Trados, WordFast, OmegaT... CAT tools just don't click with me.

[Nate is fine.]

I've recently given myself the task of making my in-house job easier/
more efficient. Not much research later, it looked like CAT tools
were gonna be just the tool for the job. My company will put out the
money if I can demonstrate with confidence that it will pay off...
That would be great because I would wind up with free OTJ training,
which I'd be able to use in potential future freelancing endeavors.

The thing is, I can't say it will pay off with any confidence.

I have typically had no trouble picking up new software. However,
even as I keep poking around with them, I find these three CAT tools
nothing but burdensome. I've even gone so far as cracking open the
manuals and following tutorials and such, to very little benefit.

I have to be missing something. Or there is a real possibility that
the nature of my work makes these tools impractical, or at least makes
other means more practical.

So maybe someone can enlighten me a little?
Here are some of the key frustrations I have with the programs. If
I'm mistaken, or if these complaint is offset by some wonderful
benefit, please clue me in. (or tell me to RTFM more closely?)

1. These programs really have no idea where to break a Japanese
(henceforth J) sentence, and are very very bad at breaking things like
section titles.

2. If a J sentence has more than one clause, and I need to reorder the
clauses in English, or worse yet, I need to reorder sentences in
English, I can't create any worthwhile data for the TM.

2.1. If I join the clauses of a multi-clause sentence to make sure my
TM is correct, the sentence has absolutely zero chance of ever
occurring again. This makes moot the whole process of using the CAT
tool.

3. Our company uses different visual styling between our E and J
documents. Undoing all the bad formatting is an awful waste of time.
Is using raw text my only alternative?

4. Juggling windows to figure out what tags are/ where in the document
I am slows me down. (Our documents can be pretty heavily formatted)

5. These programs match the number "3" to the number "5" with 98%
accuracy. I never had a single fuzzy match that didn't need a
complete rewrite.

6. The ugly, sterile environments and unnatural keystrokes make the
whole process much less pleasant than even Excel word-processing.

7. From what I can tell, the glossaries in these programs barely
interact with the program at all. I need to explicitly select a word
and enter it into the glossary, and then the CAT tool never ever
suggests the word to me again. I may as well have written it on
paper.

8. Based on what I'm hearing, any agency that requires one of the
tools uses it against the freelancer as a means to pay less... In the
case of a post I read here, the pay for the repeated phrases was
ZERO. This practice makes the software nearly worthless to the
translator, doesn't it?

I apologize for the negative tone, but I hope you CAT tool users can
show me the light.
Please feel free to comment on all or any single one of the points.
I don't need the software to light up the sky. I'm just looking for a
decent ROI, or even return on time invested.

Thank you.

Nate

[Cary Strunk]

Cary Strunk:

Thanks, Nate. Someone has needed to write this post for a loooooong time, at least in my opinion. Not much time to write now, so I will just say that (and I stress that this is only in my experience) CAT tools only pay dividends quickly for someone working on instruction manuals, legal texts, or other text types with a lot of repetition. I also think that the integration between the various parts of the tools isn't all that hot. I wish someone who designs software for Apple, Pure Digital, or Nintendo would come along and revolutionize the CAT tool "industry". Those are companies that can make pick-up-and-use software that doesn't demand that the user pore over a manual for hours on end. Ask Ryan Ginstrom, an extremely nice software developer who posts regularly on here, about his "Felix" tool. A lot of people seem to like it, and he wrote it specifically with Japanese in mind. I don't have enough experience with it to judge, but if it is as good as its reputation, that'd be the way to go.

The glossary features bother me too. Pray that you never have to work with MultiTerm.

BTW, don't apologize for a negative tone on my account! Recreational complaining can be a great catharsis.

Best wishes always,

Cary Strunk (who had to work with Trados for 3 years and still feels like he doesn't know the first thing about it)

[southwind_lan...@nihon-honyaku.com]

Following Nate's report on 2009/02/17 at 18:34,
Cary Strunk wrote on 2009/02/18 at 07:51:
> [...] CAT tools only pay dividends quickly for someone working

> on instruction manuals, legal texts, or other text types with
> a lot of repetition.

