IRC log for #koha, 2018-11-23

← Previous day | Today | Next day → | Search | Index

All times shown according to UTC.

Time Nick Message
00:13 irma_ joined #koha
02:59 mtompset Nemo_bis, because it is the most current for CentOS.
03:00 No one has written more up to date instructions.
03:00 Well, sleep is screaming my name.
03:01 Have a good day, #koha eythian Nemo_bis tcohen
04:57 irma_ joined #koha
06:34 calire joined #koha
06:47 cait joined #koha
06:51 fridolin joined #koha
06:56 m23 joined #koha
07:00 cait ah, no marcelr
07:03 ashimema Koha should move to the more traditional cpanfile for its Perl deps really
07:04 And carton
07:12 cait i tihnk there is a bug somewhere about carton ;)
07:13 * cait is trying to figure out why Outlook is evil about umluats
07:14 marcelr joined #koha
07:14 marcelr hi #koha
07:16 cait hi marcelr
07:16 thx for taking a look a tthe bug
07:16 marcelr hi cait
07:16 cant reproduce it
07:16 cait I am thinking we might do something different between the DUE and sendmail
07:16 yeah, it's outlook
07:16 which... unfortunately still a lot of people use
07:17 marcelr too bad
07:17 cait it works for DUE ontices
07:17 we have a lot of encoding things int he code for sending things via message_queue
07:18 http://paste.koha-community.org/189
07:18 something of that might make the difference, as the sendshelf looks simpler
07:18 marcelr yeah encoding is fun..
07:19 cait when you said you can't reproduce it i tried again and to my home address... so that gave another clue
07:20 marcelr outlook--
07:20 look out for outlook
07:20 cait it's a tool... but it's front end stuff
07:20 you can't tell people what to use
07:20 especially OPAC uses
07:20 marcelr fsure
07:20 s/f//
07:23 cait i wonder if this is the thing: $message->{subject}= encode('MIME-Header', $utf8);
07:25 we do that for the mails send via message_queue
07:27 magnuse \o/
07:28 cait ?
07:28 still stuck on hwo to make it work for lists :(
07:30 ashimema Is there not a bug about moving all such mail to go via the message_queue 😉
07:30 magnuse queue all the things!
07:30 cait i tihnk that would not qualify as a bug fix maybe :)
07:30 ashimema I thought we'd actually got to that point.. but clearly not
07:30 cait but we do have cod enow to send immediately via message_queue
07:30 so that we can move it
07:30 ashimema Haha, indeed
07:30 cait it's used for password recovery
07:31 ashimema Indeed
07:31 cait worth trying to throw the MIME thing in here?
07:31 ashimema I think that's the route I would take
07:31 cait 114     # Analysing information and getting mail properties
07:31 115     if ( $template_res =~ /<SUBJECT>(.*)<END_SUBJECT>/s ) {
07:31 116         $mail{'subject'} = Encode::encode("UTF-8", $1);
07:31 117         $mail{subject} =~ s|\n?(.*)\n?|$1|;
07:31 ashimema I'm sure we could get that past kidclamp between us Cait 😉
07:31 cait ashimema: down to 17.11? ;)
07:32 ashimema It's the sort of thing I would attempt to backport
07:32 But yeah, could be challenging
07:33 alex_a joined #koha
07:33 ashimema I can have a proper look in a minute it you like.. coming from phone whilst making the girls breakfast at the minute
07:33 alex_a bonjour
07:34 cait bug 21874
07:34 huginn Bug http://bugs.koha-community.org[…]_bug.cgi?id=21874 normal, P5 - low, ---, koha-bugs, NEW , Umlauts broken in list email subjects
07:34 andreashm joined #koha
07:34 cait i have to run to work :( meeting at 9:30 too
07:34 will be back later and thx for any help!
07:34 marcelr++
07:34 ashimema++
07:34 magnuse cait++
07:35 cait bbiab
07:36 andreashm morni9ng
07:42 laurence joined #koha
07:43 reiveune joined #koha
07:43 reiveune hello
07:43 AndrewIsh joined #koha
07:49 ere morning!
07:51 magnuse hiya andreashm reiveune ere
07:51 reiveune \o_
07:54 marcelr hi ere magnuse reiveune andreashm
07:57 ere++
07:57 ere: only the bug title is still somewhat misleading
07:57 about item status etc
08:13 cait joined #koha
08:20 paul_p joined #koha
08:21 alexbuckley joined #koha
08:34 ere marcelr: Thanks! Right, updated.
08:34 marcelr: and hi!
08:35 marcelr o/
08:36 ere So if anyone's interested in trying out the new Z39.50/SRU server with Elasticsearch, please head to bug 13937, try it out and report your findings :)
08:36 huginn Bug http://bugs.koha-community.org[…]_bug.cgi?id=13937 new feature, P5 - low, ---, ere.maijala, ASSIGNED , Add a search backend agnostic Z39.50/SRU daemon that can inject item status MARC subfields
08:40 marcelr ere should it be NSO ?
