The converter uses the GPL licensed library java-getopt to do some command line parsing.
If I don't use the command line I am still "suffering" from the GPL dependency. In my opinion it would make sense to separate the "core" library and the command line processing.
What do you think?
Regards,
Thomas
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Thanks for your interest in this program – I really appreciate that.
I reckon that a core library should have the fewest possible number of dependencies.
The skeleton jar you talk about would probably have one unique dependency with commons-logging (instead of three at present) which sounds good.
At the outset I wanted the converter to remain as simple as possible and that meant for me to offer one unique multi-purpose downloadable library at a time to keep the Sourceforge page as straight as possible.
It might sound kind-of simplistic model - not to say primitive - but I have a special concern for non-programmers who might be interested in the library and who might get confused by several downloadable jars in a row (even though I don’t even know if non-programmers will ever be interested in such a tool).
If I'm not mistaken you don't use the command line interface and java-getopt-1.0.13.jar happens to remain unused on your platform. Hum - it's 56K only of wasted bytes – is this worth the effort?..
Anyway I guess soon or later someone will most certainly complain about conflicts that the dependencies create with other local jars (for instance a collision between commons-logging-1.1.3.jar and another local commons-logging library).
An uber-jar would do the trick I guess, but a core jar like the one you suggest would certainly alleviate that collision risk for those who are only interested by a plain Java interface (but wouldn't completely suppress the risk because of commons-logging I imagine).
Hum - maybe a custom Maven assembly making it possible for programmers to generate the core library plus another one to generate an uber-jar, and both of them shipped in a supplemental.zip file for experts to download with the next version?
OK, OK, you got it, I push your idea on next version's ToDo list [V1.1.0 not expected before the next alignment of the 9 planets I'm afraid - Time flies by so quickly].
Regards,
Laurent
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
The converter uses the GPL licensed library java-getopt to do some command line parsing.
If I don't use the command line I am still "suffering" from the GPL dependency. In my opinion it would make sense to separate the "core" library and the command line processing.
What do you think?
Regards,
Thomas
Hi Thomas,
Thanks for your interest in this program – I really appreciate that.
I reckon that a core library should have the fewest possible number of dependencies.
The skeleton jar you talk about would probably have one unique dependency with commons-logging (instead of three at present) which sounds good.
At the outset I wanted the converter to remain as simple as possible and that meant for me to offer one unique multi-purpose downloadable library at a time to keep the Sourceforge page as straight as possible.
It might sound kind-of simplistic model - not to say primitive - but I have a special concern for non-programmers who might be interested in the library and who might get confused by several downloadable jars in a row (even though I don’t even know if non-programmers will ever be interested in such a tool).
If I'm not mistaken you don't use the command line interface and
java-getopt-1.0.13.jarhappens to remain unused on your platform. Hum - it's 56K only of wasted bytes – is this worth the effort?..Anyway I guess soon or later someone will most certainly complain about conflicts that the dependencies create with other local jars (for instance a collision between
commons-logging-1.1.3.jarand another local commons-logging library).An uber-jar would do the trick I guess, but a core jar like the one you suggest would certainly alleviate that collision risk for those who are only interested by a plain Java interface (but wouldn't completely suppress the risk because of commons-logging I imagine).
Hum - maybe a custom Maven assembly making it possible for programmers to generate the core library plus another one to generate an uber-jar, and both of them shipped in a
supplemental.zipfile for experts to download with the next version?OK, OK, you got it, I push your idea on next version's ToDo list [
V1.1.0not expected before the next alignment of the 9 planets I'm afraid - Time flies by so quickly].Regards,
Laurent