I once used OmegaT for a collection of tourism related texts that had
many exact word repetitions and where there was also a good number of
similarly worded expressions, and i can see how such a tool can be
useful on the long run for material of that sort, especially if you get
similar material again and again over time. In my case, much of what i
have been dealing with (even in regards to tourism) has been "one of
kind" text that required a creative approach.

It is, of course, somewhat oversimplified, but i might say my CAT tool
is the combination of Firefox and Google. ;-) Of course, with certain
standardized texts (family registers, etc.) i make reference to
previously translated documents that i've kept, but other than that my
first stops tend to be WWWJDIC and EIJIRO, followed by several other
online glossaries/ dictionaries and then i tap the vast expanse of the
internet as a whole, via search engines, (that way i also get, by
default, data from publicly accessible archives of translation related
mailing lists). Finally, when i still have doubts, i turn to the mailing
lists themselves, with new queries...

Regards: Hendrik

--

[Peter Clark]

Dear Unhappy CAT user,
 
What material do you translate? What applications are you using for layout? As it is tagged, I presume a DTP application like InDesign or FrameMaker.
 
Re ROI, OmegaT is free, Wordfast (Classic) can be used indefinitely for free with limitations on translation memory size, and Trados would pay itself off in a short time if you translate 50,000 words a month of the right kind of document.
 
I have used both Trados and Wordfast. I hated Trados, but that was probably mostly because I started out with Wordfast. I am still a Wordfast user, and wouldn't want to be without it even if it didn't serve anything up from the TM.
 
Sentences end in end of sentence punctuation (ESP). All CAT tools know what these are, and you can add others if you need to. Worfast has Western punctuation as the default, so you will need to add Japanese ESPs to let it know.
 
Some people translate clause by clause, segmenting each one. As you seem to have realized, these leads to either completetly meaningless English ordering or a completely meaningless TM. I don't recommend this. Translate sentence by sentence, maybe even joining the occasional sentence together for more meaningful English flow.
 
For an introduction to CATs and answers to your other questions, try http://www.your-translations.com/cat-wordfast.php
 
Peter the Great
 

---

Sell your car for just $40 at CarPoint.com.au It's simple!

[Nate is fine.]

So, no CAT zealots out there?

On Feb 18, 9:03 am, Peter Clark <peterclarkat...@hotmail.com> wrote:

> I have used both Trados and Wordfast. I hated Trados, but that was probably mostly because I started out with Wordfast. I am still a Wordfast user, and wouldn't want to be without it even if it didn't serve anything up from the TM.

Now I'm scratching my head. Without the TM coughing up the occasional
100%er, I can't see why anyone would ever use these infernal things.

I'll freely admit that I probably didn't give any of the programs
enough time. I suspect that neither a 30-day sample nor a 500 entry TM
limit is likely to pay off quickly or dramatically, given the variety
of stuff I work on around here. I had thought my work was pretty
repetitious, but maybe you're all accustomed to a considerably higher
degree of repetition.

I'm not gonna give up yet. I'll take a spin through the tutorial
Peter provided, and give Felix a try later today.

Thanks for the advice.

[Jacob Dunlap]

On Feb 18, 7:51 am, Cary Strunk <cary.str...@gmail.com> wrote:
> (and I stress that this is only in my experience) CAT tools only pay
> dividends quickly for someone working on instruction manuals, legal texts,
> or other text types with a lot of repetition.

I respectfully disagree. Although the memory function for pulling out
sentence-level matches may not fit in well with your particular line
of work, TM functionality is not the only benefit of CAT tools. I
find glossary functionality -- where you register terms in a database
and they appear more or less automatically in your translation -- to
be a life-saver, especially when working with long documents in which
terms like "image format conversion determination unit" and so on
appear hundreds of times. Being able to type those with a single
keystroke is a godsend, IMO. That's just one example.

