Received: from jimi.CS.UNLV.EDU by JIMI.CS.UNLV.EDU id aa01964;
          26 Dec 96 23:13 PST
To: jay@JIMI.CS.UNLV.EDU
Subject: bug-chimera jun 96
Date: Thu, 26 Dec 1996 23:13:15 -0800
From: Jay Nietling <jay@JIMI.CS.UNLV.EDU>


------- Forwarded Messages

Received: from ttmath.ttu.edu by JIMI.CS.UNLV.EDU id aa02722; 1 Jun 96 9:17 PDT
Date: Sat, 1 Jun 1996 11:17:05 -0500 (CDT)
From: Jake Kesinger <kesinger@math.ttu.edu>
To: Erik Corry <ehcorry@inet.uni-c.dk>
Cc: Chimera Hacker List <bug-chimera@cs.unlv.edu>
Subject: Re: Does everyone have Posix regular expressions
In-Reply-To: <m0uOXwn-0003kLC@kroete2.freinet.de>
Message-Id: <Pine.SOL.3.91.960601111558.6983A-100000@ttmath>
Mime-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII

On Wed, 29 May 1996, Erik Corry wrote:

> I'd like to use the regular expression support that is supposed to
> be in your C library accoring to Posix. But I'm not sure everyone has
> it. I know Linux has, but if people with access to other OSs could
> check that they have regcomp, regexec and regfree, I would be grateful.

Has them:
    Solaris 2.4
    AIX 3.2
    NetBSD 1.0(?)
Doesn't have them:
    Ultrix 4.4

  --Jake     _    Jake Kesinger (kesinger@math.ttu.edu), Outrageous Liar
LUBBOCK -> _|*~-  https://www.math.ttu.edu/~kesinger/
           \,  _} ``It's a damn poor mind that can only think of one way
             \(    to spell a word.'' (Andrew Jackson)


------- Message 2

Received: from cygnus.com by JIMI.CS.UNLV.EDU id aa06943; 1 Jun 96 22:26 PDT
Received: from tweedledumb.cygnus.com (tweedledumb.cygnus.com [192.80.44.1]) by cygnus.com (8.6.12/8.6.9) with SMTP id WAA17103; Sat, 1 Jun 1996 22:25:56 -0700
Received: from maneki-neko.cygnus.com by tweedledumb.cygnus.com (4.1/4.7) id AA06623; Sun, 2 Jun 96 01:25:54 EDT
Received: by maneki-neko.cygnus.com; (8.7.5/1.1.8.2/20Sep95-0235PM)
	id BAA03944; Sun, 2 Jun 1996 01:25:36 -0400
Sender: eichin@cygnus.com
To: Jake Kesinger <kesinger@math.ttu.edu>
Cc: Erik Corry <ehcorry@inet.uni-c.dk>, 
    Chimera Hacker List <bug-chimera@cs.unlv.edu>
Subject: Re: Does everyone have Posix regular expressions
References: <Pine.SOL.3.91.960601111558.6983A-100000@ttmath>
From: Mark Eichin <eichin@cygnus.com>
Date: 02 Jun 1996 01:25:36 -0400
In-Reply-To: Jake Kesinger's message of Sat, 1 Jun 1996 11:17:05 -0500 (CDT)
Message-Id: <xe1ohn2lqun.fsf@maneki-neko.cygnus.com>
Lines: 8
X-Mailer: Gnus v5.2.2/Emacs 19.31

Jake Kesinger <kesinger@math.ttu.edu> writes:
> Has them:
>     Solaris 2.4

As I mentioned privately earlier, Solaris 2.4 has a function of the
correct name -- which always returns -1, even on trivially correct
input. (Kerberos V5 has an autoconf test to detect this case rather
than only checking if the function exists...)

------- Message 3

Received: from carey.CS.UNLV.EDU by JIMI.CS.UNLV.EDU id aa09225;
          5 Jun 96 14:10 PDT
To: bug-chimera@carey.CS.UNLV.EDU
Subject: progress report
Date: Wed, 05 Jun 1996 14:10:16 -0700
From: John Kilburg <john@carey.CS.UNLV.EDU>

I'm still here...

I'm doing a bit of work on the HTML code.  I hope that the next release
will support most of HTML3.2 (although there are no plans for
<applet> and <script>).

