viewing <> constructs via the web archive

viewing <> constructs via the web archive

[Daniel Karipides]

My apologies if this has been covered before, but (at least for me)
I can't see any of the include file names or template parameters
when I view archived messages via the web.  This is due to the fact
that the <stdio.h> in "#include <stdio.h>" gets interpreted as an
unknown HTML tag and is not printed.

A quick fix would be something like this:

#include <<BLANK>stdio.h>

Note that BLANK can be replaced by any character.  A script could
easily be written that inserts this text (by replacing all <'s with
<<BLANK>, for example), however I can think of two potential problems.

(1) If the archived mail is viewed outside of a web browser, this
is not an acceptable solution.

(2) The simple script wouldn't work on templates of templates...

Is there any way around this problem?  At the moment I "view source"
a lot...but this gets tiresome.

Thanks in advance,

-Dan
----
karipid@mae.cornell.edu

Re: viewing <> constructs via the web archive

[Per Bothner]

> This is due to the fact that the <stdio.h> in "#include <stdio.h>"
> gets interpreted as an unknown HTML tag and is not printed.

The correct fix would seem for the script be to replace all instances of
'<' by '&lt;'.  The characters '&' and '>' should also be replaced the
same way.

Of course such a conversion needs to be aware of the current mine type.
If the type is html, no conversion should be done, and if the type
is richtext, the tags should be converted to html.

This is arguably a bug in hypermail (or whatever software is used),
but then I am told hypermail is full of bugs, and should be replaced
by something that is actually being maintained.  (Jason Molenda
suggests somebody should look into MHonArc.)

	--Per Bothner
Cygnus Solutions     bothner@cygnus.com     http://www.cygnus.com/~bothner

Re: viewing <> constructs via the web archive

[Richard Henderson]

On Fri, Apr 17, 1998 at 07:06:50PM -0700, Per Bothner wrote:
> > This is due to the fact that the <stdio.h> in "#include <stdio.h>"
> > gets interpreted as an unknown HTML tag and is not printed.
> 
> The correct fix would seem for the script be to replace all instances of
> '<' by '&lt;'.  The characters '&' and '>' should also be replaced the
> same way.

This was explicitly disabled so as not to horribly mung patches
sent to the list.


r~

Re: viewing <> constructs via the web archive

[Per Bothner]

On Fri, Apr 17, 1998 at 07:06:50PM -0700, Per Bothner wrote:
> The correct fix would seem for the script be to replace all instances of
> '<' by '&lt;'.  The characters '&' and '>' should also be replaced the
> same way.

Richard Henderson <rth@cygnus.com> replied:
> This was explicitly disabled so as not to horribly mung patches
> sent to the list.

I am not saying we should mung patches sent to the list;  I am saying
that you have to mung text when you convert it to html to be stored
in a mail archive.  Not doing the munging *loses* information,
because you can no longer tell html directives from text.

If the goal is that you want to be able to use an html file as
direct input to patch, well I am sorry, that is just plain bogus.
You can't do that.  Use the original un-html'd mail message,
which is still available in the base (non-hypermail) archive.

Alternatively, it is easy to write a filter for removing html markup,
recovering a file suitable for patch.  Of course, that requires a
little bit of care in the original html conversion, to make sure it
is reversible.

	--Per Bothner
Cygnus Solutions     bothner@cygnus.com     http://www.cygnus.com/~bothner

Re: viewing <> constructs via the web archive

[Joe Buck]

Per writes:
> I am not saying we should mung patches sent to the list;  I am saying
> that you have to mung text when you convert it to html to be stored
> in a mail archive.  Not doing the munging *loses* information,
> because you can no longer tell html directives from text.
> 
> If the goal is that you want to be able to use an html file as
> direct input to patch, well I am sorry, that is just plain bogus.
> You can't do that.  Use the original un-html'd mail message,
> which is still available in the base (non-hypermail) archive.

If disk space is not a problem, one possibility is to save each message
twice, as .txt and .html; munge the .html version as hypermail usually
does (with respect to <, >, & etc), and have the .html version link to
the .txt version.  Alternatively, the .txt version could be saved only
if there is a patch present (the patch program has code to detect this;
I don't know if patch can be used just as a patch-detector rather than
a patch-applier.

Re: viewing <> constructs via the web archive

[Gerald Pfeifer]

On Fri, 17 Apr 1998, Per Bothner wrote:
> but then I am told hypermail is full of bugs, and should be replaced
> by something that is actually being maintained.  (Jason Molenda
> suggests somebody should look into MHonArc.)

I can second that. hypermail is basically dead, while MHonArc is
quite up-to-date with regard to current standards (MIME) and actively
supported and developed.

It's a little bit more involved to set up than hypermail, but that's
mostly because MHonArc is much more configurable.

Gerald
-- 
Gerald Pfeifer (Jerry)      Vienna University of Technology
pfeifer@dbai.tuwien.ac.at   http://www.dbai.tuwien.ac.at/~pfeifer/

Re: viewing <> constructs via the web archive

[Jeffrey A Law]

  In message <Pine.GSO.3.96.980426005006.28077G-100000@nashira.dbai.tuwien.ac.au write:
  > On Fri, 17 Apr 1998, Per Bothner wrote:
  > > but then I am told hypermail is full of bugs, and should be replaced
  > > by something that is actually being maintained.  (Jason Molenda
  > > suggests somebody should look into MHonArc.)
  > 
  > I can second that. hypermail is basically dead, while MHonArc is
  > quite up-to-date with regard to current standards (MIME) and actively
  > supported and developed.
Jason was supposed to be looking into replacing Hypermail with
MHonArc.  I don't know what the status of that conversion is.

jeff

Re: viewing <> constructs via the web archive

[Daniel Egger]

On Fri, 17 Apr 1998, you wrote:

>(1) If the archived mail is viewed outside of a web browser, this
>is not an acceptable solution.

 Why not?
 Use #define <BLANK>

--

Servus,
       Daniel