winter - winter streak 2017

[danse]

I am thinking to do a streak this Winter, in order to dash in the direction i want to pursue. The time frame could be from the 18 of January to the 21 of March, last day of Winter. I am thinking of one-week time units. I could do a bot per week, in Haskell, from conception to deployment. That would be 8 bots in 8 weeks

[danse]

Alternatively, i could write 4 projects of 8 weeks each

[danse]

Alternatively, i could write 8 or 4 Purescript or GHCJS web applications, simple ones.

Just in that case there wouldn't be any deployment step, and any functioning service, so it would be harder to figure out the definition of done

[danse]

Since i have several technologies i want to use, i might want to try 8 technologies in 8 weeks:

• Mobile app - Clojure or Java
• Offline app - Purescript or GHCJS
• Server side web app - Yesod or Rails or Django
• Programming Interface - Postgrest
• Client side app - GraphML + React
• Bot - Twitter
• Creative coding - Open frameworks
• Data science - R or Pandas

The outcome could be a blog post for every week. Every project could fail or succeed. Success means that i end up with some software that is useful and i want to maintain and will get its own repo. This is interesting and useful but it seems too ambitious. Repeating the same kind of project would allow me to get to a more satisfying result at the end of the streak

[danse]

Alternatively, i could use these iterations to try to reach a breakthrough in topics that are really important to me, for example Purescript or visualisation. I think that being repetitive is a key in this kind of streaks, changing technology every week would feel shallow and dispersive. I could develop 8 visualisations about Europe.

I have had the idea of developing visualisations for Europe for a long time, but until now i wasn't able to produce anything. Spending a winter streak on that could allow me to get fluent with the domain, the resources and the process.

From the point of view of the goal i am motivated to work on this. Technologically, visualisations are challenging but i am already acquainted with them, so this is less motivating. I would rather do experiments on systems that allow the user to be more interactive and add their data, and on system that interact more with other services and platforms ... a bot would be a better target in this sense ... a Twitter or Facebook bot.

The first iteration would be a simple echo bot. Twitter or Facebook? I like Facebook less, so i could choose it with the intent of learning more new things. I don't like nor use it much but the world is using it a lot.

This seems to be my best option as a library

[danse]

This query ot Github shows my main library options

[danse]

It's important to go all the way down to deploying the bots, that's why it makes sense to have an echo bot as the first one

[danse]

While this idea is appealing, i have too little resources right now and i don't feel like investing most of them towards a single direction. Thanks to some new habits, in these weeks i've been catching up with projects that were lagging behind since years. This already makes me more productive, but it also allows me to stay flexible and spend my time in short or long efforts towards multiple directions. Starting a bigger commitment towards a goal decided more or less arbitrarily would feel contrived and reduce my opportunities, rather than increasing them

[danse]

Hopefully i will be able to do a similar exercise later if i will have a bigger motivation towards a single type of activity

[danse]

The gist "fun with quaternions" from Mike Bostock made me remember that it's fun once in a while to just write an interesting algorithm

[danse]

I won't adopt this format this year. Spending the mornings coding is already giving me good results, and i want to invest all my available energies in something along the lines of #34 in the next months