The table tag works but the attributes are ignored.  If anyone has a URL
with a table torture test, please let me know :)

There is a possibility of a release this weekend or next weekend.

							-john

------- Message 4

Received: from inet.uni-c.dk by JIMI.CS.UNLV.EDU id aa12051; 5 Jun 96 17:05 PDT
Received: from kroete2.freinet.de (arh170.ppp.uni-c.dk [130.228.13.170]) by inet.uni-c.dk (8.6.12/8.6.9) with SMTP id CAA13004 for <bug-chimera@cs.unlv.edu>; Thu, 6 Jun 1996 02:05:31 +0200
Received: by kroete2.freinet.de (Smail3.1.28.1 #4)
	id m0uRSYQ-0003jkC; Thu, 6 Jun 96 02:03 MET DST
Message-Id: <m0uRSYQ-0003jkC@kroete2.freinet.de>
Subject: Re: progress report
To: Chimera Hacker List <bug-chimera@cs.unlv.edu>
Date: Thu, 6 Jun 1996 02:03:58 +0200 (MET DST)
In-Reply-To: <199606052132.XAA03775@inet.uni-c.dk> from "John Kilburg" at Jun 5, 96 02:10:16 pm
Reply-To: Erik Corry <ehcorry@inet.uni-c.dk>
From: Erik Corry <ehcorry@inet.uni-c.dk>
X-Face:  5am4$s<`Jfx-,*w6$)uX,dDt3Z2w?4ZNGS-i@_w(Fzv%(<[(XtN*t\,OmV_a=lS
  <jt,R~:$--|={s;V(G/wsw[@S(Fv4{7`b/@Z7k_QW-s]J-<8oe*5)dg+S.CeqQ$~m]}jY$c
X-Mailer: ELM [version 2.4 PL24 ME8]
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Content-Length: 336       

> 
> I'm still here...
> 
> I'm doing a bit of work on the HTML code.  I hope that the next release
> will support most of HTML3.2 (although there are no plans for
> <applet> and <script>).

All you have to do to support HTML 3.2 is to _ignore_ applet and
script. That should be doable. :-)

That's great news on tables!

- --
Erik Corry

------- Message 5

Received: from cheviot.ncl.ac.uk by JIMI.CS.UNLV.EDU id aa15987;
          6 Jun 96 1:04 PDT
Received: from aire.ncl.ac.uk by cheviot.ncl.ac.uk id <JAA02554@cheviot.ncl.ac.uk>
  (8.6.12/ for ncl.ac.uk) with SMTP; Thu, 6 Jun 1996 09:04:29 +0100
Received: (njdc@localhost) by aire.ncl.ac.uk (8.6.12/8.6.x-cf revision 3 for SunOS 4.1.x) id JAA11791; Thu, 6 Jun 1996 09:02:00 +0100
From: "J.D.Coleman" <J.D.Coleman@newcastle.ac.uk>
Message-Id: <199606060802.JAA11791@aire.ncl.ac.uk>
Subject: Re: progress report
To: Chimera Mailing List <bug-chimera@cs.unlv.edu>
Date: Thu, 6 Jun 1996 09:02:00 +0100 (BST)
In-Reply-To: <m0uRSYQ-0003jkC@kroete2.freinet.de> from "Erik Corry" at Jun 6, 96 02:03:58 am
X-Organisation: University of Newcastle Computing Service
X-Phone: +44-191-222-8068
X-Fax: +44-191-222-8765
X-Mailer: ELM [version 2.4 PL23]
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Content-Length: 479       

John Kilburg wrote:
> > I'm doing a bit of work on the HTML code.  I hope that the next release
> > will support most of HTML3.2 (although there are no plans for
> > <applet> and <script>).
> 
Erik Corry wrote:
> That's great news on tables!

Seconded.

... waiting eagerly for the release ...

J

- -- 
"The Ravenous Bugblatter Beast of Traal!  Is it safe?"
"Oh yes, it's perfectly safe.  It's just us who are in trouble!"
- --
Wargame (n): battle simulation using painted numbers.

