PR closed without merging first

[formacube]

My PR has been closed without being merged.

My code is plenty of patches since begining of september because I'm waiting for you to validate the code.
You make me wait 6 months to change the formating of the code while it would take you less than 5 minutes to do it.

I have to wait for 6 month ...
and then you decide to kill everything because you have another idea ... that is not even yet implemented !

Please first merge my PR, so that I can have some code that works correctly and get rid of the patches.
Then when you have the time, there is no problem to implement your new idea.

But at least in the meantime, I can have a code that work, without monkey patch.
and also, it will be easy to test if your new idea is still functional for tasks that takes few days or few weeks to execute

[formacube]

it does not really change anything for you.
For you there is no drawbacks merging this PR which is now ok and formated according to your standards.

But for me it changes a lot of things

[formacube]

I couldn't react earlier as I was ill with covid.
I would appreciate if you stop closing all the issues and PRs before you get an answer. You can at least wait one or two weeks.

This is not very constructive, especially with all the efforts I put in it.

[jedie]

I'm really sorry about how it all happend. It went very unhappily for you.
The whole thing is due to the fact that I had too little time for the project in the past. I hope that this will change.

I couldn't react earlier as I was ill with covid.

I hope you have gotten over it well!

Please first merge my PR, so that I can have some code that works correctly and get rid of the patches.

OK, i will do this.
But you added cumulate2parents in #44 to pervert spamming the database full, isn't it? But my refactor will remove this, because then the cache is used and it's useless, isn't it?
So in the end there is a release with cumulate2parents and a later release will deprecate this again.

Is this okay for you?

[formacube]

I understand that you had little time.
It is ok with me that the changes will be deprecated at some point if we are able to satisfy the business needs.

It changes a lot of things for me, because I can have at least a version that I can use, and which can work without monkey patches.

However, the goal of cumulate2parents in #44 is not linked at all to prevent spamming the database full.
This is just to allow me to work.

As I explained in #44, there are two main different business use with sub-tasks:

1/ progress of main task = sum of progresses on sub-tasks (cumulate2parents = True) i.e.: main_task_progress = subtask1_progress+subtask2_progress+ ...

2/ progress of main task is linked to the completion of sub-tasks (mainly my case: cumulate2parents = False) for instance: main_task_progress = 1subtask1_is_complete+10subtask2_is_complete+ ...
In that case, main_task_progress cannot be simply calculted from progress of the different sub_tasks. It is the different sub_tasks that update the value of main_task_progress.

Being able to have this kind of behaviour (together with the light weight of the application) was the main reason why we decided to go with django-huey-monitor

Preventing spamming the database is only achieved through updating the value of TaskProgressModel, instead of creating a new TaskProgressModel each time we record progress. It is completely unrelated to the implementation of cumulate2parents

Hopes it clarifies better.
Thanks in advance for your understanding Jedie

[jedie]

Yes, i see, a cumulate bool is usefull and should be stay.

Please see also the new idea, here: #57 (comment)

[jedie]

I have convert this issues as a discussion ;)
#61 is on the way.

[jedie]

cumulate2parents is now releases as v0.4.3

[formacube]

Thanks a lot Jedie.
Have a nice evening