[PATCH] version 7 of tab patch is ready
Oops... I seem to have messed up the subject of my last message. Anyway, as stated... the next version of my tab patch is ready, get it at http://www.geocities.com/ikbenfrank Read my last message for more info... Cheers//Frank -- WWWWW ________________________ ## o o\ / Frank de Lange \ }# \| / +46-734352015 \ \ `--| _/ <Hacker for Hire> \ `---' \ +31-640037120 / \ frank@unternet.org / `------------------------' [ "Omnis enim res, quae dando non deficit, dum habetur et non datur, nondum habetur, quomodo habenda est." ]
Hi Great work the tab and frame support! If I try to open the reiserfs-Homepage (www.namesys.com) only the menu-frame is shown. Here is the output of dillo: [tux@tux tux]$ dillo Setting locale to C... dillo_dns_init: Here we go! Enabling cookies as from cookiesrc... Nav_open_url: Url=>about:splash< Nav_open_url: Url=>http://www.namesys.com<; Dns_server [0]: www.namesys.com is 0x81180b0 Connecting to 212.16.7.65 HTML warning: the TITLE element must be inside the HEAD section Nav_open_url: Url=>http://www.namesys.com/menu.html<; Connecting to 212.16.7.65 HTML warning: Invalid HTML syntax: too many <frame> tags in <frameset> This I get at other sites, too. Jens
On Tue, Aug 05, 2003 at 10:04:35AM +0200, Jens Arm wrote:
Hi
Great work the tab and frame support!
Thanks...
If I try to open the reiserfs-Homepage (www.namesys.com) only the menu-frame is shown. Here is the output of dillo:
[tux@tux tux]$ dillo Setting locale to C... dillo_dns_init: Here we go! Enabling cookies as from cookiesrc... Nav_open_url: Url=>about:splash< Nav_open_url: Url=>http://www.namesys.com<; Dns_server [0]: www.namesys.com is 0x81180b0 Connecting to 212.16.7.65 HTML warning: the TITLE element must be inside the HEAD section Nav_open_url: Url=>http://www.namesys.com/menu.html<; Connecting to 212.16.7.65 HTML warning: Invalid HTML syntax: too many <frame> tags in <frameset>
This I get at other sites, too.
It works here... Did you use the latest CVS version of Dillo, did the patch aplpy without problems, do you use a filtering proxy which does weird things to the HTML? Namesys' page comes up in two frames. Size of frames is wrong (but that was to be expected), content is OK. This is what I get with 'View Page Source' on Namesys' page: <html> <TITLE>NAMESYS</TITLE> <frameset cols="160,*" frameborder=0 framespacing=0 border=0> <frame name="menu" src="menu.html" scrolling="auto" noresize border=0 frameborder=0 marginheight=0 marginwidth=0> <frame name="main" src="content_table.html" noresize border=0 frameborder=0 marginheight=0 marginwidth=0> </frameset> </html> So yes, that IS invalid HTML (TITLE tag outside of HEAD), but Dillo renders it OK here. It tells me about the errors in the page: Nav_open_url: Url=>http://www.namesys.com<; Connecting to 212.16.7.65 HTML warning: the TITLE element must be inside the HEAD section Nav_open_url: Url=>http://www.namesys.com/menu.html<; Nav_open_url: Url=>http://www.namesys.com/content_table.html<; Are there any others out there who have problems like this? Cheers//Frank -- WWWWW ________________________ ## o o\ / Frank de Lange \ }# \| / +46-734352015 \ \ `--| _/ <Hacker for Hire> \ `---' \ +31-640037120 / \ frank@unternet.org / `------------------------' [ "Omnis enim res, quae dando non deficit, dum habetur et non datur, nondum habetur, quomodo habenda est." ]
On Tue, Aug 05, 2003 at 10:04:35AM +0200, Jens Arm wrote:
Hi
Great work the tab and frame support!
If I try to open the reiserfs-Homepage (www.namesys.com)
Just an addition to my previous reply... Namesys' page does show up as expected (two frames), but navigation does not work as it should yet. The current version of the frames patch ignores the <base target='main'> tag in the left frame, so all content ends up in the left frame instead. This will be fixed in the next version of the patch. Cheers//Frank -- WWWWW ________________________ ## o o\ / Frank de Lange \ }# \| / +46-734352015 \ \ `--| _/ <Hacker for Hire> \ `---' \ +31-640037120 / \ frank@unternet.org / `------------------------' [ "Omnis enim res, quae dando non deficit, dum habetur et non datur, nondum habetur, quomodo habenda est." ]
On Tue, Aug 05, 2003 at 10:23:12AM +0200, Frank de Lange wrote:
On Tue, Aug 05, 2003 at 10:04:35AM +0200, Jens Arm wrote:
Hi
Great work the tab and frame support!
If I try to open the reiserfs-Homepage (www.namesys.com)
Just an addition to my previous reply...
Namesys' page does show up as expected (two frames), but navigation does not work as it should yet. The current version of the frames patch ignores the <base target='main'> tag in the left frame, so all content ends up in the left frame instead. This will be fixed in the next version of the patch.