To Nate:

1. I personally use Wordfast, and it rarely has any trouble breaking
up sentences properly. Patent claims can be a bit tricky, as the
ordering is quite different between E and J, but I've managed to work
out a decent system even for claims. AFAIK, the user can define what
punctuation will be used to break up sentences.

2. I'm not really sure what you mean by "worthwhile". In my
understanding, the idea of a TM isn't to create data that will give
you lots of hits all over the place, but rather to create data that
will give you quality hits only when appropriate.

2.1. Well, maybe, but what about the shorter sentences that may repeat
again? As I said above, it depends on what you are expecting to get
from your TM, but I use my CAT tool with the full understanding that
many of the sentences I end up inputting will never pop up again. So
be it -- the other advantages of CAT tools outweigh that "problem" for
me.

3. Complex formatting is an issue I rarely have to deal with (as I
work with patents) so you may have a point there.

4. Same as #3.

5. TMs propose fuzzy matches based on the percentage of characters
that match within other sentences in the TM, as well as some other
criteria that is probably defined on a tool-by-tool basis. Computers
aren't thinking beings, so what seems like a perfectly reasonable
concept to a human -- that "3" is completely different from "5" --
just doesn't "occur" to the computer, so to speak. That's one of the
sacrifices of allowing a computer to "think" for you.

6. Ugly and sterile -- I'll give you that. ;) Keystrokes, though,
can be customized. I have six keys I otherwise never use -- insert,
home, page up, delete, end, and page down -- assigned to the most
common CAT tasks and it makes navigation a breeze.

7. That sounds like a problem with the tool, or some other sort of
malfunction. Or -- and don't take this the wrong way -- you may not
be using it properly. Although Wordfast, for example, has its tricks
and quirks, it generally functions without a hitch, and I imagine
Trados and other CAT tools are the same.

8. The practice of paying nothing for 100% matches is the subject of a
lot of debate. I tend to disagree with the practice, but if an agency
requires the use of a CAT tool, and a translator is willing to use it,
that translator may end up with a lot more work than s/he would have
otherwise. In such a case, the CAT tool has significant value.

Of course, I don't necessarily _advocate_ the use of CAT tools -- I
think people should work however they want to. Personally, though,
I've found Wordfast to be incredibly useful and can't imagine working
without it. Like any professional tool, there is a definite learning
curve, though, so you have to go in with the expectation that you may
have to modify your working style and spend some time to learn how to
use it properly. Then again, I imagine the same thing was being said
about computers themselves back when they first were working their way
into translators' workflows. ;)

Whew. Sorry for the length of that.

Jacob Dunlap

[Aaron Madlon-Kay]

Nate,

I've used OmegaT a bit, so I'll try to answer your questions as best I
can with the assumption you're using OmegaT as well.

> 1. These programs really have no idea where to break a Japanese
> (henceforth J) sentence, and are very very bad at breaking things like
> section titles.

OmegaT is pre-set to segment (break text into chunks to be translated)
Japanese text at the following characters:
！？...。．
That takes care of body text; if your body text isn't segmenting
correctly then there could be something wrong with your import
settings. Section titles, if they're on their own line (surrounded by
linebreaks), should be handled by the plaintext segmenting rule
(segment at \n).

> 2. If a J sentence has more than one clause, and I need to reorder the
> clauses in English, or worse yet, I need to reorder sentences in
> English, I can't create any worthwhile data for the TM.

As far as I know, reordering the clauses in English shouldn't have any
effect on whether the translation is "worthwhile" or "useful" to the
TM. The TM doesn't care or understand how the source words map to the
target words; it looks for matches between the current source and
"old" sources, then offers up the translation for the "old" source as
a hint or suggestion.

Reordering the sentences likewise should not be a problem as long as
you reorder them in the source. In other words, as long as you aren't
lying to the TM about which source sentence goes with which target
sentence, there shouldn't be a problem.

> 2.1. If I join the clauses of a multi-clause sentence to make sure my
> TM is correct, the sentence has absolutely zero chance of ever
> occurring again. This makes moot the whole process of using the CAT
> tool.