------- Message 6

Received: from aohakobe.ipc.chiba-u.ac.jp by JIMI.CS.UNLV.EDU id aa01981;
          11 Jun 96 5:22 PDT
Received: from aohakobe (localhost [127.0.0.1]) by aohakobe.ipc.chiba-u.ac.jp (8.7.5/3.4Wbeta3 (-: 95041617) with ESMTP id VAA13305 for <bug-chimera@cs.unlv.edu>; Tue, 11 Jun 1996 21:21:46 +0900 (JST)
Message-Id: <199606111221.VAA13305@aohakobe.ipc.chiba-u.ac.jp>
To: Chimera Lovers <bug-chimera@cs.unlv.edu>
Subject: Re: Japanese? 
In-reply-to: Your message of "Mon, 22 Apr 1996 22:20:55 -0400."
             <9604230256.AA23861@cuipc.ipc.chiba-u.ac.jp> 
Date: Tue, 11 Jun 1996 21:21:46 +0900
From: Yozo Toda (TELEPHONE +81-43-290-3539) <yozo@aohakobe.ipc.chiba-u.ac.jp>


this is a response to the very old mail...

> Is there any hope that chimera will someday be able to render Japanese text?
> That's one of the few things I still use Netscape for.

now I use DeleGate as a proxy server.
DeleGate can be configured to translate Japanese text to inline image.
sometimes occurs font size mismatch, but not so big problem.

- -- yozo.


------- Message 7

Received: from aohakobe.ipc.chiba-u.ac.jp by JIMI.CS.UNLV.EDU id aa02474;
          11 Jun 96 6:20 PDT
Received: from aohakobe (localhost [127.0.0.1]) by aohakobe.ipc.chiba-u.ac.jp (8.7.5/3.4Wbeta3 (-: 95041617) with ESMTP id WAA13458 for <bug-chimera@cs.unlv.edu>; Tue, 11 Jun 1996 22:19:48 +0900 (JST)
Message-Id: <199606111319.WAA13458@aohakobe.ipc.chiba-u.ac.jp>
To: Chimera Lovers <bug-chimera@cs.unlv.edu>
Subject: Re: Japanese? 
In-reply-to: Your message of "Tue, 11 Jun 1996 21:21:46 +0900."
             <199606111221.VAA13305@aohakobe.ipc.chiba-u.ac.jp> 
Date: Tue, 11 Jun 1996 22:19:48 +0900
From: Yozo Toda (TELEPHONE +81-43-290-3539) <yozo@aohakobe.ipc.chiba-u.ac.jp>


> now I use DeleGate as a proxy server.

sorry. forgot to add URL.  <ftp://wall.etl.go.jp/pub/DeleGate/>

- -- yozo.

------- Message 8

Received: from pchelwig1.mathematik.tu-muenchen.de by JIMI.CS.UNLV.EDU
          id aa24142; 12 Jun 96 12:36 PDT
Received: (from nikl@localhost) by pchelwig1.mathematik.tu-muenchen.de (8.6.12/8.6.9) id VAA02743 for bug-chimera@cs.unlv.edu; Wed, 12 Jun 1996 21:33:16 +0200
From: Gerhard Niklasch <nikl@pchelwig1.mathematik.tu-muenchen.de>
Message-Id: <199606121933.VAA02743@pchelwig1.mathematik.tu-muenchen.de>
Subject: Suggestions for 2.0 from a happy user
To: bug-chimera@cs.unlv.edu
Date: Wed, 12 Jun 1996 21:33:16 +0200 (MET DST)
Reply-To: Gerhard Niklasch <nikl@mathematik.tu-muenchen.de>
X-Mailer: ELM [version 2.4 PL24]
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Content-Length: 6161      

So let's see whether this mailing list is still alive;  the Web pages
were last touched in September last year...

First, I'm using Chimera-1.65 daily, and prefer it very much over all
sorts of Other Browsers.  First because I like fast & slim jobs, second
because some of its idiosyncrasies and unforgiving treatment of certain
HTML errors helps me a lot when chasing up bugs in my own Web pages, third
because it's the only browser I know which displays all Response Headers
along with the source, which is often useful.  Thanks a bunch, John,
and b*lls to M*z*ll*.

Here are some features I've noticed  (possibly including some bugs?)  which
I think it might be worthwhile looking into for 2.0, in no apparent
order.

(1) (HTML entities) Map &nbsp; and &copy; to their corresponding iso-8859-1
characters, and try not to break a line at a non-breaking space.  This seem
to be the only two entity-names of HTML/2.0 which are not translated.  (May
be inherited from M*saic-2.4, and gone already in the 2.0-alphas.)

