lynx-dev
[Top][All Lists]
Advanced

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

Re: LYNX-DEV numbered forms patch update


From: Klaus Weide
Subject: Re: LYNX-DEV numbered forms patch update
Date: Fri, 2 May 1997 01:01:23 -0500 (CDT)

On Thu, 1 May 1997, Laura Eaves wrote:

> Well, if I were to grab fote's changes and incorporate them, that would add 
> more
> work for me with no guarantee it would get into the devel code.
> What I'm wondering is why fote isn't making his mods directly to the devel
> code (or at least basing his mods on the devel code) instead of maintaining
> a separate version.  It makes it hard to know what version to base my mods on.
> Is Fote's code significantly different from your code?  

The development code adds chartrans, Wayne's patches, auto-config and
related changes, slang-like color with curses, two ways of parsing,
the fragment stuff Fote sez is "bugs", several other changes and
additions (Wayne's recent EXTERN, changes for WHEREIS different from
Fote's, ...), and fotemods has changes which are not (some maybe "not
yet") in the devel code -- you get an idea.

If you want to know exactly, you can alway grab both (non-)versions
and do a `diff -r'.

> I know you were both
> working on parsing problems at one point.  I thought you had merged the "best
> of both worlds".  

The devel code has Fote's parsing changes as an alternative method,
switchable with '^V', yes.  (It is not really appropriate to call that
"old parsing" or "Old DTD", since it isn't, but I haven't come up with
anything better.  Suggestions?)  

Since you are already using that code, you should test it a bit. :)

> Perhaps you should start working off the same base.

Fote has explicitly asked to not be considered an active developer,
whatever exactly that means, so accepting that, all we can do is
consider his changes for inclusion (similar to other contributions,
like yours) as they come.  Up to now this has been somewhat easier
than it might be, because his code is already on the same machine
and because of using a version control system, but as the number of 
changes increases it gets more difficult, takes a longer time, etc.

Now if you feel impatient about that, you could try to merge Fote's
latest changes with the devel code on your own. (I am not saying you
should, or even that it would be a good idea, just that you have the
possibility since the code is there to grab and modify and offer back) :) 
Alternatively, and this might be much simpler and also useful, you
could download Fote's code not _instead of_, but in addition to the
devel code, keep it separate, and test the new features, then report
if there are problems and how it works.  That might help Fote to find
problems and fix them (should there be any) before I or somebody else
gets a chance to look at his code.

> > From: Klaus Weide <address@hidden>

> > BTW I notice your patched version is sending User-Agent headers
> > of the form:
> >
> > Lynx/2.7.1ac-0.15 + numbered_forms patch + highlighting fix (4/27/97) 
> > libwww-FM/2.14
> >
> As for the long agent string --
> I put that string in LYNX_VERSION in userdefs.h so when users found
> bugs they could type "lynx -version" to send me the version number.
> I was not aware that LYNX_VERSION was also incorporated into the agent string.
> Is there some syntax that has to be used in the agent string?

The syntax for the User-Agent header is defined in the HTTP/1.0 and
HTTP/1.1 RFCs.  As I said, just put parentheses around your additions
(so that they will be comments), unless you want to make them valid
"product tokens" in the sense of the specs.

> Would this syntax cause anything to fail that wouldn't otherwise?

Probably not since tool which interpret that header cannot rely on
valid syntax anyway.  It's still nice to do The Right Thing,
especially when it is no extra work to do so.

   Klaus

;
; To UNSUBSCRIBE:  Send a mail message to address@hidden
;                  with "unsubscribe lynx-dev" (without the
;                  quotation marks) on a line by itself.
;

reply via email to

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