Here is a compilation of stuff I wrote during this day:
In the afternoon:
This isn't yet a post about XTech itself (and it's presentations), but about being here. Since I finally found out where the power outlets are (yeah, I'm stupid), I have enough battery to write about stuff like that... Until now it is being amazing to know so many interesting people, and the presentations are being good. At lunch time, before going to my new hotel to check-in and leave my luggage, I met Donna, a really nice girl from Australia that is going to give the ODF presentation I said previously I want to attend to, tomorrow. I'll tell you again that you should be checking Planet Noori to be updated about this conference between other things, since it is aggregating more stuff than Planet XTech does. The wireless connection is still sucking heavilly (my wireless setup doesn't help much), and there are being some technical issues that are making the presentation "XML, REST, and SOAP at Yahoo" being late, so I guess I'll take this time to organize the notes I've been taking during the day...
In the end of XTech's day:
I'll consider yesterday my XTech's "day 0", and now it's the end of XTech's day 1... I've just attended to a presentation called "The power of declarative thinking" made by one of the creators of ABC (the language that inspired Python). I just have to say that he's an hell of a speaker, preety cool and convincing.
As a turn-down, I'm getting really pissed off with this laptop (as a matter of fact with Fedora Core 5). Probably it needs an update (even if it was installed and presumedly updated two days ago), but most of the times that I close the monitor Xorg just goes boom and there's no way to go to the console. Besides that, while I was in the last presentation, I just lost my keyboard. Thinking that a reboot or even a hard one would solve it, I tried to shut it down which crashed most of Gnome. So I just had to shut it down on cold, pressing the power button until it was turned off. I'm not 100% sure that this is not hardware problems (although there are no logs indicating such), but way to go Fedora.
At night, in the hotel:
First of all, I don't want to give you a wrong impression: I'm loving being here in Holland. After XTech, I visited (walking) a great part of the center of Amsterdam, so I'm really tired and will be quite short in my writtings tonight. But I think this city is awsome! Found a funny thing: they have here what they call "age coins": coins with no value at all that makes tobacco machines only work after you insert one of those! :-) On the other hand, the less expensive pack of cigarretes is Pall Mall, with 19 cigarretes and a price of 4,- € (they use - where we use 00), against the cigarretes I usually smoke, that besides being better ;-) have packs of twenty and cost only 2.75 € :-P I still didn't enter in any coffee shop (that's not a thing I want to do by myself, I expect to get on XTech someone who wants to go to one with me tomorrow), but then, even if I'm curious, I allways heard saying that being in Amsterdam and not smoking a joint is worse than being caught and convicted by a murder you didn't do. Well, I guess that tomorrow I'll get in jail, 'cause I'm a non-smoker (marijuana-wise). I passed by the Red Light District, and it was kind of disappointing for me. Don't get me wrong: I liked to see it live after hearing so much talk about it, but in the whole "district" I only saw like... 10? girls showing off, and I was expecting way way more. Instead of that, what you have it's a huge (more than business allow, I would think) ammount of porno movies stores, and then sex shops, some with cabins and some live shows theaters. Well, I saw it by myself, and that's what counts.
Regarding to XTech (you must find this whole posts completely boring, and I must try to convince myself they are usefull some way :-P ), this are the presentations I saw today:
- Putting the BBC's Programme catalogue on Rails
- Collaborative Atlas: Post geopolitical boundaries
- Web 2.0 on Speed
- Publish-subscribe using Jabber
- XML, REST and SOAP at Yahoo
- The power of declarative thinking
They are six instead of the eight I wanted to since I had those problems I described in past posts, but still, it was a full day earing lot's of cool stuff.
Putting the BBC's Programme catalogue on RailsI thought I would loose this presentation, but after all I just lost it's beggining.
This presentation was made by Matt Biddulph (from hackdiary) which is a freelancer and ended his presentation saying "Hire me"... And you should.
BBC had a programme catalogue database full of information, some quite old (semi-historical), and, even knowing that there they had lot's of valuable information, they quite didn't know what to do with it. That information could be "read" (but not "surfed") in a green-screen application, but what Matt had in mind was to turn it into a Web 2.0 application, usefull not only to BBC, but to everyone (which, once again, is usefull to BBC). To do it, he used the exponencialy growing Ruby on Rails framework, because, as he almost said, everything really innovative being made on/to the Web is being done in Ruby on Rails. Of course that that isn't the only factor of decision. Rails was good because:
- You can easilly have absolute and unique human-readable links for every page (and we're talking about gazillions of them)
- Active Record makes it simple to do it
- Ajax'ing your application is so easy that is painfull (as he demonstrated)
- He draw numbers, and I wasn't quite convinced, but he said that RoR has way more performance. I won't even discuss that if you're comparing to big J2EE frameworks and stupid stuff like that, but RoR isn't the only thing you have against that kind of sollutions... I would like to discuss more about RoR performance, and maybe I will still have the opportunity to do it on this XTech.