(2) (in view of "second" above) Make pedantic or lenient treatment of HTML
a runtime settable option.  (Environment/resource/commandline or, better,
changeable on the fly.  Same goes for timeout on slow net connections.)

(3) (Cancel) Obey the Cancel button even before the "Downloading Document"
message appears, which under certain conditions may take ages to happen.
(M*saic has a similar problem at least through version 2.6.  The culprit
seems to be DNS lookup;  perhaps one would need to fork a subprocess for
that if the library calls cannot otherwise be made interruptible?)

(4) (HTML rendering) Try to pick reasonable fonts when highlighting is
nested inside a heading.  (Currently <EM>...</EM> inside a <Hn>...</Hn>
looks exactly the same as outside, or can this already be tweaked via
resources i.e. app defaults?)

(5) (HTML comments) Comments should end at the first --> after the opening
<!--, not already at the first >.  Java Applets otherwise look horrible,
as they usually contain a comparison operator sooner or later;  I'd prefer
them to look not at all. ;)

(6) (FORMs) Can we handle multiple submit buttons yet?  From some forms
out there I've tried, I'm not sure.

(7) (URL encoding) Change + signs to %hex notation (%2B) before changing
spaces to pluses, otherwise mailto URLs and the whole extension interface
will never work quite right.

(8) (mailto.pl by Micheal Kellen, Jan 1995) The URL decoding is wrong;  it
doesn't take +'s to spaces before handling the %hex stuff.  (Fixed that
one but still get into trouble because of (7).)

There still seems to be another problem, in Chimera not in the Perl script,
apparently triggered by slashes, but I haven't been able to isolate the
exact circumstances.  It must have something to do with slashes going to
%2F but not back, or vice versa, or something, or slashes affecting pattern
matches in ways they shouldn't, or something else.

(In fact, whenever a text field coming from a form contains a slash but
does not start with one, all that will be left of it after Chimera's
URL-encoding and a standard URL decoding will be the three characters,
"2F\n", sometimes with a few other characters from further down the text
surviving after the 2F, and as far as I can see with all newlines surviving;
if it starts with a slash and doesn't contain another, it will end up after
decoding with a %2F which must have fallen through the decoding, or been
decoded from something even weirder, followed by whatever else there was.
Does this help?)

[Yes I still have the sources.  Yes I could have read them.  No I haven't
got round to doing so, and you'd probably find such things faster because
you'd know where to look. ;-]

Uck, just spotting something else, and that is partly a problem with HTML
as such.  It is somewhat tricky to get any of < > & " into text to be
displayed inside a form.  TEXTAREAs are relatively unproblematic since
the standard escapes will work  (but the mailto.pl script certainly doesn't
do what it should with the signature file before passing it on to Chimera
as initial value of a TEXTAREA;  I've fixed that one for me as well).

Strings passed via attribute values inside HTML tags are worse:  There
seems to be no way of making a " appear in them, since the escapes don't
work in this context.

(This is Broken As Designed in HTML itself.  Also, there's a similar problem
with TITLEs, where you can't use < > as they are, and shouldn't use their
escapes either, although the latter seem to work with most browsers other
than lynx...)

(9) (FTP) It seems that when an FTP directory is displayed, its URL is
stored with a slash at the end, and any relative FTP URLs extracted when
clicking on an entry start with a slash, so we get double slashes in a
row in the FTP URL used for retrieving that entry.  Harmless but unnecessary.

(10) (Cache) There are still some occasional problems with the cache.
Sometimes the BACK button would skip a place or two of the true history.
Sometimes some cache files are left behind upon exit.  There might be
some stoopid off-by-one error in one loop or another when trying to
purge a few cache entries.

(11) (Cache and images) If something goes wrong with the downloading or
postprocessing of an inlined or separate image, Chimera-1.65 would cache
the (broken) result of the download or processing, and refuse to reload
it henceforth.
(It'd be nice if it were possible to reload a page _and_ the images on
it.  In the other direction, chimera often reloads images when I'm trying
to go back to a page which is still in the cache.  But this might just
be caused by my browsing in the meantime having overflowed the maximal
cache size...;)

Enough for the moment.  I've subscribed to the list so replies through it
ought to reach me.

