Base forum url?

Thanks :-)

If someone is interested in my MOD, here is:

http://punbb.org/forums/viewtopic.php?p … 32#p101432

It will be on xhtml 2.0 . Furthermore, if I can do something without Javascript, I fell more happy ^____^

PS: please post any question on the proper thread, I wouldn't split the discussion :-)

http://punbb.org/forums/viewtopic.php?p … 32#p101432

Actually I don't think I can use target as an attribute, and certainly IE6 can't interpret it.

When xhtml 2.0 will be interpreted by 99% of browsers (as xhtml 1.0 I think), I'll use target as attribute. Actually, I prefer to do a little incorrectness instead of using JS rubbish :-P

in english:

It'is for visibility. Users than follows an external link could forget the original site! It sounds stupid, but very much people do "zapping" on internet. Furthermore very much people doesn't know what really means "internet"... just imagine how "pretty easy" is open links in a new window!

Unluckily this could cause some statistical problems, because the time on the forum could result very much higher than the real one. But I don't think it's a crucial problem ^___^''

in italian:

e' una questione di visibilita'. Gli utenti che vanno su un sito esterno possono dimenticarsi del sito originale! Sembra una stupidaggine, ma ci sono un sacco di persone che fanno "zapping" tra i siti internet. Tra l'altro, molti non sanno neanche che significhi esattamente "internet" o "browser", pensa quanto dev'essere semplice aprire un link in una nuova finestra per loro! :-P

Purtroppo un sito che funziona in questo modo puo' scombinare le statistiche del sito stesso, perche' risulterebbe che la permanenza media sul forum e' molto piu' lunga di quanto non sia in realta'. Comunque non mi pare un contro assurdo ^___^

Must admit, I am guilty of opening three links on the forum in new pages/tabs. The gallery, for uploading images for posting, and the two links to the articles/downloads sites. I use the JS method. Must admit, I'm not fond of the opening new windows/tabs concept myself, but trying to constantly explain to people that a right click and open in...... was all that was required ended up getting tedious after a while, so I digressed on those three.

If you are going to open links in new tabs/windows however, I would suggest using the title= part to list that fact in the <a tag/link. At least people will be advised when they hover over the link, that way.

That way is even better. Just mentioned it because many people never mention the new window bit anywhere.

Just whilst we're on the subject again, , noticed you use a vaguely similiar method to the parser for link truncation. It screws validity in Xmhtl 1.1. Altered my URI truncation to this to over come the problem:

        if ($link == '' || $link == $url)
        {
                $pos = strpos($url, '?');

                if ($pos != '')
                {
                        $link = pun_htmlspecialchars(substr($url, '?', $pos));
                }
                else
                {
                        $link = pun_htmlspecialchars($url);
                }
        }
        else
        {
                $link = stripslashes($link);
        }

The likes of links such as Google will be stripped back before the '?', so retain usable layout and validity. Anything which is just stupidly long is left as is. You could use a mix of your original code and the above to stop URI's with arguments supplied being truncated by the original parsing method, and truncate anything that doesn't have supplied arguments using your original code. It should retain XML validity in all scenarios that way. Just an idea.

The problem occurs if a bare URI is posted in a post. For example, something along the lines of:

http://example.com?script&argument1&argument2

could end up being truncated part way through an &, which will, under normal validation in <Xhtml1.0, fail validation, but display the page. Running in Xhtml1.1, however, the XML parser would just print an error message on screen instead of displaying the page if that occured. Truncating at the ?, however, removes any entities from the printed link.

Mmmmhhhhh....... have you tried to do some test before? ^_____^

something&

Mmmmhhhhh..... I'm starting to be very confused :-P

You wrote the text of a link could be truncated at an &, invalidating an xhtml document.

For example if I have the link

somesite.org/somepath?etc&someotherqueries

the link could be posted as

<a href="somesite.org/somepath?etc&omeotherqueries">
somesite.org/somepath?etc&
</a>

BUT! If you notice, punBB already substitutes these characters with them entities. For example, in the link in my previous post, the exit code is:

<a href="http://www.noproblemswithpunbb.com/something&end">something&</a>

So the code is valid, I don't see any problems with the truncation of the text of a link at any offset :-)

It's not possible, because the truncation is before the entity conversion ;-)

in include/parser.php:

function parse_message($text, $hide_smilies)
{
    [...]
    // Convert applicable characters to HTML entities
    $text = pun_htmlspecialchars($text);

in viewtopic.php:

    // Perform the main parsing of the message (BBCode, smilies, censor words etc)
    $cur_post['message'] = parse_message($cur_post['message'], $cur_post['hide_smilies']);

On the contrary, handle_url_tag() is called in include/parser.php, not in vietopic.php :-)