05:30:27 #startmeeting i18n 05:30:27 Meeting started Mon Nov 29 05:30:27 2021 UTC. 05:30:27 This meeting is logged and archived in a public location. 05:30:27 The chair is paragan. Information about MeetBot at https://fedoraproject.org/wiki/Zodbot#Meeting_Functions. 05:30:27 Useful Commands: #action #agreed #halp #info #idea #link #topic. 05:30:27 The meeting name has been set to 'i18n' 05:30:27 #meetingname i18n 05:30:27 #topic agenda and roll call 05:30:27 #link https://fedoraproject.org/wiki/I18N/Meetings/2021-11-29 05:30:27 The meeting name has been set to 'i18n' 05:31:35 hi 05:31:38 hi 05:31:39 hi 05:31:58 hello 05:32:32 Hi all 05:32:42 hi 05:32:58 #chair pwu juhp tagoh suanand fujiwara 05:32:58 Current chairs: fujiwara juhp paragan pwu suanand tagoh 05:33:11 * mfabian ☺ 05:33:26 #chair mfabian 05:33:26 Current chairs: fujiwara juhp mfabian paragan pwu suanand tagoh 05:33:46 let's start this meeting 05:33:47 #topic Upcoming schedule 05:33:47 #info 2021-11-30 Fedora Linux 33 EOL auto closure 05:33:48 #info 2021-12-22 Change Checkpoint: Proposal submission deadline (Changes requiring infrastructure changes) 05:33:48 #info 2021-12-28 Change Checkpoint: Proposal submission deadline (Changes requiring mass rebuild) 05:33:48 #info 2021-12-28 Change Checkpoint: Proposal submission deadline (System Wide Changes) 05:33:49 #info 2022-01-18 Change Checkpoint: Proposal submission deadline (Self Contained Changes) 05:34:13 Tomorrow is EOL for F33 release 05:34:27 Change proposal dates are also coming soon 05:35:22 not much yet for F36 release milestones 05:36:03 Let's move to next topic 05:36:40 #topic Issues 05:37:02 #info #135: Fedora 33 Bug triaging (pnemade) 05:37:02 #link https://pagure.io/i18n/issue/135 05:37:25 Only 4 bugs are remaining on F33 bugs tracker 05:37:51 Shall we review them? 05:37:53 I think all have no opinion further on how to proceed. 05:38:05 Ok we can review them now 05:38:36 1) ibus bug https://bugzilla.redhat.com/show_bug.cgi?id=1751061 05:39:33 mfabian: is that still relevant for F34+? 05:39:58 ah probably not around 05:40:10 oops hi mfabian, my mistake 05:40:59 Oh, I should check on F35 ... 05:41:11 Is it fine to close culmus-fonts? 05:41:41 Thanks mfabian do check that bug today and provide your comment 05:41:53 2) ibus-kkc bug https://bugzilla.redhat.com/show_bug.cgi?id=1910907 05:42:50 No thoughts on that one 05:43:15 Okay 05:43:33 3) glib2 bug https://bugzilla.redhat.com/show_bug.cgi?id=1919760 05:44:14 fujiwara: any final thoughts? 05:45:02 What about culmus? 05:45:42 fujiwara, do provide your comment in that bug 05:45:48 4) culmus-fonts bug https://bugzilla.redhat.com/show_bug.cgi?id=1919932 05:45:52 It might be fixed but should be left until the auto-lose. 05:46:03 s/auto-lost/auto-close/ 05:46:15 okay 05:46:28 There are 2 things about this bug 1) I have not able to reproduce this 2) culmus-fonts moved ahead dropping bitmap fonts now 05:46:39 so I am not sure how to proceed further 05:46:55 Okay, let's move on then - I think the remaining bugs are okay 05:47:49 I added same in comment to culmus-fonts bug 05:48:10 Ok 05:48:17 thanks 05:48:27 if nothing more to discuss on this ticket, we can move to next issue 05:48:44 #info #137: Docs Beats for Fedora 35 (pnemade) 05:48:44 #link https://pagure.io/i18n/issue/137 05:49:01 (I think Vishal moved his 6 bugs earlier) 05:49:28 In last meeting it was decided to close this ticket but did not get closed. So I closed this now. 05:49:50 moving to next issue 05:50:38 #info #140: Get locale user stats using DNF Countme (vishalvvr) 05:50:38 #link https://pagure.io/i18n/issue/140 05:51:02 I don't see any recent comment on this ticket 05:51:19 hopefully Vishal will add some comment there 05:51:59 Yeah that one is kind of stalled, not sure how much headway we can make, though I still feel it would be useful 05:52:19 Maybe we should open an upstream ticket? 05:52:35 agree 05:53:32 I suggested it in the ticket 05:53:37 Thanks 05:53:58 Let's move to next issue 05:54:22 #info #142: Hunspell Dictionary dir change from myspell/ to hunspell/ (vishalvvr) 05:54:23 #link https://pagure.io/i18n/issue/142 05:55:58 I don't see any further comment in this ticket which tells about the test repo results 05:56:28 hopefully Vishal will add it 05:57:06 any comments? if not let's move to next issue 05:57:40 yes 05:58:30 #info #143: Tracker for Fedora 36 Changes (pnemade) 05:58:31 #link https://pagure.io/i18n/issue/143 05:59:09 if anyone is planning to submit any F36 Change then please do so early and add link to that Change page to this ticket 06:00:11 Ok let's move to next issue :) 06:00:26 and we can also discuss proposals here 06:00:39 yes 06:01:46 #info #144: Fedora 34 Bug triaging (pnemade) 06:01:47 #link https://pagure.io/i18n/issue/144 06:02:42 we need to start triaging F34 bugs soon 06:02:49 currently we have 63 bugs 06:02:56 aha 06:05:59 moving to next issue 06:06:12 #info #145: decide meeting chairperson for F36 cycle (petersen) 06:06:12 #link https://pagure.io/i18n/issue/145 06:07:10 Would anyone like to step up to chair this cycle's meetings? 