Searches on applications like this are allways a pain in the ass (we're talking about fulltext searches). To get the job easilly done, he used an external API from Yahoo! that almost makes the work for him.
He also uses some Wikipedia API (I should search more about that one) to relate content of the BBC programs with Wikipedia's content. That's preety cool, specialy if you see not only that person "foo bar" was a special guest in some debate, but who is he and why is he there. I don't know if the interaction between BBC's RoR App and wikipedia is bidirectional, but it would really cool if it was.
He then talked a little about the application deployment, and I found it quite funny since it remembered my job in some aspects I won't discuss here. Mainly, he showed us that he needed no hyper-mega-giga-cluster to run the thing, so he has one machine doing the whole job (he knows that not having redundancy isn't such a good idea, but - and that shocked me - he wasn't worried about it. I think I'll try to talk with him and ask why...). He still works for BBC two days a month, for maintainance purposes.
Finaly, he left some words that sounded to me like he was trying to say this (which I agree): doing something interesting and doing it using RoR, is doing it Web 2.0.
Collaborative Atlas: Post geopolitical boundariesIt's getting late, so I can't be so descriptive as I was to the previous talk on this one. Since this presentation is on the web, at least for now I'll just leave to you the
link to it.
Web 2.0 on SpeedPreety nice presentation. Fortunately for me someone else already
blogged about this.
Publish-subscribe using JabberSorry mate, but your presentation didn't cativate people (we've talked about this later, so you know what I mean). That's a shame, because the theme is really interested, I had a personal interest in it, and at the end you didn't went as further as I liked to on the issue.
The talk was about XMPP (Jabber) uses besides IM'ing, and what could be done. XMPP is a streaming XML technology that can be used in the Publish-Subscribe model to do some cool stuff, way cooler than what was mainly talked about (extended presence, where you say what's your mood or what you're listening to ATM) like Atom-over-XMPP (a good example given).
Here are some raw notes (I'll have to come to this later and explain them) I took during the presentation:
* What's the Publish/Subscribe scenario (pubsub)
* Types of pubsub
- Topic based pubsub
- Content based pubsub
- Notifications with change info
- Persistence
* Examples of pubsub
- GUI's
- Mailing Listgs
- Chat rooms
- Presence & contact lists
- Any system with event handlers, callbacks, signals and slots
* pubsub in Jabber
- Misuse of
- Groupchat is poor man's pubsub
- Generic pubsub system:
* Publish Subscrive Service (Mediator)
* Publisher (Subject)
* Subscriber (Observer)
* what can be done
- Extended Presence:
* Geolocation
* Address
* User Mood
* User Activity
* User Tune
- News / Weather
- 'Mailing Lists'
- Emergency Alerts
- System Administration (Spong-like)
* Who's doing it? Psi is, but there are other apps doing it... Like one used
as a newsreader
* How is it done:
News source ----- service
|
.news bot <---> jabber client
database'
`-webserver --- browser
Yeah, I had the time to do an ASCII drawing, so see how bored I was...
XML, REST and SOAP at YahooThis was undoubtly the best presentation I attended during this day. It wasn't the best theme nor the best orator, but the combination was blasting awsome. Man, I totally agree with all of what you've said, and had a couple of good laughs (and stares) with your fully-truthfull bashing on Java, J2EE and SOAP sollutions using it.
The presentation was so good that I just forgot to type when you talked, so I just took some notes about it's beggining. But this is something I
must fix, I'll write a big issue about this talking at a later time.
Sorry people, but I don't see any point of posting my notes, since they aren't "incomplete" in a sense of "short", they are only covering the first minutes of the talk... I'll try to fix it ASAP.
The power of declarative thinkingThe theme was quite different that what I was expecting (my fault, tho). Steve Pemberton, from W3C/CWI, was one of the creators of ABC (the father-language of Python), and didn't stopped with it. He's an hell of a speaker, convincing, and, well... I liked his pose while talking. He talked about the requirements for Web Apps, about the fact that we have lot's of computing power and usually let them idle'ing while we're doing our job - jobs that most of the time could be mady by the computer and not by ourselves. He talked about the inderpinnings necessary to make Web 2.0 follow the spirit and development the "first version of World Wide Web" had. It's a hard road, but with speakers like himself maybe we'll get to there.
The end...Well, for today I already tiped much, and I'm so sleepy you can't imagine (the lack of quality in the coffee might justify part of it)· So, this is it. I'll write more when I find the time to do it... Until then, if you're finding this set of posts interesting, you might want to read more about XTech
here.
Note to selfI should fix three issues of my Planet ASAP, hopefully tomorrow.
libextractor is a library used to extract meta-data from files of arbitrary type. It is designed to use helper-libraries to perform the actual extraction, and to be trivially extendable by linking against external extractors for additional file types.
