Unfortunately this post is also going to be super short. Just couldn’t keep my thoughts together and get up and write some cock sucking, bad-ass blogs as I have been doing in the beginning.



Here’s a list of accomplishments:

Got the cEP [*] merged.

But it feels like a pity merge since I wasn’t moving forward with anything. (Now that’s an accomplishment to brag about 😗). Still I am glad its over. I was done with just sitting and thinking and it was nice taking a break and doing some actual stuff. Gonna improve upon it / add implementation details later on probably in the starting of the third coding phase.

Got a Pull Request merged

(approx 600 lines). Turns out Satwik (@satwikkansal) was right about the speedy coding process after the idea is finalised. It took hardly any time but the review process was tedious. I wish we could set up some in between sort of “coding meet ups”. (I meet up with him every sunday at 7 p.m. on gitter a.k.a. regular meet-ups and he also replies as soon as he sees the notification whenever I ping him in the middle of the week, but its still tedious). He gives some reviews to me and I fix them almost immediately (yeah right, I don’t seem to be having a life like these other people and I stay online all the time on gitter, replying immediately) but then its not until the next day or 4-6 hours after, that, that I get a new set of reviews which took it almost a week to get the Pull Request finalised.

The Pull Request is about seperating the arguments to the run/create_arguments methods of bears Reference 1, Reference 2 which accept boolean values from the ones which accept any other kinds of values. They have also been categorised into optional arguments and non-optional arguments. The data is stored in a class whose object is returned to the main method for coala-quickstart. There was a slight complication in the middle where some arguments had type annotaions like

def run(my_awesome_setting: bool = None):

due to which I had to refactor everything. As earlier I was sorting everything based on the default value but now I had to consider the possibilities that:

  • a bear setting (arguments to the run or create_arguments methods) can have a default value but no type annotaion.
  • can have a type annotaion and no default value.
  • have both of them.
  • have none.

So that took some time but the task is complete and everything’s working fine. Also parsed the bear dependency tree to check recursively for non-optional settings to the bears. Considered the possibility of using dynamic programming algorithms but it felt as a kind of premature optimisation and the present bear dependencies are hardly complex in coala.

Here’s a list of unsuccesful events:

Wasted quite a few days last week

I did nothing, just literally nothing, gazing at the roof, the ceiling fan, whirring about doing its business. (You’ve got no right to judge me, we all are guilty of indulging in these kinds of day dreams)

Had decided that I could seperate all the settings before today

But hit a bottle neck over there with the contracts package giving me a hard time on adding type annotations and value checks to Type 3 settings.

That’s it. Buzz off from here. I’ve been looking forward to posting my awesome blogs instead of this non-interesting update stuff. A juicy blog is on its way this week.

[*] for newbies to the blog, more about what is a cEP over here