On 04/19/2015 11:42 PM, ggrafendorfer wrote:
> On Sunday, April 19, 2015 at 8:28:49 PM UTC+2, leif wrote:
> 
>     On 04/19/2015 06:42 PM, John Cremona wrote:
>     > Do you have a reason for having SAGE_ROOT set at all?  I don't.
> 
>     That's exactly the purpose of a *user-set* SAGE_ROOT:  To run a Sage
>     installation *somewhere else* in the filesystem hierarchy.
> 
>     It's probably a bit surprising that that even works with (e.g.)
>     './sage', but in principle intentional.
> 
> 
>     -leif
> 
> 
> Seems like we have a different understanding of the proper use of
> environment variables.
> If I type
> 
> sage-6.6$ make ptestlong
> 
> then I expect that the test is done with the "new" sage-6.6, which can
> be found in the same directory as the makefile, no matter if SAGE_ROOT
> is set (to whatever) or not.

Agreed, /that/ "use case" is rather unintended.  We'd have to check
whether SAGE_ROOT is set and points somewhere else in the Makefile to
avoid such potentially surprising behavior.  (It's less obvious that
'sage' is called from the Makefile and hence SAGE_ROOT is used /
relevant there at all.)

But e.g. './sage -t ...' (or 'sage -t ...') should IMHO still "respect"
SAGE_ROOT if it is set (since that's its purpose).


I wouldn't mind if you opened a ticket for disallowing / disabling
SAGE_ROOT in case the current working directory happens to contain a
different Sage installation though (which also covers the 'make' case).
 Presumably not as easy to implement as one might think in the first place.


-leif


-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To post to this group, send email to [email protected].
Visit this group at http://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Reply via email to