selective creativityis a blog fromflowingabout design, contexts, experiences and whatever makes our culture a better place

selective creativity

design as ultimate language

Mapless is a small framework for storing objects in a key->data fashion (i.e.: noSQL databases) without requiring any kind of object-data map. So far only MongoDB is supported. It can use Redis for reactivity (pub/sub) and cache.

It's an obscenely simple piece of opinioated software that allows you to do things like this workspace from our little experiment on real-time on the web we described here and here:

odb := MongoPool instance databaseAt: 'Reactive'.
odb do:[RTask findAt: 'c472099f-79b9-8ea2-d61f-e5df34b3ed06'].
odb do:[(RTask findAt: 'c472099f-79b9-8ea2-d61f-e5df34b3ed06') isCompleted].
odb do:[task :=  ...

Last week we made a tour into the internals of tasks, its structure and some general aspects.

In this post we are publishing a second part addressing some of the questions we received regarding to communication. So here we show more details about the models and commands coming and going between back-end and front-end:

Feel free to write to for questions, I read all the emails that are sent by humans :)

And if you enjoyed, please rate and comment.


This video shows the internals of tasks, a single page application from flowing. 

Tasks is an experiment on full-duplex client-server communication to produce distributed and multiple presentations of a common model mounted in a scalable architecture.

Also an example on how you can develop an application with Amber in the front end and Pharo in the backend.

Check this single page app visiting:

For other free pieces of code check Sebastian's public repos on github

And for more about Amber you can visit:

And for Pharo ...

A tribute to all the engineers who make things real.



The idea here is to leave some useful references on things you will need to ship software in a way that works.

With this you'll be able to design, edit and deploy your apps efficiently.

First, how to setup your own private Git server on linux. You need to put your scratch, experimental and productive code in order.

Second, you are going to use the terminal commands to use git but you're also going to use a good or great GUI git client.

Once you do that you need a method.

Here is what they claim to be a successful branching ...

our favorite insight...
Creativity is great... until is not.
and this make us think...
our main product
most popular