Using the EU DGT as a tmx
22 views
Skip to first unread message
Will Helton
Mar 24, 2015, 7:05:26 AM3/24/15
to cafetra...@googlegroups.com
I've decided to give the EU DGT a try in CT as a tmx memory and am
finding that it not only is slow to load as I move between segments
(hey, it's pretty big!), but it also cranks up my CPU usage to well over
100% (average of about 115%).
Is there a more system-friendly way to use this resource that would put
a lower load on my laptop?
Thanks,
Will
finding that it not only is slow to load as I move between segments
(hey, it's pretty big!), but it also cranks up my CPU usage to well over
100% (average of about 115%).
Is there a more system-friendly way to use this resource that would put
a lower load on my laptop?
Thanks,
Will
Hans van den Broek
Mar 24, 2015, 7:42:32 AM3/24/15
to Cafetran support
> On 24 Mar 2015, at 18:05, Will Helton <will....@gmail.com> wrote:
>
> I've decided to give the EU DGT a try in CT as a tmx memory and am finding that it not only is slow to load
> as I move between segments (hey, it's pretty big!),
> but it also cranks up my CPU usage to well over 100% (average of about 115%).
> Is there a more system-friendly way to use this resource that would put a lower load on my laptop?
http://www.screencast.com/t/95NPzHRJJIpn
(only 26 seconds)
Cheers,
Hans
--
Hans van den Broek
Schrijf-, vertaal- en redigeerwerk
Yogyakarta
Indonesia
http://cafetran4mac.blogspot.com
--
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.
Will Helton
Mar 24, 2015, 8:45:29 AM3/24/15
to cafetra...@googlegroups.com
> Hans van den Broek <mailto:ir...@indo.net.id>
> 24 March 2015 11:42
>> On 24 Mar 2015, at 18:05, Will Helton<will....@gmail.com> wrote:
>>
>> I've decided to give the EU DGT a try in CT as a tmx memory and am finding that it not only is slow to load
>
> Not really. On my old Mac, the 2 million segments DGT loads in a minute or two. 8 GB assigned to CafeTran/Java heap.
I mean when moving between segments. This takes about a minute each
>>
>> I've decided to give the EU DGT a try in CT as a tmx memory and am finding that it not only is slow to load
>
> Not really. On my old Mac, the 2 million segments DGT loads in a minute or two. 8 GB assigned to CafeTran/Java heap.
time. Longer than it takes me to translate the segment in a lot of cases.
>> as I move between segments (hey, it's pretty big!),
>
> What do you mean? Auto-assemble is just about impossible. A search* is sort of acceptable.
between segments because it contains so much data.
>> but it also cranks up my CPU usage to well over 100% (average of about 115%).
>
> Really?
>> Is there a more system-friendly way to use this resource that would put a lower load on my laptop?
>
> Yes, the external database. Importing it takes a while (4 hours on my system for 2 million segments), but that's a one time affair. After that, it's readily available - no waiting time - and search is very fast. I recorded a *screencast of a search in both the TMX and the database (though that was still with H2 database, not the new SQLite).
>
> http://www.screencast.com/t/95NPzHRJJIpn
Bedankt!
Will
Hans van den Broek
Mar 24, 2015, 8:52:28 AM3/24/15
to Cafetran support
> On 24 Mar 2015, at 19:45, Will Helton <will....@gmail.com> wrote:
>
>
> I mean when moving between segments. This takes about a minute each time. Longer than it takes me to translate the segment in a lot of cases.
> No idea what you mean.
> Yes, really. Perhaps you've not found this to be the case.
> I'll give it a try.
Will Helton
Mar 24, 2015, 9:02:35 AM3/24/15
to cafetra...@googlegroups.com
I don't use AA. It's useless for the type of work I do.
When I get caught up I'll try the upgrade and come back with what I've
found regarding importing and loading times.
Cheers,
Will
> 24 March 2015 12:52
> 24 March 2015 11:05
> (hey, it's pretty big!), but it also cranks up my CPU usage to well
>
> Will
When I get caught up I'll try the upgrade and come back with what I've
found regarding importing and loading times.
Cheers,
Will
> 24 March 2015 12:52
>> On 24 Mar 2015, at 19:45, Will Helton<will....@gmail.com> wrote:
>>
>>
>> I mean when moving between segments. This takes about a minute each time. Longer than it takes me to translate the segment in a lot of cases.
>
> With AA enabled? Yes. You could also try to pretranslate, both real-time with the TMX version, and before you start translating using Total Recall.