08:41 AndrewIsh ?
08:41 wahanui somebody said AndrewIsh was ready and waiting to jump on QA comments :)
08:41 marcelr hmm
08:41 ashimema I think at the moment it's in the 'give me feedback' stage marcelr, rather than fully seeking a signogg
08:42 AndrewIsh marcelr:  Aloha! I've seen your comment, I've just not got round to acting on it yet :)
08:42 ashimema I can grab AndrewIsh for you as soon as he's in marcelr
08:42 oh.. there he is
08:42 marcelr no big deal
08:42 thought he was in already
08:43 ashimema: the mime stuff in Letters is not something for 18.11 any more
08:43 imo
08:43 ashimema firs point release I reckons ;)
08:44 marcelr hmm this is really bad stuff'
08:44 ashimema ?
08:44 fridolin https://24timezones.com/horloge_mondiale.php#/map
08:44 marcelr nasty code
08:44 ashimema oh the email stuff in general
08:44 yeah
08:44 fridolin this is really nice to follow people local time ;)
08:45 marcelr encode and mail::sendmail and mime
08:45 hi fridolin
08:45 ashimema there's some really funky handling
08:45 marcelr and do not comment such things..
08:46 cait marcelr: i agree this looks not nice
08:47 i remember in a very early version we had to fix the due notices subject
08:47 the more that i think of it
08:47 marcelr funny if you search for MIME in the git log, you have very old hits only
08:47 ashimema I don't understand why there's so many versions of encoding the subject line
08:47 cait i have another failing mail client now - so it's 2 : 3
08:47 my web.de provider's gui shows it broken too
08:47 f you search for broken umlauts and subject you get quite a few things
08:48 marcelr one of the last main changes was in 3.20 with bug 11944
08:48 huginn Bug http://bugs.koha-community.org[…]_bug.cgi?id=11944 major, P5 - low, ---, jonathan.druart, CLOSED FIXED, Cleanup Koha UTF-8
08:48 AndrewIsh marcelr:  I'm a bit confused with bug 21833. It has come from magnuse rather than me. I had a brief discussion with magnuse about it. I believe it's a non-issue. Bug 20941 switched away from displaying medium in the UI since we seemingly don't use it any more, we store the material type as an illrequestattribute, so I think any patches related to medium are no longer needed.
08:48 huginn Bug http://bugs.koha-community.org[…]_bug.cgi?id=21833 normal, P5 - low, ---, koha-bugs, Signed Off , Request type is not displayed on ILL request detail page
08:48 Bug http://bugs.koha-community.org[…]_bug.cgi?id=20941 normal, P5 - low, ---, andrew.isherwood, RESOLVED FIXED, Displaying requests does not display request material type
08:49 cait bug 4420
08:49 huginn Bug http://bugs.koha-community.org[…]w_bug.cgi?id=4420 normal, P5 - low, ---, koha-bugs, CLOSED FIXED, umlauts or diacritics in subject of notices are deleted
08:49 marcelr thx AndrewIsh could you add a comment on the bug
08:49 AndrewIsh marcelr:  Yep, will do
08:49 marcelr and change status ;)
08:49 wahanui marcelr: that doesn't look right
08:49 wahanui joined #koha
08:50 marcelr hah
08:50 ashimema `subject => $string`
08:50 `subject => Encode::encode('UTF-8', $string)`
08:50 `subject => encode_qp(Encode::encode('UTF-8', $string)`
08:50 ere marcelr: Not yet, I want to add some documentation and tests and make sure all the necessary field mappings are in place before NSO
08:50 marcelr ok
08:51 ashimema: it is messy
08:51 like try this o try that
08:51 ashimema indeed it is
08:51 cait apart from it being messy... any clue for a quick fix that might work?
08:52 i am aware we should fix this 'right' but having something quick to test/verify woudl be nice too
08:52 marcelr a quick fix might be breaking other things that we do not see immediately
08:52 this is a bad area
08:52 cait it is... but the sendshelf is pretty isolated
08:53 marcelr ok convince me with solid code
08:53 :)
08:53 cait i guess if we require a complete rewrite this is not going to get fixed
08:53 because i am not able to do that
08:54 marcelr we agree about fixing it
08:54 cait I think touching the code for queued mails would be quite dangerous
08:54 even if it looks quite messy
08:54 marcelr sure
08:54 cait it took a few iterations to make it work, however it works...
08:55 marcelr well it might break one of these days with a perl library update in debian
08:55 or so
08:55 cait it could
08:57 i think encoding utf-8 is not enough if i understand correctly you have to add the encoding in front of the subject line
08:57 subject: =?utf-8?Q
08:57 something like this (without the Q)?
08:58 marcelr hmm not sure i like that, we are sending that to mail::sendmail later on
08:59 but feel free to try
08:59 cait i am trying to take a look at the 'raw' email but failing in outlook (of course...)
08:59 i am not sure what it looks like now really atm
08:59 marcelr use another email address..
