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


This page is part of the web mail archives of SRFI 103 from before July 7th, 2015. The new archives for SRFI 103 contain all messages, not just those from before July 7th, 2015.

On Thu, 2009-10-08 at 07:02 +0300, Abdulaziz Ghuloum wrote:
> On Oct 7, 2009, at 11:23 PM, Derick Eddington wrote:
> >>
> >> SRFI-59 and SLIB use the environment variable SCHEME_LIBRARY_PATH to
> >> configure their single Scheme-source library directory.  If SRFI-103
> >> appropriates SCHEME_LIBRARY_PATH for its use, it will cause SRFI-59
> >> and SLIB supporting implementations (of which there are many) to  
> >> fail.
> I'm not sure there is a conflict in practice if this SRFI uses the name
> SCHEME_LIBRARY_PATH since the intersection of users/implementations that
> use both R6RS and SLIB is practically 0.0.
> Also, this SRFI specifies that files with extension ".sls" are used in
> these paths.  So, if there's this one person in the world who might be
> using both R6RS and SLIB, that person can still put both libraries and
> SLIB files in the same place with no problems.

SRFI-59's reference implementation's SCHEME_LIBRARY_PATH can be only one
path, but this SRFI's can be multiple paths.  If something using
SRFI-59's gets a (semi)colon-separated sequence of paths, it will break
trying to use it as a single path.

> The whole idea of using "SCHEME_LIBRARY_PATH" is to say where your
> Scheme files are.  

I agree, and I think it would be better if both SRFI-59/SLIB and this
SRFI could use it.  But that doesn't seem possible.

> >> Please don't adopt the environment variable SCHEME_LIBRARY_PATH.
> I think SCHEME_LIBRARY_PATH adequately describes what this SRFI is
> about and does NOT describe what SRFI-59 is about.  Plus, in that
> SRFI, this name is only mentioned in the reference implementation
> and is nowhere in the text.

Persuasive points.

> > Okay, this SRFI will use a different environment variable name.   
> > Should
> > it be SCHEME_PATH?  Or something else?  SCHEME_PATH, being shorter,
> > seems even more prone to conflict with others choosing it.
> Definitely not SCHEME_PATH because you don't want to go chasing after
> everybody who comes after you and wants to use SCHEME_PATH (which is
> Aubrey's situation here for picking a name that does not adequately
> fit the purpose of that SRFI).

Yeah, it needs to be something which will not require chasing-after
forever into the future.  Something that, if someone else thinks of
using it, the semantics should be compatible, because they should match
the name.

> Frankly, the fact that SLIB did not use a proper name like SLIB_PATH
> or SLIB_LIBRARY_PATH or SLIB_VICINITY_PATH or whatever doesn't change
> much about what name this SRFI should use.

Part of me is inclined to agree, but the other part doesn't want to
conflict with widely established SLIB.

> You can call it R6RS_LIBRARY_PATH but I won't be happy with it.

I don't want the name to imply it's specific to a particular dialect of
Scheme, because this SRFI intends to possibly be used by other dialects.

> I would keep it as it is because SCHEME_LIBRARY_PATH is appropriate
> for this SRFI.

I don't know what to do.  I invite more discussion/arguments/fights
about this and I'll arbitrate.

or ??? could be used...

: Derick