Yet another addition... Namesys' page uses a <base> tag without a href: (http://www.namesys.com/menu.html: ... <base target="main"> ... This is in violation of the HTML specs (href is a required parameter for <base>). What should Dillo do in this case? Currently it only checks for the target parameter after it has found a href in the tag (literal interpretation the definition of <base>), but this breaks pages like Namesys' which violate the standard. I can: a) unconditionally change the code so these pages work b) add a compile option (--pedantic-mode or --quirks-mode or whatever) to disable/enable these changes c) leave it the way it is For users a and b are the best choices, for developers c is the best. It is probably infeasible to trap all the crufty HTML out there, Namesys' site alone is good for I don't know how many problems... I personally prefer b, which gives everyone a choice. Choice is good. Any thoughts? Cheers//Frank -- WWWWW ________________________ ## o o\ / Frank de Lange \ }# \| / +46-734352015 \ \ `--| _/ <Hacker for Hire> \ `---' \ +31-640037120 / \ frank@unternet.org / `------------------------' [ "Omnis enim res, quae dando non deficit, dum habetur et non datur, nondum habetur, quomodo habenda est." ]
On Tue, Aug 05, 2003 at 11:39:59AM +0200, Frank de Lange wrote:
This is in violation of the HTML specs (href is a required parameter for <base>). What should Dillo do in this case? Currently it only checks for the target parameter after it has found a href in the tag (literal interpretation the definition of <base>), but this breaks pages like Namesys' which violate the standard. I can:
a) unconditionally change the code so these pages work b) add a compile option (--pedantic-mode or --quirks-mode or whatever) to disable/enable these changes c) leave it the way it is
I personally prefer b, which gives everyone a choice. Choice is good.
Having just worked my way through the XHTML-1.0 specifications (previous work was based on 3.2 and 4.01), I suggest the following: The next version of the patch will include a compile option '--xhtml-dtd' which can be set to 'strict', 'transitional' and 'frameset'. The 'strict' setting follows XHTML-1.0-Strict (and thus disables frames and frame-related code). With 'transitional' iframe is supported, frameset is not (leaving out the frameset-related code from the binary). Use 'frameset' for full frame support. Default setting will be 'frameset' to get the behaviour most people expect. Any suggestions/comments to my suggestion? Cheers//Frank -- WWWWW ________________________ ## o o\ / Frank de Lange \ }# \| / +46-734352015 \ \ `--| _/ <Hacker for Hire> \ `---' \ +31-640037120 / \ frank@unternet.org / `------------------------' [ "Omnis enim res, quae dando non deficit, dum habetur et non datur, nondum habetur, quomodo habenda est." ]
On Wed, 6 Aug 2003, Frank de Lange wrote:
On Tue, Aug 05, 2003 at 11:39:59AM +0200, Frank de Lange wrote:
This is in violation of the HTML specs (href is a required parameter for <base>). What should Dillo do in this case?
Just as stated in our "HTML parsing policy" inside [Project Notes]!
Currently it only checks for the target parameter after it has found a href in the tag (literal interpretation the definition of <base>), but this breaks pages like Namesys' which violate the standard. I can:
a) unconditionally change the code so these pages work b) add a compile option (--pedantic-mode or --quirks-mode or whatever) to disable/enable these changes c) leave it the way it is
I personally prefer b, which gives everyone a choice. Choice is good.
We never take a) -- because of our policies, neither b) -- explained at http://ometer.com/features.html (available from our [Links] section) sometimes c) -- When we don't have time! :) In a nutshell: we try to follow the standard, not to crash on bad pages, and ideally send a warning when facing bad HTML.
Any suggestions/comments to my suggestion?
As stated above. Cheers Jorge.-
On Wed, Aug 06, 2003 at 01:13:03PM -0400, Jorge Arellano Cid wrote:
a) unconditionally change the code so these pages work b) add a compile option (--pedantic-mode or --quirks-mode or whatever) to disable/enable these changes c) leave it the way it is
I personally prefer b, which gives everyone a choice. Choice is good.
We never take a) -- because of our policies, neither b) -- explained at http://ometer.com/features.html (available from our [Links] section) sometimes c) -- When we don't have time! :)
In a nutshell: we try to follow the standard, not to crash on bad pages, and ideally send a warning when facing bad HTML.
The new patch adds a new compile option: --enable-xhtml-dtd, which can be set to 'strict', 'transitional' or 'frameset'. Setting it to either of these configures the parser for that xhtml-dtd. I take it this is consistent in the current practice. Cheers//Frank -- WWWWW ________________________ ## o o\ / Frank de Lange \ }# \| / +46-734352015 \ \ `--| _/ <Hacker for Hire> \ `---' \ +31-640037120 / \ frank@unternet.org / `------------------------' [ "Omnis enim res, quae dando non deficit, dum habetur et non datur, nondum habetur, quomodo habenda est." ]
participants (3)
-
Frank de Lange -
Jens Arm -
Jorge Arellano Cid