HOWTO: Get started, gain macports-foo, make bad first impression

[Rainer Müller]

Ryan Schmidt wrote:
> The other problem with parallel builds is that they are not identical  
> each time you run them. The build may succeed 3 times, then fail the  
> 4th. I personally haven't had the time to try to build each of my  
> ports several times with parallel build on to see if they repeatably  
> build correctly.

If the port does not build correctly the dependencies in the Makefile 
are wrong. If it ever fails for someone and he/she files some ticket it 
can be disabled. Usually there is no need to test it multiple times or 
something like that.

Locally I use a patched version since this option was introduced which 
has "use_parallel_build yes" as default. And so far port failed only on 
one occasion due to parallel building which was the python frameworks.

And as a side note, I am still in favor of making parallel building 
opt-out and enable it for every port when buildmakejobs is set to a 
value greater 1. Of course the default value of buildmakejobs should 
still be 1 because this really depends on the machine it runs on.

[snip]
> Maybe we should revisit this topic, though like you say, if we get  
> binary packages, then we don't need build success/failure statistics  
> from each user, but only from the build servers.

As I said before in other emails, we will not be able to provide 
packages for each and every variant combination (that would be 2^n for a 
port with n variants). So building custom ports on the end-user side 
will still happen even once we ship binary packages for default_variants.

Rainer