simon <si...@simonfoley.net> added the comment:

Thanks 

I have found teh root cause of the problem ...

--with-openssl=[my_dir]

The configure scripts has an assumption you are compiling against a binary 
packaged version of openssl and that there is a /lib folder under [my_dir]. 
This simply does not exist under any of the source code releases of openssl. So 
after I compiled the openssl source code I had to create the lib folder under 
my openssh build directory and symlink the *.so libraries there for the 
configure script to work

This is still an issue even if you edit Setup correctlty to compile the module.

>> This is a problem for people like me who are institutional users that have 
>> cross platform enterprise softwre deployment platforms (e.g. BladeLogic). 
>> There are restricted policies on what packages you can install on a server. 
>> In most cases especially for in house developed software) you need to build 
>> all dependencies seperatly and bundle them into a package (e.g. /opt RPM) 
>> that includes all required depencencies rather than rely on distribution 
>> library packages that are hard to manage at an Enterprise level and where 
>> you may be sharing the same OS.


To make the code more robust should it not 1st check under the root of [my_dir] 
before assuming [my_dir]/lib exests or at least report teh full path with the 
/lib added onto teh end of {my_dir} so you know where confiure has gone wrong ?

Is this not a fair expectation?

no lib folder
checking for openssl/ssl.h in /home/BD7046/openssl... no
checking whether compiling and linking against OpenSSL works... no

with lib folder
checking for openssl/ssl.h in /home/BC7046/openssl... yes
checking whether compiling and linking against OpenSSL works... yes


Thanks for all your help 
#PortingPerltoPython

----------

_______________________________________
Python tracker <rep...@bugs.python.org>
<https://bugs.python.org/issue34028>
_______________________________________
_______________________________________________
Python-bugs-list mailing list
Unsubscribe: 
https://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com

Reply via email to