Re: lynx, pine, metamail...(mailcap&mime.types) netbase?
In message <65JqZG6zcsB@khms.westfalen.de>, Kai Henningsen writes:
>mdorman@lot49.med.miami.edu (Michael Alan Dorman) wrote on 20.03.96 in <m0tzU
>Ly-00027gC@lot49.med.miami.edu>:
>> In message <65BFTpdjcsB@khms.westfalen.de>, Kai Henningsen writes:
>> >Potential problem alert. You should probably check that nothing bad
>> >happens if you reference an X program there while not having X running.
>> I'm understand what you're saying, but you seem to be thinking of a
>> specific situation that you don't mention, so I worry that I'm missing
>> something.
>I thought I did explicitely mention the situation above. Umm, on
>rereading, I still think I did.
When I say _specific_ situation, I'm thinking along the lines of "I
had xv in my mailcap/mime.types, and I was running from a terminal
when I tried to read a piece of mail using mh that had a .gif attached
and as a result my mail folder got trashed" or some such. I would
categorize what you said as describing a _generic_ situation.
It seemed to me, though, that your generic situation might have had,
as its genesis, a specific incident in which something bad happened.
Based on this feeling---combined with the fact that I thought I might
be able to address a specific incident better than trying to produce a
proof that having some unknown X-dependent program being run under
some unknowable circumstances will never cause a problem---I said what
I said.
>> I mean, by default if you don't have X running, the program you attemt
>> to start should tell you that it couldn't get a connection to your
>> display---does this count as something bad happening? And if so, what
>> would you rather have happen?
>I'd call it "something bad" if it leads to losing data, getting the
>display trashed, or similar. If you only get an error message, that's
>probably fine. Note I said "Potential problem" and "You should
>probably check"; I don't know of a specific bad event, I'd just like
>to make sure this gets checked in advance.
The problem is, unless Debian has the resources to check every single
combination of every single viewer against every single program that
uses mime.types/mailcap, viewing every single type of potential
document, there's no way we can guarantee that "something bad" won't
happen.
It would seem, then, that the only alternative based on your criterion
is to do nothing with mime.types and mailcap. And I think you'd agree
that this is not the way to create a reputation for doing a good job
of configuring the system to best exploit the installed resources.
Mike.
--
"Don't let me make you unhappy by failing to be contrary enough...."
Reply to: