MacOSX compile problems

Bastian Doetsch bastian.doetsch at
Sat Jan 17 20:02:17 CST 2009


First I'd like to thank you for trying to solve this :).  
Unfortunately, I still haven't found a solution.

> Mads is looking at the latest unstable Mercurial which does things a
> little bit differently. I'm pretty sure you're using the stable branch
> as you still seem to have a

I'm running (afaik) latest crew-tip, so it should be the same that  
Mads uses:

$ hg log -l1
changeset:   7664:3cc74ee75b0d
tag:         tip
user:        Patrick Mezard <pmezard at>
date:        Tue Jan 13 20:28:06 2009 +0200
summary:     diffstat: don't fail on merges

> The old way involved calling mercurial/ which ran  
> 'hg
> identify' and then recorded its own identity. This was a little bit
> over-complicated.
>> hg id -it works in the mercurial directory:
>> $ hg id -it
>> 3cc74ee75b0d tip
> So now you have to figure out why it doesn't work when called from
> It almost certainly involves the environment variables  
> PATH or
> PYTHONPATH. You may also find the 'which' command enlightening. I also
> suspect if you run 'hg version' -outside- the hg directory, it won't  
> be
> very happy because it doesn't happen to have the libraries it's  
> looking
> for -in the current directory-.

Having been enlightened at work a few years ago, I already tried the  
"which" command. As I posted before, the install seems to work fine:

$ which python

$ which hg

My path is:


Actually, I can run hg version in every directory I choose - it  
doesn't complain and hg just works. Only the make/install doesn't...

> Or you could re-read my previous email.

... and come to what conclusion? After re-reading and not spotting  
anything I overlooked, it seems like I don't understand correctly,  
what you were saying.

Thanks for the help, though

> -- 
> : development and support for Mercurial and Linux

More information about the Mercurial-devel mailing list