One thing that has irritated me a lot with the iPhone is the fact, that there is always "on app for that". Like to use Skype. You have to go and install an application. And to use it, you have to keep it open (and of course without multitasking this is... )

I really like the maemo approach, where skype is not an application, but rather a feature integrated on contacts. Same way you can send pictures to the flicker or other services directly from image gallery etc. I would really like this kind of appoach to be eaven wider. For example integrated note taking application to contacts, calendar etc. So you could be viewing your image, then select "add note" and select your favourite note -application to write a note - or xournal to draw it down.

I was watching the video of Hermes App that updates contact information from facebook etc. AFAIK it is a standalone application for manual update. It would be a lot more functional being able to integrate this kind of application directly in contacts: after installing Hermes, the provided services appear under contacts "synchronise" -button. Then all services would work under the same automated sync etc.

IMHO this kind of development is something that would make Maemo a (even more) special in the market. In two years, instead of having 65000 different applications, it could have 45 000 applications and 20 000 installable features/implementations.


How well current API makes this possible at the moment? (I'm a very light-level coder, so I guess this information COULD be found in development -section, but .... ) How much it is controlled/maintained by nokia? In file sharing there are now few services like Ovi and Flicker. I have heard that "More will be coming". Are they made 100% by nokia, or is it also possible for 3rd party developers to do them?

Perhaps. However, the Maemo API doesn't allow for arbitrary applications to enhance arbitrary menus in other applications: there are a certain number of defined places (Telepathy accounts, sharing services) that have extensible APIs.

It might be possible to get Hermes doing that (but then you'd have to manually merge your contacts ;-p)

There is a sharing service API, yes. This is used by PixelPipe, for example.

I'm with Jaffa; as long as "if its implemented" includes standrized APIs and the openness to fix them when needed, then such a usage/development case becomes compelling enough for the platform to grow and evolve for all users.

Me three.

This goes against basic *nix tenets if implemented wrong.

*brief flashes of IE interwoven with Windows 98* *shudders*

Which would lead into a disaster. But implementing would not need to go so far. Only affect through thouse standardised APIs: In program "Start share" -> Then the share API decides, what content to show in next menu.

And my question is: how many are these places and is this something that can and should be in focus when developoping maemo further on?

And therefore it should be controlled by an able group. So that the core stays clean, functional and well documented (so that they can also be found and used properly).


Example: Basic recipe program
You can add a title and write text. (A "Hellow World" of a -program ).

More value through APIs
Then i use contact API to add a contact (person who cave me that receipe) and share API to add a share -button. And then I give this program to be used by calendar.

Use case
Now I have a calendar entry for upcoming bitrhday, for which I add a receipe. When this birthday comes, I can open a calendar entry and open the receipe. When I get into trouple I can tap the person who came it to me and send SMS/start skype call. When my other friend asks me two weeks later about that delicious food I made, I can open calendar event, open the receipe, tap share and send it to him/her via bluetooth. Then I tap the persons avatar and send a message, that the receipe is a hit!

Now that is not a "Hello world" program anymore... And when someone makes MMS support, my program can send the receipe through MMS as well, without need to make any changes in the program.


How far from reality is this at the moment? Or can it never be reality?

Can't comment on that, just wanted to say I fell the same way. For a really seamless experience I would want to have access to or integration with a map application whenever or in whichever application I encounter an adress or location, be it in a geotagged photo, in the contacts list or on a webpage. Similarly I would want an intuitive connection to calendaring when I encounter a date, the photobrowser when I encounter a photo, the contacts application when I encounter a person etc.

On the other hand I wouldn't want to be restricted to the built-in core applications. But my feeling is the same as yours (if I got it right) that a lot of this could be accomplished with a thought-through API. I am no way near proficient to say if that's right though...

Less isolation, more integration!