2018-11-14 GnuCash IRC logs
00:44:49 *** Mechtilde has joined #gnucash
00:58:59 *** gncbot has joined #gnucash
00:59:44 *** warlord has quit IRC
00:59:58 *** warlord has joined #gnucash
01:12:12 *** ChanServ sets mode: +qo warlord warlord
01:12:14 *** warlord sets mode: +o gncbot
01:20:31 *** jotrago has quit IRC
01:28:22 *** WebManOfFesto has quit IRC
01:28:38 *** WebManOfFesto has joined #gnucash
01:29:42 *** mdforbis_ has joined #gnucash
01:31:34 *** mdforbis has quit IRC
01:34:20 *** jralls_afk has joined #gnucash
01:36:01 *** warlord has quit IRC
01:47:10 *** jralls_afk is now known as jralls
01:50:31 *** gjanssens has joined #gnucash
01:50:31 *** ChanServ sets mode: +o gjanssens
01:50:36 <gjanssens> .
01:52:32 <jralls> gjanssens: You might be amused to know that I just pushed a change to gdk-quartz that shrinks the font size to match better with what native apps display.
01:52:50 *** Simon has quit IRC
01:53:47 <gjanssens> jralls: nice. That will reduce the number of complaints/requests on our user list
01:54:32 *** Mechtilde has quit IRC
01:54:54 <gjanssens> Except for those that already had compensated for it in css ;)
01:54:57 <jralls> Should do. Depending on when Mattias Classen does the next gtk release I may have to patch it for GC3.4.
01:55:18 <jralls> They can get reading glasses. ;-)
01:56:22 <gjanssens> LOL
01:57:41 *** Simon has joined #gnucash
01:58:27 <jralls> But that's only one scaling issue. Changing screen scaling messes with mouse event coordinates and we've gotten a bunch of bug reports about that. Now I'm studying the Apple docs trying to figure that out.
02:01:31 <jralls> There seem to be several coordinate systems in play: User, Backing Store, and Device. User is always scaled in printer's points. Device is scaled at whatever Quartz thinks about the physical display, and backing store is either printer's points or 1/2 points (ie 144/inch) depending on whether the display is scaled as Retina.
02:03:23 <jralls> Plus NSViews, which conceptually correspond to GdkWindows, have their own origins and translation functions.
02:14:32 *** jotrago has joined #gnucash
02:24:53 *** jralls_ has joined #gnucash
02:26:32 *** jralls has quit IRC
02:26:34 *** jotrago has quit IRC
02:27:00 *** jotrago has joined #gnucash
02:30:59 *** JayC has quit IRC
02:31:20 *** JayC has joined #gnucash
02:36:34 *** suntan has joined #gnucash
03:06:57 *** warlord has joined #gnucash
03:30:04 <warlord> .
03:31:53 *** Mechtilde has joined #gnucash
03:40:42 *** gour has joined #gnucash
03:40:46 *** Mechtilde has quit IRC
04:39:41 *** jotrago has quit IRC
04:40:19 *** jotrago has joined #gnucash
04:47:58 *** fell has joined #gnucash
05:04:37 *** fabior has joined #gnucash
05:05:54 *** jotrago has quit IRC
05:13:21 *** Aussie_matt has quit IRC
05:44:45 *** maschinenhans has joined #gnucash
05:47:24 *** jotrago has joined #gnucash
05:48:02 *** maschinenhans has quit IRC
05:54:47 *** fabior has quit IRC
06:05:02 *** fabior has joined #gnucash
06:13:02 *** fabior has quit IRC
06:29:00 *** ncv has joined #gnucash
06:40:39 *** User has joined #gnucash
06:41:03 *** gncbot sets mode: +o fell
06:50:35 *** User has quit IRC
06:51:15 *** gour_ has joined #gnucash
06:52:43 *** gour has quit IRC
06:54:37 *** Jimraehl1 has joined #gnucash
06:55:11 *** Jimraehl1 has left #gnucash
06:58:35 *** User has joined #gnucash
07:00:56 *** gour_ is now known as gour
07:12:15 *** User has quit IRC
07:16:08 *** User has joined #gnucash
07:42:54 *** chf has quit IRC
07:44:15 *** chf has joined #gnucash
08:07:14 *** oozer has joined #gnucash
08:08:01 *** jralls_ has quit IRC
08:12:00 *** mdforbis_ has quit IRC
08:25:11 *** Unhammer has quit IRC
08:27:42 *** Unhammer has joined #gnucash
08:41:06 *** storyjesse has quit IRC
09:16:13 *** fekepp has quit IRC
09:16:34 *** fekepp has joined #gnucash
09:24:46 *** fekepp has quit IRC
09:25:11 *** gncbot has joined #gnucash
09:26:43 *** fekepp has joined #gnucash
09:27:53 <gjanssens> fell: The last couple of days I have been working on the proof of concept to handle our documentation translations via po files instead of separate sets of xml documents.
09:28:05 <gjanssens> You can check my work so far here: https://github.com/gjanssens/gnucash-docs/tree/itstool
09:28:37 <gjanssens> That's an experimental branch on the gnucash-docs repository
09:29:15 <gjanssens> In it I have
09:29:35 <gjanssens> 1. reduced the English and German guide and help files to a smaller test set to limit the test scope somewhat
09:30:37 <gjanssens> The set I have retained was chosen to be big enough to be able to merge sunfish62's PR in which he moved large chunks from help to guide
09:30:58 <gjanssens> To allow us to evaluate how a big change affects the translations
09:31:48 <gjanssens> 2. Next I have lined up the xml tags in the English and German xml files
09:32:19 <gjanssens> The goal in this step is ensuring itstool extracts the exact same msgid structure from the English and German set
09:32:23 *** tienne has joined #gnucash
09:32:36 <warlord> gjanssens: cool
09:32:48 <gjanssens> This is influenced by tag order and punctuation/capitalization of msgids.
09:34:01 <gjanssens> 3. If the msgid order can be made exactly the same, a script will generate a real po file from the msgcatalogs extracted by itstool in step 2
09:34:26 <gjanssens> 4. Next I have written a number of extra makefile rules to help in this process. The commit log gives all the details.
09:34:50 <gjanssens> With all of that I have made a first version of de.po and committed it on the branch.
09:35:40 <gjanssens> Keep in mind this only deals with the limited subset of files. If the po route proves viable, we have to extend this to include the messages from all xml files
09:36:17 <gjanssens> 6. Then I have merged in sunfish62' changes to estimate its impact on de.po
09:36:59 <warlord> I'd be concerned about how to keep the msgid order consistent. I wonder if there is a "po.sort" function that could provide a "canonical" order to po files so you don't have to worry about the xml files.
09:37:29 <gjanssens> As it's a huge PR, this results in 670 translated messages, 314 fuzzy translations, 282 untranslated messages, where the original de.po had only one unstranslated message.
09:40:36 <gjanssens> warlord: msgid order in a pot file is pretty consistent as long as the order of the input files passed to itstool remains unchanged
09:40:53 <warlord> gjanssens: oooh... msgcat --sort-output
09:44:19 <gjanssens> I quickly tested that command. It only changes line wrapping compared to my output from itstool
09:44:40 <gjanssens> In this particular case that only makes things worse, so I'll not use it.
09:44:44 <fell> gjanssens: I am this week out of town, but who knows if i get some spare time here.
09:44:53 <warlord> gjanssens: but take two different files with different include-order and see if msgcat will canonicalize them in the same way?
09:45:29 <gjanssens> fell: no worries, I won't have much time either in the next few days, but I wanted to advance this issue at least a bit.
09:45:58 <warlord> did you verify that itstool will generate different po files if the xml tags are different in the sources?
09:46:01 <fell> Abiut the sort order: in code it's by file path, which is in most cases nice.
