SpazBlog

I almost forgot. Back in June or July, I went to ATI-Revolution. It was a big (~600 ppl) LAN party on the south shore. Fifth biggest ever in North America, or so I understand.

Anyhow, while I was there I interviewed one of the organizers who has been big in the LAN scene in Montreal. He was doing sponsorships for ATIR. I transcribed the thing ages ago, but recently I finally finished editing it and went over it with him. So now I'm able to let other people read it.

It's a bit long, but I think there is a lot of interesting stuff in there about what it takes to put on such a huge LAN party:

http://teknews.net/roch.php

I am going to try to spread it around to some news sites like anandtech and tom's hardware (Tom's did a writeup of the LAN), but I'm probably going to do some security updates and performance tests before I do that.

When Google Analytics was first announced, I set it up on TekNews and promptly forgot about it.

Well, I logged into my account today to take a look. The results are pretty surprising. Since its launch, TekNews has had pretty stable traffic of about 800 to 1200 pageviews per day. It's been like that for over a year. I didn't really know who was viewing the site, or where from.

Turns out that most users are in the US, followed closely by Germany. And the city with the most users by far is Flemington, Australia. Second place goes to Madrid (Spain). It looks like overall most users are in Europe. I also seem to have all continents covered except Africa. And obviously Antarctica.

I'm still somewhat surprised that the number of users of TekNews hasn't really changed. I haven't really put any development into it for many months, save a quick new feature allowing independent selection of RSS refresh times when Slashdot put a 15 minute limit on their RSS feed. Despite this, people keep using it. I guess that is because the site runs itself; all of its RSS sources are still valid, and it just keeps pulling them and updating itself.

There isn't much that could use changing/fixing anyhow (Though of course there is always extra stuff you could add to a site). About the only thing that comes to mind is better detection of duplicate articles. Currently it assumes an article is the same if it has the same title, OR the same blurb, but not both. This is imperfect since sometimes they update both the title and blurb of an article and TekNews picks up a default. They're not that frequent though, only happening every few days, so I haven't really bothered.

Speaking of updating itself, TekNews has managed to gather 1,628 articles from its RSS sources, along with 21,172 news posts. All searchable, which was part of the original intention.

This morning started out pretty crummy. On tuesday/thursday, I normally take the 8:38AM downtown. This morning I missed it and had to take the one about 20 minutes later. I ran into Claudiu at the train station, who apparently normally takes the later train.

We were going over some system hardware stuff on the train. Bit shifting, binary operations, the like. In my sleep deprived state I stupidly confused AND and OR and tried to explain AND to Claudiu as outputting a one when either of the inputs are 1.

The guy sitting accross from us looked up from his magazine (WIRED, I think), and mentioned that I was describing OR. I was somewhat surprised that a stranger on the train would know anything about binary operations, but it turned out he was a senior programmer at GameLoft, Ubisoft's cell phone game division. We had quite an interesting, albeit short discussion about the use of assembly in modern programming, as well as cell phone development platforms.

He actually mentioned that Jamdat (their direct competitor) had written translators to translate code between J2ME, Symbian, and BREW. They write code in either Java or C++ for any of the three platforms, and their framework handles translation to the other two platforms, including a language port if required (C++ --> Java or Java --> C++). I'm surprised they are able to pull that off, but I guess they do.

He also mentioned something else I've read before, which is that J2ME's primary "advantage", write-once-run-everywhere, is totally nonexistant. Developers have to maintain seperate versions of their J2ME apps for every single cellphone, even from the same company. Oh, and J2ME isn't JIT compiled (with very few exceptions), so it's dog slow compared to BREW and Symbian apps too. But we already knew that ;)

Of course, the more practical information from the discussion was that Ubisoft and GameLoft hire a lot of people from Concordia, and take a lot of Concordia co-op students who are in their last year.

When does the current waterloo co-op semester end? It's been a few months since I last visited NITI, so I think I'd like to swing by some time before Steve and Adrian leave.

Normally I don't block ads on web pages. The things pay for the sites I enjoy. I've relied on them to cover costs myself, and still get the occasional check from Google. But when a site puts up so many flash advertisements that are so processor intensive that Firefox locks up whenever I visit the page, well, the sympathy very quickly dissipates and AdBlock gets turned on.

Seriously, people. Google has proven that you don't need rich media ads to make a heck of a lot of money with advertisements. Do you really need sixteen animation-heavy flash ads on one page? It's called sign up for AdSense. You'll probably make more money, and your users will be a heck of a lot less annoyed.

Turns out it was an entrance scholership for $2000. It was a very nice surprise, though I'm surprised they don't announce them before the semester starts as an incentive to attend the school.

Apparently I've won some kind of award from Concordia. I got an email from the financial aid & awards office informing me that I should swing by to pick up the letter with the details tommorow. Some sort of mass-email that went out to all recipients of all awards.

Since I have done absolutely nothing noteworthy to date while at Concordia, I can only conclude that this is some sort of entrance scholarship. The odd thing is I would have imagined they'd tell you about those BEFORE you started school, in order to entice you to go there.

Oh well, whatever it is, it will be nice. That, along with the $3000 the Quebec government is going to give me for finishing a professional CEGEP program on time will help pay for university tuition, and I suppose my eventual laptop replacement, to happen in one or more years.

Was reading an interesting article on flash-based HDDs. It seems they're starting to gain some traction in the high-end enterprise market. Proponents are citing durability (No moving parts, obviously), speed (Higher speed flash is already way faster than HDDs by several times, and even lower end flash has almost nil seek times), and reliability.

It's that last one that interests me. Flash has a limited rewrite count before the chips fail. With HDDs, it's not so well defined. It seems that flash chips are up to about 1 million writes before they fail, which leads to four or five years in servers. Obviously it's way higher in consumer products due to lower disk activity. Still, it's not quite high enough. They're claiming that four or five years is already more reliable than most magnetic HDDs, but I'd give it a few years before I'd feel comfortable with that claim.

Another claim they're making is that flash capacity is going up about 35% per year, and prices down by 35% per year. Looks like moving entirely to flash for storage in a general purpose PC isn't so far off after all. The idea of hybrid drives is already floating around (Using flash as an enormous multi-gig cache on magnetic drives), and that's a bit more likely to happen before full-on flash drives become popular.

There are alreay some flash-based IDE hard drives (In 2.5" notebook drive form factor as a drop-in replacement for a magnetic drive, I'm not talking about a DOM here) on the market, but I don't think they come bigger than 16 gigs or so. They cost something like $100 to $150 per gig anyhow.

Still, I drool about the day when flash drives take over. Flash has significantly higher transfer rates than magnetic drives (with better future potential too), comparatively non-existant seek times, and way lower power consumption. The seek times mean that disk fragmentation becomes effectively meaningless, because there is no penalty for grabbing another fragment from elsewhere on the disk.

I imagine the disk defrag software vendors aren't too happy about that, though :P

The original article can be read here: http://www.tgdaily.com/2005/11/04/ssd_harddrives/