Technicky jde o upravený Wordpress. Žádný kouzla, žádný nákupní košík ani placení kreditkou. Prostě místo příspěvků do blogu, krátké anotace knih a navíc jednoduchý formulář pro objednání. Easy, easy, teď mě jen zajímá, za jak dlouho si to na sebe vydělá.

Now that Google Mapplets have been released for quite some time I decided to write one for GeoURL:


var map = new GMap2();
var overlay = null;
GEvent.addListener(map, "moveend", function() {
map.getCenterAsync(function(center) {
var overlayTmp=overlay;
overlay = new GGeoXml('http://geourl.org/near/?lat=' + center.lat() + '&long=' + center.lng() + ';format=rss10');
if (overlayTmp!=null) {
map.removeOverlay( overlayTmp );
}
map.addOverlay( overlay );
});
});

This was not super hard and solves the visualization problem nicely. Once installed to Google MyMaps it displays the nearest sites to map center wherever you go on Google Maps. It can be used together with other Mapplets so one can see GeoURL sites next to flickr images or any other geocoded information provided by mapplets.

To instal this GeoURL Mapplet use this link or search for GeoURL on Mapplets gallery.

This is how Prague center looks like after instalation:

Personally I like this map much more than the one provided by Czech Golf Federation. It does not filter by course size or region, but viewed in Google Earth or Maps it can be used to find directions and zoom in to see actual course if Google has bought satellite map good enough for the location. As an example, this is a GE screen shot of Golf Hostivař (including the new clubhouse under construction).

View the original map in Google Earth, Google Maps or embedded underneath.

View Larger Map

Facebook does not officially support OpenID. But would the Facebook Platform make it possible for 3rd party developers to implement OpenID App on top of it? I’ve been thinking about this lately while reading the API documentation. I think it is not possible to implement OpenID consumer, so those of you who wanted a way how to log in to Facebook with their existing OpenID might now stop reading.
OpenID provider on the other hand should be possible. According to Authentification description in the documentation:

In order for a Facebook API client to use the API, the user of the client application must be logged in to Facebook. To accomplish this, direct your users to: http://www.facebook.com/login.php?api_key=YOUR_API_KEY&v=1.0, which will prompt the user to log in if necessary.

After successfull login it is possible to retrieve all user date with facebook.users.getInfo method and thus verify the user’s identity.

So the complete workflow looks like this:

1. User enters OpenID enabled site (OpenID consumer) such as this blog, Zooomr, Jyte or whatever.
2. He enters his OpenID URL such as http://www.f8ID.org/700107342. The number is his facebook profile number and the domain is a domain of the hypothetical OpenID App service provider (the domain is free as of writing this).
3. The consumer site redirects him to f8ID according to OpenID protocol.
4. f8ID redirects him to Facebook login page.
5. User logs in with his Facebook password.
6. Facebook redirects him back to f8ID.
7. f8ID calls the Facebook getInfo method and verifies that this session has the the same profile number as is in the OpenID URL. If they are equal f8ID redirects back to original OpenID consumer site according to OpenID protocol.

This might seem complicated and unintuitive. But as with OpenID in general, the theoretical description puts off average internet user, but practical use is much easier to understand and use. I practice the user is involved only in step 2 and the first time of the day also in step 5. The rest is transparent to him.

So technically implementing this is possible, but would anybody actually use it? I’m not sure. There are many OpenID providers to chose from. This app would have just one advantage - one can use just one password for OpenID, one password that you need to have anyway. In principle it is very similar to idproxy.net, which turns Yahoo ID into OpenID. I can’t find the numbers, but I guess that only a very small fraction of Yahoo users use idproxy.

Please share your opinion. Would you consider using Facebook as OpenId server?

All I have is one positive experience: I have been to the states only once for a few months. I stayed in New York only about a 4 days of the time. I have walked down the Central Park, across the Brooklyn Bridge and I have seen the Flat Iron Building. I also remember eating several meals at one pizza place across the street from the hostel. It is probably a place like hundreds others, but I liked it a lot. Especially the fact, that the waitress spoke worse English than me.

I took a photo of the place from the hostel room. Now, 5 years later, while trying out the new google feature I became curious if I still can find the place and if so, if google took a photo too. And guess what! I can and they did!

There is only one thing missing in the puzzle - a photo of the waitress. I can’t google it, it’s neither on flickr nor on Zooomr. Anybody going there for pizza, who would send me the photo?

Takže ve zkratce: Deep Tagging je tagování částí obrázků, zvukových nahrávek a videa a odkazování na tyto části. Něco jako #relativní odkazy v HTML, které lze navíc otagovat.

Kromě veotag (výše zmíněná investice) se tímto konceptem zabývá více webů. flickr a Zooomr umožňují tagovat části obrázků. Pro video se mi zdá nejlepší služba SceneMaker, která umí otagovat scény z videí na Google Video, YouTube, Metacafe a DailyMotion, takže je možné otagovat filměčky už dávno uploadnuté bez jejich migrace na jinou službu à la veotag. Například scénu z 66 Seconds, kde hraju.

I’m not going to promise any regular post from now on, but still I don’t want to close this site down. Why? Because if I did, I may start to wonder again: “Shouldn’t I start a blog”.

I still have one or two ideas for posts, but not much time to work on them. So maybe this year I will write a few more little posts. And that’s a reason not to close this down too.

I have always wanted this kind of openness in social networks. I thought that FOAF will eventually evolve in this kind of network (I even wrote a scraper transforming LinkedIn account data into FOAF). Now I have to admit, that maybe ClaimID’s approach is more user friendly and also more secure and verified (after all, anyone can write FOAF about anyone, but stealing OpenID is difficult).

The only disadvantage compared to FOAF distributed files is the little lock-in. In ClaimID, you don’t have to lock yourself or your contacts in the service, but you still lock the relationships in. If one day ClaimID disappears, so will the relationships. Even worse, if you want to build similar competing service, although both will be open to all OpenID users, the relationships will never mix-every user will have some of them in one service and other ones in the other service never seeing them all at once.

XFN solves the problem a bit. Still I think marking the relationships with FOAF and RDFa would make the service even more transparent, allowing anyone to make SPARQL queries over the social network and possibly several social networks following the same approach, making the social networks part of the semantic web.

Even without FOAF/RDFa integration, ClaimID seems like a very good social network offering. The problem is, most of the people I know, don’t use OpenID, so my ClaimID will stay contact-less until OpenID goes massive in this part of the world. Considering most of my friends are only recently discovering LinkedIn and using it as their first social network, this might take a long time. Or does the sudden interest in LinkedIn among my peers have something to do with the actual situational relevance as most of them are now (like me) graduating and starting careers?

Update po 24 hodinách: dneska to vypadá tak, že na technorati padl nový rekord-1500 blogpostů o twitteru za dnešek a na twittermaps jsou 4 twitterposty z Čech-50% Praha, 25% Roudnice nad Labem, 25% Kladno a 0% rodina. Lehká progrese, ale zatím předčasné z toho dělat závěry.