This page is part of the web mail archives of SRFI 84 from before July 7th, 2015. The new archives for SRFI 84 contain all messages, not just those from before July 7th, 2015.
Neil Van Dyke writes:
> I think taxonomies for classifying libraries for human location are
> (or should be) more ephemeral than the universal identifiers of the
> libraries themselves.
I think you're right about that.
The only disadvantage I can think of for having a flatter structure is
that perhaps there might be more conflicts with people wanting the
same name for their library.
If the name of the Spiffy web server is "web:server:spiffy", then
someone else can have a Spiffy graphics drawing library if they want
to.
I agree with you that a more hierarchical or taxonomy based naming
scheme will probably result in a greater number of name changes over
time.
I expect some amount of library name changes is inevitable.
Library name changes might not be a problem to the extent we can avoid
reusing the old name.
If I decide to change the name of my library, perhaps I could release
a version of the library under the new name and a version under the
old name which forwards imports to the new library name:
;; forward old to new
(library andrewwilcox.name::neato r6rs
(import new-place::neato)
(export a b c d e f ...))
If something like this would be a workable solution, and library name
changes were not a big deal, then perhaps the cost of an increased
number of library names changes due to using a hierarchical taxonomic
structure would be ameliorated.
Andrew