Update timestamps each time a build is done

[modos189]

Found a way to avoid code duplication

Closes #740 #770

[github-actions]

🤖 Pull request artifacts

file	commit
IITC_Mobile-test.apk	dd3877b
test-0.39.1.20241103.111649.zip	dd3877b

See build on website

[nexushoratio]

What needs to be done to move this forward?

[modos189]

What needs to be done to move this forward?

If you confirm that this PR works properly, I will merge it

[nexushoratio]

If you confirm that this PR works properly, I will merge it

Ahh, sorry. Didn't realize this was up to me. :->

[nexushoratio]

Embarrassing git novice question (I've not done this step before):

How do I get this locally using git instead of gh ?

From reading https://docs.github.com/en/pull-requests/collaborating-with-pull-requests/reviewing-changes-in-pull-requests/checking-out-pull-requests-locally, I thought I could something like:
git fetch origin pull/772/head:testing

But I just git an error:

fatal: couldn't find remote ref pull/772/head

[modos189]

How do I get this locally using git instead of gh ?

This should work:

git fetch origin pull/772/head:fix/build-timestamps
git switch fix/build-timestamps

[nexushoratio]

git fetch origin pull/772/head:fix/build-timestamps

That's what I was trying but got the couldn't find remove ref error. At least I was on the right track. I must have something misconfigured locally.

[xscreach]

git fetch origin pull/772/head:fix/build-timestamps

That's what I was trying but got the couldn't find remove ref error. At least I was on the right track. I must have something misconfigured locally.

Thats because you are fetching from your fork, not the original repository.
For that you'd need to add the original remote and then do the fetching from that remote instead of "origin"

I'd really only suggest checking out the fix/build-timestamp branch.
Just make sure your branch is based on latest master

[nexushoratio]

Thats because you are fetching from your fork, not the original repository.

I'd just figured that out. Like I said, novice. :-/

[xscreach]

Maybe, catch us on telegram, we can discuss these things better/faster there...

https://t.me/screach

[nexushoratio]

What needs to be done to move this forward?

If you confirm that this PR works properly, I will merge it

Looks good to me.

Thanks!

[xscreach]

Would it be better or more readable to call these functions from the get_current_timestamp method and have the actual building here?

[modos189]

Do you propose to get rid of get_current_timestamp(), move the code to build_timestamp() and build_date()?
Maybe. I like both options, there are pros and cons to both

[nexushoratio]

Both have a race condition. If the second ticks over between the calls to gmtime(), the results will be different.

I don't know if there is any code that depends on those values being identical.

[xscreach]

Do you propose to get rid of get_current_timestamp(), move the code to build_timestamp() and build_date()?
Maybe. I like both options, there are pros and cons to both

Well, no... do not remove the get_current_timestamp.
Rather to have the time.strftime(...) inside of the "build" methods which would get called from get_current_timestamp with the same "time" parameter and have this parameter default to time.gmtime() in case its called separately.

The point is to not calculate both when we need only one of those.

(This is just about polishing the code, I'm quite ok with the current state, it just feels weird to call method to calculate several values and then use only one 🤷‍♂️)

[xscreach]

No need to block the PR for this, but if it could be improved I'd be glad 😉

[modos189]

I've updated the code to get rid of a possible time unsynchronization within the same build

[xscreach]

Maybe just remove the return from generate timestamp, otherwise ok

[nexushoratio]

Thanks!