Born On The Web

After I learned how to use PHP to move files to S3. I was looking for a way to do this in a transparant way, so you’ll still be able to use your own domain name when storing your files on S3.

Twitter uses S3 for storing its user avatars, try loading the timeline and you’ll see the amazonaws.com url coming up in your browser status bar. While this is just a small and visual annoyance, it becomes an issue when people start linking directly to your files stored on Amazon S3. The direct link to a file on S3 will look something like http://aws.amazonaws.com/your_bucket_name/your_file_name . This is not exactly pretty and you’re missing out on a lot of Google juice. In this post I’ll show you how to avoid this in a few simple steps. Only thing you need is an S3 account and access to your DNS config file.

Read the rest of this entry »

I’m an avid public transportation user: train, bus, tube, … The only thing I hate is waiting too long for a train to come, or missing a train because you were seconds late.

This could be solved by a simple location based service. (I’m not completely sure, but I think both Mobistar and Proximus are already providing these kind of services, correct?) When you come close to a train station, say like in a range of 200m / 600m (you should be able to adjust this yourself). The service detects you are nearing a train station and simply sends you a message with the schedule for trains going to your home station. So a 100% push medium, which is triggered by your location. I wouldn’t mind paying like 50 cents per message, because it adds value. It let’s me know if I need to start running or can go grab a bite before catching my train.

But since the nmbs (belgium railway company) doesn’t even provide a mobile site. I’m guessing it’s going to take a while before seeing this kind of services …

I talked about this before, but if the NMBS would open up it’s train schedule database, a whole new range of services can exist:

• more services for the people who travel with the train, you make their lifes easier
• an extra business model if you charge for commercial use of the web service
• build an ecosystem around the brand NMBS

Really, somebody should wake them up … we live in the year 2008, not 1998. Damnit !!

You RSVP (It is standard practice to reply to an RSVP request whether confirming attendance or declining) to an event at last.fm, facebook, upcoming, meetup, …

Wouldn’t it be easy to have an universal RSVP web service… which is coupled to your openID. So if you RSVP to an event on last.fm, the event is automatically added to your events on facebook and other …

I have to get this of my chest, I wanted to do a BarCamp presentation about it, but I can’t wait … So here is my small rant!

In Belgium we have a public transportation company, NMBS(trains). The NMBS has a pretty good website, so it’s relatively easy to find the correct times of trains and since I’m an avid train traveler. It’s really necessary being able to quickly check what time your train leaves or at what time you’re going to arrive somewhere.

But mobility is strangled into our daily lives, strangled into our professional and private life. So why isn’t it possible to integrate the mobility data of the trains in more life-managing-applications like Outlook, Google Calendar, … But also more specific needs for example school lesson schedules, contact information of companies, … So integration of relevant travel info with time data would be incredibly useful.

If you have a good database (which I hope nmbs.be has), it shouldn’t be to difficult to build a web service on top of it. Or am I wrong? The data is already public, so you don’t really have security issues.

These are the first 3 methods I would love to play with: 

GetStationInfo

Parameters: Stationname & Time

Returns: all the trains on a certain time, different directions with all the stations which are reachable. Maybe station location info.

GetTravelRoute

Parameters: Start station name & Destination station name

Returns: route & possible times of departure.

GetStationList

Parameters: State

Returns: List of stations in a certain state.

Of course this is really basic, just to give you an idea how cool this could be.

I think it would really be an added value for a lot of applications. I’m just going to list a few example mashup/applications I’m thinking about: Decent mobile interface (do they currently have on?), iPhone interface, google maps mashup, integration on the contact information of company websites, school websites (integration in lesson schedule), integration on the website of Brussels airport, … I really think the possibilities are endless.

So my message to the NMBS web staff, please guys … open up your database, give us a nice web service to play with. The mashups will be stunning!

What would you make with the NMBS web service?

Thanks to the native interoperability between google maps and flickr. I can now embed maps with recent geotagged flickr photos on it. Sweet huh! Making mashups has become easy…

Photos from Leuven:

View Larger Map

Photos from Brussels:

View Larger Map

Photos from Gent:

View Larger Map

Or even my personal geotagged photo map:

View Larger Map

Would you rather read this site en Français, 中国 or auf Deutsch? Not a problem. Apparently Google also translates complete webpages.
All links on the page point to translated pages too. So you can just go on and keep on surfing sites that were made in another language.

A service like this could prove to be very handy for translating the millions of Japanese, Chinese, Spanish or whatever sites out there that we can’t understand. I’m sure there must be excellent content on these non-English parts of the internet that we’re missing out on.