06:08:03 Dunno if we need another week... to conclude 06:08:30 I think it is a good chance to gain experience chairing the meeting 06:09:28 let's discuss this offline to find new chair before next meeting 06:13:21 moving to next issue 06:13:30 #info #146: switch @fonts and langpacks-pa to google-noto-sans-gurmukhi-vf-fonts (petersen) 06:13:30 #link https://pagure.io/i18n/issue/146 06:13:54 If we all are okay for this change then I can proceed and will make the change in langpacks 06:14:17 and in comps 06:14:19 Please 06:14:28 It seems an oversight 06:14:41 The change required in package has already been done in rawhide btw. 06:14:59 tagoh, you mean in comps file? 06:15:01 Thanks 06:15:33 paragan, no, I meant changing the priority for fontconfig file for google-noto-sans-gurmukhi-fonts and -vf-fonts. 06:15:43 aha 06:15:45 thanks 06:16:28 I will make the required change in comps and langpacks then 06:17:06 #topic Open Floor 06:17:12 anything to discuss here? 06:18:27 Recently I noticed that when enchant2-aspell and aspell-en are installed, python3-enchanged uses aspell instead of hunspell for English automatially. Is aspell “better” than hunspell? 06:20:26 I think hunspell is better than aspell but some people still prefer to use aspell 06:20:55 if aspell is getting preferred, we can think of changing the preference if its possible 06:21:16 It is hardcoded somewhere in enchant. 06:21:34 It is possible to change it when using it from python3-enchant though. 06:21:35 I wonder how nuspell compares 06:22:10 I just googled and saw some old reddit thread saying aspell is better (algorithm) 06:22:58 Maybe aspell is less maintained? 06:22:59 I was undecided which was better, aspell shows many more suggestions usually, although these didn’t seem to be so useful to me: 06:23:10 Okay 06:23:11 $ python3 06:23:11 Python 3.10.0 (default, Oct 4 2021, 00:00:00) [GCC 11.2.1 20210728 (Red Hat 11.2.1-1)] on lin 06:23:11 ux 06:23:11 Type "help", "copyright", "credits" or "license" for more information. 06:23:14 >>> import enchant 06:23:17 >>> d = enchant.Dict('en_US') 06:23:20 >>> d.suggest('kamel') 06:23:25 ['Jamel', 'Camel', 'camel', 'Jamal', 'gamely'] 06:23:26 >>> 06:23:30 06:23:33 That is aspell. 06:23:39 hunspell would suggest only "Camel" and "camel". 06:23:47 mfabian, both have different dictionary base, hence suggestions will also change 06:23:47 aha 06:24:13 Well seems the algorithms are also different 06:24:13 mfabian, some languages may have good apsell dictionary coverage whereas hunspell is less 06:24:16 paragan, typically the number of suggestions using aspell is much bigger. 06:24:54 mfabian, you may want to check how many words are listed in its aspell and hunspell packages 06:25:08 aha 06:25:13 aspell dictionaries are apparently not plain text files like hunspell dictionaries are, so thy are not so easily readable. 06:25:28 tools are there 06:25:45 some packages converted aspell to its hunspell in Fedora 06:27:30 e.g. hunspell-csb, hunspell-az 06:27:32 I saw some statistics after googling that aspell more often has the correct word in the suggestions than hunspell, but if aspell just always suggests a lot more words, that makes it more likely that the correct word is included, even though the long list might not be so helpful. 06:28:33 Last offical hunspell release seems from 2018, aspell 2019, maybe both are active. 06:28:34 I still think it should not be aspell vs hunspell rather which dictionary per lang have better coverage aspell or huspell 06:28:56 The defaults in enchant are different per language. 06:29:15 For some languages enchant uses hunspell by default, for some languages aspell. 06:29:34 interesting 06:29:53 if available? 06:31:00 http://pyenchant.github.io/pyenchant/tutorial.html#brokers 06:32:43 “The following example states that for American English the MySpell provider should be tried first, followed by the aspell provider. For all other languages, the ordering is reversed:” 06:32:54 b = enchant.Broker() 06:32:54 06:32:54 b.set_ordering("en_US","myspell,aspell") 06:32:55 06:32:58 b.set_ordering("*","aspell,myspell") 06:33:01 06:33:04 b.request_dict("en_US").provider 06:33:05 :) 06:33:07 06:33:10 06:33:14 b.request_dict("en_GB").provider 06:33:17 06:33:33 :) 06:33:38 So if one does not change this, it has default orders to try, for English it tries aspell first by default but if aspell-en is not installed, it tries hunspell. 06:34:20 That’s how I found it because some guy on Debian had different results in the Typing Booster unit tests, that was because he had aspell-en installed and I had not. 06:35:24 I dunno - overall sounds like hunspell might be more effective generally for Typing Booster? 06:35:49 Depends on what for maybe. Spellchecking and completing is sort of independent. 06:36:20 For completing I do not use aspell dictionaries at the moment as they are not plain word lists. 06:37:36 mfabian, do you have any proposal here? 06:38:10 Not at the moment, it just surprised me that there are so many spellcheckers and that enchant chooses automatically depending on what is installed. 06:38:25 Okay 06:38:39 I may try to find out whether I can access the aspell dictionaries for completion somehow, currently I don't know how. 06:38:41 if you need any change, you can report bug against enchant 06:38:59 Ok 06:39:02 Thanks 06:39:17 any other thing to discuss, otherwise we can close this meeting 06:40:03 Thanks for joining this meeting 06:40:05 #endmeeting