09:00 cait i still need a gui that lets me look at it
09:00 Subject: Ihre Liste: Jugendbücher
09:00 it loks like we might be missing something
09:00 marcelr but i understood that you also have local changes
09:01 cait nope
09:01 marcelr so how are they affecting things
09:01 cait indexes
09:01 stuff like that
09:01 nothing in the perl and pm
09:01 marcelr ok
09:01 that makes it easier
09:01 cait ok, Search.pm... but that shoudl not be the problem :)
09:01 marcelr nothing in Letters
09:01 cait exactly
09:01 not that crazy
09:01 marcelr :)
09:02 cait the problem is i haven't found a broken display that is easy to share yet
09:02 ashimema we override content-type in sendshelf as compared to elsewhere
09:02 cait outlook and web.de... pretty specific
09:02 ashimema in SendAlerts it's set to `text/plain; charset="utf-8"`
09:03 cait SendAlerts is another issue we haven't tested that
09:03 ashimema here we set it to `multipart/mixed; boundary=\"$boundary\""`
09:03 cait shelf is not using SendAlerts
09:03 yeah, because it has an attachement
09:03 the list and cart emails include the marc file attatchements
09:03 at least that's what I reasoned
09:04 ashimema wonder if adding `; charset="utf-8"` there may be the magic bullet
09:04 no.. but I think SendAlerts is working correct right?
09:04 cait not sure, the email shows in outlook with Charset: utf8
09:04 what i read so far
09:04 you have the content-type which is the content
09:04 encoding
09:04 marcelr does alert have attachments ?
09:04 cait but the subject encoding has to be specified separately
09:04 no attachements for alert
09:05 https://en.wikipedia.org/wiki/Unicode_and_email
09:05 o use Unicode in certain email header fields, e.g. subject lines, sender  and recipient names, the Unicode text has to be encoded using a MIME "Encoded-Word" with a Unicode encoding as the charset.
09:06 https://tools.ietf.org/html/rfc2047
09:07 ashimema right
09:07 found it
09:07 marcelr double encodign?
09:07 ashimema line 1286 in C4/Letters.pm
09:07 * cait anxiously waits for the solution
09:08 marcelr but we are talking about sendshelf ?
09:08 ashimema we need to encode it as a 'MIME-Header'
09:08 cait ashimema: you are slightly in the wrong place
09:08 but that#s my guess too
09:08 ashimema yeah.. I was looking at how it's working elsewhere
09:08 marcelr adding that to sendshelf, you mean
09:08 ashimema I think so yeah
09:09 cait ah
09:09 marcelr only a bit of trial and error, but ok
09:09 ashimema it's a really odd thing
09:09 cait agreed
09:09 it seems uncessearily hard
09:10 ashimema decode as MIME-Header, re-encode as MIME-Header then re-encode again as UTF-8
09:10 that's a really odd set of events
09:10 marcelr thats weird stuff i commented on
09:10 cait i get the encode... but why the decode first?
09:11 marcelr the decode looks like a noop
09:13 cait something like this? https://mensuel.framapad.org/p/bX5LJBCDeI
09:14 marcelr i think that subject needs a encode_qp call in sendshelf ashimema
09:14 the mime stuff goes via encode_qp in sendshelf
09:15 but subject seems forgotten
09:15 ashimema I think I'd try both of those cait
09:15 marcelr line 139
09:15 ashimema try your change
09:15 marcelr $mail{subject} = Encode::encode("UTF-8", $mail{subject});
09:15 ashimema then try marcelr's suggestion
09:16 marcelr add encode_qp
09:16 ashimema I 'think' they may actually be doing the same thing.. just via different routes
09:16 marcelr i should try one at the time
09:16 cait i found this in a german perl posting:
09:16 $subject = encode_qp(encode("UTF-8", $subject));
09:16 btw, you can change the framapad if you want :)
09:16 marcelr yeah thats what i said
09:16 add qp
09:17 cait second option added
09:17 marcelr touch only one line in sendshelf thats isolated
09:17 cait btw, not sure why it uses Encode::encode
09:17 can someone explain that?
09:17 it seems it has a use Encode at the top
09:17 marcelr np cait its more specific
09:17 cait use Encode qw( encode );
09:17 but not really needed, right?
09:17 marcelr i like Encode::encode
09:17 but its not needed
09:17 cait so more a style thing
09:17 ok
09:18 we have to change this on a server so i have an email sending system
09:18 will get back to you
09:18 marcelr i will test it too
09:18 cait cool thx
09:19 just remembered... we have sendshelf and opac-sendhself
09:20 code looks identical
09:20 in that part...
