PacketHistory and memory efficiency?

[Marek-mk]

Hello.
I'm new here. I see sophisticated language constructs in code of PacketHistory.xxx (the little comlete thing), and as I get to know the code almost everything seems to be written that way, what completely explains why nodes are crashing later or sooner in cities with >200nodes that try to be reachable.
If I change one little configuration constant usually something breaks. Something which have only the system memory management in common.
So am I wrong - reagrding PacketHistory - to be efficient in allocationg 16Bytes (as stated in comment inside sources) used are techniques that need to allocate 40Bytes of heap per every new 16Byte element?

I try to write this module from scratch, but I don't know how to properly commit.
Can someone point me to guide to easy steps of how to contribute I will be glad to create pull request.

Regards, Marek

[GUVWAF]

Hi Marek,

So am I wrong - reagrding PacketHistory - to be efficient in allocationg 16Bytes (as stated in comment inside sources) used are techniques that need to allocate 40Bytes of heap per every new 16Byte element?

Yes, indeed due to the unordered set, the 16-byte PacketRecord might actually take up quite some more heap.
Would you be able to briefly explain how you would implement it?

To properly commit, you have to create a fork of the repository: https://docs.github.com/en/pull-requests/collaborating-with-pull-requests/working-with-forks/fork-a-repo

Then do your changes there, push to your fork and create a pull request: https://docs.github.com/en/pull-requests/collaborating-with-pull-requests/proposing-changes-to-your-work-with-pull-requests/creating-a-pull-request-from-a-fork

[fifieldt]

Yes please!

[Marek-mk]

My first pull request. Comments appreciated.
#7034
Regards /Marek

[Marek-mk]

Ok, I'm a bit stuck. I adopted 2 of 3 commented changes, and commited them. Some actions run on github, and I'm confused what's happening, and what's not. A hard first timer.
(I also managed to carefully break github's tutorial on 'reviews'. If something wants to break, it will break on me the first time I touch it ;)

But - I have up-to-date version of upstream on github and in VScode, and with my commits went thru too to my branch - so it seems good.

And the question - to keep my private mx_LOG_.. loggers handy (for me when needed), while not convoluting commited code - how to do it the proper way of not losing them completely for my purposes?

And when I do this - what are next steps?

Regards /Marek

[GUVWAF]

Yes, committing and pushing to the same branch will update the pull request. Seems to be working fine!

If you want to keep your changes, you could create a new branch at this point. Then only remove it in the branch you created the pull request.

[Marek-mk]

Thanks for guidance! :-)