Alcides Fonseca

Community & Customer Service – The video of an interesting talk with Matt Mullenweg from Wordpress, Gina Bianchini from Ning, Tara Hunt from Citizen Agency and Patty Roll from Timbuk2.

Ruby and RDF - What about creating RDF statements using Ruby's beatiful syntax?

Why Joost will loose to Miro – A nice article showing why in the near future Miro is the best choice for P2P video.

Why I Unfollow People Who Use Hashtags On Twitter – A negative view of Twitter microformat Hashtags. I do use them sometimes, but most of the time, I guess I’ll agree with Dave Coustan.

DataPortability and me, JB – Another video by John Breslin on DataPortability and a bit of RDF and SioC. I really liked it!

We have old Gravatar, hAvatar, pAvatar (discussed here) and now we have Openvatar.

This last one is based on OpenID, but uses email as the identifier… This makes no sense!(Read comments) and is centralized! OpenID is all about using your own URL as you main identity (although email seems possible too) so you should fetch your avatar based on your URL. Both pAvatar and hAvatar do that correctly, but hAvatar uses another piece of technology I love (and makes all sense together with OpenID) that are the microformats (hCard in this case).

So I’d like all this not-that-good solutions to stop coming, and that we all focus on adopting the one that makes sense to the majority of people.

Yeah, there’s missing only Microsoft for having all the buzzwords in this post’s title. But I won’t be discussing that polemical bid.

Yahoo has launched their 0penID support and I hope the beta there means it’s going to be improved!

I’m doing some OpenID-related stuff and I’m wondering if I should recommend Yahoo as a OpenID provider. First of all, you need to activate OpenID in you account (just like with Blogger). So allowing the users just to enter their URLs and login would be perfect, but I guess they are just doing this for some stats or so. But for regular users having to active something may make them think twice (could be good or not).

One thing that a multi-website entity like yahoo did was to allow users to choose their OpenID url. So I can pick between me.yahoo.com/alcides and my flickr URL. That’s very sweet from Yahoo since they are admitting having flickr in your URL is cooler than the crazy Yahoo. But I find a problem here. Isn’t the “id” in OpenID for identity? http://flickr.com/photos/alcidesfonseca doesn’t have anything to do with my identity, it’s just a place to display my photos. But http://flickr.com/people/alcidesfonseca does! It has all my personal information, even my hCard! So, Mr Yahoo/Flickr Folks, why do we use the photos page instead of our profile page? I know you have a rel=“me” link to the profile page, but it’s just not the same thing!

Ideias3 is now one year old and this weblog will also be in a few days. Since its beginning, I’ve been using it as the center of my web activity. It has my hCard, it is my OpenId, it has my XFN friend list, it has me links to all my social profiles. Although typing in your browser’s address bar “alcides fonseca” would return this website, I guess that having my identity outside ideias3 was important, and now my web presence is pointing to alcidesfonseca.com.

Although the old url is still active and redirecting to here, I’d appreciate you link to this new address now. I’ve just changed a few lines in a few templates (and APML and FOAF) and this line helped a lot regarding the database. Now I am in the quest of changing all my web profiles with the new email and website. I’m curious to see how technorati deals with it.

Changing your personal details everywhere is always painful (I’m hating it, and I’m fearing changing my address when I move to my new house two streets away from current one) but I believe the social web may help this. I had this conversation with Bruno and I guess it could work in a year or so: I have a new email and url. I change my hCard in this new website. I ping technorati’s kitchen or pingerati for both urls. Both those services propagate the pings to popular services like twitter, flickr, etc… They notice that alcides.ideias3.com is redirecting to a new url, so they change it from my profile, and then they update my details from the hCard, since I use alcidesfonseca.com as my main identity page (or as my OpenId). And that way, everything is updated everywhere, with just one change and one ping!

Just a final note: I now enter my email address without any spam protection. I guess it is already somewhere in the web my plain email address, so I don't care anymore. Gmail works fine as a Spam blocker (I've stopped using my MSN address because I still get too much spam).

Brad has announced Google’s Social Graph API. It is indeed cool, but not that amazing. There are a lot of FOAF and XFN parsers around, but being able to poll the web cache they have for google search. This really gets your requests time down to one (the API). Of course, now you don’t get real-time parsing. I really hope Google will work with Technorati’s Kitchen Ping Service but it’s more probable that they launch their own microformat search.