The sentence doesn't have to occur again. *Similar* sentences need to
occur again. Similarity is based on both the words themselves and
their order, so the chances of any given source-target pair being
"useless" is very low.

> 3. Our company uses different visual styling between our E and J
> documents. Undoing all the bad formatting is an awful waste of time.
> Is using raw text my only alternative?

It's usually fairly trivial to *remove* text styling--just copy and
then paste "as plain text," or Save As plain text in Word or
OpenOffice.org, etc. However, it's true that OmegaT seems to work
best in text-based formats (plain text, HTML, etc.) so *reapplying*
the styling after translation might be a pain. OmegaT does support
OpenDocument format, so you could convert from .doc to .odf, translate
with OmegaT, and then convert back to .doc.

> 4. Juggling windows to figure out what tags are/ where in the document
> I am slows me down. (Our documents can be pretty heavily formatted)

This I can't comment on. When I've used OmegaT my workflow has been
Excel > plaintext > OmegaT > plaintext > Excel, so I never have tags
in my source.

> 5. These programs match the number "3" to the number "5" with 98%
> accuracy. I never had a single fuzzy match that didn't need a
> complete rewrite.

The goal of a TM is usually to suggest *similar* translations, not
exact drop-in matches. I *never* have had a fuzzy match that didn't
need significant changes, except where the match was 100% (the source
strings were identical). Even if the strings are identical, you may
need to make changes based on context, etc. I think you may have been
mislead as to what exactly a TM does.

> 6. The ugly, sterile environments and unnatural keystrokes make the
> whole process much less pleasant than even Excel word-processing.

Well there's no arguing with taste, but as someone who works mostly in
Excel, I have to say that *almost anything is better than Excel.* I
would kill (or maim, at least) for a wholly plaintext-based workflow,
but it's just not going to happen for me.

Keystrokes may be customizable to some degree. I haven't really
looked into it.

> 7. From what I can tell, the glossaries in these programs barely
> interact with the program at all. I need to explicitly select a word
> and enter it into the glossary, and then the CAT tool never ever
> suggests the word to me again. I may as well have written it on
> paper.

I think you're confused about what glossaries do. Quick overview:
You create glossaries manually. Glossaries define terms that appear
in the *source*. In OmegaT, relevant glossary terms are shown in a
separate pane from the fuzzy matches.

I assume you're doing J > E. Perhaps you defined an E term in the
glossary; in that case, OmegaT will never show it to you because the E
term will of course never appear in your J source text. You need to
define the J term; here's a sample from one of my glossaries (the
format is [Source][tab][Target][tab][Comment].)
孫仲謀 Sun Quan Alternate name
馬孟起 Ma Chao Alternate name
黄漢升 Huang Zhong Alternate name

Note that due to encoding issues I had to name my glossary files
"gossary.utf8" to force OmegaT to recognize my glossaries as UTF-8.
Also be sure you're putting your glossary files in the appropriate
folder ([Project]/glossary).

> 8. Based on what I'm hearing, any agency that requires one of the
> tools uses it against the freelancer as a means to pay less... In the
> case of a post I read here, the pay for the repeated phrases was
> ZERO. This practice makes the software nearly worthless to the
> translator, doesn't it?

If your choice is between

A. using CAT tools and getting paid X, and
B. not using CAT tools and still getting paid X,

then by all means use whatever workflow you're most comfortable with.
But it sounds more like the situation is

B. not using CAT tools and not getting paid at all (by that particular
client)

If you can live without that client's business, then you're all good.


Now, with all the above said, I have also had trouble fitting CAT
tools (again, specifically OmegaT) into my workflow. I think the
concept is great, but unfortunately everything I do is so entrenched
in Excel that external tools end up being more trouble than they're
worth. I have to agree with Cary that these tools are best suited to
repetitive texts, and that the tools themselves are still hard to use
and poorly integrated.

Anyway, I hope that was at least somewhat helpful.

-Aaron

[Nate is fine.]

Jacob,

Thank you very much for that well though out response.

You never need to apologize for proving a pessimist wrong.