>
>> No idea what you mean.
>
> Same here.
>
>> Yes, really. Perhaps you've not found this to be the case.
>
> No. I monitored things regularly, but even though CPU usage hit even higher than 150, no nasty consequences.
>
>> I'll give it a try.
>
> Please report. Also importing time.
>
>
> Cheers,
>
> Hans
>
>>
>>
>> I mean when moving between segments. This takes about a minute each time. Longer than it takes me to translate the segment in a lot of cases.
>
> With AA enabled? Yes. You could also try to pretranslate, both real-time with the TMX version, and before you start translating using Total Recall.
>
>> No idea what you mean.
>
> Same here.
>
>> Yes, really. Perhaps you've not found this to be the case.
>
> No. I monitored things regularly, but even though CPU usage hit even higher than 150, no nasty consequences.
>
>> I'll give it a try.
>
> Please report. Also importing time.
>
>
> Cheers,
>
> Hans
>
> Hans van den Broek <mailto:ir...@indo.net.id>
> 24 March 2015 11:42
>> On 24 Mar 2015, at 18:05, Will Helton<will....@gmail.com> wrote:
>>
>> I've decided to give the EU DGT a try in CT as a tmx memory and am finding that it not only is slow to load
>
> Not really. On my old Mac, the 2 million segments DGT loads in a minute or two. 8 GB assigned to CafeTran/Java heap.
>
> 24 March 2015 11:42
>> On 24 Mar 2015, at 18:05, Will Helton<will....@gmail.com> wrote:
>>
>> I've decided to give the EU DGT a try in CT as a tmx memory and am finding that it not only is slow to load
>
> Not really. On my old Mac, the 2 million segments DGT loads in a minute or two. 8 GB assigned to CafeTran/Java heap.
>
>> as I move between segments (hey, it's pretty big!),
>
> What do you mean? Auto-assemble is just about impossible. A search* is sort of acceptable.
>
>
> What do you mean? Auto-assemble is just about impossible. A search* is sort of acceptable.
>
>> but it also cranks up my CPU usage to well over 100% (average of about 115%).
>
> Really?
>
>
> Really?
>
>> Is there a more system-friendly way to use this resource that would put a lower load on my laptop?
>
> Yes, the external database. Importing it takes a while (4 hours on my system for 2 million segments), but that's a one time affair. After that, it's readily available - no waiting time - and search is very fast. I recorded a *screencast of a search in both the TMX and the database (though that was still with H2 database, not the new SQLite).
>
> http://www.screencast.com/t/95NPzHRJJIpn
>
>
> Yes, the external database. Importing it takes a while (4 hours on my system for 2 million segments), but that's a one time affair. After that, it's readily available - no waiting time - and search is very fast. I recorded a *screencast of a search in both the TMX and the database (though that was still with H2 database, not the new SQLite).
>
> http://www.screencast.com/t/95NPzHRJJIpn
>
> (only 26 seconds)
>
>
> Cheers,
>
> Hans
>
> Will Helton <mailto:will....@gmail.com>
>
>
> Cheers,
>
> Hans
>
> 24 March 2015 11:05
> I've decided to give the EU DGT a try in CT as a tmx memory and am
> finding that it not only is slow to load as I move between segments
> (hey, it's pretty big!), but it also cranks up my CPU usage to well
> over 100% (average of about 115%).
>
>
> Is there a more system-friendly way to use this resource that would
> put a lower load on my laptop?
>
> Thanks,
> put a lower load on my laptop?
>
>
> Will
Will Helton
Mar 24, 2015, 9:11:30 AM3/24/15
to cafetra...@googlegroups.com
Hans,
Before I do this, how large a database does it create?
Cheers,
Will
> 24 March 2015 12:52
> (hey, it's pretty big!), but it also cranks up my CPU usage to well
>
> Will
Before I do this, how large a database does it create?
Cheers,
Will
> 24 March 2015 12:52
>> On 24 Mar 2015, at 19:45, Will Helton<will....@gmail.com> wrote:
>>
>>
>> I mean when moving between segments. This takes about a minute each time. Longer than it takes me to translate the segment in a lot of cases.
>
> With AA enabled? Yes. You could also try to pretranslate, both real-time with the TMX version, and before you start translating using Total Recall.
>
>> No idea what you mean.
>
> Same here.
>
>> Yes, really. Perhaps you've not found this to be the case.