So what’s the feature I miss so much? Contact merger. When I ask for my friends I get this huge list of all my contacts in all my social networks. But I have friends that are in more than one social network, and then being repeated there. It’s possible to merge them, but I would require the maximum of N-1 queries extra from the API. This will really slow it down for friendly people like me. Can’t powerful Google support this in its API?

Hail king of the python parsers! Will you (or anyone who is cool enough) port Mofo (and Hpricot) to Python?

Until now, there was already this one that I couldn’t even put to work. And its old and lacks of many microformats.

I have found this one in beta but it has too many dependencies and sucks to have that all in shared hosting :/

I am really liking Mofo, but not liking ruby that much! But fits in all my current projects (I won’t release its source so you wont make fun of my ugly code!)

Some interesting links for the week:

Folksr is a project that intends to gather people’s votes through their microformated links. You login with your openID, and from then on just add rev=“vote-for” and that’s a vote! Of course this is only for geeky people. But an interesting proof of concept anyway!

MOAT is a way of giving meaning to tags (via URLs). Its a cute project, but I don’t know if it will be useful. Anyway, there are already public servers and a Drupal implementation. Wordpress already in the TODO list.

Gravatar is attached to your email address and is not open nor distributed. The future is URL-based and distributed. There already two alternatives: pAvatar that seems more (unnecessarily) complex to implement, and hAvatar is just the photo on your hCard. Pretty simple hein? I just got one problem: When more than one hCard are present in the page, which one is the author’s? Tantek uses the class author, but I don’t find it in any other website. Any tips?

I’m sure you are familiar with the MyBlogLog concept. It’s a list of people that have recently visited your website. It works through a widget that reads a Cookie in your computer and saving it for the site owner to check.

Well, I am conducting my own experiment to do the same, but instead of requiring a MyBlogLog/Yahoo account, users just need an OpenId. As I don’t use openid for logging in or commenting (yet), only users who have OpenID integration in the browser will be recorded. This works fine with Seatbelt Firefox Extension, so if you are using it, you are being watched by the Big Brother.

According to OpenID documentation, form elements for openid should be called “openid_url” in order for browsers to fill the url automatically. I take advantage of this feature, and when loading the webpage, it that hidden field is not empty, I send a AJAX request passing the visitor’s openid as a parameter that is saved in a txt file (for now).

Yes, it’s as simple as that. I also have a simple script for printing the OpenIDs and I am working in order to fetch the hCard in the URL. For now I am using an external link, but when I got the time, I will make everything work together. It is a simple ruby CGI script (ruby’s CGI module is awesome!) and uses Mofo for scrapping the microformats.

For now, it justs let me know who visits me that uses Seatbelt and OpenId, but in a near future with browsers having natively this future, it will Rock! Oh, and it helps building whitelists of openids.

After the RSS feeds being restored, I took the chance to add some new stuff to my online identity. I use this URL as an endpoint to all my online activity. My hCard information is here, my XFN friend list is here, the links to all my profiles in several online services/social networks are here. It even works as my OpenId.

For the past days, I’ve been checking up on FOAF. There is already one opensource social network accepting this format to import all your information. There is a bit of discussion between XFN+hCard versus FOAF since the two of them represent your personal information and the relationship you have with others. I believe more in the microformat way of displaying data, but FOAF follows RDF standards.While Dan Brickey, one of the authors of FOAF, believes there is space for the both of them, I still have my doubts… Either way, I did my own FOAF file with basic info and then linking to the FOAF files extracted from my last.fm, twitter and flickr profiles. If you want to make yours, you can write by hand or you could use FOAF-O-Matic. For wordpress users, there is already one plugin.

If you are interesting in learning about RDFa (a formal way of defining everything! in semantic web) and a bit about FOAF, make sure you see this simple video.

Next, I decided to implement my own APML file from all the tags I use in my blog. The result is already linked in my head tag and I am waiting to use it in some projects. Imagine filtering through your feeds based on what you post (that is probably what you like!). Yes, there is also a wordpress plugin for this too. And then I just adapted the code to serve as a tag cloud in the sidebar. But the evening hacking wasn’t finished until I add the last delicious entry for each tag page.