On Feb 18, 10:13 am, Jacob Dunlap <jakedun...@gmail.com> wrote:

> 2. I'm not really sure what you mean by "worthwhile". In my
> understanding, the idea of a TM isn't to create data that will give
> you lots of hits all over the place, but rather to create data that
> will give you quality hits only when appropriate.
>
> 2.1. Well, maybe, but what about the shorter sentences that may repeat
> again? As I said above, it depends on what you are expecting to get
> from your TM, but I use my CAT tool with the full understanding that
> many of the sentences I end up inputting will never pop up again. So
> be it -- the other advantages of CAT tools outweigh that "problem" for
> me.

It's likely a quirk of my material, but my source sentences are rarely
fewer than three (essentially unrelated) clauses long. Six or more
clauses worth of し、し、し、 isn't all that uncommon either. I don't expect
any full sentence matches any time soon on these.

That isn't to say we don't have repetition or near repetition. Quite
the contrary. We have certain documents that are so nearly identical
from project to project that we really only need to pull out the last
report and replace proper nouns. Like I said, CATs may not be the MOST
practical solution.

> 7. That sounds like a problem with the tool, or some other sort of
> malfunction. Or -- and don't take this the wrong way -- you may not
> be using it properly. Although Wordfast, for example, has its tricks
> and quirks, it generally functions without a hitch, and I imagine
> Trados and other CAT tools are the same.

Far from taking it the wrong way... This is exactly the sort of
information I wanted to hear. I'll have to look further into
glossary/ terminology usage. At my current position, we don't
typically have a lot of "image format conversion determination unit"-
type phrases, and those that we do have are so often repeated that
they roll off of my fingers (or have been abbreviated).

I do have a concern about glossary usage though: does it not stand in
the way of actually memorizing frequently recurring words? God knows
I would never have remembered J words like "Pleistocene" or "parapet
flashing" if I hadn't needed to. I probably would be a lot less
likely to consider whether the context allows/requires some liberty/
leeway on semi-technical terms as well.

Again, thanks very much. Like you say, if I can get some productive
use out of the glossary tool, that alone may make using CAT
worthwhile.

Nate

[Peter Clark]

 

> So, no CAT zealots out there?

JC Helary posted about an OmegaT seminar (or something) in Tokyo. Go have a chat with him.

>
>
> Now I'm scratching my head. Without the TM coughing up the occasional
> 100%er, I can't see why anyone would ever use these infernal things.
 

If I never got served another 100% match, I would still be benefitting from an increase in speed and accuracy. Jacob has already spoken about glossaries. Get to know the Wordfast glossary system, set it up properly and let it show you what it can do.
The other benefit with segmenting CATs is that you get your source text directly above where you are typing your translation. If it is a simple text document, there is no chance you will miss a piece of the original. If it is a complex document, the chances you will miss something are greatly reduced.

> I'll freely admit that I probably didn't give any of the programs
> enough time. I suspect that neither a 30-day sample nor a 500 entry TM
> limit is likely to pay off quickly or dramatically, given the variety
> of stuff I work on around here. I had thought my work was pretty
> repetitious, but maybe you're all accustomed to a considerably higher
> degree of repetition.

The chances, in the great shceme of things, of a particular combination of words being repeated are pretty slim. Even in a manual, the names of buttons, functions, or whatever change. The repetition is around them, so if you are using your CAT right, you can be done with that kind of sentence in a very short time (placeables, non-translatables, tagging, etc.).
But the chances of coming across something very similar to something you have translated in the past are quite good (same or similar clients). Thats where things like context, concordance, and reference come in handy.
 
If you are doing financial reports, you may only see the benefit once a year. First year, spend a week translating, second year three days, charge the same.

>
> I'm not gonna give up yet. I'll take a spin through the tutorial
> Peter provided, and give Felix a try later today.
 

You could also look at the various forums, Wordfast and Trados in Yahoo groups, and the knowledge base at wordfast.com.
 
 
 

---

Find out: SEEK Salary Centre Are you paid what you're worth?

[Nate is fine.]

Aaron,

Thank you for your comments. I'm eager to see how everyone else is
coping or thriving with these tools.

