Why is my Trados TagEditor so slow? Thread poster: Bjorn Candel
|
|---|
I've got Trados TagEditor 6.5.5.441 and yesterday I got a translation job that consisted of 97 percent 100% matches, which meant that the TagEditor mostly worked through the text looking for the next fuzzy match. However, it moved down through the text very very slowly, mostly line by line, and because of this the job took hours, although I only got paid for a few hundred words. The project manager told me that the TagEditor should go "directly" to the first fuzzy match.
Can I do so... See more I've got Trados TagEditor 6.5.5.441 and yesterday I got a translation job that consisted of 97 percent 100% matches, which meant that the TagEditor mostly worked through the text looking for the next fuzzy match. However, it moved down through the text very very slowly, mostly line by line, and because of this the job took hours, although I only got paid for a few hundred words. The project manager told me that the TagEditor should go "directly" to the first fuzzy match.
Can I do something about this? I've got 768 MB RAM, a 2.0 GHz Celeron chip and a 16 GB harddrive (with an external USB harddrive of 60 GB).
Any help would be much appreciated! ▲ Collapse
| | | | Jalapeno
Local time: 13:24
English to German
Bjorn Candel wrote:
I've got Trados TagEditor 6.5.5.441 and yesterday I got a translation job that consisted of 97 percent 100% matches, which meant that the TagEditor mostly worked through the text looking for the next fuzzy match. However, it moved down through the text very very slowly, mostly line by line, and because of this the job took hours, although I only got paid for a few hundred words. The project manager told me that the TagEditor should go "directly" to the first fuzzy match.
Can I do something about this? I've got 768 MB RAM, a 2.0 GHz Celeron chip and a 16 GB harddrive (with an external USB harddrive of 60 GB).
Any help would be much appreciated!
Hi Bjorn,
two things come to mind:
1. Maybe you have to adjust the penalties. Maybe Trados opens every segment because the matches only count as 97% matches because of some alignment penalty or something.
2. Have you tried opening a segment and then pressing Alt and x on the numpad of your keyboard? That should get you to the first fuzzy match...
Regards,
Jalapeno
| | | | Bjorn Candel 
Local time: 15:24
English to Swedish
TOPIC STARTER | The penalties are set at default value | Oct 20, 2005 |
Hello Jalapeno,
Thanks for your reply.
1. The TagEditor doesn't open all the segments, only 99% matches and lower. My penalties are set to 1,2,2,3,15, which I think is the default setting.
2. Using Atl+Num* instead of the Translate to fuzzy icon makes no difference. I've tried.
From your reply it sounds like you expect the TagEditor to run faster, which is good to know! In which case I'll keep looking for a solution to this problem.
Bjorn
Jalapeno wrote: Bjorn Candel wrote:
I've got Trados TagEditor 6.5.5.441 and yesterday I got a translation job that consisted of 97 percent 100% matches, which meant that the TagEditor mostly worked through the text looking for the next fuzzy match. However, it moved down through the text very very slowly, mostly line by line, and because of this the job took hours, although I only got paid for a few hundred words. The project manager told me that the TagEditor should go "directly" to the first fuzzy match.
Can I do something about this? I've got 768 MB RAM, a 2.0 GHz Celeron chip and a 16 GB harddrive (with an external USB harddrive of 60 GB).
Any help would be much appreciated! Hi Bjorn, two things come to mind: 1. Maybe you have to adjust the penalties. Maybe Trados opens every segment because the matches only count as 97% matches because of some alignment penalty or something. 2. Have you tried opening a segment and then pressing Alt and x on the numpad of your keyboard? That should get you to the first fuzzy match... Regards, Jalapeno
| | | | Jerzy Czopik
Germany
Local time: 13:24
Member (2003)
Polish to German
+ ...
| Pretlanslate the file(s) with Workbench | Oct 20, 2005 |
In TAG Editor then open the first segment as usuall with ALT+HOME and then use ALT+CTRL+NUM+ (close and open next no 100%). This will allow you to skip all pretranslated (and NOT paid) segments.
Your problem is the HDD.