09:22 * ashimema reads back up.. got kids in the ears
09:22 cait thx for the discussion and help :)
09:23 * andreashm reads with interest, but has nothing to add
09:24 ashimema man, the internals of encode_qp are 'complex' to say the least
09:24 cait i wonder why the other codes doesn't seem to use it at all
09:25 http://git.koha-community.org/[…]=grep&s=encode_qp
09:25 only used in sending lists and the cart
09:25 and this code is copied back and forth so not too surprising
09:26 ashimema indeed
09:26 I'm not sure
09:26 that's what lead me to think encode_qp was pretty much equivilent to encode('MIME-Header', $string)
09:26 cait it could be
09:26 standardizing would be nice
09:26 cleaning up email might be a project for 19.05
09:27 too much code is repeated there as well
09:27 ashimema I want to grab at least some of bug 8000 too
09:27 huginn Bug http://bugs.koha-community.org[…]w_bug.cgi?id=8000 enhancement, P5 - low, ---, charles.farmer, Failed QA , Test mode for notices
09:27 ashimema it was fairly rigorous about making sure email construction went through Koha::Email
09:28 cait yes, we funded that
09:28 when adding the replyto/returnpath
09:30 i hope nothing new has crept in not using it
09:31 marcelr hmm Koha::Email is just a few lines?
09:31 nothing special
09:32 ashimema indeed
09:32 marcelr we should consolidate there
09:32 ashimema but it was the start of standardisation ;)
09:32 agreed
09:32 marcelr move all sendmail calls
09:32 ashimema +++
09:32 interesting reading cait: https://metacpan.org/pod/Encode::MIME::Header
09:33 your memory of =?UTF-8?B? being prepended was sort of close
09:34 that's what the encode('MIME-Header', $string) call does.. only it also base64 encodes the string
09:34 encode_qp I 'think' doe the exact same thing
09:34 as I suspected
09:34 it certainly base64 encodes the string
09:34 not sure about adding the prefix though.. still digging through the perl version fo the code (I don't read C)
09:35 they impliment different RFCs
09:35 one is RFC2045 and the other is RFC2047
09:38 marcelr cait: Your list : perl J=C3=BCgend =E9=81=BA=E8=B7=A1=E7=99=BA=E8=A6=8B =C3=A1=C3= in gmail
09:38 hmm no unicode problems :)
09:38 it really is qp now
09:39 we need to tell it somewhere else too
09:43 ashimema :confused: is subject part of the message body or a message header
09:43 lol
09:43 cait i am about to test qp
09:43 can#t change stuff ont he server myself so it takes a little longer
09:43 we are testing qp first - is this what youfound working marcel?
09:44 ashimema it's a header.. reading the RFC's then I think encode_qp is the wrong route
09:44 that's for message bodies acording to the corresponding RFC
09:45 cait might explain why it was left out from the subject in the first place
09:45 btw the code IS slightly different in opac-sendshelves.... because of course it is
09:46 marcelr yeah only qp is not enough
09:46 cait encode_qp breaks stuff
09:46 marcelr ^^
09:46 cait should we try the other solution?
09:46 Ihre Liste: Jugendb=C3=BCcher=20=
09:46 or are we looking for version 3?
09:46 marcelr :)
09:47 ashimema I think try version 1 from your framapad cait
09:47 cait i have to slightly change it for opac sec
09:47 jan joined #koha
09:48 marcelr body,,header..subject..isofile
09:48 cait hm?
09:48 marcelr it is certainly mixed
09:49 ashimema encode_qp encodes according to RFC2045 (message bodies), encode('MIME-Header') encodes according to RFC2047 (and RFC822) (message headers)
09:49 sameee joined #koha
09:49 ashimema it's supper silly that they all have different ways of encoding the data!
09:49 marcelr i think we are really confusing Mail::Sendmail here
09:50 which is kind of broken already
09:50 ashimema haha.. hilarious.. email headers and sms messages are encoded in the same way
09:50 * ashimema giggles
09:50 marcelr lets start koha encoding
09:50 cait can someone check my change wuickly?
09:51 marcelr submit a patch
09:51 cait heh
09:51 marcelr i do not know what you changed
09:51 cait will od if we figure it out
09:51 just adding this line before line 139
09:51 $mail{subject} = encode('MIME-Header',$mail{subject});
09:51 ashimema we should really use Email::Simple.. it's the defacto recommended perl email handler now
09:51 marcelr ok cait
09:52 cait ok, back to changing sever files brb
09:56 marcelr cait: looks good so far on gmail and outlook
09:56 ashimema marcelr++
09:56 cait++
09:57 well done guys
09:57 marcelr not too fast
09:57 cait testing now too
09:57 marcelr why does it work?
09:57 ashimema haha.. I'm waiting for more test results..