I keep running into hints that CATs may not be good for my current
work. Still they could be useful in some potential future work, so
I'm trying to take in everyone's hints/ advice.

> Section titles, if they're on their own line (surrounded by
> linebreaks), should be handled by the plaintext segmenting rule
> (segment at \n).

I'll look into this. What I typically translate has tables within
tables, checklists, myriad tab settings, usw.

> > 5. These programs match the number "3" to the number "5" with 98%
> > accuracy. I never had a single fuzzy match that didn't need a
> > complete rewrite.
>
> The goal of a TM is usually to suggest *similar* translations, not
> exact drop-in matches. I *never* have had a fuzzy match that didn't
> need significant changes, except where the match was 100% (the source
> strings were identical). Even if the strings are identical, you may
> need to make changes based on context, etc. I think you may have been
> mislead as to what exactly a TM does.

Fair enough. But my expectation was not for it to create an extra
step for me. Basically, I need to review the suggested sentence and
determine that it is too far off. Then I have to choose whether to
make sweeping edits to the suggestion, or just type my translated
sentence from scratch. Since the only suggestions that even come
close are usually for short sentences, I usually find the later to be
much faster.


> > 7.

> I think you're confused about what glossaries do. Quick overview:
> You create glossaries manually. Glossaries define terms that appear
> in the *source*. In OmegaT, relevant glossary terms are shown in a
> separate pane from the fuzzy matches.
>
> I assume you're doing J > E. Perhaps you defined an E term in the
> glossary; in that case, OmegaT will never show it to you because the E
> term will of course never appear in your J source text. You need to
> define the J term; here's a sample from one of my glossaries (the
> format is [Source][tab][Target][tab][Comment].)
> 孫仲謀 Sun Quan Alternate name
> 馬孟起 Ma Chao Alternate name
> 黄漢升 Huang Zhong Alternate name
>
> Note that due to encoding issues I had to name my glossary files
> "gossary.utf8" to force OmegaT to recognize my glossaries as UTF-8.
> Also be sure you're putting your glossary files in the appropriate
> folder ([Project]/glossary).

I appear to have fudged up my implementation of glossaries. They may
be the saving grace of it all for me.

>
> If your choice is between
>
> A. using CAT tools and getting paid X, and
> B. not using CAT tools and still getting paid X,
>
> then by all means use whatever workflow you're most comfortable with.
> But it sounds more like the situation is
>
> B. not using CAT tools and not getting paid at all (by that particular
> client)
>

I guess the fact of the matter is that CAT use is going to benefit
someone somewhere along the line: the end user, the translator or the
semi-unscrupulous agent... or a combination of the three. However, if
the time saved by the translator is rendered moot (non-pay for
repeats), the software doesn't pay for itself; especially when you
factor in the learning process and hoops jumped through to use them.
That is to say, in the non-pay scenario, it's only beneficial for a
translator if he/she has a shortfall of work... an employers market.

Again, thanks.

Nate

[Jacob Dunlap]

On Feb 18, 10:44 am, "Nate is fine." <notn...@gmail.com> wrote:
> That isn't to say we don't have repetition or near repetition.  Quite
> the contrary.  We have certain documents that are so nearly identical
> from project to project that we really only need to pull out the last
> report and replace proper nouns. Like I said, CATs may not be the MOST
> practical solution.

Indeed, if that's the case, a CAT tool could help you significantly.
(Not trying to sell any one particular tool; just making a
recommendation.) At my previous job, we had batches of 10 or so
patents come in from time to time that were hundreds of pages long,
and something like 75% of the content was identical. Needless to say,
doing such work without some sort of automation would have been sheer
torture (and was, until I started promoting the use of CAT tools; up
until that time, the translators had been doing it all manually!).

> I do have a concern about glossary usage though: does it not stand in
> the way of actually memorizing frequently recurring words?  God knows
> I would never have remembered J words like "Pleistocene" or "parapet
> flashing" if I hadn't needed to.  I probably would be a lot less
> likely to consider whether the context allows/requires some liberty/
> leeway on semi-technical terms as well.

