[18804] in Athena Bugs
Re: linux 8.4.20: htmlview
daemon@ATHENA.MIT.EDU (Greg Hudson)
Wed Feb 21 09:55:51 2001
Message-Id: <200102211455.JAA10284@egyptian-gods.MIT.EDU>
To: "t. belton" <tbelton@MIT.EDU>
Cc: bugs@MIT.EDU
In-Reply-To: Your message of "Wed, 21 Feb 2001 08:36:23 EST."
<Pine.GSO.4.30L.0102210828190.23545-100000@department-of-alchemy.mit.edu>
Date: Wed, 21 Feb 2001 09:55:44 -0500
From: Greg Hudson <ghudson@MIT.EDU>
> So which is better: a switch for htmlview to always force "HTML
> mode" for those who want it, or to have htmlview make assumptions
> based on the document suffix (i.e. if it's .htm or .html)?
There is no reason for htmlview to make decisions based on the file or
URL suffix. lynx already does that.
For http: and https: URLs, there is already a much better way to
decide what the file type is than anything we can do; the web server
will tell us. We definitely shouldn't be passing -force_html in that
case.
For local files, the issue is murkier. The browser has less
information to work with, and it's unclear why someone would use
htmlview on a local file if they didn't want to view it as HTML. But
there are reasons why the latter might happen (the filename might have
been entered in a file-or-URL context), and I'm still leery of blindly
passing in -force_html.
For ftp: URLs and the like, the issue is similarly murky but less so.
The browser still has less information to work with, but it's much
clearer why someone might want to use htmlview, as a standard pager
won't work.
> The latter is truly horrid, but I also wonder if having a switch
> (which few people will know or remember to use) is adequate ... if
> this is a severe problem.
A command-line switch should be quite adequate. htmlview is most
commonly used by scripts and Athena programs, not end users. Also, it
is most commonly used to view URLs (it would have been more aptly
named "urlview", in hindsight) for which a switch would be
unnecessary. The bug Jake is reporting arises from the specific case
of using htmlview as a viewer for a local file known to be an HTML
file, and a switch for that case makes good sense.
> See, I'm not clear on what Lynx tries to do to an HTML document when
> it displays it and *isn't* in HTML mode. I only run Lynx
> occasionally, so forgive me if I'm missing something obvious ....
It acts as a very fancy pager.