>
> No. I monitored things regularly, but even though CPU usage hit even higher than 150, no nasty consequences.
>
>> I'll give it a try.
>
> Please report. Also importing time.
>
>
> Cheers,
>
> Hans
>
>>
>>
>> I mean when moving between segments. This takes about a minute each time. Longer than it takes me to translate the segment in a lot of cases.
>
> With AA enabled? Yes. You could also try to pretranslate, both real-time with the TMX version, and before you start translating using Total Recall.
>
>> No idea what you mean.
>
> Same here.
>
>> Yes, really. Perhaps you've not found this to be the case.
>
> No. I monitored things regularly, but even though CPU usage hit even higher than 150, no nasty consequences.
>
>> I'll give it a try.
>
> Please report. Also importing time.
>
>
> Cheers,
>
> Hans
>
> Hans van den Broek <mailto:ir...@indo.net.id>
> 24 March 2015 11:42
>> On 24 Mar 2015, at 18:05, Will Helton<will....@gmail.com> wrote:
>>
>> I've decided to give the EU DGT a try in CT as a tmx memory and am finding that it not only is slow to load
>
> Not really. On my old Mac, the 2 million segments DGT loads in a minute or two. 8 GB assigned to CafeTran/Java heap.
>
> 24 March 2015 11:42
>> On 24 Mar 2015, at 18:05, Will Helton<will....@gmail.com> wrote:
>>
>> I've decided to give the EU DGT a try in CT as a tmx memory and am finding that it not only is slow to load
>
> Not really. On my old Mac, the 2 million segments DGT loads in a minute or two. 8 GB assigned to CafeTran/Java heap.
>
>> as I move between segments (hey, it's pretty big!),
>
> What do you mean? Auto-assemble is just about impossible. A search* is sort of acceptable.
>
>
> What do you mean? Auto-assemble is just about impossible. A search* is sort of acceptable.
>
>> but it also cranks up my CPU usage to well over 100% (average of about 115%).
>
> Really?
>
>
> Really?
>
>> Is there a more system-friendly way to use this resource that would put a lower load on my laptop?
>
> Yes, the external database. Importing it takes a while (4 hours on my system for 2 million segments), but that's a one time affair. After that, it's readily available - no waiting time - and search is very fast. I recorded a *screencast of a search in both the TMX and the database (though that was still with H2 database, not the new SQLite).
>
> http://www.screencast.com/t/95NPzHRJJIpn
>
>
> Yes, the external database. Importing it takes a while (4 hours on my system for 2 million segments), but that's a one time affair. After that, it's readily available - no waiting time - and search is very fast. I recorded a *screencast of a search in both the TMX and the database (though that was still with H2 database, not the new SQLite).
>
> http://www.screencast.com/t/95NPzHRJJIpn
>
> (only 26 seconds)
>
>
> Cheers,
>
> Hans
>
>
>
> Cheers,
>
> Hans
>
> I've decided to give the EU DGT a try in CT as a tmx memory and am
> finding that it not only is slow to load as I move between segments
> (hey, it's pretty big!), but it also cranks up my CPU usage to well
> over 100% (average of about 115%).
>
>
> Is there a more system-friendly way to use this resource that would
> put a lower load on my laptop?
>
> Thanks,
> put a lower load on my laptop?
>
>
> Will
Hans van den Broek
Mar 24, 2015, 9:19:17 AM3/24/15
to Cafetran support
> On 24 Mar 2015, at 20:02, Will Helton <will....@gmail.com> wrote:
>
> I don't use AA.
So why/how do you go from one segment to another?
Hans van den Broek
Mar 24, 2015, 9:23:26 AM3/24/15
to Cafetran support
On 24 Mar 2015, at 20:11, Will Helton <will....@gmail.com> wrote:
Before I do this, how large a database does it create?
Not large. I added a 2.5 million segments DE-NL, another 2.5 segments EN-NL DGT, and one IATE termbase to it, and the grand total is:
The previous indexed database (H2) contained less than half that, and inflated to around 20 GB.
The previous indexed database (H2) contained less than half that, and inflated to around 20 GB.
Cheers,
Hans
--
Hans van den Broek
Schrijf-, vertaal- en redigeerwerk
Yogyakarta
Indonesia
http://cafetran4mac.blogspot.com
Will Helton
Mar 24, 2015, 9:23:51 AM3/24/15
to cafetra...@googlegroups.com
By advancing as normal from one segment to the next: ctrl+alt+down arrow.
