Hi, While this is probably a bit generic Perl question, I do intend to use ActivePerl and PerlApp so I figure I can sneak it in here...:-)
My objective is to use ActivePerl+PerlApp to build an app from Perl source in at least three flavors (Windows, Linux, Solaris). On the theory that I'd like to keep strict control on configurations, and that 'less redundancy is better', Ideally, I'd like it all to reside in CM controlled space - i.e. 1) an ActivePerl for each platform, 2) a PerlApp for each platform, 3) a single PERLLIB area (for all the modules I expect to download and install) and 4) our own source code. Beginning from the back, #4 is a no-brainer of course. Leaving #3 for a second, I believe #1 and #2 is simply to messy to handle this way. If anyone has information to the contrary, I'd be interested to hear about it, especially how to deal with the issue of things in the AP tree having hardcoded paths to the installation (as this will be incorrect using a CM system where the tree can be checked out anywhere). Also, would it be possible/easy to factor out just the differences between the things? On #3 however, I wish to confirm that I have understood it correctly that a single tree can be shared. As I've understood it, this will work due to 1) most modules are just plain text files anyway, and 2) modules that need binaries will have their stuff installed in separate dirs by platform. Correct? So for a 'simple' module, I should be able to install it once and then use it for any platform. For 'complex' modules, I'd need to compile it and install it on each platform. Would there be any caveats regarding line endings here? ...or should I just forget the whole thing and just maintain a Perl installation for each platform, installing a module in all installations...? Any advice and opinions welcome...! TIA, ken1 _______________________________________________ ActivePerl mailing list [email protected] To unsubscribe: http://listserv.ActiveState.com/mailman/mysubs
