Re: lynx-dev "gettidy.sh": why must one define proxy before evoking Lynx?

[Klaus Weide]

On Mon, 16 Aug 1999, Henry Nelson wrote:

> > > but HN may still have a problem which has a small programming solution.
> > 
> > Nobody has ever _asked for_ recognition of arbitrary *_proxy in lynx.cfg.
> > It seems all normally used foo_proxy settings are recognized.  The
> > need for arbitrary_foo just hasn't arisen.  The xhttp_proxy, xxhttp_proxy
> > etc. are a trick that happens to work (across many lynx versions), nothing
>              ^^^^^^^^^^^^^^^^^^^^^^^^^^
> I'm sorry now I even asked, but I do like to ask on the list since others
> might benefit or get a hint.

Fully agreed, and why should you be sorry.

> I don't "have a problem", but I do have
> something I want to do that has a "small programming solution."  The point
> is though that it is all based on a "trick that happens to work."  We,
> well, I at least, don't want this kind of stuff going into a general-use
> Lynx.

You lost me here - are you still talking about anything related to
"gettidy.sh" or the foo_proxy "trick", or is that "something [you] want
to do" something completely different?  It is not clear to me what you
mean with "this kind of stuff" (nor, actually, with "going into" - being
documented?).

> If it happened to turn out to be really useful, I would announce it,
> but I would strongly oppose any efforts to integrate it into Lynx.  I,
> and I hope a majority of Lynx users, want a program that follows standards,
> not a Netscape clone, or more recently, an MSWORD clone.

But for many things there are no standards (or even "standards"),
or standards can be ambiguous, or there are several to choose from.

The "trick that happens to work", in the concrete case of foo_proxy
variables, doesn't violate any standards that I am aware of.  It just
isn't officially sanctioned by any documentation, so in theory it
could stop working in dev.Tomorrow.  If it happens to turn out to be
really useful, the 'non-official' status can be changed (without
violating any standards afaik) or some new interface (configuration
option or whatever) created.

> Seems to me I heard someone say in a very recent post: "we seem to be
> suffering from fidgety programmers."  I couldn't agree with you more.

You seem to have a very general uneasiness (to put it mildly) with
the way some things are done.  It would help more to talk about concrete
cases.

   Klaus