09:57 cait interesting question... the why
09:57 ashimema that ++ was more for teamwork++ :)
09:58 cait actually that is a fun Friday activity
09:58 apart fromt eh bug part :)
09:58 marcelr you should not try to encode to utf8 first
09:58 cait oooh
09:58 nice in outlook
09:59 trying all 3 other guis now i have access to
09:59 ashimema I 'think' the initial encode to utf8 actually just sets the perl internal utf8 flag
10:00 marcelr this line in Letters is bad: my $subject = encode('UTF-8', $message->{'subject'});
10:00 but luckily it does nothing
10:00 since you converted to mime already
10:00 it is a noop
10:01 ashimema mmm, I think you're right
10:01 cait i can't test thunderbird form here, but everywhere else it's nice now
10:01 we only added the bold line here: https://mensuel.framapad.org/p/bX5LJBCDeI
10:02 marcelr hmm no bold but i only picked your line above and it seems to work
10:02 just debugging it now
10:02 cait ok, i will wait
10:03 if you want to submit i will sign off
10:03 in theory... i think the change might be neede for the cart too and we never ran into it becuase the german translation of the hard-coded subject has no umlauts
10:03 ashimema there's a bold line?
10:03 cait 'ihr warenkorb' or something like that
10:03 shows for me in the pad
10:03 ah... lost connection
10:04 sorry, not using that often, hsould be bold now
10:04 ashimema aha.. see it now
10:04 yeah.. that's the magic bullet :)
10:04 marcelr Your list :  =?UTF-8?B?IHBlcmwgSsO8Z2VuZCDpgbrot6Hnmbr​oposgw6HDqcOtw7PDuiDpgbHokZHnmbE=?=  =?UTF-8?B?6KaC?=
10:04 ashimema man this whole area needs a few rounds of code golf
10:05 whose up for 16 holes somewhere next cycle
10:05 ;)
10:05 marcelr cait so that is your mime header
10:05 ashimema :(
10:05 sameee good evening everyone
10:05 marcelr encoded with ?UTF-8?B
10:05 ashimema did marcelr just say it's still borked?
10:05 marcelr no it is fine
10:05 just debugging to see what is going on
10:06 ashimema coolioes
10:06 marcelr cait: submit this change, i will sign and qa; i can understand the why now
10:06 but do we need in a few places ?
10:07 * ashimema is happy to SO or QA if you want the three different people ;)
10:07 marcelr even better
10:07 ashimema I think we probably need to do it all over at some point as you say
10:07 perhaps open a little series rather than holding this one specifically up?
10:07 thoughts cait?
10:08 marcelr cait fell from her chair
10:17 cait no just in the kitchen making tea
10:17 reading back
10:17 marcelr did you see the same for cart ?
10:18 cait the cart has no umlauts
10:18 marcelr ok
10:18 cait hm but i cant try with one of the goethe libraries - russian or so
10:18 give me a minute?
10:18 marcelr yeah it should be the subject line
10:19 cait trying twith greek
10:19 if i can figure out where the cart is..
10:20 i'd say yes... that's borked too: Το καλάθι σας
10:20 in an intersting way
10:20 i'll go and write ap atch for cart and list then? or should we deal with them separately?
10:20 marcelr one report is fine now
10:21 ashimema brill
10:21 marcelr ashimema: still a theoretical question about mime header
10:21 look at =?UTF-8?B?IHBlcmwgSsO8Z2VuZCDpgbrot6Hnmbr​oposgw6HDqcOtw7PDuiDpgbHokZHnmbE=?=  =?UTF-8?B?6KaC?=
10:21 ashimema one bug is great.. I was just being pragmatic thinking you wouldn't nessesarily have time to fix all cases this moment
10:21 marcelr why did the encode generate two parts here ?
10:22 it is about this fix
10:22 cait 2 words?
10:22 ashimema interesting
10:22 marcelr you see two =?UTF-8 parts
10:22 the last part is one chinese char
10:22 ashimema yeah.. I see the two parts
10:22 cait oh
10:22 * ashimema reads the RFC to see if there's a reasonably reason
10:22 marcelr but the first part contains chines chars too
10:22 cait is there a space?
10:23 marcelr this was my subject : perl Jügend 遺跡発見 áéíóú 遱葑癱覂
10:23 cait wow
10:23 marcelr more spaces than one
10:23 cait you went all out :)
10:23 marcelr it is my default utf8 test
10:24 cait dont be distracted by this thing
10:24 alex_a_ joined #koha
10:24 marcelr just theoretical curiosity
10:24 laurence joined #koha
10:25 ashimema marcelr:
10:25 * ashimema sent a long message:  < https://matrix.org/_matrix/med[…]aQUxwSCizBufrANin >
10:25 ashimema that's straight from the RFC
10:26 I bet you hit the character limit
10:26 marcelr ok you found it
10:27 ashimema :)
10:28 marcelr lets agree on that now ;)
10:28 * ashimema started reading RFC's when he had to build a json module a while back and had allot of help from he Mojolicious guys
10:28 ashimema :)
10:29 marcelr otherwise i will be testing mime the rest of the day..
10:29 ashimema hehe
10:29 it's a deep hole that one ;)
10:29 as i say.. we should eventually switch to Email::Simple.. then a rather large amount of this 'fun' just goes away
10:29 rjbs has done a pretty darn good job of cleaning up that area of perl
10:30 kohaputti joined #koha
10:30 cait witing up the patch for cart and lists
10:33 marcelr yeah we should consolidate here
10:35 cait i am giving no technical explanation in the commit message
10:35 please feel free to amend/change
10:36 ashimema sure thing
10:36 cait hm
10:36 my devbox is broken
10:36 ashimema Shall I SO thne marcelr?
