[Nipy-devel] Enthought.traits and nipy

[Jonathan Taylor]

Hi Jarrod,

Is there an earlier mail on this thread? Don't know if I got it.
----

I think Jonathan add it to the codebase in order to simplify
installation.  I personally don't think its
necessary, but don't have any strong objection.

----

One reason I did this is for Windows -- enthon does not have a numpy 
version of traits
right now. While it is possible to build traits from enthon's svn 
repository, this seems like overkill. So, I put a copy there which I 
would like to remove when enthon has a numpy version of traits....

Anyways, I don't really care where the traits goes. For now,
I put it in "neuroimaging.extra" and it ends up being installed as 
"neuroimaging.extra.enthought.traits"

I copied roughly how it was done in matplotlib, this meant a few changes to
some of the modules, but these changes could be undone and I suppose it 
could be in the
"lib" directory, and installed as "enthought.traits".

If someone wants to do that, it is fine with me.

I have a few other questions:

1) What is the status of "lib/neuroimaging/refactoring"? It fails a 
number of its tests -- is there a plan for what has to be done there?

2) I have installed nipy on Windows, with a lot of the tests passing, 
but there is something severely wrong with the data files in the svn 
repository -- my guess is that windows SVN didn't use binary transfer 
mode or some silly thing like that. Anyone know how to fix this?

3) I'm not sure if we really need all that data in the svn repository -- 
it makes for a long download. I understand that it is nice to have it 
ultimately in the svn tree, but why is it in the
svn repository?

-- Jonathan

Jarrod Millman wrote:
> I stumbled into some problems earlier, since I didn't know we needed
> the svn version of traits as well as numpy and scipy.  And the
> installation directions for fedora specified the stable release of
> traits, which ended up causing me the problems.  But I updated the
> install page once I found out that I needed an svn copy of traits
> http://projects.scipy.org/neuroimaging/ni/wiki/DevelopmentInstallFedora
> I didn't update the other install sites since I haven't had time to
> fix them yet.
>
> I haven't had time to look into how Jonathan local copy of traits
> works in detail yet, but my first thought was that it doesn't belong
> in lib/neuroimaging/ but should be in the parent directory, lib/.  I
> thought one of the ideas of the neuroimaging subdirectory was to
> separate out the nipy specific libraries from generic stuff, which
> might end up somewhere else.  Am I missing something?
>
> Jarrod
> _______________________________________________
> Nipy-devel mailing list
> Nipy-devel at neuroimaging.scipy.org
> http://neuroimaging.scipy.org/mailman/listinfo/nipy-devel
>   

-- 

------------------------------------------------------------------------
Jonathan Taylor                           Tel:   650.723.9230
Dept. of Statistics                       Fax:   650.725.8977
Sequoia Hall, 137                         www-stat.stanford.edu/~jtaylo
390 Serra Mall
Stanford, CA 94305

-------------- next part --------------
A non-text attachment was scrubbed...
Name: jonathan.taylor.vcf
Type: text/x-vcard
Size: 329 bytes
Desc: not available
Url : http://projects.scipy.org/pipermail/nipy-devel/attachments/20060628/aa95ed41/attachment-0002.vcf