[Ecls-list] About the myth of slow starting
Konovalov, Vadim (Vadim)** CTR **
vadim.konovalov at alcatel-lucent.com
Wed Jun 9 11:04:12 UTC 2010
> > https://sourceforge.net/news/?group_id=30035&id=287636
> SBCL has quite a large core, so unless it's cached by FS loading it
> isn't so fast.
> With caches dropped (echo 3 > /proc/sys/vm/drop_caches) ECL
> is actually
> faster than SBCL at start.
IMO playing with caches to find cases when ECL is faster is not good way to go.
Messing pure CPU speed with HDD speed is just doing numbers less relevant to actual execution speed: one can have slower HDD, or, on the contrary, disk emulated in RAM - and numbers will fluctuate
IMO more relevant numbers are those when both large SBCL core and smaller ECL core are cached by OS.
Then it is obvious that ECL spends more CPU ticks on pure-lisp boot process, compared to SBCL - this conclusion comes into mind when reading the URL. cited in initial message.
Attempts to hide a problem under some cache deviance are just non-effective.
As a side note, I consider ECL startup time as perfectly acceptable, and any speed improvements are welcomed by ECL users, obviously.
More information about the ecl-devel