|
|||
|
.NET, Delphi, ... - Sisulizer Localization Tool Support > Technical Support (You need to be registered at the forum to write) > Wish list for software localization tool > Localizing Delphi binary files – translation review and cost
|
| Moderated by: Renate.Reinartz, Markus.Kreisel, Jaakko.Salmenius, Ilkka.Salmenius |
|
||||||||||||||||
| Localizing Delphi binary files – translation review and cost - Wish list for software localization tool - Technical Support (You need to be registered at the forum to write) - .NET, Delphi, ... - Sisulizer Localization Tool Support | |||||||||||||||||
| Author | Post | ||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
|
|||||||||||||||||
|
Barbara Member
|
Dear Sisulizer Team, I use your software since August last year for localizing programs written in Delphi. Because of the fact that all those programs include many common units I decided to localize all 41 binary files using a single Sisulizer project. This approach allowed for fast and good translation of all the words by combining Sisulizer translation modes. Unfortunately, after the first iteration “my company – translator – my company” I noticed that your software doesn’t support translation review. Moreover, due to missing functionality it causes additional unnecessary translation costs. What I mean exactly: first of all I created single slp project including already mentioned 41 files. I sent this project to the translator and got it back, filled with the translation (ca. 90000 words). I tried to review the translation, but it was really a hard work – I had to review many units more than once, because they were included by more than one binary file (missing functionality no.1: marking the words as “reviewed” is not propagated within the whole project; company cost no. 1: my time spent on superfluous reviewing). Additionally, we had to pay our translation agency for 90000 words (despite of replications) – we were told they had to check all auto translated units (company cost no.2: cost of superfluous translation and review on translator side). After this first iteration I wrote some tool that removes superfluous units from Sisulizer project file: - in case of dfm files (-> Forms) only one occurrence of the file is allowed in a slp project - in case of pas files (-> Strings) only one occurrence is allowed, but it has to be a form merged from all occurrences of the same pas file Second project based on the same 41 binary files was definitively smaller – ca. 45000 words! I imported the translation from the “big” project and then I could meaningfully review it (company cost no.3: my time spent on writing the tool). Our translation agency liked the idea of smaller Sisulizer project and since then I use my tool every time when I send updates to it. Of course I wrote a second tool making the “small” project “big” again (company cost no.4: time spent on writing this tool). In my opinion the workarounds I had to make to protect company interest points to weakness of Sisulizer. It would be very nice to have two options like “minimize final project” (company -> translator) and “maximize translated project” (translator -> company). The project files contain all information you need. Kind regards, Barbara
|
||||||||||||||||
| |||||||||||||||||
| |||||||||||||||||
|
Markus.Kreisel Administrator
|
Hi Barbara, >>I use your software since August last year for localizing programs written in Delphi. Because of the fact that all those programs include many common units I decided to localize all 41 binary files using a single Sisulizer project. This approach allowed for fast and good translation of all the words by combining Sisulizer translation modes. Unfortunately, after the first iteration “my company – translator – my company” I noticed that your software doesn’t support translation review. Moreover, due to missing functionality it causes additional unnecessary translation costs.<< If you set Tools- General Settings - Translations - When modifying to "For Review" then you will get automatically this translation status. If you then filter for items to be reviewed you can set items to "Complete" after you reviewed them. >>What I mean exactly: first of all I created single slp project including already mentioned 41 files. I sent this project to the translator and got it back, filled with the translation (ca. 90000 words). I tried to review the translation, but it was really a hard work – I had to review many units more than once, because they were included by more than one binary file (missing functionality no.1: marking the words as “reviewed” is not propagated within the whole project;<< You could filter out duplicates and review everything only once. After you completed these you can set the "For Review" duplicate rows to "Complete". >>Additionally, we had to pay our translation agency for 90000 words (despite of replications) – we were told they had to check all auto translated units (company cost no.2: cost of superfluous translation and review on translator side). << that is really not fair play. They charged you the same for stings automatically translated by Sisulizer? And you need to filtered out duplicates before you send the files to them so the do not see the duplicates and so can not charge you? Is that what happend to you? I hear the first time that an agency does charge for duplicates. Nevertheless Sisulizer might have the right settings for you. (See screenshot). In the Exchange Wizard you find some setting to filter out duplicates ("All duplicate rows"). If you use this only the first string from the duplicate group becomes visible to the translator. After you get back the files you import them. Don't forget to check back the setting Tools- General Settings - Translations - When Importing. Then filter for "For Review". After you finished that you can use Project - Translate Duplicates. Now you filter for duplicates because you should also give the duplicates a short review. If I understood you right did you want to skip the review of the duplicates because you have so many of them. If so, then simply filter for duplicates, mark all rows in the translation grid (click the root in the translation tree), use Edit - Select All. Now use Row - Translation Status - Complete and you are done. Hope this helps Markus Attachment: 0132.JPG (Downloaded 37 times)