Actually, I've found the simple fact that I work with data, which I
can just copy and paste into my dictionary, has presented somewhat of
an obstacle to my Japanese reading ability -- for all intents and
purposes, I don't have to learn how to actually _read_ a kanji
compound -- I only need to learn what it means. Considering that
dynamic, the glossary issue you describe is pretty much non-existent
for me. Both of these issues can of course be solved by a little
diligence on the part of the translator, of course. ;)

Jacob Dunlap

[Jean-Christophe Helary]

On mercredi 18 févr. 09, at 10:43, Peter Clark quoted Nate etc:

>> So, no CAT zealots out there?
>
> JC Helary posted about an OmegaT seminar (or something) in Tokyo. Go
> have a chat with him.

Indeed, I'll be giving my time for free on Friday and Saturday...

Free as in "speech", you still have to pay me a beer or something to
get free user support...

>> Now I'm scratching my head. Without the TM coughing up the occasional
>> 100%er, I can't see why anyone would ever use these infernal things.

"translation _memory_". You don't have to remember the stuff you've
translated, the computer does that for you.

Even if matches don't pop up as often as you'd like, the whole paired
corpus of what you've translated is memorized, ie, you can do
concordance searches whenever you like.

All the rest in only bells and whistles. They are nice to have
sometimes, they can be annoying gadgets other times.

It depends on your workflow mostly.

Jean-Christophe Helary

------------------------------------
http://mac4translators.blogspot.com/

[Nate is fine.]

Jacob,

Thanks again.

On Feb 18, 11:18 am, Jacob Dunlap <jakedun...@gmail.com> wrote:
> > That isn't to say we don't have repetition or near repetition....

> Indeed, if that's the case, a CAT tool could help you significantly.
> (Not trying to sell any one particular tool; just making a
> recommendation.)  At my previous job, we had batches of 10 or so
> patents come in from time to time that were hundreds of pages long,
> and something like 75% of the content was identical.  Needless to say,
> doing such work without some sort of automation would have been sheer
> torture (and was, until I started promoting the use of CAT tools; up
> until that time, the translators had been doing it all manually!).

Actually because of the reformatting and all that my work requires,
it's much easier just to use the most recent report and edit it in
word. Using a CAT here would be akin to retranslating a form letter,
rather than just retyping the address line.

I get the feeling I can benefit from glossary tools right away, but I
don't think TMs are going to be especially useful to me in my current
position. Like any pessimist though, I want nothing more than to be
proven wrong.

Nate

[Jean-Christophe Helary]

On mercredi 18 févr. 09, at 11:37, Nate is fine. wrote:

> I don't think TMs are going to be especially useful to me in my
> current position.

I am pretty sure William Weaver has _never_ used a CAT tool in his
entire life.

That does not keep him from producing fabulous translations.

[S. Patrick Eaton]

Having worked with quite a few CAT tools myself, I can say that you
probably won't find one that does everything well. Each has its own
strengths and weaknesses, and it is really up to the translator to
select the best CAT tool for the job at hand.

In my own use of these tools, the biggest benefit is simply that your
TMs can generally be shared with others, meaning that work you do today
can help someone else translate something similar tomorrow. Similarly,
shared glossaries are great for making sure that everyone on your team
is translating the same terms the same way.

If you are thinking of using these tools as a freelancer working solo,
you might find that the someone else you are helping out is yourself six
months later, when a revised or updated version of a document you
translated previously comes back for a bit of touching up.

I think all of the mainstream CAT tools can contribute to your
productivity if used properly, but some are better at certain tasks than
others, so you would be well advised to select your CATs carefully.

Also bear in mind that you don't need to slavishly use a CAT tool just
for the sake of building up a translation memory. If you have some jobs
that are CAT friendly and others that are not, there's no reason to
force the unfriendly ones into your preferred CAT tool's way of doing
things if it is going to create more problems than it solves. Save the
CAT tools for jobs where they are most likely to help you.

Best regards,

Sako Eaton