Dan Makovec’s blog

Skype goes gold

Oooer, after Skype released v1.0 (actually 1.0.0.10) the day before yesterday, they generously gave me another $5 USD in my account. I say generously because I had only put in $10 in the first place, and only used $2. So now I’ve got more than I started with.

Apparently it was in thanks for using the beta of SkypeOut. That’s great PR, well done guys!

Viri and blogs

Very curious,

I just discovered that unique hits to my (empty) website have tripled in the past month compared to all previous months. There hasn’t been a gradual buildup, just an unprecedented overnight spike. Maybe I should put something on there
Ah now I see… The huge spike has been for Bron and Adrian’s RSS feed of their trip. Curiously, I noticed that a whole bunch of spam comments got posted in there.

I think what’s happening is the following: some worm/virus/whateverucallit scans text files on infected machines for URLs. Two such files would be the IE/Mozilla Bookmarks file and the newsfeeds file for whatever news reader the user runs. Somebody with a feed watching Bron and Adrian’s Europefest blog got infected, it found the RDF URL, then sucked it down and parsed it for entries.

The RDF also lets the virus know that the blog is run on MovableType, and this information was used to publish comments advertising “reality porn” etc to the blogs. Guys - if you received a whole bunch of porn spam disguised as blog comments yesterday you’ll know why now I’ve turned commenting off so it can’t happen any more.

The curious thing about it is that my test blog got hit as well. This means one of two things, either:

a) The virus is smart enough to know how to find other blogs hosted on the same machine (I don’t even know how to do that manually (my test blog is not referenced anywhere)

or

b) The person whose machine got hit by the virus had also subscribed to my test blog. I’d only passed the address out to about half a dozen people, so this narrows it down a bit. Guys, run a *new* virus checker pronto just in case!

Paris photos

It’s hard to believe it’s been two weeks already. It’s just as hard to believe that I haven’t blogged anything about it!

Still, Brian and I picked up our discounted Eurostar tickets and made for Paris over the weekend. Despite the fact that it was the last travelling stop on Worldfest for me and I’d already spent a week there, I found plenty to do as a tourist second time round and decided that this would definitely be a cool town to live in.

This is actually a great time of year to head to Paris. The weather is warm (although it did rain a bit) and the city is really under-crowded, since as a particularly friendly Parisian lady told us, all the Parisians head to the coast for their Summer holidays.

About the only downer is the horde of rich spoiled-rotten white teenagers from New York and California who invade the city as their parents offload them by sending them away on Summer vacation and spend all their time whining about the fact that nobody understands English. Oh btw, here’s a tip kids - ALL Parisians under the age of 50 speak English perfectly well. It’s just that they figure that if you can’t be bothered putting in the fraction of effort required to learn a few basic French phrases slightly more complex than “parlez vous Anglais” while visiting France, then they can’t be bothered putting in the effort to talk to you. Fair enough in my opinion. Think about it, you’re exactly the same when a foreigner visits your town.

Still, not even their whining was enough to spoil a magical weekend spent in the most romantic and certainly one of the most beautiful cities in the world. Loads of time was spent climbing Tours d’Eiffel and Arcs de Triomphe. We took in the sights of the Grand Arche and Place De La Defence (can’t *believe* we missed that back in ‘02!) and the churches of Basilique du Sacre Coeur and Notre Dame, using a metro system that was quick, quiet and worked, eating the kind of food that superlatives don’t do justice to and taking in the sights and sounds of what nightlife we could before collapsing in bed completely exhausted.

Anyway, we had a quiet weekend just gone, so Brian spent the time putting up a few (ermm.. does 100 count as a few? I only just realised there are that many. He doesn’t do things by halves) photos on Webshots of our adventures in Paris.

More space tracking

This site, provided by NASA, provides more realtime info on the space station and all the other satellites orbiting the earth. There’s even a cute 3D display that lets you spin the world around or speed up time to see what’s going to be flying above your head in the next few minutes. I had *no idea* there were so many machines flying around up there!

Monday tidbits

Did you know that the International Space Station lost 30km in orbital height over the past 12 months?

The German Site Heavens Above lets you keep track of it’s precise location at the moment, along with maps of the skies from your current location, where the next Iridium satelite will fall out of the sky, and other such interesting spacey information

Question of the day

Your Sunday morning TV viewing is made complete by the fact that LivingTV is screening The Golden Girls, Charmed and a Barbera Streisand movie. Does this make you gay?

Apple releases Rendezvous for Windows and Java

This could be interesting - Rendezvous on the PC. Prerelease only at this stage. Also available for Linux/BSD/Unix.

Lack of motivation

I’ve just learned that our much-hyped phase 2 has been delayed until SEPTEMBER, thereby stripping me of the remainder of the motivation I had to do any work whatsoever.

In the meantime, I’m continuing my exploration of new things net, the latest being Bit Torrent. Judging by the amount of sites out there, Bit Torrent has been around for some time, although I’ve never played with it. If you’re like me and don’t know anything about it, here’s a brief synopsis:

- It’s a peer to peer file-sharing protocol, a la Kazaa and Bearshare

- While there is one “official” client, it was released as open source from the start, so myriads of other versions have been developed by others (thereby removing the risk of lawsuits to the original author once its use reaches its logical extension

- The protocol works on the basis of one central server which acts as a simple addressbook of which clients are online. A client is only online if it’s uploading or downloading - it doesn’t just sit idle and doesn’t share files that it’s not in the process of actively transferring. This server can be set up anywhere, so in actual fact there are probably hundreds of these servers around the world which don’t know about each other.