10:36 cait yay
10:36 imight attatch it the old fashioned way
10:36 ashimema lol.. once the patch is there
10:36 hehe
10:36 * ashimema goes for a quick :tea:
10:38 marcelr change Letters later like bug 21875
10:38 wahanui marcelr: that doesn't look right
10:38 wahanui joined #koha
10:38 huginn Bug http://bugs.koha-community.org[…]_bug.cgi?id=21875 enhancement, P5 - low, ---, koha-bugs, NEW , Handling subject line in Letters.pm
10:39 cait it's allgood... but i got the wrong bug... will fix
10:42 alex_a joined #koha
10:42 cait r3eady
10:42 marcelr ok
10:43 cait now i will try and fix mx devbox :(
10:43 marcelr cait: we are not there yet
10:43 ashimema I'm not so sure about bug 21875 marcelr... as you say.. in the majority of cases one would hope the initial decode is a noop. `Text parts of the input string which do not contain MIME encoded-word stay unmodified in the output string`.. however.. if someone happens to pass an already MIME encoded header into the subject then this would prevent a double encode?
10:43 huginn Bug http://bugs.koha-community.org[…]_bug.cgi?id=21875 enhancement, P5 - low, ---, koha-bugs, NEW , Handling subject line in Letters.pm
10:43 ashimema or have I misunderstood
10:43 cait marcelr: oh, what did i miss?
10:44 marcelr cait: we should not encode to utf8 after encoding to mime
10:44 so we should remove that line
10:44 in sendbasket
10:45 a subject line should never be utf8
10:45 do you want me to correct or do you ?
10:46 cait hm it always has the encode utf8 bit
10:46 we tested with it
10:46 marcelr i tested without
10:46 cait i trust you
10:46 we can repeat the test later
10:46 i am rewriting
10:46 marcelr hmm
10:46 ok
10:46 cait or we can do another test first
10:47 marcelr no i want to remove them
10:47 cait ok, I am rewriting
10:47 marcelr and then you can test again
10:47 cait so
10:47 marcelr this was my change:
10:47 -    $mail{subject} = Encode::encode("UTF-8", $mail{subject}); +    $mail{subject} = encode('MIME-Header',$mail{subject});
10:47 cait we get rid of the encode UTF-8 bits
10:47 right?
10:48 marcelr ^
10:48 cait that's just the other way i proposed and tested it :)
10:48 we tested mime first, utf 8 after
10:48 ah sorry
10:48 missed the -+
10:48 ok, on it
10:48 marcelr yeah they are two lines on one line
10:48 patch formatted lines
10:49 no cr
10:50 ashimema: the initial decode is plain wrong
10:50 if somehting is not a mime header, you should not decode it as such
10:50 happily it does no harm
10:51 ashimema but what if something is a mime header?
10:51 marcelr than we have no problem i guess
10:51 we should show it as it is
10:51 we could try with such subject lines they should not be changed to normal text
10:51 cait updated patch uploaded
10:51 marcelr cool
10:52 ashimema what I'm saying is that I think the decode does no harm to normal strings and would correctly decode if it's passed a mime header in the string
10:53 but if we remove it.. if we pass a mime header I think the encode would end up encoding that again
10:53 * ashimema writes a test for that hypothesis
10:53 marcelr well i do not completely understand your point probably
10:53 but quite convinced that the decode should be removed
10:54 subject is just text no mime
10:54 so should be handled as text not mime
10:55 cait looks good final test
10:55 cait we will do another too, just to mae sure
10:55 marcelr hmm still seeing something but maybe correct later
10:56 $mail{subject} =~ s|\n?(.*)\n?|$1|; on sendshelf intranet
10:56 yeah thats wrong
10:57 do you correct or do I?
10:57 cait ^
10:57 ashimema ok.. I'm happy now marcelr
10:57 marcelr always fine
10:58 ashimema no.. hang on
10:58 I'm not
10:58 marcelr i will so on cait patches if you do qa after that
10:58 21875 is for later
10:58 ashimema http://tpcg.io/aJeCgR
10:58 cait i didn't touch that i think?
10:58 ashimema okies :)
10:59 cait marcel, please correct
10:59 marcelr nice find ashimema three parts
10:59 ashimema hehe
10:59 it just shows the possability to double encode
11:00 I sohuld write more in that test to show the decode does no harm.. but the removal of it can do harm.. if that makes sense now
11:00 cait ah... we should move the line, is it that?