AA is of no value for the sort of work I do (marketing, advertising, recruitment, etc.) where things have to be very colloquial, catchy and dare I say, clever. :)
Will
AA is of no value for the sort of work I do (marketing, advertising, recruitment, etc.) where things have to be very colloquial, catchy and dare I say, clever. :)
Will
Will Helton
Mar 24, 2015, 9:24:28 AM3/24/15
to cafetra...@googlegroups.com
Cool. I like, I like! :)
> 24 March 2015 13:23
> believed to be clean. --
> You received this message because you are subscribed to the Google
> Groups "CafeTranslators" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to cafetranslato...@googlegroups.com
> <mailto:cafetranslato...@googlegroups.com>.
> For more options, visit https://groups.google.com/d/optout.
> 24 March 2015 13:23
>
>
> Not large. I added a 2.5 million segments DE-NL, another 2.5 segments
> EN-NL DGT, and one IATE termbase to it, and the grand total is:
>
>
>
>
> Not large. I added a 2.5 million segments DE-NL, another 2.5 segments
> EN-NL DGT, and one IATE termbase to it, and the grand total is:
>
>
>
> The previous indexed database (H2) contained less than half that, and
> inflated to around 20 GB.
>
> Cheers,
>
> Hans
>
> --
>
> Hans van den Broek
> Schrijf-, vertaal- en redigeerwerk
> Yogyakarta
> Indonesia
> http://cafetran4mac.blogspot.com
>
>
>
>
>
> --
> This message has been scanned for viruses and
> dangerous content by *MailScanner* <http://www.mailscanner.info/>, and is
> inflated to around 20 GB.
>
> Cheers,
>
> Hans
>
> --
>
> Hans van den Broek
> Schrijf-, vertaal- en redigeerwerk
> Yogyakarta
> Indonesia
> http://cafetran4mac.blogspot.com
>
>
>
>
>
> --
> This message has been scanned for viruses and
> believed to be clean. --
> You received this message because you are subscribed to the Google
> Groups "CafeTranslators" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to cafetranslato...@googlegroups.com
> <mailto:cafetranslato...@googlegroups.com>.
> For more options, visit https://groups.google.com/d/optout.
Hans van den Broek
Mar 24, 2015, 9:37:35 AM3/24/15
to Cafetran support
> On 24 Mar 2015, at 20:23, Will Helton <will....@gmail.com> wrote:
>
> By advancing as normal from one segment to the next: ctrl+alt+down arrow.
> AA is of no value for the sort of work I do (marketing, advertising, recruitment, etc.) where things have to be very colloquial, catchy and dare I say, clever. :)
Michael Beijer
Mar 24, 2015, 10:31:09 AM3/24/15
to cafetra...@googlegroups.com
Totally agree. AA is completely useless when it comes to stuff that's gotta sparkle.
Michael
--
You received this message because you are subscribed to the Google Groups "CafeTranslators" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cafetranslators+unsubscribe@googlegroups.com.
Michael Beijer
Mar 24, 2015, 10:34:46 AM3/24/15
to cafetra...@googlegroups.com
PS: AA can be used for the little pieces of the puzzle, but not to construct the final/entire sentence in all its radiant sparkling beauteousness.
Michael Beijer
Mar 24, 2015, 10:37:02 AM3/24/15
to cafetra...@googlegroups.com
PPS: I've been experimenting with Dragon lately for stuff that has to flow and am getting great results! It has actually revealed a whole new/old dimension to me: thee spoken wordicle.
Hans list
Mar 24, 2015, 1:18:38 PM3/24/15
to cafetra...@googlegroups.com
On Tuesday, March 24, 2015 at 3:31:09 PM UTC+1, Michael Beijer wrote:
Totally agree. AA is completely useless when it comes to stuff that's gotta sparkle.
I think that the peoples are right here. On the other hand: what word price is appropriate for such struck by genius texts? Or better: a price per hour? Like 100 € or so?
I'm serious here.
Hans list
Mar 24, 2015, 1:22:03 PM3/24/15
to cafetra...@googlegroups.com
That's soooo obvious, Will.
How about:
- Using the EU DGT as a letter opener
- Using the EU DGT as a toaster
- Using the EU DGT as quirl
Much more fascinating!
Will Helton
Mar 24, 2015, 1:50:27 PM3/24/15
to cafetra...@googlegroups.com
Well, I'd like to say a percentage of the profits derived from the given
campaign, but I'm not that naive.