Thanks again to John and all contributors,
Gerhard

- --
* Gerhard Niklasch <nikl@mathematik.tu-muenchen.de> *** Some or all of the con-
* https://hasse.mathematik.tu-muenchen.de/~nikl/ ******* tents of the above mes-
* sage may, in certain countries, be legally considered unsuitable for consump-
* tion by children under the age of 18.  Me transmitte sursum, Caledoni...  :^/

------- Message 9

Received: from inet.uni-c.dk by JIMI.CS.UNLV.EDU id aa24872; 12 Jun 96 13:20 PDT
Received: from kroete2.freinet.de (arh174.ppp.uni-c.dk [130.228.13.174]) by inet.uni-c.dk (8.6.12/8.6.9) with SMTP id WAA10210 for <bug-chimera@cs.unlv.edu>; Wed, 12 Jun 1996 22:20:29 +0200
Received: by kroete2.freinet.de (Smail3.1.28.1 #4)
	id m0uTwNP-0003jiC; Wed, 12 Jun 96 22:18 MET DST
Message-Id: <m0uTwNP-0003jiC@kroete2.freinet.de>
Subject: Re: Suggestions for 2.0 from a happy user
To: Chimera Hacker List <bug-chimera@cs.unlv.edu>
Date: Wed, 12 Jun 1996 22:18:50 +0200 (MET DST)
In-Reply-To: <199606121933.VAA02743@pchelwig1.mathematik.tu-muenchen.de> from "Gerhard Niklasch" at Jun 12, 96 09:33:16 pm
Reply-To: Erik Corry <ehcorry@inet.uni-c.dk>
From: Erik Corry <ehcorry@inet.uni-c.dk>
X-Face:  5am4$s<`Jfx-,*w6$)uX,dDt3Z2w?4ZNGS-i@_w(Fzv%(<[(XtN*t\,OmV_a=lS
  <jt,R~:$--|={s;V(G/wsw[@S(Fv4{7`b/@Z7k_QW-s]J-<8oe*5)dg+S.CeqQ$~m]}jY$c
X-Mailer: ELM [version 2.4 PL24 ME8]
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Content-Length: 2009      

> First, I'm using Chimera-1.65 daily, and prefer it very much over all
> sorts of Other Browsers.  First because I like fast & slim jobs, second
> because some of its idiosyncrasies and unforgiving treatment of certain
> HTML errors helps me a lot when chasing up bugs in my own Web pages, third
> because it's the only browser I know which displays all Response Headers
> along with the source, which is often useful.  Thanks a bunch, John,
> and b*lls to M*z*ll*.
> 
> Here are some features I've noticed  (possibly including some bugs?)  which
> I think it might be worthwhile looking into for 2.0, in no apparent
> order.

These are all useful comments, and should probably be checked through
before 2.0 is released, but note that 2.0 is an almost complete rewrite,
so these stale old bugs may well have been replaced by new and exciting
ones.

> Uck, just spotting something else, and that is partly a problem with HTML
> as such.  It is somewhat tricky to get any of < > & " into text to be
> displayed inside a form.  TEXTAREAs are relatively unproblematic since
> the standard escapes will work  (but the mailto.pl script certainly doesn't
> do what it should with the signature file before passing it on to Chimera
> as initial value of a TEXTAREA;  I've fixed that one for me as well).

> Strings passed via attribute values inside HTML tags are worse:  There
> seems to be no way of making a " appear in them, since the escapes don't
> work in this context.
> 
> (This is Broken As Designed in HTML itself.  Also, there's a similar problem
> with TITLEs, where you can't use < > as they are, and shouldn't use their
> escapes either, although the latter seem to work with most browsers other
> than lynx...)

I think the correct way to do this is with entities, &amp; &gt; &lt;
and &quot;. That _should_ work everywhere, and it certainly works with
Chimera now. I don't see any reason why you shouldn't use these escaped
forms.

- -- 
Erik Corry ehcorry@inet.uni-c.dk https://inet.uni-c.dk/~ehcorry/

------- Message 10

Received: from citi.umich.edu by JIMI.CS.UNLV.EDU id aa25605;
          12 Jun 96 14:22 PDT
Received: from citi.umich.edu by citi.umich.edu for bug-chimera@cs.unlv.edu with SMTP; Wed, 12 Jun 96 17:21:45 -0400
From: Jim Rees <rees@umich.edu>
To: Chimera Lovers <bug-chimera@cs.unlv.edu>
Date: Wed, 12 Jun 1996 17:21:42 -0400
Subject: Re: Suggestions for 2.0 from a happy user 
Sender: rees@citi.umich.edu
In-Reply-To: Gerhard Niklasch, Wed, 12 Jun 1996 21:33:16 +0200

  First, I'm using Chimera-1.65 daily, and prefer it very much over all
  sorts of Other Browsers.