____________________ http://www.sisulizer.com - Three simple steps to localize |
||||||||||||||||
| |||||||||||||||||
|
|||||||||||||||||
|
Barbara Member
|
Hello Markus, thank you for the answer. Unfortunately the workflow you described doesn’t work in our case. Let’s translate the German word “Laufzeit” into English. Depending on the context there can be many various translations – delay, elapsed time, maturity… In Sisulizer it is not possible to filter out the duplicates in a context-dependent way. However, there should be such a possibility. Modification of the translation status should be context dependent too. If the list with duplicate rows contains over 400 entries (“Update Duplicate Strings” dialog), it is really hard work to find the relevant ones. >> (…) I hear the first time that an agency does charge for duplicates. << Unfortunately it is not only our reality. >> Now you filter for duplicates because you should also give the duplicates a short review. << Yes, I have to do it. And in this way I have to review 45000 words… As I mentioned last time, “minimize” and “maximize” functions would be very nice to have. Kind regards, Barbara
|
||||||||||||||||
| |||||||||||||||||
| |||||||||||||||||
|
Markus.Kreisel Administrator
|
Hi, >>Let’s translate the German word “Laufzeit” into English. Depending on the context there can be many various translations – delay, elapsed time, maturity… In Sisulizer it is not possible to filter out the duplicates in a context-dependent way. However, there should be such a possibility.<< so it would be helpful for you, to separate duplicate strings from duplicate strings who also have a similar context? While you have to check duplicate strings you would not need to check duplicate strings having a similar context. Is that right? You already did that in your applications. How similar is your context in two sources. How did your application made the pairs. In Sisulizer's projects the source is part of the context. So the context can be only similar. Can you give me some sample please how you filtered. While I don't see a chance to get minimized and maximized functionality in Sisulizer I can imagine that our R&D adds some helpful filter for duplicate context. Markus
____________________ http://www.sisulizer.com - Three simple steps to localize |
||||||||||||||||
| |||||||||||||||||
|
|||||||||||||||||
|
Barbara Member
|
Hello Markus, hello Sisulizer Team, I wish you all the best in the new year 2009! In my first post above I already mentioned the two basic filtering rules. However, below you can find some simple example how my "minimize" application works. To make this example more readable I use simplified slp notation. Let's assume we have 3 applications written in Delphi - A, B and C. Input slp project file: A.Forms.A1 B.Forms.B1First of all - when minimizing slp files my application ignores program names (in our case A, B and C). They are very important when maximizing, but for now only "Forms" and "Strings" nodes are relevant. Nodes "Forms": - Based on names, four forms are marked as globally unique (A1, A2, B1, B2). - Based on names, three forms are marked as potentially duplicated (CommonForm1, CommonForm2, CommonForm3). If all occurrences of a single form (for example CommonForm1) contain the same set of strings, only the first occurrence is copied to the final slp file. Nodes "Strings": - Based on names, four units are marked as globally unique (AA1, AA2, BB1, BB2). - Based on names, the unit CommonStrings1 is marked as duplicated. Because its every occurrence contains another set of strings, all the sets are merged. Output slp project file: A.Forms.A1In the final project file every duplicated string contains another context, which is globally unique. Due to it, it is easier to translate and review all the strings in a context-dependent way. Kind regards, Barbara
|
||||||||||||||||
| |||||||||||||||||
| |||||||||||||||||
|
Markus.Kreisel Administrator
|
Hi Barbara, so basically it would help you if there would be some Project - Translate Duplicates that honors the value plus a definable part of the context. In the moment it always uses by value only. I could think that the function shows some dialog to allow you to switch on "by context" and define a rule to define what part of the context should be ignored. In your case A|B|C.Forms.CommonForm1 A, B, and C should be ignored but Forms.CommonForm1 needs to be the same to make the match. Off course the original string (by value) needs also to be the same. I'll ask R&D if they like that idea. Markus
____________________ http://www.sisulizer.com - Three simple steps to localize |
||||||||||||||||
| |||||||||||||||||
| Current time is 08:20 am | |
| .NET, Delphi, ... - Sisulizer Localization Tool Support > Technical Support (You need to be registered at the forum to write) > Wish list for software localization tool > Localizing Delphi binary files – translation review and cost | |
Sisulizer software localization tool - Three simple steps to localize