11:00 ashimema hope it clarified what I was getting at
11:00 marcelr i will cait
11:00 ashimema: will have a look
11:00 ashimema hehe
11:00 cait we had tested withthe opac part that is slightly different
11:00 marcelr true cait
11:01 i did too
11:01 will test it still
11:01 cait thx
11:02 looks good stil for me without the utf-8 encode bit
11:02 marcelr it should
11:02 cait just confirming
11:02 :)
11:04 keeping my hands off it now, right?
11:04 marcelr yes please
11:04 still testing so no status change yet
11:05 cait np, got to fix this devbox and tons of otther things... not boring here :)
11:07 ashimema I've updated my repl for you marcelr.. should make it more clear
11:07 coyas joined #koha
11:08 kohaputti_ joined #koha
11:08 cait marcelr++ ashimema++ :)
11:09 quick lunch bbiab
11:10 marcelr cait++
11:11 ashimema cait++
11:26 marcelr ashimema: go ahead with bug 21874
11:26 huginn Bug http://bugs.koha-community.org[…]_bug.cgi?id=21874 normal, P5 - low, ---, katrin.fischer, Signed Off , Encoding broken in list and cart email subjects
11:26 marcelr what did you mean with I've updated my repl for you marcelr.. should make it more clear
11:26 not so clear :)
11:28 ashimema take a look at http://tpcg.io/aJeCgR with regards to bug 21875
11:28 huginn Bug http://bugs.koha-community.org[…]_bug.cgi?id=21875 enhancement, P5 - low, ---, koha-bugs, NEW , Handling subject line in Letters.pm
11:28 ashimema that's all it meant ;)
11:29 hopefully it clarifies why I'm worried about removing the decode line
11:29 marcelr what did you change? i think i saw that before
11:30 yeah diff zz zz2 gives no changes
11:30 will study it more carefully
11:30 have to go now
11:30 ashimema didn't know you could digg codeingground
11:30 marcelr bye #koha
11:31 magnuse have fun #koha - see you in december!
11:38 coyas joined #koha
12:46 cait back
12:49 tcohen morning
12:52 cait morning tcohen :)
13:08 tcohen hi :-D
13:10 ashimema tcohen
13:10 :)
13:11 tcohen so, can I help you with those selenium tests?
13:11 :-D
13:11 ashimema perhaps.. pretty please
13:11 tcohen I will launch it again
13:12 on jenkins
13:12 so I can see the screenshot
13:12 ashimema see the screenshot...
13:12 I forgot I'd backported that
13:13 tcohen they expired already
13:13 * ashimema has entirely forgotten where they go
13:14 tcohen when you see the tests output
13:14 there will be a screenshot link
13:14 they get stored (if possible) in https://framapic.org
13:14 ashimema indeed
13:15 where is the link though.. in the console output somewhere
13:15 ;)
13:39 koha-jenkins Project Koha_18.05_U18 build #125: STILL UNSTABLE in 26 min: https://jenkins.koha-community[…]ha_18.05_U18/125/
13:39 kmlussier joined #koha
13:51 Archie^ joined #koha
13:51 Archie^ yo Koha!
13:52 Hey guys any idea why Bulkmarcimport.pl is giving out error at the end of every loaded file? Use of uninitialized value in concatenation (.) or string at /usr/share/perl5/MARC/File/XML.pm line 397, <GEN25> chunk 1001.
13:56 cait bad data
13:57 Archie^ is it tho? it happens even if i load xml file with only 1 record
13:57 cait maybe you are missing something
13:57 does it load ok with the gui?
13:57 Archie^ ya
13:58 i mean, it loads fine with the bulcmarcimport
13:58 it just gives this error :D
14:03 tcohen @later tell paul_p ping
14:03 huginn tcohen: The operation succeeded.
14:08 pastebot "tcohen" at 172.22.66.9 pasted "ashimema" (10 lines) at http://paste.koha-community.org/190
14:10 calire left #koha
14:13 paul_p pong
14:34 coyas joined #koha
14:57 mtompset joined #koha
14:57 mtompset Greetings, #koha
14:57 @seen tcohen
14:57 huginn mtompset: tcohen was last seen in #koha 54 minutes and 17 seconds ago: <tcohen> @later tell paul_p ping
14:57 mtompset @later tell tcohen Thanks for the help yesterday. We discovered a bug in the build script. :)
14:57 huginn mtompset: The operation succeeded.
14:58 tcohen a
14:58 mtompset bug 21876
14:58 huginn Bug http://bugs.koha-community.org[…]_bug.cgi?id=21876 minor, P5 - low, ---, mtompset, ASSIGNED , Untracked changes prevent building
14:58 mtompset The everything_is_committed function is broken.
14:59 I'm working on a fix.
14:59 With a little more meaningful messages.
14:59 tcohen great, mtompset
14:59 mtompset I have a quick and dirty fix, but that would only frustrate forgetful me when it turns out it is untracked, and not uncommitted. :)
15:00 tcohen we hate quick and dirty fixes
15:00 * mtompset smirks, "Sure we do."
15:02 groans, "So many old branches to delete. I need to keep them better maintained."
