into because of the vocabulary. But unlike Simmons, who was pretty much done explaining terms by the end of the first chapter, Vinge struggles to push out ever more esoteric terms nearly to the very end. So instead of focusing on the story as I should have by about 300 pages in, I spent most of my time trying to play guess the word. Compared even to a peer like Reynolds, who is willing to stop the jargon spew by the first third of the novel, Vinge seems determined to make it nigh impossible to concentrate on the story by introducing words right up to the end.

What I will say is this: Vinge is a visionary. Wikipedia says that an initial version of what became Rainbows End was published all the way back in 2002; in 2002, I would have scarcely believed that what he refers to as the “Secure Hardware Environment” was possible – or likely. However, given the recent maneuverings around ACTA, the broad authority courts in across the world have given executive branches to breach into the sanctum of a computer, as well as continued action on the part of the assholes promoting DRM, I’m beginning to suspect that it will be only a matter of time before we do see something like SHE implemented as a combination of national security, fear of terrorism and Internet security. I expect that before the next decade is out, it will be damn near impossible to buy hardware that is not bugged at the hardware level, either purposely in order to comply with government directives, or by governments where chip assembly takes place.

Second, his ideas for overlaying the real world with augmented bits and pieces is an idea as old as sci-fi itself, but his version is more believable than most. The implementation is pretty much unique in my experience – via contacts or electronic paper, rather than the usual neural implants that dominate other novel story lines. Indeed, we’re beginning to see this sort of augmented reality come into being in the last two years or so – but currently, the ability to do so is in the realm of heavy, powerful, dedicated machines like the Xbox 360 and the PS3. Ultimately, Moore’s Law will prevail, and we will indeed see chips as powerful as the tri-core PowerPC in the Xbox or the Cell in the PS3 in tiny, watch-phone sized devices. (From what I’ve been able to tell, the novel takes place in 2025, which is approximately speaking nine Moore Cycles away, meaning that the 221 sq. mm Cell chip should be doable in about 0.5 sq mm, so not a bad guess on Vinge’s part for a watch-sized augmented reality chip.) I can’t say I pretend to understand how the contacts would work power-wise, but I suppose that’s why I’m an economist and not an engineer.

But for all of this really well-thought out future, Vinge falls short when it comes to the actual implementation of the world. I struggled to understand “belief circles” and “arXiv references”, not because I was unfamiliar with the terms themselves, but because I think he used the terms incorrectly. This is easier to illustrate with an example. The formal name for the on-screen arrow controlled by the mouse is a “cursor“, or “mouse pointer”. If the pointer is not appearing on screen, technical people might say “the pointer is not showing up on screen”. People unfamiliar with the formal terms, however, would likely say “I can’t see the mouse” or something pretty similar to that. I, like most technical people, would likely be baffled initially by this statement – particularly if we’re in the same room and can see the mouse sitting beside the keyboard. Some amount of parsing later, I’d likely conclude that what the person is saying is that the pointer is not on screen, and work from there. As far as Rainbows End goes, it’s probably worse because Vinge is using a term that has a specific meaning today, and I’m attempting to understand a sentence in the context of the current meaning. It would probably – in fact, definitely – be easier if he just invented a new term and vocabulary. Instead, I’m left with this hybrid where I have to think “wait – is this a future meaning of the term ‘foo’ or is he using ‘foo’ in the same context as we are today?” And if anyone could tell me what a Scooch-a-mout or Dangerous Knowledge is, I’d be eternally grateful.

That’s not to say it’s not an entertaining story. It’s the coming of age story of a very, very old man – a man who has lost his marbles once, and been given them back. So there’s definitely a healthy and energetic plot underlying the story. There’s also a remarkable clarity of vision (even though, as I’ve mentioned, it is a pretty dystopian future) and enough unsolved mysteries and questions at the end of the novel that you’ll beg for a sequel to find out what happened; fortunately, he’s writing one.

I just kept jarring back into reality from the story. That made the story difficult to understand, and difficult to read. More than once, I considered giving up and coming back when I was more “ready” for it. I’m glad I persevered, but it took a lot longer and was a lot more exhausting than I had hoped for.

Verdict: borrow. Required reading to get half the tech jokes on the Internet, but you’re not going to read it more than once, unless it’s reading for comprehension’s sake.

Wow. I thought I’d seen it all.

I’ve not bought or helped set up a new Windows computer in a very long time. So I’m slightly behind the times on finding malware and crapware (and quite frankly, crapware is malware by another name, so I’m simply going to refer to it all as malware) that HP and other computer manufacturers try to “offer”. Here’s where I’ve seen malware hiding to ensure that it starts up:

• The Windows registry (this is a pretty obvious place).
• The Windows startup folder (also depressingly obvious).
• As a service (less obvious, but since Microsoft and legitimate services, like anti-virus programs provide a description, it’s pretty easy to spot).
• The Win.ini and System.ini files – from way back when (less obvious, but an old timer like me would know about it).
• As a login or startup script (quite clever, since it’s never clear what Windows is doing while it is “preparing your desktop”).

However, today I was introduced to a new place to store malware, and I must say, I’m both impressed with the ingenuity of it, and angry that it was done this way. Mainly the latter, some of the former.

HP decided to include “Internet Offers”. Of course, any sensibly designed program would notice that the computer was already connected to the internet and bugger off. HP, being HP, has no quality control, and is probably paid a commission for installing malware and for every signup that the malware brought in to the service provider.

