On Sunday, December 13, 2015 at 1:31:22 AM UTC-5, Mouse wrote:
>
> IMHO the two don’t have to be mutually exclusive. Ideally, one should be 
> able to “make test” that runs *in the build directory* *before* “make 
> install”, but then one should be able to run “cryptest.exe v” *after* the 
> installation.
>
> I suspect the right way of addressing this is making “cryptest.exe” aware 
> of potential locations of “TestData/”, etc.
>
> Comments?
>

I agree with you, but I was not going to hold up things based on it.

Right now, the only place 'make check' works is in the build directory. I 
think adding CRYPTOPP_DATA_DIR is a step in the right direction because 
things will work in both cases. Users will do nothing, and it will continue 
to "just work" for them. Maintainers will set CRYPTOPP_DATA_DIR and it will 
"just work" for them.

So what would you like to do:

 (1) do nothing
 (2) add the CRYPTOPP_DATA_DIR define patch
 (3) integrate CRYPTOPP_DATA_DIR define into DataDir() class patch

(2) is what Debian and Fedora do.

(3) is kind of what we have, but not quite. We require 
"PREFIX=$CRYPTOPP_DATA_DIR" during compile and install. Its not currently 
compatible with Debian and Fedora's -DCRYPTOPP_DATA_DIR="..." during 
compile. So it will require a little work.

(3) also has a little risk associated with it. It means we are introducing 
an attacker controlled input after an install by always including CWD or 
PWD. I think the risk is minor, but its still more than was previously 
present.

Jeff

-- 
-- 
You received this message because you are subscribed to the "Crypto++ Users" 
Google Group.
To unsubscribe, send an email to [email protected].
More information about Crypto++ and this group is available at 
http://www.cryptopp.com.
--- 
You received this message because you are subscribed to the Google Groups 
"Crypto++ Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
For more options, visit https://groups.google.com/d/optout.

Reply via email to