I hope in less than 6 months I have a daily usage of this stuff I’m working on. It would be a shame if when I register in a website it doesn’t import all my public info :(

Ever wanted to host your own Facebook or Hi5? Have the benefits of a social network, but without outsourcing it to 3rd parties like Crowdvine or Ning?

I’ve already posted on creating your own social network and mentioned Elgg, an OpenSource SocialNetwork that includes features such as:

• Blogging
• File repositories
• Podcast and full RSS support
• Tagging
• Profiling
• Communities
• Multilingual

And two that caught my attention:

• OpenID support
• Importing content

OpenID is a obvious advantage comparing to hi5, Facebook or even other social networks. It allows you to have your own Identity shared by all your social networks (if you want). But what about importing content? I really thought they were thinking XFN, hCards and other microformats, but no. They are using another standard: FOAF. The Friend of a Friend standard is a XML-RDF way of expressing connections between people and information about them. And it’s even compatible with OpenID.

But personally I believe more in the hCard+XFN solution. It allows you to have your data encapsulated in your content, and not a separated XML file. You write the code once, and it’s readable by both people and machines. I guess those two tribes should gather and pick on a real solution. It’s been so hard for people to start adopting this kind of standards that if there are two to choose from, we will never have that decentralized social network we all dream of.

No, I haven’t upgraded this blog’s engine to wordpress vX.X.X, simply because it does not run on wordpress, nor blogger, nor any other engine. It was coded by me using my microframework pungi. It now supports tagging (using pungi N-N relations) and blog posts are now archived by date.

For those who don’t understand a word of Portuguese, there is now a feed containing english-only posts, so feel free to change or subscribe it.

Apart from the design changes, I added more microformats, this time XFN (XHTML friends network so I can use this website as my portable social network (besides the already implemented openID). If you are interested in learning about this, try to run this website address under XFN Graph.

As every single webdeveloper that doesn’t work for M$, I hate internet explorer and the fact that doesn’t follow standards and rendering changes a lot between versions. Well, here are a few news regarding version 8:

• Office 2007-like interface (SUCKS!)
• Full support for RSS, AJAX e CSS (Is this new?)
• Microformats ( Thank god something new and useful! But wait… microformats or some proprietary Microformat$? With M$ we never know…)
• Customizable Interface (Let’s see what this means… Skins maybe?)
w3c compliance (This is great news, and will be very nice if everyone had to upgrade to IE8, even if they’re not original copies!)

Well, just a “let’s keep up with firefox features” new version. Microsoft is going down more and more…

(src)

Finalmente já não poderão dizer “Ah, perdi aquele concerto! Gostava tanto de ter ido, mas só soube hoje!”. O Palco Principal lançou ontem ainda em fase beta, a sua página de eventos, que reúne os concertos de todas as bandas registadas e também concursos, demonstrações, workshops e outros eventos que mereçam destaque.

Na minha opinião o site está bastante usável com um calendário ao topo para ver os eventos que existem em cada dia. Existe também uma tag cloud para restringir os eventos por zonas. No meu caso, foi só procurar coimbra que é uma cidade mais ou menos grande e clicar lá. Assim vi todos os eventos que estão para acontecer aqui ao lado.

Outra das coisas que me deixou mesmo contente, foi o facto de existir um feed RSS e um calendário iCalendar para podermos estar sempre a par de tudo usando o nosso software de preferência. E gostei ainda mais quando vi ambos disponíveis para cada zona, o que é bastante interessante para mim, que quando estiver sem nada para fazer, vejo no PDA os concertos aqui ao lado.

Em conversa com o João Carvalho, o gestor do projecto, foi falada a opção de disponibilizar os eventos num formato hCalendar que permitirá a quem tem software que leia microformats (Operator, Tails Export, ou brevemente o Firefox3), adicionar apenas um evento a que tencione existir. Fica para breve esta alteração e muitas outras.

My website was a bit ugly so I started to make a new layout. It’s kinda weird this one, but I think makes this more readable. Other pages moved to the sides so you have all the information in the front page.

Now the website is heavier and confused to mobile browsers, so when I got time, I’ll do a mobile version.