Re: Virtual packages (was Re: Bug#64006:)
>>>>> On 17 May 2000 10:04:04 -0700, Chris Waters <xtifr@dsp.net> said:
Chris> cwitty@newtonlabs.com (Carl R. Witty) writes:
>> Chris Waters <xtifr@dsp.net> writes:
>> > Julian Gilbey <J.D.Gilbey@qmw.ac.uk> writes:
>> > > But a package which Recommends: www-browser needs no standard
>> > > interface whatsoever, for example.
>> > I believe they all fit this template:
>> > command-line: <package-specific-program-name> <url>
>> Is there a way to run "w3" (the Emacs Lisp web browser) from the
>> command line? I don't know if there is; if there isn't, I don't
>> think that should prevent emacs20 from "Provide"ing www-browser.
xemacs -eval "(w3-fetch \"http://www.memepool.com\")"
Chris> Ok, that seems reasonable. But my point is that we should
Chris> document *whatever* it is that we expect from packages that
Chris> provide a virtual package. That way, if I have a program that
Chris> expects to be able to lauch an URL from the command line
Chris> (like, e.g. xchat), I can know whether or not the virtual
Chris> package is going to fit my needs.
Chris> If what you say is true, then a dependency on www-browser is
Chris> not adequate for xchat. And it would be nice to know for
Chris> sure, one way or the other.
Chris> My original statement: "we should document the APIs provided
Chris> by virtual packages."
Maybe when it makes sense have things that say provide www-browser
(dealt with through alternatives of course) provide a script
www-browser that does the right thing in this case
#!/bin/sh
exec xemacs -eval "(w3-fetch \"$1\")"
?
Jim
--
@James LewisMoss <dres@ioa.com> | Blessed Be!
@ http://www.ioa.com/~dres | Linux is kewl!
@"Argue for your limitations and sure enough, they're yours." Bach
Reply to: