Not sure if I’m willing to accept that 4% per year is not huge. Accumulated over 10 years, we’re at a bit under 50%, which is an okay number just for running the compiler again?
Yes, buying new hardware gives more, but a) to fully use the hardware, we probably need to recompile anyways, b) it’s way more expensive than recompiling.
cr | 23 hours ago
Compilers never cease to amaze me, what clever thinking went into optimization
I’m wondering if some old software compiled by modern compilers will get a huge performance boost
lcapaldo | 20 hours ago
If you believe https://proebsting.cs.arizona.edu/law.html the answer is no, at least depending on “old” and “huge”.
skade | 19 hours ago
Not sure if I’m willing to accept that 4% per year is not huge. Accumulated over 10 years, we’re at a bit under 50%, which is an okay number just for running the compiler again?
Yes, buying new hardware gives more, but a) to fully use the hardware, we probably need to recompile anyways, b) it’s way more expensive than recompiling.
fanf | 19 hours ago
Evidence for Proebsting’s law is mixed: