That would be like if here in the States we had to start using metric. I can use metric , sure, I had to deal with lots of physics in Engineering, but do I think in metric no. If you said āit is 10 feet by 10 feetā I would intuitively know how big that is. If you said itās ā8 meters by 8 metersā Iād have to stop and think to do the math in my head to convert to feet.
Like I previously stated itās now just a single measurement file per language. Well, actually 4, because you have to include the seamly2d & ts qm files that would include ui texts like color / colour. Without changing the qm loading in the translations.pri file you have to load both the measurement and seamly2d qm files based on the language.
At any rate, the issue of having to deal with 56 ts files PER language that Peter experienced has been reduced to a single measurement ts per language.
Some of my early schooling was in mechanical drafting. Since machining is done in 100āths or 1000āths of inches we had to use engineering scales - which are inches divided by 10 rather than 2, 4, 8, 16, 32, 64, etcā¦ so itās not metric nor is it your standard imperial system.
I do have to say though, metric or engineering sure makes it easier to proportionatly scale something since theyāre both base 10 like our number system.
I went through all 19 measurement ts files and verified they all contain the source text entries from the source code, and removed all the vanished source texts as well as removing a bunch of CRās that shouldnāt be there. BTWā¦ for those interestedā¦ en_INā¦ is India English. There is no translation for UK English which would be en_GB.
Our language filenames can drop the nation prefix, with one exception ā Letās keep the zh_CN (Chinese Simple) file name because its very different from traditional Chinese, and is universally used. And if we want a UK translation, that would be en-GB for āEnglish (United Kingdom).ā
ISO Translation Code Table.
Current Language filename
New Language filename
Language Name
Best file selection
Description
cs_CZ
cs
Czech
p0
only 8 translations available. Use p0
el_GR
el
Greek
p998
Half the terms translated, p998 has the most
en_CA
none
Canadian English
Drop
English to English
en_IN
none
Indian English
Drop
English to English
en_US
none
American English
Drop
English to English
es_ES
es
Spanish
p0
Half the terms translated
fi_FI
fi
Finnish / Suomi
p0
negligible number of translations
fr_FR
fr
French
p998
three complete translations, p998 most consistent & concise
he_IL
he
Hebrew
Drop
no translations
id_ID
id
Indonesian
p0
one term translated
it_IT
it
Italian
p0
less than half translated
nl_NL
nl
Dutch
p998
Two full translations. Minor differences p0 and p998, p998 seems best
pt_BR
pt
Portuguese
p0
limited set of terms translated
ro_RO
ro
Romanian
p0
half translated only in p0
ru_RU
ru
Russian
p0
more than half translated. Only p0
uk_UA
uk
Ukrainian
p0
one third translated. Only p0
zh_CN
zh_CN
Chinese (Simple)
p0
two terms translated. Only p0
de_DE
de
German
p998
three complete translations. P998 most consistent & concise
No. We would still use the ISO HTML / XML codesā¦ the translation code table is just to identify what code is what.
And just a note for future referenceā¦ I would highly recomend that anyone editing / translating the ts files to use Linguist. I had to cleanup up a few other glitches besides the CRās in some ts files that are only going to get there by editing the xml files directly. Also some files had some sources out of order from the base file - which I didnt bother fixing at this time as it doesnāt affect translations. ā¦ it just makes it harder to diff 2 different files. Theoretically they all should be the identical except the translations.
As well as adding the British flag icon for use in the prefs.
So I was just checking the stats for the PR after the tests passed, and Iām thinking how did I change 1087 files? Oh yeahā¦ I forgot - 1045 useless ts files deleted.
@Douglas , you are faster than your shadow! Itās great that this problem has been solved. Making a translation now becomes a lot more feasible than it was! Nevertheless it will still require some effort to find people who can translate into all the target languages.
@slspencer, would this not be the right time to work out the mapping between measures used in the patternmaking systems and the standard Seamly2D names? If you can share your database, I can propose a mapping table including mapping onto the translated names (so that you can also use Aldrich in the Dutch version of the program).
Ah @Bernadette, als je de Nederlandse versie gebruikt en je botst op een rare vertaling, geef een gil. Alles is dan wel vertaald, maar hier en daar zitten nog scherpe randjes!
Well, Iāve been working with the code for like 5 years, so next to RT I probably know it better than anyone. In this case, while I generally knew how the measurement translations were handled, I didnāt pay that close attention to them. Once I understood what āsystemā part of the tsās was, it was like I said easier to just do than try to explain how to do. BTWā¦ TY for bringing up the issue.
Thatās always the problem. Some of the UI translations would be easy for anyone to do using Google Translate. For example Open, Close, Save, Save Asā¦ standard terms any application will have. Itās the sewing / drafting specific terms that requires some expertise with the terminology.
For what purpose? To translate? To do that weād be right back where we were with 1045+ more ts files. Besides, other than a setting in the Preferences, the whole PMS is not implemented. To implement translated pms terms would require a whole bunch more codingā¦ again the measurement translations are not generated by running Lupdate, but rather specifically calling the translator on harddcoded terms based on the localeā¦ and formerly the system.
@Grace had the right ideaā¦ in that specific pms measurements should just be a template .vit you loadā¦ or should just be left to users mapping their own custom vitās in what ever language. Actually IMO the whole PMS thing should just be removed from the app (for now) as it does nothing. Not to mention Iām tired of explaining to new users what it doesnāt do.
@Douglas, with the āmappingā I intended purely adding to the documentation. That is, seamly2D has its own standard naming system for measurements, but the user who is studying one of the patternmaking systems is helped in the documentation with a mapping table showing how the measurements from this system translate to the Seamly2D names. This would involve no changes to the code. Users can then decide to adopt the standard Seamly2D names, or make their own variables directly from the PMS, or adopt a dedicated .vit file if these exist. And there will be all the freedom to remove reference to the PMS in the code, as it has indeed no function and that is difficult to explain.
Aaah. That makes sense to me. This would uncouple the pms terms from the app code making it way easier to maintain. Sounds like a spreadsheet to meā¦ which could be also be translated.
@peterh & @Douglas, a very good idea. Perhaps a table in the wiki would be a great place for it and when people ask, they can be referred to the table to check out their PMS codes and the table can be included in the manuals.
Further to the translatingā¦ A page was started in the wiki for people to add the translations of a few of the terms in different languages hereā¦ Perhaps it can be added to so that users can translate the sewing terms?
Absolutley. I would guess though that for the most part most pms terms will be the sameā¦ for exā¦ I donāt think thereās 56 different ways to say āchestā. It would seem what one will find are terms like āchest scaleāā¦ which is not a direct measure, but rarher the chest measurement / 2. This is because youāre only drafting 1 /2 the pattern.
Iāve started an Excel doc & have entered 6 of the PMSās that I can find. I think we should have ALL the codes with their names & descriptions in it, but at least we have a startā¦
It would have been one thing if the translations were actually being used, but all the app was doing was loading 56 of 1096 qm files at run time or with a change in language. Ohā¦ and verifying that there were 1096 ts files between 56 pmsās and 19 languages in the translastions test at build time.
BTWā¦ in your opinion can I just remove all the PMS related code?
Well, I can provide a list of measurements for several ststemsā¦ including Master Designer for Men and Boys, and the Women & Childrenās systems. The Croonberg Supreme systems, The Poulin System (not on the list) and a few others that might not be in the PMS list.