[reportlab-users] Installing ReportLab on Mac

Reed O'Brien Reed.L.OBrien at nasa.gov
Fri Jan 2 10:16:24 EST 2009


On Jan 2, 2009, at 9:41 AM, Andy Robinson wrote:


> 2009/1/2 Dirk Holtwick <dirk.holtwick at gmail.com>:

>> Just to drop another name into the round I would like to suggest to

>> have a

>> look on "pip" ;) Maybe this could at least be a solution for some

>> dependency

>> troubles.

>>

>> http://pypi.python.org/pypi/pip

>

> This actually looks to me like a big step in the right direction. I

> wish setuptools had worked this way from day one. And if it got

> popular we could add a 'setup_pip.py'. But we won't worry about it

> for our 2.3 release. I'm taking the view that having a distutils

> "setup.py" is the python standard until Guido says otherwise; but we

> will happily include alternatives contributed by others for them to

> use at their own risk.

>

> IMHO "Real Men" unzip the package, put it on their path, and know what

> extensions they are compiling. Unix should seek to educate the

> developer. Sadly there are very few left.


Some of my, um, apps/systems, have dozens of interdependent packages
as well as several dozen python packages. It would be near if not
impossible to have half a dozen people unzip configure and compile all
of these things manually.

Additionally there are multiple target installation environments: OSX,
RHEL, Fedora (and sometimes ubuntu), Solaris and sometimes FreeBSD. I
can't imagine the level of effort required to maintain debs, rpms,
ports, fink, solaris whatevers and so forth.

zc,buildout (which has its own warts) with setuptools and a privately
maintained index allows maintaining a system that builds development
and production configurations, repeatably and reliably, for all of the
above platforms with minimal headache. I am not arguing others should
do it this way, but I thought I would share why I use non "Real Men"
methods.

~ro


>

>

> - Andy

> _______________________________________________

> reportlab-users mailing list

> reportlab-users at reportlab.com

> http://two.pairlist.net/mailman/listinfo/reportlab-users




More information about the reportlab-users mailing list