If I worked for Saatchi and Saatchi, I'd hate to think what I could
charge, but alas, I'm just me and my rates are about as aggressive as I
reckon I can make them and still get work.
Or maybe if you fellas are up for it, we could have a private chat about
rates as I'm sure there are a number of points we could all make on this
topic.
Nou, wat zeggen jullie?
Will
> Hans list <mailto:hans...@gmail.com>
> 24 March 2015 17:18
> <mailto:cafetranslato...@googlegroups.com>.
> 24 March 2015 14:30
> <mailto:cafetranslato...@googlegroups.com>.
> 24 March 2015 13:23
> 24 March 2015 13:02
campaign, but I'm not that naive.
If I worked for Saatchi and Saatchi, I'd hate to think what I could
charge, but alas, I'm just me and my rates are about as aggressive as I
reckon I can make them and still get work.
Or maybe if you fellas are up for it, we could have a private chat about
rates as I'm sure there are a number of points we could all make on this
topic.
Nou, wat zeggen jullie?
Will
> Hans list <mailto:hans...@gmail.com>
> 24 March 2015 17:18
> --
> You received this message because you are subscribed to the Google
> Groups "CafeTranslators" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to cafetranslato...@googlegroups.com
> You received this message because you are subscribed to the Google
> Groups "CafeTranslators" group.
> To unsubscribe from this group and stop receiving emails from it, send
> <mailto:cafetranslato...@googlegroups.com>.
> For more options, visit https://groups.google.com/d/optout.
> Michael Beijer <mailto:mic...@beijer.uk>
> 24 March 2015 14:30
> Totally agree. AA is completely useless when it comes to stuff that's
> gotta sparkle.
>
> Michael
> gotta sparkle.
>
>
>
>
> --
> You received this message because you are subscribed to the Google
> Groups "CafeTranslators" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to cafetranslato...@googlegroups.com
>
>
> --
> You received this message because you are subscribed to the Google
> Groups "CafeTranslators" group.
> To unsubscribe from this group and stop receiving emails from it, send
> <mailto:cafetranslato...@googlegroups.com>.
> For more options, visit https://groups.google.com/d/optout.
> Will Helton <mailto:will....@gmail.com>
> 24 March 2015 13:23
> By advancing as normal from one segment to the next: ctrl+alt+down arrow.
>
> AA is of no value for the sort of work I do (marketing, advertising,
> recruitment, etc.) where things have to be very colloquial, catchy and
> dare I say, clever. :)
>
> Will
>
>
> AA is of no value for the sort of work I do (marketing, advertising,
> recruitment, etc.) where things have to be very colloquial, catchy and
> dare I say, clever. :)
>
> Will
>
> Hans van den Broek <mailto:ir...@indo.net.id>
> 24 March 2015 13:19
>
> So why/how do you go from one segment to another?
>
> Cheers,
>
> Hans
>
> Will Helton <mailto:will....@gmail.com>
> So why/how do you go from one segment to another?
>
> Cheers,
>
> Hans
>
> 24 March 2015 13:02
Will Helton
Mar 24, 2015, 1:50:56 PM3/24/15
to cafetra...@googlegroups.com
Considering how hot it makes my MacBook Pro run, a toaster certainly can't be ruled out!
Hans list wrote:
> That's soooo obvious, Will.
>
> How about:
>
> * Using the EU DGT as a letter opener
> * Using the EU DGT as a toaster
> * Using the EU DGT as quirl
>
> Much more fascinating!
>
Hans list wrote:
> That's soooo obvious, Will.
>
> How about:
>
> * Using the EU DGT as a toaster
> * Using the EU DGT as quirl
>
> Much more fascinating!
>
Hans List
Mar 24, 2015, 2:14:25 PM3/24/15
to cafetra...@googlegroups.com
I'm in.
> Op 24 mrt. 2015 om 18:50 heeft Will Helton <will....@gmail.com> het volgende geschreven:
>
> Nou, wat zeggen jullie?
> Op 24 mrt. 2015 om 18:50 heeft Will Helton <will....@gmail.com> het volgende geschreven:
>
> Nou, wat zeggen jullie?
Will Helton
Mar 24, 2015, 3:30:56 PM3/24/15
to cafetra...@googlegroups.com
Cool. Hopefully Michael will be interested when he's finished moving and maybe even Hans vdB!
Any other takers?
Will
Sent from my iPhone
Verstuurd vanaf mijn iPhone
Any other takers?
