Jump to content

Technical problems with uTS


Firon

Recommended Posts

Hi everyone,

We've rolled out some updates to the new translation system and there's naturally going to be issues with it.

Please report those issues here to keep the discussion in one nice, central location.

Known issues:

  • UTF-8 files with a BOM will not upload
  • Uploading langfile is slow
Link to comment
Share on other sites

  • Replies 87
  • Created
  • Last Reply

Firon, having an open online translation system may appear as optimal in order to have the translations updated as early as possible, but there is a big problem attached to that. As of now, it seems that the online translation system can be used by anyone, which will decrease the quality of the translations. This has already begun.

I have downloaded today the Romanian translation file after receiving yesterday the e-mail that it needed an update. To my surprise there are strings in it that were not translated by me. And what is worse, those strings were written incorrectly. You see, the Romanian language uses some particular diacritics (Ăă, Îî, Şş, Ţţ), like almost all the other european languages. The Romanian diacritics designate specific sounds, like all the other letters, for example Ţ is for the sound written in english as TZ and Ş for the sound written in english as SH. But because of pure laziness, many Romanians do not use these diacritics, using instead the letters that ressemble them the most. So, instead of using Ă, many Romanians use A; instead of Ţ, they use T; instead of Ş, they use S; instead of Î, they use I. This is completely wrong, because in Romanian there are also the letters A, I, S, T, which are pronounced completely different than Ă, Î, Ş, Ţ. Letting inexperienced Romanian translators to have access to the online translation system will have the result of having words written with A for both A and Ă, I for both I and Î, S for both S and Ş, T for both T and Ţ. This creates a lot of confusion and again, is not at all correct. All the good Romanian translations have diacritics. It is not in the interest of uTorrent to have a Romanian translation with strings written incorrectly.

So now I must check the Romanian language file string by string to catch and correct all the translations that were written incorrectly, without diacritics. This is extra work that I have not done until now, when I was the only one to modify the translation. Please do not allow online access to the Windows Romanian translation to other persons than me.

Link to comment
Share on other sites

Today I have tested the new system and like always try to first download and then upload the same file. I know that with the new system this is not necessary, but I tried: but don't work ok. The translation I download is the penultimate one (27206). The downloaded file is ok and is utf-8 with boom. When I try to upload that file the uTS says "Uploaded file encoding is wrong! File needs to be encoded in UTF-8! Click here to convert it" and then ask the source encoding. I think that there is a problem with this, because the file is already utf-8.

Link to comment
Share on other sites

Well, Schnurlos, your hypothesis could be true if it weren't for the fact that I have been translating uTorrent since 2007 and my translation was made from scratch. So, all this time in the Romanian language file there were no strings translated by someone else, and no strings written without diacritics. But as I said, in the file for the 27434 build there are strings translated by someone else and written incorrectly, without diacritics. My first thought was that the new translation system allows anyone to add translations. But since you have told me that this is not the case and that I am the only registered Romanian translator, there must be another explanation. If it's not a human error (a bad translator adding strings), probably it's a software error, a bug in the translation system. I have investigated the thing further and this is what I have discovered:

1) The Romanian language file for the 27434 build has a size of 190 KB, whereas the previous file downloaded by me had a size of 111 KB. So, it seems that a lot of strings were added, but for the most part they are already translated, because the system says that the translation is 98% completed. This is why I thought that someone else had access to the translation system and translated them.

2) As it turns out, a lot (if not the vast majority) of these extra strings are duplicates of the existing strings. But, they are duplicates mainly with regard to the the english string, because the translated string has two different versions. It seems like the translation system has retrieved form the database some old translations and inserted them into the file alongisde the existing ones. For example the new file contains many instances of identical english strings, with different translations, put one after another, like in the example below:

=============================

Run this program when the download finishes:

La terminarea descarcarii, porneste acest program:

Run this program when the download finishes:

Rulează acest program la terminarea descărcării:

=============================

The first translation was not made by me and is written incorrectly, without diacritics. The second translation is the one made by me. There are a lot of such string pairs (duplicated english string and two different translations) in the new file, and this explains its bigger size. The previous translation file had 5426 lines and the new one has no less than 15839. There must be a lot of duplicated strings inside it. This duplication is very easy to spot. For example I have found the following succession (sequence) of strings:

==============================

Maximum number of connected peers per torrent:

Maxim parteneri conectati per torrent:

Maximum number of connected peers per torrent:

Numărul maxim de parteneri conectați per torent:

Maximum number of connected peers per torrent:

Numărul maxim de parteneri conectați per torent:

Maximum number of connected peers per torrent:

Numărul maxim de parteneri conectați per torent:

Maximum number of connected peers per torrent:

Numărul maxim de parteneri conectați per torent:

================================

The first pair of strings is not translated by me, and of course is written without diacritics. However, the four other string pairs have my translation. In the previous language file only 1 instance of that string was present, whereas in the latest file there are 5 instances of it, including one with an old translation.

