Please see bug 387480 comment 74 to 77 for the discussion that lead to the creation of this bug. In short: - the new add-exception dialog that got added with bug 387480 currently uses xmlhttprequest to obtain the cert - necko blocks access to many ports, including all standard mail server ports, so currently it's impossible to add exceptions for mail servers. I'm proposing a workaround, that will make the add-exception dialog work, as soon as you've visited the broken site. The "real" solution (do not require to visit bad server first) will be more difficult to implement. (But in my opinion, even the "real" solution will benefit from the code I'm proposing for this workaround. The major problem for the real solution is sites like STARTTLS, that use a protocol dependent plaintext communication. Therefore, the real solution will most likely involve changes to protocol dependent configuration UI (like SMTP server configuration). In that context, a button could initiate a protocol connection to the server, in order to obtain the server certificate. With the code I'm proposing, it will be sufficient to open a connection. The protocol specific code won't have to deal with obtaining and storing the cert) I'll attach the patch that I had initially attached to bug 387480 comment 73.

This should block 1.9, because without it, we make life really difficult for mail users. We have a patch already.

Comment on attachment 284016 [details] [diff] [review] Patch v1 r+ I've reviewed this once already;).

Comment on attachment 284016 [details] [diff] [review] Patch v1 Requesting approval for this patch to make mail users happy (both Thunderbird / SeaMonkey)

checked in, marking fixed.

V.Fixed between [Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.9a9pre) Gecko/2007101503 SeaMonkey/2.0a1pre] (nightly) (W2Ksp4) and [Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.9a9pre) Gecko/2007101611 SeaMonkey/2.0a1pre] (nightly) (W2Ksp4) The U.I. now lets me add <imap/ssl>:993 and <smtp/ssl>:465 entries. (See (duplicate) bug 398534.)

Reopening bug. All patches that got checked in to trunk yesterday are being backed out, because it's unclear which patch has caused a performance regression.

checked in again, marking fixed.