Hi [email protected], Steffen Mueller wrote: > Now, renaming par.pl to something else (par-archive would be fine in > principle) is something that might break user's code. Simply renaming > the file certainly breaks even the PAR-Packer code since a copy of > par.pl is packaged into binary executables. So some patching would be > necessary there. > > I'm not entirely sure what the implications of a rename may be for user > code. I am sending a copy of this mail to the PAR mailing list, so it > can be discussed.
There haven't been any objections to this. Does this mean that renaming par.pl (with the associated modifications to the module code) should be okay? Can you think of a use case of PAR(::Packer) which relies on par.pl being named "par.pl" exactly? Splitting up PAR into PAR and PAR::Packer has caused lots of confusion and still does sometimes. I still think it was a sensible move. Similarly, I don't like the thought that the par.pl tool will have different names on different platforms, but I am afraid of adding more confusion. I'd love to have some feedback on this. I won't touch it otherwise. While I'm at it: I was away for a couple of weeks without nntp access to the mailing list. There have been quite a few interesting bug reports lately, but it will be some time before I can address them since I'm moving house. If anybody would like to start hacking the core PAR and Module::ScanDeps code, this would be a great time to get involved! Best regards, Steffen