09:46:24 <gjanssens> warlord: we can worry about this when the conversions are done
09:47:16 <gjanssens> At this point the script I wrote to generate a po file from the two itstool generated files will choke if a msgid spans more than one line.
09:47:23 <warlord> gjanssens: okay. I figure it would be easier to test now, versus trying to ensure the xml remain consistent across languages.
09:47:25 <fell> Until now I worked on doc translation with 2 diff editor windows:
09:47:35 <gjanssens> itstool is nice enough to never write such lines.
09:47:45 *** Mechtilde has joined #gnucash
09:48:23 <fell> In the first I see for changes in C and in the second I compare C with de and translate,
09:48:30 <warlord> gjanssens: maybe you need --strict, or --no-wrap
09:48:34 <gjanssens> warlord: I did verify that itstool will generate different po files (pot files actually) if the xml tags are different
09:48:56 <fell> They are nice synced by the sect-ids.
09:49:51 <gjanssens> warlord: about "trying to ensure the xml remain consistent across languages"
09:50:10 <gjanssens> The whole point of the exercise is to eliminate language specific xml (source) files
09:50:30 <gjanssens> This consistency is only needed once to convert the current files into po files
09:50:58 <warlord> Oh, if it's a one-off then maybe it doesn't matter.
09:51:10 <gjanssens> Once that's done, the language specific xml files will be a build product
09:51:38 <gjanssens> At that point only one set of documentation sources will exist.
09:51:51 <warlord> AH. Okay, then I rescind my concerns .
09:51:53 <gjanssens> This will be largely in English with a few language specific sections
09:52:24 <gjanssens> These language specific sections will only be included in the built xml files for the target language
09:52:29 <fell> One weak point of po based translation:
09:54:12 <fell> If the souce is bad designed like our htdocs, the change of an URL like bugzilla.gnome -> bugs.gnucash invalitades many strings.
09:55:14 <gjanssens> That's true. On the bright side all languages would know this string requires updating.
09:55:39 <gjanssens> In the other scenario with independent xml sets per language, such changes might go unnoticed.
09:55:44 <warlord> gotta run.. Back in a bit.
09:55:45 *** warlord has quit IRC
09:55:58 <fell> So one preparation should be to put all URLs in entities.
09:56:29 <gjanssens> yes and no.
09:56:41 <gjanssens> Uh, yes
09:56:53 <gjanssens> That's one solution and a useful one
09:58:04 <gjanssens> I'm not sure whether we should do it before or after conversion to po though.
09:58:23 <fell> I have to think, if we should also put other parts like the content of codes glossary in entities.
09:59:02 <gjanssens> What's codes glossary ?
09:59:17 <fell> (for pure xml files you can change URLs over all LANGs)
10:01:03 <fell> https://wiki.gnucash.org/wiki/Translation#The_glossary_file
10:01:13 <gjanssens> In general I'm tempted to use entities sparingly.
10:01:41 <gjanssens> If there are too many, contributors may not remember all of them.
10:02:34 <fell> We have there ~200 terms
10:02:53 <gjanssens> How is that glossary file typically used in say PoEdit ?
10:04:23 <fell> I did not explore, how to use them in the different translation tools. It is also on my todo list.
10:04:42 <gjanssens> Ok
10:05:03 <fell> I saw somo of our editors put them as comment at top in their po file.
10:05:29 <fell> editors-> translators.
10:07:19 <fell> I know KTranslate can use additional files, but don't know in which formats.
10:09:11 *** benoitg has quit IRC
10:09:58 *** KevinDB has quit IRC
10:13:24 *** KevinDB has joined #gnucash
10:19:46 *** fekepp has quit IRC
10:28:36 *** warlord has joined #gnucash
10:36:28 *** finster has left #gnucash
10:38:05 <gjanssens> fell: fyi I have started collecting info on this experiment here: https://wiki.gnucash.org/wiki/Po_Based_Documentation_Translations
10:54:04 *** fabior has joined #gnucash
11:18:17 *** jerryq_ has quit IRC
11:24:57 *** Mechtilde has quit IRC
11:47:06 *** fekepp has joined #gnucash
11:54:01 <gjanssens> warlord: Martin has replied and he has uploaded the stable packages in his redmine instance. So we can use those links for the flatpak builds.
