mozilla :: #maildev

13 May 2017
11:01WaltS48Slight delay in testing 52.1.1 on Windows 10 for me. They decided to update me to the Version 1703. Later!
11:24aceman!seen ratty
11:24firebotratty was last seen 75 days and 20 hours ago, saying 'IanN: I think symlinks confuse fsmonitor' in #maildev.
11:24aceman!seen rattyaway
11:24firebotrattyaway was last seen 75 days and 22 hours ago, changing nick to Ratty .
11:52marcoagpintoaceman!
11:58acemanhi
13:03WaltS48Still downloading the Windows 10 Creators update. The good news is I have my morning routine done and can concentrate on testing in about another ???
13:05acemanhow large is it?
13:06WaltS48How the heck do you tell? All I know is when I do Windows 10 updates it takes a long time.
13:08WaltS4894% downloaded and it has been over 2 hours.
13:08marcoagpintoI installed manually the Win10 Creators and it hurt my VMs
13:08marcoagpintobut they have release some three cumulative updates since
13:09marcoagpintobut I still don't have Internet on the VM with Kubuntu
13:10WaltS48I got some Thursday, and updated LibreOffice, VLC, Geany, Vivaldi browser, Firefox Beta and Nightly and Daily. That killed about 3 hours of my day.
13:11WaltS48Preparing to install now. Progress is being made. :)
13:12WaltS48Hopefully things work after the restart.
13:14WaltS48Daily just nagged me to update. Bye!
13:17acemanwsmwk: is 52.1 out or not?
13:47acemanjcranmer: hi, what is the replacement for MailServices.headerParser.makeMimeAddress() ? it is marked deprecated
15:55jcranmeraceman: calling jsmime
16:09WaltS48So, what is supposed to happen when I have the POP account configured to only receive messages under 50 KB, I send one larger and click the link in testing 52.1.1?
16:11acemanthe full message loads again and should replace the trucated one
16:11acemanyou can still see the truncated version of you view the mbox file
16:12acemanjcranmer: is there an inverse function to makeMimeAddress?
16:12WaltS48oops, I configured the accounts for maildir. I like to push boundaries. :)
16:13jcranmeraceman: use jsmime to parse the mime header and grab the first element if there is one?
16:14acemanI mean splitting the string produced by makeMimeAddress back into the name and email part
16:15jcranmermakeMimeAddress makes a MIME header, so you'd have to parse it as a MIME header
16:15jcranmerstrings are a really sucky way to pass this information around
16:16acemanyes
16:17acemanin the compose window, mailing lists are encoded as &quot;name <name>&quot; or &quot;name <description>&quot;
16:18acemanI wonder how we determine whether it is a recipient or a mailinglist if we allowed @ in descriptions
16:18acemananyway, we have a case where it gets encoded as &quot;name&quot; <name>
16:18acemanbecause the name contains e.g. dots
16:19acemanand now I need to determined if that is actually a valid encoding
16:20jcranmerif you want to be clever, encode a mailing list as &quot;name&quot;: :;
16:20jcranmerno question as whether or not that&#39;s a mailnig list or an address
16:20jcranmerexcept I don&#39;t think the C++ API for jsmime supports mailing list syntax
16:21jcranmerour present mailing list hack is not legal RFC 5322
16:21acemanrewriting that syntax would touch a lot of places
16:21jcranmersince addr-spec does need an @
16:23jcranmeryeah, I agree it would touch a lot of places
16:24acemanthen if the name of the list contains quotes , the result is &quot;name&quot; <&quot;name&quot;>
16:26acemanand for some cases with special characters in name (without quotes), the encoding again produces &quot;name&quot; <&quot;name&quot;> (not &quot;name&quot; <name> as in the first case)
16:26acemanso if I want to get the name of the list, I have a hard time to guess when to strip the quotes
16:29acemanit seems to me it works when sending because e.g. nsMsgCompose::LookupAddressBook() we just check all mailinglists in AB and see if one matches the string
16:31acemanbut I&#39;d like a way without hitting the databases
16:35jcranmerwith the current syntax, there is none
16:36jcranmerhave I mentioned how sucky the current way of doing things is?
16:46acemanwe have MailServices.ab.mailListNameExists()
16:46acemanbut that again hits the database
16:46acemanand it seem I&#39;ll have to ask it for both versions (whether name or &quot;name&quot; exists)
16:48acemanha, naming a list to have quotes around the name does not work
16:48acemanI mean sending to the list does not work
16:49acemanprobably some of the C code removes the quotes and then the name is not found :)
16:54jcranmerthe C++ code parses the mailing list as a MIME header IIRC
17:30acemanquotes inside the name get escaped: &quot;na\&quot;me&quot;
17:30acemanbut quotes around are not
17:30jcranmerthere&#39;s probably a reason for that
17:32acemanok, how did you propose to parse this to get na&quot;me back?
17:33acemanthis is created by makeMimeAddress
17:35jcranmernsIMsgHeaderParser::parseDecodedHeader() should work
17:35jcranmeror you can use MimeParser.parseHeaderField directly
17:38acemanshould the flag be MimeParser.HEADER_ADDRESS ?
17:38jcranmeryep
17:41acemanit returns some object
17:47acemanyeah, array of { name, email ]
17:47jcranmeralmost, not quite
17:47jcranmerit returns an array of objects that are either {name, email} or {name, group}
17:49acemanand what is a group?
17:49jcranmeran array of {name, email} objects
17:49aceman:)
17:49acemanso when does that happen?
17:50jcranmerit&#39;s the mailing-list: ... ; syntax
17:51acemanbut we do not use it normally
17:51acemanso there is some support for it?
17:53jcranmerthe parser API supports it
17:56acemanso thanks, this does decode properly
17:56acemanit also removes the quotes from the list names surrounded with quotes
17:56acemanso actually decodes the name wrongly, but at least it is the same as when sending
18:33acemanjcranmer: so thanks for the help
18:52Fallensousmangoosta: there are a few, I don&#39;t know a best pick for the moment. I can check blame. Can you remind me of the bug number?
18:53sousmangoostayep
18:53sousmangoosta1362611
18:53sousmangoostathx Fallen
18:54Fallensorry for teh delayed response :)
18:56Fallensousmangoosta: glandium is a good pick
18:57sousmangoostano pb, ok thx u Fallen
19:38wsmwkaceman: 52.1.0 has been out
19:39acemanthanks
20:35wsmwkPaenglab: have you noticed the circle+? in Firefox prefs panes?
20:39Paenglabwsmwk: What do you mean with circle+? ? The help link?
20:39wsmwkPaenglab: yes
20:42wsmwkI see it in 0416 nightly. hadn&#39;t noticed it before today
20:44PaenglabThis would need help pages. And with the prefs in dialog there isn&#39;t much space to place such. It&#39;s since long time in FX. before it was on every page on the top right.
20:47wsmwkMight be a good idea for us. all the prefs dialings that have tabs certainly have space for it
20:51wsmwkJoeS: ^
20:54acemanhelp pages?
20:54acemaninternal files, or just link to web pages?
20:56Paenglabweb pages.
14 May 2017
No messages
   
Last message: 97 days and 20 hours ago