I stumbled on this while browsing through the very impressive list of Googles services to make your life better.

I’ve always liked the way flickr build their URL’s. Now that they just launched geoRSS feeds, it will show you how the URL of a service has become a kind of command line for that service.

The URL is structured in the following way:

flickr.com/services/feeds/geo/ countrycode / province or state / place

It will get clear if I give you some examples:

• http://www.flickr.com/service/feeds/geo/ will give you all geotagged photos from all over the world.
• http://www.flickr.com/services/feeds/geo/BE/ will give you all geotagged photos from Belgium, (BE is the country code).
• http://www.flickr.com/services/feeds/geo/BE/vlaams-brabant/ will give you all the photos from the province “Vlaams-Brabant “in Belgium.
• http://www.flickr.com/services/feeds/geo/BE/vlaams-brabant/Meise will give you all the photos from a town called ”Meise” in the province of Vlaams-Brabant, Belgium

Cool huh! But that’s not all. Now surf to google maps and just copy past a flickr geoRSS(a click for the lazy people). It should display the images nicely on the map. This is what interoperability between to services is all about.

Great work !

When I first heard of openID, I taught, great that’s going to come in handy. But when you think about it the possibilities are much more than just a way to skip the register page. It makes your network of trusted people (friends) much usable in every web applications. I’m going to show you what the real power of openID is, using my wordpress blog as an example.

This is an example how a new technology like openID can have huge online social consequences. I like that! :)

But cut the crap let’s get our hands dirty:

1. the problem

Well everyone has a lot of comment spam, the longer you blog, the more spam you get. Askimet catches lots of spam, still this remains a huge problem.

2. the solution

Social Whitelisting (idea from Tom Coates):
I trust most of my blog readers, because I know them and they hate comment spam as much as I do. I want them to be able to moderate comments on this blog. If they see a spam comment, they should be able to delete it. So now my readers can sign in to my blog with their openID. I get an email, I check if I know this person and if I trust him. If yes, I give him the rights to moderate comments. It’s as simple as that. So if you like to test it out, simply login and I’ll give you the rights to moderate my comments. Just to give you something extra, you will also be able to edit my published posts, jep, I make a lot of spelling mistakes ;-) AND make drafts.

3. How ?

I used 2 wordpress plugins:

1. openID plugin
2. role manager plugin

It still needs some tweaking. The role manager plugin is acting a bit strange. But I’ll figure it out.

4. Next steps

This is great and it works, but still I need to approve every new member manually. I want to be able to have a list of people who I trust, not only to moderate my blog but also to view my private photos at flickr for example. How can I do this? Well I can make a list of people who I always trust by making a group on jyte, people in my group will always be given the correct rights.
This is a plugin I plan to write, when somebody signs in on my blog with an openID, I check if he’s part of my network of trusted people (with jytes API) and give him the permissions he needs to moderate comments.

With jyte you get the possibility to decentralize your friends list. Make your friends list cross-platform and give them the permissions they need to make your life a bit easier.
It’s one step closer to a decentralised social network and that, my friends, is the real power behind openID.

Resources:

• six cool things you can build with openID
• social whitelisting with openID
• Simon Willison’s talk on FOWA

I’m still looking for a good way to divide me posting over here and at my employers’ blog i-wisdom. It’s seems so stupid to cross post it. I think a lot of readers over here, read i-wisdom to? not? Anyway, here the link to my article about geotagging photos on flickr.

Over here I do have the tendency to post more about techy related topics. Maybe it’s good I can keep this blog for my geek side.

I wanted to post this a while ago, but never came to it. During the world cup this year we held a small competition at Pickdee. Gary from Pickdee mailed me last week with a nice new feature he implemented:

Gary from Pickdee here, hope you’re enjoying the football.

I just came across your post on pickdee [ after having being introduced to your blog from the pickdee post ] — I also thought the facebook API opening up was an interesting event. It’s definitely of a different nature than most of the other web API’s.

It let me build this:

http://www.pickdee.com/facebook/

in pretty short order, it lets any facebook user use their login there to authenticate with pickdee and play one of the two new American Football games…

Funny seeing your two posts together.

A lot of mashup kids like Gary (from Pickdee) and me (sometimes) clearly see the advantage of building further upon existing services. He plugs his system into facebook, he doesn’t have to make a registration page, although he already made one now. The value of a facebook account goes up because you can use it in Pickdee.

The web is evolving to a nice ecosystem, where every platform has its specific task and focuses on doing that specific task very good.