[Top][All Lists]

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

Re: lynx-dev Lynx 2.7.1 and 2.8 refuse to render certain HTML documents

From: Bela Lubkin
Subject: Re: lynx-dev Lynx 2.7.1 and 2.8 refuse to render certain HTML documents
Date: Thu, 7 May 1998 07:36:51 -0700

Foteos Macrides wrote:`

> Foteos Macrides <address@hidden> wrote:
> >Bela Lubkin <address@hidden> wrote:
> >>[...]
> >>Why not use '\', i.e. the SOURCE command?  That makes the SOURCE command
> >>more symmetric -- a general toggle between handling data as text/html
> >>vs. text/plain.
> >
> >     That's already overloaded for such a purpose.  [...]
>       Another thing that's wanted is a way to wrap long lines when
> source ('\') is toggled for text/html and you don't plan to pass that
> to an external text/plain viewer.  One way to handle everything that's
> wanted without using up another command key is to have '\' invoke a
> prompt which offers you the relevant options when the document is
> thought to be text/html versus text/plain based on a Content-Type
> header, suffix map, or default assumption.  Since the source toggle
> isn't used really often, interposing a prompt wouldn't be a terrible
> pain in the neck (no worse than the clicking for pull-downs to make
> homologous choices with GUI browsers :).

I like that suggestion very well.

As to the whole caching issue:

  1. David Wooley says "This is better done with an external cache,
     which is easy for Unix."

     Um.  I use Lynx on about 20 different Unix systems.  I own two of
     them; I have grudgingly granted root access on one other; I share
     several others with co-workers and have root access but should be
     careful; and I use about a dozen which are maintained by a jealous
     and incompetent IS staff.  There are network issues which would
     preclude setting up a single proxy server of my own and using it
     from all the others.

     Lynx's lack of a document source cache causes me real, ongoing,
     not-easily-solved pain.

  2. I *FREQUENTLY* do things which cause Lynx to re-get a page, because
     it has failed to cache the source.  '\' isn't the only thing.
     There's '^V' (try-other-DTD); '*' (toggle image links); "'" and "`"
     (comment handling); changing the assumed charset; and several other
     such actions.

  3. Even if I used a proxy, it wouldn't be as fast as an in-core cache
     inside Lynx's memory space.  The machines I'm using are mostly
     memory-rich.  I don't want the "cache" to involve disk or network
     access.  Lynx's internal cache of rendered pages is quite
     satisfactorily fast; I want the same performance for changes in my
     requested rendering strategy.

The implementation I'm hoping for is simple.  Lynx currently caches
rendered documents.  I want it to _optionally_ cache source documents as
well.  The existing structure becomes two.  The source cache is very
simple, because it should be a pure bytestream as received from the
server, suitable for re-parsing by any newly requested strategy.

This should be optional, so the user can choose whether memory or
re-rendering speed is more important.

The option of keeping only the source might also be offered.  Of course
the *current* document would be in both forms, but non-current documents
might be kept only in source.  I'm not sure the difference would be
noticable on a fast CPU.  That might also offer a memory advantage: I
suspect the parsed representation is much larger than the source
(because it's stored in complex linked structures).


reply via email to

[Prev in Thread] Current Thread [Next in Thread]