I suppose you are using a notebook, as 16 GB is very few for a desktop system. Notebook HDDs are slow. And if you store data via USB it is even slower, with USB 1 somwhere between 8 and 12 mbit per second data rate, with USB 2 theoretically 40 mbit/s data rate ... See more In TAG Editor then open the first segment as usuall with ALT+HOME and then use ALT+CTRL+NUM+ (close and open next no 100%). This will allow you to skip all pretranslated (and NOT paid) segments.
Your problem is the HDD.
I suppose you are using a notebook, as 16 GB is very few for a desktop system. Notebook HDDs are slow. And if you store data via USB it is even slower, with USB 1 somwhere between 8 and 12 mbit per second data rate, with USB 2 theoretically 40 mbit/s data rate (AFAIR). This is extremly slow, if you take into considaration, that Trados is permamently searching the databae up and down. A good HDD (fast) system is a must for Trados.
Regards
Jerzy ▲ Collapse
| | |
|
|
|
Bjorn Candel
Local time: 15:24
English to Swedish
TOPIC STARTER | Is the harddrive design the reason? | Oct 20, 2005 |
Hello Jerzy,
Thanks for your reply.
No, the TagEditor is alreay skipping all the pretranslated (and not paid for) segments. It's just that the skipping takes such a long time. In fact, I just ran one of the files I've already translated once again, which means it now consists entirely of 100% matches. When I clicked on Translate to Fuzzy, it ran through the entire file (1.3 MB) without opening any segments, and it still took over 17 minutes for the TagEditor to reach t... See more Hello Jerzy,
Thanks for your reply.
No, the TagEditor is alreay skipping all the pretranslated (and not paid for) segments. It's just that the skipping takes such a long time. In fact, I just ran one of the files I've already translated once again, which means it now consists entirely of 100% matches. When I clicked on Translate to Fuzzy, it ran through the entire file (1.3 MB) without opening any segments, and it still took over 17 minutes for the TagEditor to reach the end of the file. With 9 files to translate, that's over two hours I'm not getting paid for!
Yes, I've got a laptop but I only use the C drive for translation work, not the USB. According to the documentation, I've got a "High-capacity enhanced-IDE". Are you saying that the TagEditor won't work properly on a laptop because of the design of the harddrive? I'm not sure I understand why the harddrive would need to figure in all this, if the file I'm translating is 1.3 MB, my memory files are 4 MB, and I've got 768 MB of RAM.
Best regards,
Bjorn
Jerzy Czopik wrote:
In TAG Editor then open the first segment as usuall with ALT+HOME and then use ALT+CTRL+NUM+ (close and open next no 100%). This will allow you to skip all pretranslated (and NOT paid) segments.
Your problem is the HDD.
I suppose you are using a notebook, as 16 GB is very few for a desktop system. Notebook HDDs are slow. And if you store data via USB it is even slower, with USB 1 somwhere between 8 and 12 mbit per second data rate, with USB 2 theoretically 40 mbit/s data rate (AFAIR). This is extremly slow, if you take into considaration, that Trados is permamently searching the databae up and down. A good HDD (fast) system is a must for Trados.
Regards
Jerzy ▲ Collapse
| | | | Jerzy Czopik
Germany
Local time: 13:24
Member (2003)
Polish to German
+ ...
| You are using the "Translate to fuzzy" command (ALT+*) | Oct 20, 2005 |
And therfore TAG Editor is opening the next segment, then it considers it as 100% match, closes it and takes the next. Believe me.
Try the other command ALT+CTRL+NUM+
Regarding the HDD - per design notebook HDDs must be slower than other desktop HDDs, as theier rotation speed is much slower. Desktop HDDs rotate nowadays with 7200 rpm or if you are using SCSI up to 15000 rpm, when notebook HDDs rotate with something about 4000 up to 5400 rpm. This makes the access time much lon... See more And therfore TAG Editor is opening the next segment, then it considers it as 100% match, closes it and takes the next. Believe me.
Try the other command ALT+CTRL+NUM+
Regarding the HDD - per design notebook HDDs must be slower than other desktop HDDs, as theier rotation speed is much slower. Desktop HDDs rotate nowadays with 7200 rpm or if you are using SCSI up to 15000 rpm, when notebook HDDs rotate with something about 4000 up to 5400 rpm. This makes the access time much longer, even if enhaced data transfer rate can be reached. When searching a database access time is the key factor.
