Safari rendering bugs in <table> ?

Safari 1.2 appears to have rendering bugs with <table> directives that use the "groups" rule. See below for a specific case but in general the problem is that Safari completely ignores <colgroup> and <tbody>. If the table itself has a border it always puts a border around every cell, no matter what. The html fragment below should appear more or less like this: |-------------------------| | COL1 | COL2 COL3 | COL4 | |-------------------------| | 1,2 | 1,2 1,3 | 1,4 | | 2, 1 | 2,2 2,3 | 2,4 | |-------------------------| | 3,1 | 3,2 3,3 | 3,4 | |-------------------------| Camino and iCab handle this properly. Safari does not. Neither does Omni5, which as I understand it uses the same rendering enging as Safari. Here's the html, which as far as I know is valid (the doctype is somewhat arbitrary). <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Frameset//EN" "http://www.w3.org/TR/REC-html40/frameset.dtd"> <html> <head> <title>Safari Bug</title> </head> <body> <table border="1" rules="groups"> <colgroup> <colgroup span="2"> <colgroup> <thead> <tr> <th><b>COL1</b></th> <th><b>COL2</b></th> <th><b>COL3</b></th> <th><b>COL4</b></th> </tr> <tbody> <tr> <td>1,1</td> <td>1,2</td> <td>1,3</td> <td>1,4</td> </tr> <tr> <td>2,1</td> <td>2,2</td> <td>2,3</td> <td>2,4</td> </tr> <tbody> <tr> <td>3,1</td> <td>3,2</td> <td>3,3</td> <td>3,4</td> </tr> </table> </body> </html>

In article <2uv03oF2ac9goU1@uni-berlin.de>, respower <respower@studiors.org> wrote: > Safari 1.2 appears to have rendering bugs with <table> directives that > use the "groups" rule. See below for a specific case but in general the > problem is that Safari completely ignores <colgroup> and <tbody>. If the > table itself has a border it always puts a border around every cell, no > matter what. > > The html fragment below should appear more or less like this: > > |-------------------------| > | COL1 | COL2 COL3 | COL4 | > |-------------------------| > | 1,2 | 1,2 1,3 | 1,4 | > | 2, 1 | 2,2 2,3 | 2,4 | > |-------------------------| > | 3,1 | 3,2 3,3 | 3,4 | > |-------------------------| > > > Camino and iCab handle this properly. Safari does not. Neither does > Omni5, which as I understand it uses the same rendering enging as > Safari. > > Here's the html, which as far as I know is valid (the doctype is > somewhat arbitrary). > > <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Frameset//EN" > "http://www.w3.org/TR/REC-html40/frameset.dtd"> > <html> > <head> > <title>Safari Bug</title> > </head> > > <body> > > <table border="1" rules="groups"> > > <colgroup> > <colgroup span="2"> > <colgroup> > > <thead> > <tr> > <th><b>COL1</b></th> > <th><b>COL2</b></th> > <th><b>COL3</b></th> > <th><b>COL4</b></th> > </tr> > > <tbody> > <tr> > <td>1,1</td> > <td>1,2</td> > <td>1,3</td> > <td>1,4</td> > </tr> > <tr> > <td>2,1</td> > <td>2,2</td> > <td>2,3</td> > <td>2,4</td> > </tr> > > <tbody> > <tr> > <td>3,1</td> > <td>3,2</td> > <td>3,3</td> > <td>3,4</td> > </tr> > > </table> > > </body> > </html> And you're posting this to the group because... Honestly, Apple employees don't typically wander around this group. You should send this problem directly to Steve Jobs. He'll take care of it right away. -- DeeDee, don't press that button! DeeDee! NO! Dee...

In article <2uv03oF2ac9goU1@uni-berlin.de>, respower <respower@studiors.org> wrote: <clip> > Here's the html, which as far as I know is valid (the doctype is > somewhat arbitrary). <clip> Per BBEdit: Line 8: Document type does not permit element �<body>� in content of element �<html>�. Line 13: Element �<colgroup>� implicitly closed here. Line 14: Element �<colgroup>� implicitly closed here. Line 16: Element �<colgroup>� implicitly closed here. Line 24: Element �<thead>� implicitly closed here. Line 38: Element �<tbody>� implicitly closed here. Line 46: Element �<tbody>� implicitly closed here.

