With bug 31906, we can recognize nested quotes and display the in HTML style. WeE should quote them in HTML style as well. This helps tremendously with (re)wrapping, is easier to recognize and looks better. Code bascially works already with patches for bug 31906, try it out by setting "mail.quoteasblock" to true. Remaining problems: - Too much whitespace: Remove all linebreaks at the beginning and and of a <pre> block before inserting into the editor. - Some other minor problem I don't remember right now. - Test, test, test (already did for ~1 day), e.g. with other mailers. You (listeners/lurkers) can help here.
Akk, J-F, what do you think about this?
add rhp to the cc list
I think, the first issue I mentioned above is not really a stopper for enabling the feature, because the whitespace around quotes is still messed in all modes anyway (e.g. see bug 38492). Of course, that doesn't mean, I won't fix the extra whitespace. I just don't think, we should stop this feature because of it and risking its inclusion into final. The code has been checked in with bug 31906. You can enable it with |user_pref("mail.quoteasblock", true);|. I urge everybody to try it out and so to help testing interoperability (and operability in general :) ). OTOH, I think, it is a big improvement for HTML mail. Should we enable this by default now to get wider testing or should I ask for testers on .mailnews first and wait 1 or 2 weeks until enabling it as default? (BTW: For those concerned about the safety of the recognition: Lines must start with a (mixed) sequence of ">" or "> ", while the last ">" must have a space after it. E.g. > > bla >> bla >> > bla are recognized, but not > bla >bla > >bla I consider this being pretty safe.)
I'm assuming the OK of the mailnews team here, of course. I'll speak with putterman.
One more problem: 4.x' editor doesn't like the <div>s. I'll just remove them for now (until we have a better solution). Minor tweak. As for the "too much newlines" problem: I don't even know if this is actually a bug. The newlines around <blockquote>s the editor puts out is odd anyway. This bug needs to be fixed before I know if something/what has to be done. Until then, we may just have 2 empty lines between 2 quotes - currently a minor problem. Summary: IMO, we can enable this now. Waiting for decision.
OK, no answer till now, so I'm assuming OK unless somebody objects. Will create patch. Akk, would you then review, please?
The "Remove <div>s"-patch is for libmime, but is just removing the div in the quoting case, adding quotes around the class in the other cases and changing some comments. akk, rhp: Could one of you please review it (and deny, if not)? The "Change default pref"-patch is the "real" change, it turn the feature on by default. Akk, could you please review that?
Ducarroz, if you want to review, that is fine, too.
Will the removal of <div> stop the "This message contains HTML. How do you want to send it?"-message for pure text messages?
Daniel, yes. See also bug 44552.
Rich is the right persone to review this fix. Rich can you do it?
Rich reviewed the change in mimetpla.cpp. Leaving is "just" the change of the default pref. As this pref defines with what we feed the composer (by default), I would argue, an editor person is the best reviewer, not?
I interpret <quote src="http://www.newsreaders.com/gnksa/gnksa.txt, Requirements 14"> To get well-readable articles, the user SHOULD be provided with the possibility to rewrap excessively long lines of quoted text, respecting quotation -- i.e. have the option to correct `inherited' bad formatting. </quote> to support this bug.
a=alecf. Checked in. Yippie!
Changing qa assign.
Verified on win32 and linux using the following builds: win32 commercial seamonkey build 2001-010821-mtrunk installed on P500 Win98 linux commercial seamonkey build 2001-011911-mtrunk installed on P200 RedHat 6.2 Note: per Ben "It should change the quote in the *HTML* composer, not the plaintext composer, after replying to a real plaintext msg." but failed on macos. The user pref "mail.quoteasblock" as no effect on mac. I will open separate bug to track.
Logged parity bug on Mac http://bugzilla.mozilla.org/show_bug.cgi?id=66349