If you have a desktop computer, try to use this for that specific task.
Do you have the lates patch for Trados 6.5? AFAIR there was an issune with very slow TAG Editor, adressed with some earlier version of Trados, don´t remember which. Consider upgrading to Trados 7, the offered functionality is much greater than with 6.5.
Regards
Jerzy ▲ Collapse
| | | | Bjorn Candel
Local time: 15:24
English to Swedish
TOPIC STARTER | Ah, Translate to Fuzzy is not the same as Next no 100% Open Get! | Oct 20, 2005 |
Oh, sorry! I didn't pay enough attention to your last entry. (You were saying "Alt+Ctrl+Num+" and not "Alt+Num*".) Yes, I've tried Alt+Ctrl+Num+ and the TagEditor just jumps through the text, very quickly! Wow, I wish I'd known about this yesterday!
So, what is the difference between "Translate to Fuzzy" and "Set Close Next no 100% Open Get"? When I go through the same document as before, but use Alt+Ctrl+Num+ instead, the TagEditor stops and opens a number of segments that are show... See more Oh, sorry! I didn't pay enough attention to your last entry. (You were saying "Alt+Ctrl+Num+" and not "Alt+Num*".) Yes, I've tried Alt+Ctrl+Num+ and the TagEditor just jumps through the text, very quickly! Wow, I wish I'd known about this yesterday!
So, what is the difference between "Translate to Fuzzy" and "Set Close Next no 100% Open Get"? When I go through the same document as before, but use Alt+Ctrl+Num+ instead, the TagEditor stops and opens a number of segments that are showing up as 100% matches in the Workbench/memory (as I've already translated the text and these segments are now in the TM). Does this mean that if I use "Set Close Next no 100% Open Get", the TagEditor "remembers" the segments that were not 100% matches before I translated the text?
Anyway, I'm very grateful for your help!
Bjorn
Jerzy Czopik wrote:
And therfore TAG Editor is opening the next segment, then it considers it as 100% match, closes it and takes the next. Believe me.
Try the other command ALT+CTRL+NUM+
Regarding the HDD - per design notebook HDDs must be slower than other desktop HDDs, as theier rotation speed is much slower. Desktop HDDs rotate nowadays with 7200 rpm or if you are using SCSI up to 15000 rpm, when notebook HDDs rotate with something about 4000 up to 5400 rpm. This makes the access time much longer, even if enhaced data transfer rate can be reached. When searching a database access time is the key factor.
If you have a desktop computer, try to use this for that specific task.
Do you have the lates patch for Trados 6.5? AFAIR there was an issune with very slow TAG Editor, adressed with some earlier version of Trados, don´t remember which. Consider upgrading to Trados 7, the offered functionality is much greater than with 6.5.
Regards
Jerzy ▲ Collapse
| | | | Jerzy Czopik
Germany
Local time: 13:24
Member (2003)
Polish to German
+ ...
| There is a big difference | Oct 20, 2005 |
Bjorn Candel wrote:
Oh, sorry! I didn't pay enough attention to your last entry. (You were saying "Alt+Ctrl+Num+" and not "Alt+Num*".) Yes, I've tried Alt+Ctrl+Num+ and the TagEditor just jumps through the text, very quickly! Wow, I wish I'd known about this yesterday!
So, what is the difference between "Translate to Fuzzy" and "Set Close Next no 100% Open Get"? When I go through the same document as before, but use Alt+Ctrl+Num+ instead, the TagEditor stops and opens a number of segments that are showing up as 100% matches in the Workbench/memory (as I've already translated the text and these segments are now in the TM). Does this mean that if I use "Set Close Next no 100% Open Get", the TagEditor "remembers" the segments that were not 100% matches before I translated the text?
Anyway, I'm very grateful for your help!
Bjorn
When you have a new document, which is not pretranslateg, the ALT+CTRL+NUM* (Close Open next not 100%) does not work or better has the same effect, as translate to fuzzy.
Translate to fuzzy means, that Trados (with any editor you are using, either TAG Editor and Word and some T-Windows) opens every single segment and applies the 100% match automatically, then closes the segment and opens the next, until it finds no 100% match.