15:20 coyas hello
15:31 mtompset Greetings, coyas
15:32 coyas hi mtompset
15:43 fridolin left #koha
16:09 koha-jenkins Project Koha_18.05_D8 build #122: STILL UNSTABLE in 23 min: https://jenkins.koha-community[…]oha_18.05_D8/122/
16:12 reiveune bye
16:12 reiveune left #koha
16:17 koha-jenkins Project Koha_18.05_U18 build #126: STILL UNSTABLE in 30 min: https://jenkins.koha-community[…]ha_18.05_U18/126/
16:22 ashimema WHAT!
16:22 it passes for me now
16:26 cait have a nice weekend all!
16:27 cait left #koha
16:27 cfhong2000 joined #koha
16:28 ashimema enjoy cait
16:34 koha-jenkins Project Koha_18.05_D9 build #123: STILL UNSTABLE in 45 min: https://jenkins.koha-community[…]oha_18.05_D9/123/
16:35 ashimema ack
16:35 it's failing differently
16:35 and I can't replicate it still
16:41 mtompset ashimema, untracked files?
16:41 I blame untracked files for everything now. ;)
16:41 ashimema lol
16:41 shouldn't be.. considering it's on Jenkins ;)
16:42 mtompset can't hurt to check. :) But you are probably right. :)
16:43 ashimema nah.. my repo is clean here
16:43 and the tests all pass here
16:43 so it's not that I missed committing the fix
16:43 at least that's one thing ruled out
16:43 mtompset Untracked...
16:43 no non-committed.
16:43 different
16:44 git status --porcelain
16:44 " M" uncommitted
16:44 "??" untracked/uncommitted
16:45 laurence left #koha
16:47 coyas joined #koha
16:47 coyas hi
16:48 why i need to install PazPar2(metasearch tool)
16:48 i am installing koha for source code
16:48 from source code
16:49 * mtompset in funny voice, "Just do eeeeeet!"
16:49 koha-jenkins Yippee, build fixed!
16:49 Project Koha_18.05_D8 build #123: FIXED in 23 min: https://jenkins.koha-community[…]oha_18.05_D8/123/
16:49 ashimema WHAT.. why did it now pass
16:49 I didn't change sqat
16:50 mtompset the tiny magical dwarves are listening.
16:55 koha-jenkins Project Koha_18.05_U18 build #127: STILL UNSTABLE in 27 min: https://jenkins.koha-community[…]ha_18.05_U18/127/
17:08 cait joined #koha
17:31 coyas i need explanation on SRU/Zebra https://justpaste.it/1r3kd
18:03 mtompset SRU... let's you share your koha info with other people via z39.50, if I recall correctly.
18:03 @later tell coyas SRU... let's you share your koha info with other people via z39.50, if I recall correctly.
18:03 huginn mtompset: The operation succeeded.
18:04 kmlussier joined #koha
18:06 dpk_ joined #koha
19:58 mtompset @seen eythian
19:58 huginn mtompset: eythian was last seen in #koha 22 hours, 52 minutes, and 28 seconds ago: <eythian> ithan: probably fine, yeah
19:58 mtompset @later tell eythian So, how do I ignore the spelling of Sheat? in the Koha QA tools?
19:58 huginn mtompset: The operation succeeded.
20:00 mtompset Anyone know how to tell the QA Test Tools, that Robin's last name is not spelt poorly?
20:17 ashimema Lol
21:34 sameee joined #koha
21:39 koha-jenkins Project Koha_18.05_U18 build #128: STILL UNSTABLE in 26 min: https://jenkins.koha-community[…]ha_18.05_U18/128/
21:40 Project Koha_18.05_D9 build #124: STILL UNSTABLE in 26 min: https://jenkins.koha-community[…]oha_18.05_D9/124/
21:49 rh_sam joined #koha
21:50 rh_sam hi i need help  ---- while removing koha instance got error ERROR 1133 (28000) at line 1: Can't find any matching row in the user table
21:50 please help
21:52 is anyone there?
22:33 mtompset @seen ashimema
22:33 huginn mtompset: ashimema was last seen in #koha 2 hours, 16 minutes, and 7 seconds ago: <ashimema> Lol
22:34 mtompset Are you still lurking here, ashimema?
22:35 @later tell tcohen I put something up on bug 21876.
22:35 huginn mtompset: The operation succeeded.
22:37 mtompset @later tell ashimema If you are curious about Robin's last name being triggered in a QA Test tool... check out bug 21876. You'll have to rename it to .pl to get the tools to pick it up, but that's what triggered my comment.
22:37 huginn mtompset: The operation succeeded.
22:57 mtompset bug 21876
22:57 huginn Bug http://bugs.koha-community.org[…]_bug.cgi?id=21876 minor, P5 - low, ---, mtompset, ASSIGNED , Untracked changes prevent building
22:57 mtompset Yes, still haven't fully run through testing to my satisfaction. :)
22:58 Have a great weekend, #koha.

← Previous day | Today | Next day → | Search | Index

koha1