I'm still using 1.65.  I'll switch to 2.0 when it gets true table support.

Many of the bugs you complained about have been fixed.  I know I've fixed a
few of them myself, and posted the fixes to this list.  Has anyone been
keeping an archive of 1.65 bug fixes?  I've got them all, but not in a
convenient format.

Your complaint about comment delimiters affects javascript, not applets.
Still, it's annoying.  I had thought that this was a bug in the javascript
specification, but now I'm not too sure.

------- Message 11

Received: from haymarket.ed.ac.uk by JIMI.CS.UNLV.EDU id aa04195;
          13 Jun 96 5:11 PDT
Received: from aiai.ed.ac.uk (skye-alter.aiai.ed.ac.uk [192.41.104.6]) by haymarket.ed.ac.uk (8.6.13/8.6.12) with SMTP id NAA18301; Thu, 13 Jun 1996 13:10:58 +0100
Received: from subnode.aiai.ed.ac.uk (scarp) by aiai.ed.ac.uk; Thu, 13 Jun 96 13:08:35 BST
Date: Thu, 13 Jun 96 13:08:35 BST
Message-Id: <13897.9606131208@subnode.aiai.ed.ac.uk>
From: Tim Bradshaw <tfb@aiai.ed.ac.uk>
To: Jim Rees <rees@umich.edu>
Cc: Chimera Lovers <bug-chimera@cs.unlv.edu>
Subject: Re: Suggestions for 2.0 from a happy user 
In-Reply-To: <199606122140.WAA14237@haymarket.ed.ac.uk>
References: <199606122140.WAA14237@haymarket.ed.ac.uk>

* Jim Rees wrote:

> Your complaint about comment delimiters affects javascript, not applets.
> Still, it's annoying.  I had thought that this was a bug in the javascript
> specification, but now I'm not too sure.

The original person was right about comments: the SGML (and therefore
HTML) comment syntax is:

	<!-- 	start comment
	-->	end comment

(note syntax is different in dtds).

chimera 1.65 gets this wrong, which used to bite me as I had headers
which went: <!-- ... <tfb@ed.ac.uk> ... stuff seen by chimera ... -->

- --tim

------- Message 12

Received: from cheviot.ncl.ac.uk by JIMI.CS.UNLV.EDU id aa26243;
          17 Jun 96 6:44 PDT
Received: from aire.ncl.ac.uk by cheviot.ncl.ac.uk id <OAA27348@cheviot.ncl.ac.uk>
  (8.6.12/ for ncl.ac.uk) with SMTP; Mon, 17 Jun 1996 14:44:30 +0100
Received: (njdc@localhost) by aire.ncl.ac.uk (8.6.12/8.6.x-cf revision 3 for SunOS 4.1.x) id OAA23901; Mon, 17 Jun 1996 14:44:16 +0100
From: "J.D.Coleman" <J.D.Coleman@newcastle.ac.uk>
Message-Id: <199606171344.OAA23901@aire.ncl.ac.uk>
Subject: Re: Suggestions for 2.0 from a happy user
To: Chimera Mailing List <bug-chimera@cs.unlv.edu>
Date: Mon, 17 Jun 1996 14:44:15 +0100 (BST)
In-Reply-To: <199606121933.VAA02743@pchelwig1.mathematik.tu-muenchen.de> from "Gerhard Niklasch" at Jun 12, 96 09:33:16 pm
X-Organisation: University of Newcastle Computing Service
X-Phone: +44-191-222-8068
X-Fax: +44-191-222-8765
X-Mailer: ELM [version 2.4 PL23]
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Content-Length: 4381      

> (11) (Cache and images) If something goes wrong with the downloading or
> postprocessing of an inlined or separate image, Chimera-1.65 would cache
> the (broken) result of the download or processing, and refuse to reload
> it henceforth.
> (It'd be nice if it were possible to reload a page _and_ the images on
> it.  In the other direction, chimera often reloads images when I'm trying
> to go back to a page which is still in the cache.  But this might just
> be caused by my browsing in the meantime having overflowed the maximal
> cache size...;)

Hmm, the no image reload has been bugging me for a while too.  I have my
cache keep set to 6 days and usually leave Chimera running until it eats
up too much memory!  Also, I often want to reload images - pages with inline
status graphs updated every 30 mins.  So ...

