Re: Thoughts on limiting the parallel build load, suggestion for a new "-j auto" option

[R. Diez]

> -l is utterly useless. Load average is computed too slowly; by the time
> it passes any useful threshold the actual make load will have spiralled
> out of control.

If that's the case, exactly that reasoning should be written in the 
documentation.


> Go and shoot the morons who wrote those scripts. The make -j value
> should never be encoded in any file; it should only ever be set by the
> user invoking the actual top level make command.

Again, such a recommendation should be in the manual.

However, for practical reasons, I still think GNU Make should offer some 
simple heuristic for people without the necessary time or inclination to 
calculate their own optimum value (that would be most humans).

That simple heuristic need not be totally fixed. The user could provide 
their own factor for a simple "CPU * <factor>" formula, something like 
"-jcpufactor 1.5" could do the trick. I just want to spare the user the 
effort to retrieve the CPU count, which is always a small, 
system-dependant pain.

Please copy me on the answers, as I'm not on this list.

Regards,
  R. Diez