JulioBiason.Net

When I started using Linux, around the year 2000, you could use a very simple window manager or you could use GNOME. GNOME, at the time, had this very very cute window manager called Enlightenment, which was also a royal pain in the ass to use. It required certain settings on your X to be enabled or you couldn’t use some features (like key-bindings.)

The thing about Enlightenment is that it was really nice on the eyes. Its themes were extremely good looking and there were lots to choose from. I could say that Enlightenment was the Vista of that time.

One of the things that actually did happen a few years later was that GNOME changed its default window manager from Enlightenment to Sawfish. Sawfish, although not that good looking, was way more configurable and, apparently, the author was willing to make it more integrated to GNOME than Enlightenment. No biggy, GNOME changed window manager, but Enlightenment had its own fan base, so they took separated ways.

More years later, the Enlightenment team announced the start of release 17, also called E17. They plans were big: they would use a lot of new libraries and it would be fast and you would get even more eye candy, with shadows and real transparency and real time updates on icons and such (almost what you get today using Compiz.)

The biggest problem with E17 is that it didn’t survived its own promises. Every step forward in development was followed by two steps back. Features added were moved to yet another library and everything had, apparently again, to be refactored again. All libraries were being constantly hacked and never had any releases.

It was almost a year after the E17 announcement when the XFree team announced some new features they were planing, which would allow any window manager to use features like real transparency and real drop shadows and almost everything that the E team promised. At this point, anyone would think “So the E team joined the big dogs and help them to develop something for the community.” Well, wrong. They decided to keep their plans and don’t look around.

Even more time later, XFree announced XDamage and XRender, two features that paved the way to the current compositor-enabled window managers. Even weirder, there were two projects that managed to do that, one lead by RedHat and another by SUSE. Problems? No, they decided to talk and found a way to merge their projects into a single entity, not fragmenting the community and giving a fair change to everyone.

So, since the E17 announcement, we had a major release of X, now forked into X.org, opening it to the great community (XFree was not that open with other people suggestions), several libraries and almost every single desktop is using these new features and getting some very nice eye candy every day.

And what did happen with E17?

Last week I downloaded gOS, a light-weight distribution which uses E17 as window manager and desktop environment, using the sources from the repository, as there are no formal release yet.

What I saw was a alpha release of… something. The keybindings still don’t work properly, the themes are more proof-of-concept than usable (too much animation and very few helpful things), their widget set sometimes decides to ignore the current theme and falls back to the default one and, still, it feels like the window manager is always trying to get in your way and annoy you in the worst possible way and never ever help you.

More than 7 years and still no stable release. In that time, every single desktop environment managed to slim down and get more eye candy and be more user-friendly.

I don’t know about you, but I think E failed.

• Thanks to Parallels and X11 (Xquatz), now I really *hate* OS X. #
• Oh, nice, pissing-like weather. Actually reminds me of buying an umbrella… #
• Brilliant analisys of GlaDOS: http://www.game-ism.com/?p=91 #
• Listening to "Coming Back to Life" again (a song that some people called "emo" *cough*@dsturnbull*cough*) #
• Hm… It looks like the delete button on Mitter is not working… #
• My playlist: "Coming Back to Life", followed by Tenacious D "Beelzeboss". #
• Doesn’t beat "I Hope You Die", Bloodhound Gang followed by "Don’t Worry, Be Happy" (my playlist at UNI some long time ago.) #
• "Following: 25,000+ people". Are you stupid or what? #
• Ok, how about "Crazy" by Alanis Morissette followed by "Eric the Half a Bee" by Monty Python? That makes sense to me… #
• I like that my OS X is all weird and now my Windows is showing signs of brokenness too… NOT! #
• Hm… svn commit to Twitter. That would be really cool. #
• Why the fuck iTunes simply pause the song with no apparent reason? #
• A little bit more of white hair in the current places and I’ll look like Mr. Fantastic. #
• Is it me or the fonts on Ubuntu look *better* than on OS X? #
• Loading pictures in the backgorund was easier than I thought. Just need to fix some glitches and the major issue for Mitter 0.2 is closed. #
• I was planing to fix the image download class in Mitter (write it properly), but I’m too tired to do that right now. #
• @pilgerowski Wow! Mais um Social Network, Web 2.0 thingy! Exatamente o que a gente ta’ precisando! #
• @pilgerowski "Poutz! esqueci a senha…" Ha, e e’ dos profissional ainda! #
• "Banco de Jobs" Porque a gente somos muito retardados pra escrever em portugues, mas a gente queremos parecer gente fina. #

Powered by Twitter Tools.