So they decided to make it impossible to find. How do you do that on Vista? Well, pretty simple. You find the most difficult to understand, most worthless piece of crap program that comes with Vista, and put it there.

Yes, the Task Scheduler.

So after I spent hours scouring through the bowels of the registry, the services list and miscellaneous startup bits, I found myself trying to comprehend the New! Improved! Impossible to Understand! Task Scheduler. And sure enough, buried under layers of folders called “Microsoft”, I found… “HP Internet Offers”!

For that reason alone, I’ve ruled out buying a HP 2133 and changed my tune on buying HP. They may make decent computers – sorry, they used to make decent computers – but thanks to the Carla “screw our customers” Fiorina mentality that now pervades every HP department, I can’t recommend an HP to anyone in good conscience.

Screw you HP.

(And now, to install Ubuntu via Wubi…)

According to Download Squad, Trillian is eight years old today. Of which, approximately, four years have been spent on waiting for Trillian 4 to come out.

I kid.

Mostly.

Back in the day, I was an ardent Trillian fan. I remember using it on my parents’ old, slow Toshiba laptop, because it just worked and worked pretty fast, considering how far behind the curve that laptop was. I tried a couple of other programs too – ranging from GAIM (slow beyond belief, and ugly to boot) to the native clients (ads – talking ads!!!) and settled on Trillian. I don’t even remember how I first found it, but in the days of a 120 MHz Pentium with first 16MB and then 48MB of RAM, it was a godsend.

When I got my first personal laptop in 2002, I was eager to install Trillian – until I discovered to my chagrin that features which were earlier free were now pay. I stuck with the free version for a year, until Trillian Pro 2.0 came out in 2003, at which point, I paid up for a licence. For the next two years, I was an ardent Trillian fan – going so far as to buy friends Trillian Pro licences for their birthday gifts, because I believed in Trillian so much. When 2004 came around, I renewed my licence for Trillian 3.0 and bought another round for my friends.

And then… Trillian died.

In retrospect, I should have seen the warning signs. The developers would just disappear as a matter of routine. People would report major connectivity issues, and a reply, if any would be in the form of “we’re working on it, don’t worry” and two-three months later, an update to fix a connectivity issue. If your next thought is, “I bet it broke a week later again”, you are psychic. The grumblings on the forums increased. People publicly started decrying Cerulean Studios as frauds, and around that time, I left.

Two things happened.

First, Google happened. In February 2005, my school took its entire mail server system down for the second time in as many years. I should add there was precisely a one week warning (though that was better than the previous crash and burn, which was announced ex post facto) and it was during mid-term exams. Really, really smart. Having given up on a reliable communication there, I and a number of people encouraged people to move to Gmail, and route all emails there. There was no IM at the time, but using Google set me up for the second thing, which was that…

Second, in the middle of my winter exams in March 2005, my Toshiba laptop mysteriously died. It powered down fine, but when I came back the next day after a two hour nap, it refused to start. At all. An emergency trip to buy a computer resulted in my installing only what I needed to write my term papers and study for my exams; everything else was left till the summer of 2005. I got around to installing Trillian Pro again, but it never worked with any reliability. So in August 2005, just before I went back to school again, I was delighted to find Google Talk – a lean, clean client that not only just worked, it worked elegantly with the people I talked with most. Victory! So through my final year at school, Trillian, now the relative memory hog, stayed off.

A year or so after I left, I stumbled upon the new version of Trillian, codenamed Astra. Hoping that this would be the rebirth, I was in the first few batches of alpha testers. Astra is currently on Alpha build 81. I first used Astra, build 25 or so in February 2007. And it seemed like they had been working on it as far back as 2005. To put it in context, it’s taken three people working full-time three years to write an IM program; it took 2000 Microsoft programmers to build Windows Vista in that much time. Vista alpha testing took about a year; Astra is well into its second year of alpha testing. With luck, by 2010, we might see a beta.

In the meantime, I stopped using it. Not only did GAIM catch up in features, it bettered Trillian in resource usage. The last Trillian build I used ate nearly 45MB – without any accounts enabled. GAIM uses 31MB with nearly half a dozen accounts. Trillian never drops below 10% processor usage; GAIM rarely exceeds 10%. GAIM was cross-platform by design so I could use exactly the same way under Windows, Mac OS, and Linux, and its open nature led to some of the most interesting offshoots I’ve ever seen – Digsby. It was a direction others and I encouraged as far back as 2004 on Trillian’s forums. Pure IM is nearly dead – it’s almost all about social networks and building communication capability into social networks: witness MySpaceIM and Facebook Chat, in addition to the private messages both networks offered from the get-go. Digsby does this with aplomb, all while keeping IM as a key focus. It’s amazing. On the web, a client like Meebo does everything you could want an IM client to do – and it’s universally accessible from anything with a network connection, a screen and a JavaScript interpreter. I think belated recognition of these different modes of using IM forced a mad scramble at Cerulean Studios to build an online version as well.

So, looking back at the few happy years of Trillian usage, I am glad that there was a decent IM client when I needed it. Graduation also killed off IM usage – there’s only so many hours in a day one is home to use it, and collaboration rarely happens at the office over IM.

It’s time to say bye to Trillian. I know Astra may eventually be released, and there may even be a successor. But Trillian seems to be an idea whose time came and went. Unless there’s a drastic reinvention – not one of Cerulean’s trigger-happy “complete rewrites” – Trillian will be one of those amazing products that will be scarcely merit a footnote in the early history of IM.

Good luck Trillian and Cerulean Studios, but also, good bye.