On 04/11/2004 10:27 AM, respower wrote: > Safari 1.2 appears to have rendering bugs with <table> directives that > use the "groups" rule. See below for a specific case but in general the > problem is that Safari completely ignores <colgroup> and <tbody>. If the > table itself has a border it always puts a border around every cell, no > matter what. > > The html fragment below should appear more or less like this: > > |-------------------------| > | COL1 | COL2 COL3 | COL4 | > |-------------------------| > | 1,2 | 1,2 1,3 | 1,4 | > | 2, 1 | 2,2 2,3 | 2,4 | > |-------------------------| > | 3,1 | 3,2 3,3 | 3,4 | > |-------------------------| > > > Camino and iCab handle this properly. Safari does not. Neither does > Omni5, which as I understand it uses the same rendering enging as > Safari. > > Here's the html, which as far as I know is valid (the doctype is > somewhat arbitrary). > No, this code does not validate. Check it again.

On 2004-11-04, Chris Moore <reply_to@newsgroup.INVALID> wrote: > In article <2uv03oF2ac9goU1@uni-berlin.de>, respower ><respower@studiors.org> wrote: > ><clip> > >> Here's the html, which as far as I know is valid (the doctype is >> somewhat arbitrary). > ><clip> > > Per BBEdit: > > Line 8: Document type does not permit element �<body>� in content of > element �<html>�. Framesets don't have bodies, which is presumably BBEdit is complaining. Browsers don't really care about this though. So in and of itself, this is not the source of the problem. If it were Safari wouldn't display anything at all. Every browser I tried displays something more or less plausible. It's the details that are sometimes wrong. Feel free to plug in any doctype you like, my guess is you'll see exactly the same behavior. If not, and in particular if you find a doctype that causes Safari to display this example properly, please post! That would be very useful information. Fwiw I also tried: <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN" "http://www.w3.org/TR/REC-html40/loose.dtd"> > Line 13: Element �<colgroup>� implicitly closed here. > Line 14: Element �<colgroup>� implicitly closed here. > Line 16: Element �<colgroup>� implicitly closed here. > Line 24: Element �<thead>� implicitly closed here. > Line 38: Element �<tbody>� implicitly closed here. > Line 46: Element �<tbody>� implicitly closed here. The close tags for these are defined to be optional. These are just warnings. So practically speaking nothing's really wrong with the html. But I appreciate the response. Possibly, although it doesn't seem at all likely, this is a true Safari bug that's never been reported before. If this is so, I'll report to Apple. But I'd be amazed if that were the case. More likely, I'm misunderstanding something, or some very specific doc-type would work. That's why I'm posting. Only a fool would instantly report this as a bug without asking around a little first. Btw the terminal browser 'links' gets it exactly right. Here's its output, which is better than my handdrawn ascii version: +--------------------------+ | COL1 | COL2 COL3 | COL4 | |------+------------+------| | 1,1 | 1,2 1,3 | 1,4 | | 2,1 | 2,2 2,3 | 2,4 | |------+------------+------| | 3,1 | 3,2 3,3 | 3,4 | +--------------------------+

On 2004-11-04, clvrmnky <clvrmnky-uunet@coldmail.com.invalid> wrote: > No, this code does not validate. Check it again. That's just the doctype dtd reference, which every browser I know will happily ignore. That's why I said in the original post that the doctype was chosen arbitrarily. Pick any standard html doctype you like or even omit a doctype altogether, I think you'll see the same behavior. If you don't, let me know.

In article <2uvgv8F2f766cU1@uni-berlin.de>, respower <respower@studiors.org> wrote: > On 2004-11-04, clvrmnky <clvrmnky-uunet@coldmail.com.invalid> wrote: > > No, this code does not validate. Check it again. > > That's just the doctype dtd reference, which every browser I know will > happily ignore. That's why I said in the original post that the doctype > was chosen arbitrarily. Pick any standard html doctype you like or even > omit a doctype altogether, I think you'll see the same behavior. If you > don't, let me know. It's interesting that the W3C designed application "Amaya 8.6" shows the same problem. -- Tom Stiller PGP fingerprint = 5108 DDB2 9761 EDE5 E7E3 7BDA 71ED 6496 99C0 C7CF

On 2004-11-04, Tom Stiller <tomstiller@comcast.net> wrote: > It's interesting that the W3C designed application "Amaya 8.6" shows the > same problem. So the only browsers that are known to handle this properly are the gecko-based ones (at least Camino and Firefox), iCab and links? Is this feature of html just too new? I thought it had been around for some time and could safely be assumed to be supported in current browsers. That's the reason I was so surprised that Safari didn't handle it.

