lua-users home
lua-l archive

[Date Prev][Date Next][Thread Prev][Thread Next] [Date Index] [Thread Index]


I have the perception that >90% of the rocks available on the primary
repository provide modules and <10% provide applications, and the
suspicion that 10% is a very generous estimate.

Yet I find it hard to believe that 90% of the time of a typical Lua
programmer is spent in developing tools and 10% in developing
applications.

I agree to almost everything that has been said to this already. However I'll add another thing: as far my observation goes -- without doing statistics -- this is a common phenomena across languages. So in Lua it's due to arguments posted just a little stronger as elsewhere.

I also agree with the points you brought up, and would just formulate them differently. For code to be usefully shareable it needs to be generic enough so it's useful in a various setting. Much code is written on the fly for the one use case in a specific setting, it doesn't even need to be confidential, but when you write for example some custom script to pull your data from and older database to convert it to a newer, this is not code that adapts to other use cases.

Libraries as building blocks are way easier to make general enough to get a coder audience, applications are a way, way harder job to make them general enough so they work in various settings / systems / setups etc. For the coder doing some custom stuff it may sooner seem beneficial to group some generic code in a library, share it, and possible get some feedback or even bug fixes.

( This argument is based on the idea from I think chapter 1 of "The mythical man-month" )

PS: Lua offering itself as extension language is itself taking the "library only" approach being generic / useful as a extension library, avoiding the harder task to be offered as a full standable scripting application / eco-system.