back
7 comments
All told, the entire stack is comfortably under 200 lines of code

I'm almost willing to bet that's not counting the number of lines of code needed for the additional layers of abstraction underneath. Unfortunately, we can't read more about those because all the vpri.org links seem to have 404'd since this was written 14 years ago.

...but I did find someone who looked at it in more detail, and it is not under 200 lines:

https://dercuano.github.io/notes/small-tcp.html

Would you count the compiler against their LOC?

The article describes the VPRI goal as a full system in 20k LOC, including graphics rendering, multiprocessing, etc. The infrastructure for this specific piece (IIRC, mostly Ohm) was used across many different parts of the system. So probably you can allocate some %, but certainly not all

My count there didn't count the main compiler against their LOC. IIRC the version I looked at predated Ohm.
Hmm, too bad I never reformatted that into proper Markdown.
(2008)

20 Comments, 2009 https://news.ycombinator.com/item?id=846028

Added. Thanks
Do we have access to the sauce code of the Viewpoints team?

Sounds like they had a very effective method to generate communication stacks.