- When a user wants to share a file, he runs a utility which creates a metafile (filename.torrent - about 32kb in size). The existance of this metafile is then announced via IRC, a newsgroup, email or a web page.

- Other users who want to retrieve the file download the metafile by clicking the web page link, downloading via IRC or opening their email attachment. It’ll fire up the BT client, which normally looks like nothing more than a download window.

- The metafile appears to contain simple information about the file: filename and checksum, server to use as a directory (tracker). When users open the file in their BT client, the client contacts the nominated directory server and polls every user listed by the server, asking “hey, do you have any part of this file?”

- All the users who are currently sucking down the file report back “yep, I’ve got this part of the file” and start transferring it.

- Consequently, the file is delivered on each user’s machine in a fragmented format that is gradually assembled as more parts are downloaded.

- As each user downloads a part of a file, they can upload part of the file they’ve already downloaded. That way the more users that are downloading at any given time, the more opportunities there are to grab the file. Quite scalable because in theory at least, more people downloading means faster transfers. It also means file downloads can be stopped and resumed at any point in the future. Another corollary of this is that the larger the file, the more people are likely to be transferring it at a given time. Hence you can imagine the type of files that are best suited to this sort of application (movies, CD/DVD ISOs, etc)

All in all a rather interesting protocol. Naturally the pirates have taken to it with much gusto (I’ve already seen DVD ISOs for Garfield the Movie and other upcoming features available - though I can’t be bothered downloading them since they’re probably filmed on a dodgy handycam in a cinema with crap sound somewhere west of BF Idaho), although some companies are making legitimate use of it (Linspire.com uses it to distribute ISOs of their Linux product for example).

Anyway, more info on the protocol is available here.

GUI separation from code

I was chatting with Vidar today about application development, and in general, how crap Java Swing GUIs generally are at handling user input and responding in a timely fashion.

He frustratedly exclaimed (as he does on numerous occasions about topics like this) that developers spend too much time hacking around with repetitive GUI code when they could just get on with writing their applications.

I thought about this for a bit and thought of how separating presentation code from application logic makes sense. Take Mozilla for example. Once they perfected the XUL toolkit, applications like Firefox and Thunderbird came along in leaps and bounds as the GUIs just *worked* and silly little rendering/layout/event handling bugs disappeared. I thought how cool it would be if we could do something like this with Java.

I hate writing Swing. It looks pretty and gives you a cushy API for getting widgets laid out, but writing code to handle mnemonics, localization, event handling and layout is tedious and often distracts me from getting to the guts of what it is I’m trying to do.

When I’m working with it I often just give up and don’t achieve anything in a programming session because I’ve screwed around with widget layout, compiled and executed over and over. Maybe this is why I hardly ever complete Java GUI app work!

So I consulted the holy Internet Oracle and came up with SwiXML, which seems to do exactly what I want. I whip up a quick XML descriptor of the UI which defines the layout and add hooks for the events to be triggered, then go back to my beloved Java and write the program itself.

As the UI is defined outside of the code, its definition is parsed at runtime. Whipping through an XML DOM is just as fast as running through dozens of lines of hand coded panel.add()’s. Also as a side benefit I can dynamically alter the UI while executing the program to run tests. I was really impressed by this Demo page where I could quickly muck around with GUIs on the fly.

This is actually something I could get into. Must… start… coding… again…

Chatter chatter blog blog

Graz asked the other day what people are using for reading blogs. I don’t know whether the whole RSS bubble has burst already - the amount of blog reading I do has really dropped off so I’m not much of a gauge - but it does appear that RSS readers really haven’t progressed much in the past year.

On the Mac, everybody still loves Net Newswire. On the PC, the greatest achievement known to blog-kind still seems to be Newsgator, which lets you browse RSS feeds in Outlook like mail folders *yawn*. Mozilla’s got the Sage project underway to give the same capability to Firefox et al, but there’s nothing very inspiring out there.

I hear Awasu is still the creme de la creme of standalone browsers, but in all seriousness I dont’ see why it’s functionality can’t reside in a simple Mozilla sidebar.

For the most part I read my feeds on my BSD box at work - which means I either use KNewsTicker (which is so limited in what it does that I only have about half a dozen subscriptions on it) or Bloglines.

The thing that most (all) of these readers lack is the ability to define views for feeds. Let’s say I subscribe to Slashdot, CPAN, the Register, SecurityFocus and a couple of news sites. With all the above software, I’ll get some boring old tree view listing the feeds I’m subscribed to, writing their name in bold if they contain unread items, and if I’m lucky a little number in brackets telling how many unread items there are. Yay.

Now, in my current(* - see my post on June 28th) I am a MySQL DBA, Perl, PHP and C++ coder, online payment processing engineer and unofficial configuration management specialist (they haven’t quite got the hang of CM here). For RSS feeds to be any more useful than a simple toy, I’d like the ability to define my treeview as follows:

- any articles from any source to do with credit cards on the internet go into a view called “Credit card security”, but can also be viewed in the folder relating to the site that supplied it

- SecurityFocus OR Mysql.com releases news on a vulnerability in MySQL; Apache has a buffer overload problem: I want it going into a “Things to patch” view. If on the other hand Apache or MySQL announce a general upgrade, it goes into my “Upgrades to consider” view for when I’ve got time on my hands

- Any articles from any source mention my home town or country and IT jobs: put it in my “Potential work opportunities” view

Opera do this for mail (don’t know how well it works with IMAP tho), and apparently Google do too. I’d love to see it actually being extended to RSS tho. Hmm - a possible project for getting me back into Java?