On 2004-11-06, Shawn Hearn <srhi@comcast.net> wrote: > Try upgrading Safari. It renders fine for me. You mean 1.2.4 (v125.11), from 10.3.6? I just tried it, it looks exactly the same as before: the renderer still pays no attention to the cell groupings. Unless my update was somehow incomplete it didn't fix this problem.

In article <2uvlfeF2e72cjU1@uni-berlin.de>, respower <respower@studiors.org> wrote: > On 2004-11-04, Tom Stiller <tomstiller@comcast.net> wrote: > > It's interesting that the W3C designed application "Amaya 8.6" shows the > > same problem. > > So the only browsers that are known to handle this properly are the > gecko-based ones (at least Camino and Firefox), iCab and links? Is this > feature of html just too new? It's part of HTML 4, which is from april 1998... I don't know about you, but to me that doesn't count as "new by a long shot. Unfortunately browser vendors are wasting an enormous amount of resources on getting their code to make sense out of all the crappy HTML/CSS/javascript out there. If Web publishers would bother to learn their job Web browsers might evolve again. > I thought it had been around for some > time and could safely be assumed to be supported in current browsers. Hell no. No browser I know of supports all of HTML 4. Even iCab doesn't support everything. <COL align="char" char=","> for instance. Then again, no browser does AFAIK... -- Sander Tekelenburg, <http://www.euronet.nl/~tekelenb/> Mac user: "Macs only have 40 viruses, tops!" PC user: "SEE! Not even the virus writers support Macs!"

On 2004-11-06, Sander Tekelenburg <user@domain.invalid> wrote: > It's part of HTML 4, which is from april 1998... I don't know about you, > but to me that doesn't count as "new by a long shot. I don't think so either. That's why I was surprised at the non-support, and even more surprised that this has apparently never come up before in the Mac newsgroups (according to Google). >> I thought it had been around for some >> time and could safely be assumed to be supported in current browsers. > > Hell no. No browser I know of supports all of HTML 4. True but <table>s with groupings are pretty widely used -- this is not exactly an obscure feature of html 4. Even 'links' supports it. Btw, some pointless fuss was made earlier over the fact that my original sample as posted here happened to include the wrong doctype. Browsers don't generally care about the doctype, certainly not in this case. As I suggested earlier, the 4.0 transitional type would have been the more appropriate one to use, and indeed validator.w3.org validates the file with that type. Safari, and every other browser I'm familiar with, does exactly the same thing as it did with the other type. <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN" "http://www.w3.org/TR/REC-html40/loose.dtd"> <html> <head> <META HTTP-EQUIV="Content-Type" CONTENT="text/html; charset=ISO-8859-1"> <title>Safari Bug</title> </head> <body> <table border="1" rules="groups"> <colgroup> <colgroup span="2"> <colgroup> <thead> <tr> <th><b>COL1</b></th> <th><b>COL2</b></th> <th><b>COL3</b></th> <th><b>COL4</b></th> </tr> <tbody> <tr> <td>1,1</td> <td>1,2</td> <td>1,3</td> <td>1,4</td> </tr> <tr> <td>2,1</td> <td>2,2</td> <td>2,3</td> <td>2,4</td> </tr> <tbody> <tr> <td>3,1</td> <td>3,2</td> <td>3,3</td> <td>3,4</td> </tr> </table> </body> </html>

In article <2v42tgF2fmqiiU1@uni-berlin.de>, respower <respower@studiors.org> wrote: > On 2004-11-06, Sander Tekelenburg <user@domain.invalid> wrote: > > It's part of HTML 4, which is from april 1998... I don't know about you, > > but to me that doesn't count as "new by a long shot. > > I don't think so either. That's why I was surprised at the non-support, > and even more surprised that this has apparently never come up before in > the Mac newsgroups (according to Google). Well, this group might not be that relevant to this issue, so those search results may not mean much. Groups like <news:comp.infosystems.www.authoring.html>, <news:comp.infosystems.www.browsers.misc> or <news:comp.infosystems.www.browsers.mac> are probably more relevant. -- Sander Tekelenburg, <http://www.euronet.nl/~tekelenb/> Mac user: "Macs only have 40 viruses, tops!" PC user: "SEE! Not even the virus writers support Macs!"