On Sun, Oct 15, 2017 at 08:41:45PM -0400, Maxim Cournoyer wrote:
> >> [0] https://golang.org/doc/code.html#ImportPaths
> 
> I just read that link, and while it's true that they recommend using the
> source repository domain as the base path of the library, it is by no
> mean an obligation, as noted:
> 
>     In practice you can choose any arbitrary path name, as long as it is
>     unique to the standard library and greater Go ecosystem.
>
> I personally fail to see how using github.com gives much more uniqueness
> to a library name (especially since I expect that most go stuff would be
> hosted there) and find it equally disturbing. How hard would it be to go
> against this de facto standard? Maybe we could have a procedure that
> would strip any domain name from the libraries import paths?

Using the domain name as part of the *upstream* library name is useful
for upstream authors because of how Go's built-in dependency management
tools work. Go integrates dependency management into the language and
the `go` tool itself. Re-using the upstream library name is useful
because they have already disambiguated for us.

I don't intend to be rude, but I'm not going to put much effort into
responding to further comments that are not based on knowledge of how Go
handles package / dependency management with its built-in tools, or
modular programming in Go, in general. Already I used tons of my free
time to learn this stuff, just so I could make Guix packages of Go
software. Please meet me where I am.

Again, I don't see an ethical problem here, so any motivation for me to
participate in this discussion, as a volunteer, must be technical. If
it's *wrong* to name the packages in this way, I will behave
differently.