Re: lynx-dev your 2 weeks' window

[David Woolley]

> 
> If the browser cannot cache at least the page it is presently showing, so that
> display of source doesn't have to hit the open 'net, I consider it broken. The

It wouldn't hit the open net with a local cache.

> fact I can work around it (and have, since 2-4-2) doesn't make it acceptible,
> it only means I was too lazy to say anything about it.

>   * we ship for lots of platforms... what is easy in one may not be in 
> another.
>     Where do we UNIX types get off telling the VMS, Mac and Windows people 
> what
>     is easy to make their OS do seperately?

The CERN cache is based on largely the same code as Lynx, it ought to be 
as easy to port as Lynx.  The only one reservation I have is that I can't
remember whether it stores last accessed date explicitly, or relies on
the filesystem.  In fact, by using the Lynx version of libwww, you might
be able to bring it up to date with HTTP 1.1.

>   * if an end-user cannot use the app barebones, he or she may never bother 
> to 
>     again.  (Willingness to continue using, or try again to use, a 

You are then competing with Microsoft, who produce low capability but
easy for first time user applications.  I don't think Lynx can really
claim to be competing in the same market; people who are after ease of
first time use are always going to go for the GUI system.

If you want plug and play, create a Linux system with a PPP and
ethernet configurer and include Lynx and a cache in it.  It should
come out something like 2 - 3MB as compressed archive, about an order
of magnitude less than just the MSIE 4 part of a Microsoft system.
You can put the cache in RAM disk or purge it every time if you don't
want them to worry about disk space management.

Alternatively, you can create a bundle of Lynx and cache.  Unfortunately,
this won't plug and play for most people. Please, though, don't just keep
bloating Lynx.