Hi, if I'm interpreting http://www.w3.org/Addressing/rfc1808.txt correctly (check "5.1. Normal Examples"), dillo does not handle relative urls with queries correctly. In case of queries it truncates the filename. Anchors are handled correctly. You can see the difference here: <a href="#foo">#foo</a> <a href="?foo=bar">?foo=bar</a> Please check attached patch. Cheers, Johannes
On Thu, Feb 12, 2009 at 12:23:22PM +0100, Hofmann Johannes wrote:
Hi,
if I'm interpreting http://www.w3.org/Addressing/rfc1808.txt correctly (check "5.1. Normal Examples"), dillo does not handle relative urls with queries correctly. In case of queries it truncates the filename. Anchors are handled correctly. You can see the difference here:
<a href="#foo">#foo</a> <a href="?foo=bar">?foo=bar</a>
Yes, dillo doesn't follow rfc1808.txt. Dillo follows rfc2396 which updates rfc1808. This particular case was changed by it. It took me some time to review the issue. A long time ago I made a test program for the URL resolver that tests all the examples in rfc2396. The riddle became clear once I reviewed its tests and the result for this case was different. ;) -- Cheers Jorge.-
On Fri, Feb 13, 2009 at 10:07:36AM -0300, Jorge Arellano Cid wrote:
On Thu, Feb 12, 2009 at 12:23:22PM +0100, Hofmann Johannes wrote:
Hi,
if I'm interpreting http://www.w3.org/Addressing/rfc1808.txt correctly (check "5.1. Normal Examples"), dillo does not handle relative urls with queries correctly. In case of queries it truncates the filename. Anchors are handled correctly. You can see the difference here:
<a href="#foo">#foo</a> <a href="?foo=bar">?foo=bar</a>
Yes, dillo doesn't follow rfc1808.txt.
Dillo follows rfc2396 which updates rfc1808. This particular case was changed by it.
The nice thing about RFCs is that there is so many to choose from :) Just found this here: http://www.communities.hp.com/securitysoftware/blogs/jeff/archive/2007/12/19... After reading it I would suggest to follow RFC 3986 which reintroduces the old (rfc1808) behaviour which is also consistent with e.g. firefox. Cheers, Johannes
It took me some time to review the issue. A long time ago I made a test program for the URL resolver that tests all the examples in rfc2396. The riddle became clear once I reviewed its tests and the result for this case was different. ;)
-- Cheers Jorge.-
_______________________________________________ Dillo-dev mailing list Dillo-dev@dillo.org http://lists.auriga.wearlab.de/cgi-bin/mailman/listinfo/dillo-dev
On Fri, Feb 13, 2009 at 02:18:06PM +0100, Hofmann Johannes wrote:
On Fri, Feb 13, 2009 at 10:07:36AM -0300, Jorge Arellano Cid wrote:
On Thu, Feb 12, 2009 at 12:23:22PM +0100, Hofmann Johannes wrote:
Hi,
if I'm interpreting http://www.w3.org/Addressing/rfc1808.txt correctly (check "5.1. Normal Examples"), dillo does not handle relative urls with queries correctly. In case of queries it truncates the filename. Anchors are handled correctly. You can see the difference here:
<a href="#foo">#foo</a> <a href="?foo=bar">?foo=bar</a>
Yes, dillo doesn't follow rfc1808.txt.
Dillo follows rfc2396 which updates rfc1808. This particular case was changed by it.
The nice thing about RFCs is that there is so many to choose from :)
Just found this here: http://www.communities.hp.com/securitysoftware/blogs/jeff/archive/2007/12/19... After reading it I would suggest to follow RFC 3986 which reintroduces the old (rfc1808) behaviour which is also consistent with e.g. firefox.
Given that info, it looks like RFC 3986 is the one to follow. I'll work on it. -- Cheers Jorge.-
On Fri, Feb 13, 2009 at 02:18:06PM +0100, Hofmann Johannes wrote:
On Fri, Feb 13, 2009 at 10:07:36AM -0300, Jorge Arellano Cid wrote:
On Thu, Feb 12, 2009 at 12:23:22PM +0100, Hofmann Johannes wrote:
Hi,
if I'm interpreting http://www.w3.org/Addressing/rfc1808.txt correctly (check "5.1. Normal Examples"), dillo does not handle relative urls with queries correctly. In case of queries it truncates the filename. Anchors are handled correctly. You can see the difference here:
<a href="#foo">#foo</a> <a href="?foo=bar">?foo=bar</a>
Yes, dillo doesn't follow rfc1808.txt.
Dillo follows rfc2396 which updates rfc1808. This particular case was changed by it.
The nice thing about RFCs is that there is so many to choose from :)
Just found this here: http://www.communities.hp.com/securitysoftware/blogs/jeff/archive/2007/12/19... After reading it I would suggest to follow RFC 3986 which reintroduces the old (rfc1808) behaviour which is also consistent with e.g. firefox.
After a more detailed review of rfc3986 there turned out to be more cases with a different resolution than in rfc2396. A new patch was committed. It passes all the tests in rfc3986 plus some more, and also the below quoted page to resolve as Firefox: <HTML> <HEAD> <TITLE>Relative URL test</TITLE> <BASE href="http://a/b/c/d;p?q#f"> </HEAD> <BODY> BASE ="http://a/b/c/d;p?q#f"<br> <a href="#foo">#foo</a><br> <a href="?foo=bar">?foo=bar</a><br> <a href="?foo=bar#frag">?foo=bar#frag</a><br> <a href="">empty</a> </BODY> </HTML> It was very good you spotted this bug! Now this kind of links will work in dillo. e.g. those in http://hg.dillo.org/dillo ;-) @all: Please test. -- Cheers Jorge.-
participants (2)
-
jcid＠dillo.org -
Johannes.Hofmann＠gmx.de