11:54:05 <gjanssens> I'm testing right now.
11:56:05 *** jerryq_ has joined #gnucash
11:59:39 <gjanssens> warlord: that worked fine. I have pushed the changes so you can retry when you find the time
12:20:23 *** Mechtilde has joined #gnucash
12:29:30 *** User has quit IRC
12:42:34 *** kael has joined #gnucash
12:43:55 *** Spash has joined #gnucash
12:53:45 *** fabior has quit IRC
13:00:32 *** KaiForce has joined #gnucash
13:04:52 *** calvinct has joined #gnucash
13:35:56 *** pilottage has joined #gnucash
13:53:33 *** ArtGravity has joined #gnucash
13:55:14 *** Unhammer has quit IRC
13:55:49 *** Mechtilde has quit IRC
14:08:17 *** jralls has joined #gnucash
14:08:24 *** frakturfreak has joined #gnucash
14:09:56 *** Unhammer has joined #gnucash
14:11:18 *** jralls has quit IRC
14:17:41 *** calvinct has quit IRC
14:20:08 *** silviu has joined #gnucash
15:12:54 *** jotrago has quit IRC
15:21:06 *** Mechtilde has joined #gnucash
15:35:18 *** kael has quit IRC
15:35:19 *** kael1 has joined #gnucash
15:37:48 *** kael1 is now known as kael
15:45:28 *** gour has quit IRC
15:49:06 *** jralls has joined #gnucash
16:03:22 *** bowe has joined #gnucash
16:28:48 *** gjanssens has quit IRC
16:33:53 *** Aussie_matt has joined #gnucash
16:38:36 *** calvinct has joined #gnucash
16:41:16 *** bowe has quit IRC
16:41:26 *** bowe has joined #gnucash
16:41:51 *** Aussie_matt_ has joined #gnucash
16:42:54 *** Aussie_matt has quit IRC
16:47:18 *** jralls has quit IRC
16:49:32 *** Mechtilde has quit IRC
16:49:45 *** Mechtilde has joined #gnucash
16:49:56 *** bowe has left #gnucash
16:50:58 *** frakturfreak has quit IRC
16:57:34 *** calvinct has quit IRC
16:59:33 *** ArtGravity has quit IRC
17:30:22 *** jralls has joined #gnucash
17:34:32 *** Mechtilde has quit IRC
18:10:20 *** kael has quit IRC
18:15:52 *** jerryq_ has quit IRC
18:44:32 *** bertbob has quit IRC
18:45:36 *** bertbob has joined #gnucash
19:43:48 *** jerryq_ has joined #gnucash
19:44:19 *** tienne has quit IRC
20:03:11 *** emilyprincess has joined #gnucash
20:29:16 *** fell has quit IRC
20:39:40 *** jralls has quit IRC
20:39:46 *** jralls has joined #gnucash
20:40:27 *** jralls has quit IRC
20:40:31 *** jralls has joined #gnucash
20:41:18 *** jralls has joined #gnucash
20:42:00 *** jralls has quit IRC
20:42:06 *** jralls has joined #gnucash
20:42:56 *** jralls has joined #gnucash
20:43:34 *** jralls has quit IRC
20:43:38 *** jralls has joined #gnucash
20:44:20 *** jralls has quit IRC
20:44:24 *** jralls has joined #gnucash
20:45:11 *** jralls has joined #gnucash
20:45:53 *** jralls has quit IRC
21:03:29 *** badger92 has joined #gnucash
21:48:46 *** oozer has quit IRC
22:46:19 *** gnomey has quit IRC
22:50:18 *** gnomey has joined #gnucash
23:30:29 *** jralls has joined #gnucash
23:33:29 *** jralls has quit IRC