I've had a poke around the source and have a partial fix for this - if you
'reload', *all* images in the document will be reloaded.  The reason I say
*all* is that, if a page contains multiple instances of an image, the image
will be loaded multiple times.  This is because the reload flag is global
and my partial fix just forces every inline to inherit that.  I thought
about looking at the cache times but that doesn't allow for variable speed
download times and/or instantaneous updates.  I suppose a better way would
be to carry around state of which images have just been reloaded - but I
can't see an elegant way of doing this.  If I get a chance, I'll have another
look.

I've appended my fix below.  It also fixes a minor bug where the global
reload flag in AuthOKCallback() wasn't reset.  The line numbers may be
a bit out as the diff is against the released 1.65 but I have also applied
the patches from the list.

J

- -- 
"The Ravenous Bugblatter Beast of Traal!  Is it safe?"
"Oh yes, it's perfectly safe.  It's just us who are in trouble!"
- --
Wargame (n): battle simulation using painted numbers.

 ---8<---------------------------- Cut here ---------------------------->8---

*** src/main.c.dist	Sat Apr  1 02:45:19 1995
- --- src/main.c	Fri Jun 14 16:40:15 1996
***************
*** 656,682 ****
    Document *d;
    URLParts *tup = NULL, *nup = NULL;
  
!   if (root.dlist != NULL && reload == 0)
    {
!     if (root.dlist->base != NULL)
      {
!       URLParts *kup;
  
!       kup = ParseURL(root.dlist->base);
!       if (kup != NULL)
        {
! 	tup = nup = MakeURLParts(up, kup);
! 	DestroyURLParts(kup);
        }
        else nup = up;
      }
!     else if (root.dlist->up != NULL)
      {
!       tup = nup = MakeURLParts(up, root.dlist->up);
      }
      else nup = up;
    }
- -   else nup = up;
  
    if (!root.cacheOff && !reload) d = ReadCache(nup);
    else d = NULL;
- --- 661,702 ----
    Document *d;
    URLParts *tup = NULL, *nup = NULL;
  
!   if (reload == 0)
    {
!     if (root.dlist != NULL)
      {
!       if (root.dlist->base != NULL)
!       {
!         URLParts *kup;
  
!         kup = ParseURL(root.dlist->base);
!         if (kup != NULL)
!         {
! 	  tup = nup = MakeURLParts(up, kup);
! 	  DestroyURLParts(kup);
!         }
!         else nup = up;
!       }
!       else if (root.dlist->up != NULL)
        {
!         tup = nup = MakeURLParts(up, root.dlist->up);
        }
        else nup = up;
      }
!     else nup = up;
!   }
!   else /* reload == 1 */
!   {
!     if (root.dlist != NULL)
      {
!       if (root.dlist->up != NULL)
!       {
!         tup = nup = MakeURLParts(up, root.dlist->up);
!       }
!       else nup = up;
      }
      else nup = up;
    }
  
    if (!root.cacheOff && !reload) d = ReadCache(nup);
    else d = NULL;
***************
*** 770,775 ****
- --- 790,797 ----
  
    XtDestroyWidget(w);
  
+   root.rflag = False;
+ 
    return;
  }
  
***************
*** 1814,1819 ****
- --- 1897,1903 ----
    Visual *v;
    int depth;
    int stype;
+   int reload;
    RealmInfo *r;
  
    if (root.cancelop || url == NULL)
***************
*** 1825,1832 ****
    tup = ParseURL(url);
    if (tup == NULL) return(NULL);
  
!   if (delay) d = LoadDoc(tup, 0, 1);
!   else d = LoadDoc(tup, 0, 0);
  
    DestroyURLParts(tup);
  
- --- 1909,1919 ----
    tup = ParseURL(url);
    if (tup == NULL) return(NULL);
  
!   if (root.rflag) reload = 1;
!   else reload = 0;
! 
!   if (delay) d = LoadDoc(tup, reload, 1);
!   else d = LoadDoc(tup, reload, 0);
  
    DestroyURLParts(tup);


------- End of Forwarded Messages