The other method, Close Open Next not 100% can be applied only to pretranslated text and means, that the segment is closed and the next segment, which will be opened, is not necessarilly the next segment in your document, but simply the next no 100% match. This means TAG Editor does skip all pretranslated segments, as they are 100% matches. So the process is much quicker.
Regards
Jerzy
| | |
|
|
|
Bjorn Candel
Local time: 15:24
English to Swedish
TOPIC STARTER | Thanks, this has been a great help! | Oct 21, 2005 |
Dear Jerzy,
This has been enormously helpful! I've gone through old pretranslated files I've worked on in the past (both the original files as well as the translated files), using both Alt+Num* and Alt+Ctrl+Num+ and I've noted the difference. I've also done this with a clean TM. I realise now that the difference between these two functions (Alt+Num* and Alt+Ctrl+Num+) is crucial when I get a pretranslated file but no current TM from the client (or my own TM is not updated). The TagE... See more Dear Jerzy,
This has been enormously helpful! I've gone through old pretranslated files I've worked on in the past (both the original files as well as the translated files), using both Alt+Num* and Alt+Ctrl+Num+ and I've noted the difference. I've also done this with a clean TM. I realise now that the difference between these two functions (Alt+Num* and Alt+Ctrl+Num+) is crucial when I get a pretranslated file but no current TM from the client (or my own TM is not updated). The TagEditor just flies through the file!
Best regards,
Bjorn
Jerzy Czopik wrote: Bjorn Candel wrote:
Oh, sorry! I didn't pay enough attention to your last entry. (You were saying "Alt+Ctrl+Num+" and not "Alt+Num*".) Yes, I've tried Alt+Ctrl+Num+ and the TagEditor just jumps through the text, very quickly! Wow, I wish I'd known about this yesterday!
So, what is the difference between "Translate to Fuzzy" and "Set Close Next no 100% Open Get"? When I go through the same document as before, but use Alt+Ctrl+Num+ instead, the TagEditor stops and opens a number of segments that are showing up as 100% matches in the Workbench/memory (as I've already translated the text and these segments are now in the TM). Does this mean that if I use "Set Close Next no 100% Open Get", the TagEditor "remembers" the segments that were not 100% matches before I translated the text?
Anyway, I'm very grateful for your help!
Bjorn When you have a new document, which is not pretranslateg, the ALT+CTRL+NUM* (Close Open next not 100%) does not work or better has the same effect, as translate to fuzzy. Translate to fuzzy means, that Trados (with any editor you are using, either TAG Editor and Word and some T-Windows) opens every single segment and applies the 100% match automatically, then closes the segment and opens the next, until it finds no 100% match. The other method, Close Open Next not 100% can be applied only to pretranslated text and means, that the segment is closed and the next segment, which will be opened, is not necessarilly the next segment in your document, but simply the next no 100% match. This means TAG Editor does skip all pretranslated segments, as they are 100% matches. So the process is much quicker. Regards Jerzy ▲ Collapse
| | | | Vito Smolej
Germany
Local time: 13:24
Member (2004)
English to Slovenian
+ ...
SITE LOCALIZER | The moral of the story: "pretranslation can be fun";) | Oct 23, 2005 |
I realise now that the difference between these two functions (Alt+Num* and Alt+Ctrl+Num+) is crucial when I get a pretranslated file but no current TM from the client (or my own TM is not updated)...
Pretranslate your files. And/or create a new TM from the pretranslated input.
regards
smo
| | | | To report site rules violations or get help, contact a site moderator: You can also contact site staff by submitting a support request » Why is my Trados TagEditor so slow? | CafeTran Espresso | You've never met a CAT tool this clever!
Translate faster & easier, using a sophisticated CAT tool built by a translator / developer.
Accept jobs from clients who use Trados, MemoQ, Wordfast & major CAT tools.
Download and start using CafeTran Espresso -- for free
Buy now! » |
| | Anycount & Translation Office 3000 | Translation Office 3000
Translation Office 3000 is an advanced accounting tool for freelance translators and small agencies. TO3000 easily and seamlessly integrates with the business life of professional freelance translators.
More info » |
|
| | | | X Sign in to your ProZ.com account... | | | | | |
Login to reply/comment 