Will
Sent from my iPhone
Verstuurd vanaf mijn iPhone
> --
> You received this message because you are subscribed to the Google Groups "CafeTranslators" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to cafetranslato...@googlegroups.com.
> You received this message because you are subscribed to the Google Groups "CafeTranslators" group.
Igor Kmitowski
Mar 25, 2015, 2:55:40 AM3/25/15
to cafetra...@googlegroups.com
Hi Will,
If you load a huge TMX to RAM, pretranslation is a must. See Translation >
Pretranslate menu items.
Give it a head start (a minute or two) and then take a next segment. You
can start translating even before the Pretranslation is completed but do
not race with the computer Pretranslation process. :)
However, I would strongly recommend using the new Total Recall system
instead.
Cheers,
Igor
--
Igor Kmitowski
Translator and Java developer
CafeTran website: http://www.cafetran.com
CafeTran support: cafetran...@gmail.com
If you load a huge TMX to RAM, pretranslation is a must. See Translation >
Pretranslate menu items.
Give it a head start (a minute or two) and then take a next segment. You
can start translating even before the Pretranslation is completed but do
not race with the computer Pretranslation process. :)
However, I would strongly recommend using the new Total Recall system
instead.
Cheers,
Igor
Igor Kmitowski
Translator and Java developer
CafeTran website: http://www.cafetran.com
CafeTran support: cafetran...@gmail.com
Will Helton
Mar 25, 2015, 4:20:41 AM3/25/15
to cafetra...@googlegroups.com
Hi Igor,
I'm in the process of upgrading and using the new system. I'll hopefully
have all my work out of the way and things set up by the weekend
(barring any new work coming in, that is).
Many thanks,
Will
> Igor Kmitowski <mailto:cafetran...@gmail.com>
> 25 March 2015 06:55
> 24 March 2015 11:05
I'm in the process of upgrading and using the new system. I'll hopefully
have all my work out of the way and things set up by the weekend
(barring any new work coming in, that is).
Many thanks,
Will
> Igor Kmitowski <mailto:cafetran...@gmail.com>
> 25 March 2015 06:55
> Hi Will,
>
> If you load a huge TMX to RAM, pretranslation is a must. See
> Translation > Pretranslate menu items.
>
> Give it a head start (a minute or two) and then take a next segment.
> You can start translating even before the Pretranslation is completed
> but do not race with the computer Pretranslation process. :)
>
> However, I would strongly recommend using the new Total Recall system
> instead.
>
> Cheers,
> Igor
>
>
>
> Will Helton <mailto:will....@gmail.com>
>
> If you load a huge TMX to RAM, pretranslation is a must. See
> Translation > Pretranslate menu items.
>
> Give it a head start (a minute or two) and then take a next segment.
> You can start translating even before the Pretranslation is completed
> but do not race with the computer Pretranslation process. :)
>
> However, I would strongly recommend using the new Total Recall system
> instead.
>
> Cheers,
> Igor
>
>
>
> 24 March 2015 11:05
Hans List
Mar 25, 2015, 4:27:56 AM3/25/15
to CafeTran Google Group
Hello Will, as an addition some extra information:
On 25 Mar 2015, at 09:20, Will Helton <will....@gmail.com> wrote:However, I would strongly recommend using the new Total Recall system instead.
The impact on your system will be less with TR, since only those segments that are relevant to your source document are taken out of the DGT and saved in a relatively small TMX file. The relevance is base on fuzzy matching. So, if in the 4 million TUs in the DGT only 2000 could be of any relevance to you, only those TUs are taken out.
Is this interpretation correct, Igor?
Igor Kmitowski
Mar 25, 2015, 4:45:29 AM3/25/15
to cafetra...@googlegroups.com
Hello Hans,
> The impact on your system will be less with TR, since only those
> segments that are relevant to your source document are taken out of the
> DGT and saved in a relatively small TMX file.
It is correct.
> The relevance is base on fuzzy matching. So, if in the 4 million TUs in
> the DGT only 2000 could be of any relevance to you, only those TUs are
> taken out.
The relevance is based on the matching of project words to the contents of
your Total Recall base.
Igor
> The impact on your system will be less with TR, since only those
> segments that are relevant to your source document are taken out of the
> DGT and saved in a relatively small TMX file.
> The relevance is base on fuzzy matching. So, if in the 4 million TUs in
> the DGT only 2000 could be of any relevance to you, only those TUs are
> taken out.
your Total Recall base.
Igor
Reply all
Reply to author
Forward
0 new messages