Skip to main content


#curl source code age (take 2)

How large share of the code was written after a certain date, over time.

#curl
in reply to daniel:// stenberg://

how is it possible that the graph doesn’t add up to 100% ? Has it been modified ? Is some period missing ?
in reply to Étienne Parmentier

those are the few lines we committed before 2000-01-01
This entry was edited (3 weeks ago)
in reply to daniel:// stenberg://

That's fascinating. Are all projects so evenly fresh? And what's it look like in LOC vs time vs the % shown here?
in reply to Stephen Shankland

lines of code have exploded over the decades:
This entry was edited (3 weeks ago)
in reply to Stephen Shankland

@stshank it's actually quite fascinating I think and it will be curious to see for how long it can continue., I mean logically it needs to plateau at some point!
in reply to daniel:// stenberg://

My scripts for this are of course available and open source and you can run them against your project easily.

Just needs perl, git, gnuplot - and patience.

github.com/curl/stats/blob/mas…

github.com/curl/stats/blob/mas…

in reply to daniel:// stenberg://

I like how this paints a very different picture than the percentile graph earlier.
in reply to daniel:// stenberg://

instead of percentage. Can you do it for total lines of code so we can also see the growth?
in reply to daniel:// stenberg://

"Look, if we drill deep enough we'll get to the 2011 crunch code!"

Some (paired) years seem to have a bigger impact than others on persistence in the code base - interesting plot!

in reply to daniel:// stenberg://

Cool chart! The 100% are code base today? Does it mean curl is not growing significantly since around 2005? What happened there?
in reply to Daniel Nordhoff-Vergien

@dve 100% is the entire code base of the time. It has kept growing ever since before 2000.

But that confusion risk is also why I won't keep this version of the graph, I will go with the "take 4" I posted later.