First of all, yes, I use Digg and Reddit. For those who don’t know, these are two of the most well-known “Web 2.0″ sites. People send links about some news (or something interesting) and other people vote for such links. The most voted links go to the main page, from time to time. It is interesting to see where the collective mind goes, even when you find yourself in one of the retarded corners of the internet.

The biggest problem with them is that there is no way to point that one history is the same as another posted already. While I’m not sure about it, I hope the both sites would consider that, if someone post a link that was already posted, it would count as a vote for the first story and not create a new thing so people have to vote for it again.

Even with link dup checking, there are still some problems: Imagine that I find something interesting in the web, so I post some very small description on this website with a link to the original story and post my blog link on Digg/Reddit. Then I just sit down and wait for people to come to my site and I get a lot of money from Google Ads (no, I don’t have Google Ads on this site.) This is called “link-hijack” by the Reddit community and, more than once, I saw links with “Non-Link-Hijacked” in the title, which means someone decided to pick the original link and post it instead of someone else blog.

There are some link-hijacks that are a little bit more complicated to catch. First, let’s say someone find an interesting image. You can post it on tinyimg, imageshack or even in your hosted Gallery, with all those lovely Google Ads. There is no link for the original story and, most of the time, since the image appears on several different places at the same time, you lose the track of the original. Also, there are some stories which are, actually, part of the same big event. One of the examples that comes to my mind is the “This is cool”, which appeared on Reddit (sorry, but it is quite hard to find anything on Reddit after more than one week.) It was a photo of Barack Obama pointing to something. Someone photoshopped it, putting some sunglasses and added the link with the title “More cooler”. Then it started: People added an explosion on the background (you can see it here, which was posted under the story “Cooler”), and people put a nanchuck on Barack’s hand (see it here) and posted under the title “Coolerer”.) So, the joke spawned over several different links, and over several different links. There is no way an engine would recognized them as the same thing (and, most importantly, are they the same thing?)

That brings a question: what it is more interesting, something that people say “this is interesting” or something that gets a lot of attention on the web (like the original link-hijack)? Personally, I think that the even behind such stories and links is the main factor. Posting a link which explains climate change is destroying the environment and another link where scientists make pretty graphs showing that there is no relationship between global warming and the decline of pirates are, in fact, different stories, but they are linked by the same event. And that’s the problem with those aggregators: they care about links, not events.

Now, to be completely honest, I don’t think anyone would come an easy solution for that. It is easier to track links than events. And how would you check if link X is really related to link Y? Again, you have to trust that the community would take care of showing that X and Y are linked (or not) by some mechanism (tags? direct dragging links to say that they are related?) The first think that comes to my mind is something like “Human Brain Cloud” does to create the relationship between two words: the more the people link those two, that relationship becomes stronger and all the other ones, weaker. The problem is: would you really expect that people would sit down and say that link X is related to link Y? Over and over again? Instead of just clicking an arrow that points up or down? No, I don’t think so. You’ll have to search the current links, see of there is anything related and create the links.

But, in the end, I can see that cool things would emerge. Like you could be seeing some news report about google, which points to another news about how the energy usage is going up in the world, which is related to another story about Finland hoping that big datacenters move there where it is cooler (so no need of air-conditioning) and energy is plenty. Too bad we can’t expect that people would actually sit down and relate stories.

Google car, originally uploaded by Julio Biason.

I saw it a few days ago, but it was raining and there was a bucket on the camera thing.

Today, a sunny day, the car was there again, without the bucket. Maybe I’ll be on Google Maps in the future!