It is obvious that the new translation system has a bug which manifests itself by duplicating the translated strings and also inserting old translations from the database.

###################################

My first thought that there was another Romanian translator who had inserted his bad translations in the file was triggered by the fact that a little more than a year ago, someone had modified a little my translation for AIMP (an audio player) and tried to replace me as the Romanian translator for that software, then tried the same thing for uTorrent. Fortunately, both Artem Izmaylov (the author of AIMP) and Firon understood the situation and quickly reinstated me as the official translator. But for AIMP it didn't quite end there, because the translation thief opened a thread on the AIMP forum and began badmouthing me and complaining that he was unjustly punished by having his translation (in fact my translation, slightly modified) removed from the installation package. I was asked by Artem Izmaylov to give my opinion on the thread, so from September 2011 to February 2012 I have posted several messages explaining my point of view not only about the attempted theft, but also about the situation of software translation in Romania. It was a heated discussion, and maybe some uTorrent translators would want to read it in order to be prepared if they would find themselves in a similar situation in the future. The discussion is in english.

http://www.aimp.ru/forum/index.php?topic=29579.0

Link to comment
Share on other sites

@muntealb: I think that the new system of translation keep all the strings that was ever translated (from start of translation?), also if it is the same string. This is because before there was an ID basis of string-translation, this is why every time the id changed, also if the string was already translated, you have to retranslate it. But now the basis is not more the string id, but the string itself. Also if I don't understand why there is a lot of duplicated strings (for very common strings also 10 or more time)...

Link to comment
Share on other sites

So Firon should modify the uTS in that way, that just the (in the actual last build) used strings are available/visible/translateable. 1+

Now with uTS 1.3.3 there are 263 pages!

(Sorry, I did not recognize this before) :rolleyes:

EDIT: partially solved

Link to comment
Share on other sites

  • 2 weeks later...
Today I have tested the new system and like always try to first download and then upload the same file. I know that with the new system this is not necessary, but I tried: but don't work ok. The translation I download is the penultimate one (27206). The downloaded file is ok and is utf-8 with boom. When I try to upload that file the uTS says "Uploaded file encoding is wrong! File needs to be encoded in UTF-8! Click here to convert it" and then ask the source encoding. I think that there is a problem with this, because the file is already utf-8.

Today I try to upload an previously download translation and the error is always the same: Uploaded file encoding is wrong!

Link to comment
Share on other sites

Today I have tested the new system and like always try to first download and then upload the same file. I know that with the new system this is not necessary' date=' but I tried: but don't work ok. The translation I download is the penultimate one (27206). The downloaded file is ok and is utf-8 with boom. When I try to upload that file the uTS says "Uploaded file encoding is wrong! File needs to be encoded in UTF-8! Click here to convert it" and then ask the source encoding. I think that there is a problem with this, because the file is already utf-8.[/quote']

Today I try to upload an previously download translation and the error is always the same: Uploaded file encoding is wrong!

same here, can't import exported file. if i save exported file without BOM, uts hangs(it parses, parses that file and throws 404 error)...

Link to comment
Share on other sites

Try putting in the strings normally for now. I need to figure out why it's so slow - think it's something wrong with the SQL server configuration.

Also, I fixed the issue where using "download language" gave a lot of duplicate strings.

Link to comment
Share on other sites

Not sure if this has already been addressed, but for example my Latvian "Status of translation" is stuck on 99.94%, in spite of all lines being translated.

Maybe this is because under "Display strings from build:" latest built I can get is 27434, while the actual latest stable is 27568 (or there are just no new strings).

Link to comment
Share on other sites

Hi everyone,

We've rolled out some updates to the new translation system and there's naturally going to be issues with it.

Please report those issues here to keep the discussion in one nice, central location.

Known issues:

  • UTF-8 files with a BOM will not upload
  • Uploading langfile is slow
Link to comment
Share on other sites

Norsk - Norwegian (Nynorsk) - nn-NO Translation

Today I got this message:

Text from a new build of µTorrent for Windows is now available for translation.

-But the status of translation for build 27434 shows 100.00 % done.

However the earlier build 27206 and older has about 4 pages with untranslated strings and some of them with no text or only the words «The», «your», «simply» and incomplete sentences like «Your computer and».

There are also untranslated strings in µTorrent 3.2 that have appeared translated in earlier releases.

Best Regards

Vestrvegr/Bjørn I.

Translator of Norsk - Norwegian (Nynorsk) nn-NO

Link to comment
Share on other sites

i can confirm bug which reported muntealb, in his case diacritics are missing, but problem is deeper: there was version 27206 with it everything is ok and later in version 27434 strings was restored(imported) from previous old version. strings for 27434 was imported at least from 26753.

EDIT: it can be and vice versa, maybe problem is in export system. maybe export system exported not that version which i asked and then i uploaded that old file.

P.S. current Lithuanian translation is my personal backup, so it's not an example of bug.

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.


×
×
  • Create New...