New to DVX3. How can I …

Selcuk Akyuz wrote:

I rarely use 2 or more TMs/TBs. Normally, each segment goes to Big Mama (main TM) and each new term goes to Big Papa (main TB). But Client and Subject metadata are essential for me.

I use a hierarchical client list, e.g.

1. Agency 1 (subject: General)
2. Agency 2 (subject: Fashion)
3. Agency 3 for Sony translations (subject: Computers)
4. Agency 3 for Asus translations (subject: Computers)
5. Agency 3 for DHL translations (subject: General)
6. Agency 4 for Levi's translations (subject: Fashion)
7. Agency 4 for Sony translations (subject: Computers)
...

So when doing translation for No. 6, I get results in my AutoSearch window from No. 2 as well, both in the Fashion industry. It applies to TB as well.

Keeping separate TMs/TBs for each client/subject is a waste of time for me, and maintenance is not easy.

Of course, I do have some reference TMs/TBs, some from clients. I attach them to my projects, if needed. But generally I open them in a second tab or in another instance of DVX.

Michael Joseph Wdowiak Beijer wrote:

I assume you have edited the strange & wonderful list of subjects that comes with DVX3:

Generalities
Prolegomena. Fundamentals of knowledge and culture. Computer science
Science and knowledge in general. Organization of intellectual work
etc.

Any idea how big these big Mamas and Papas can get?

Michael

Michael Joseph Wdowiak Beijer wrote:

But you said you use:

subject: Computers
subject: Fashion

… which I couldn't find in the list. I did find "fashion", but only in "391: Costume. Clothing. National dress. Fashion. Adornment"

… and "computers" in:

004: Computer science and technology. Computing
0042: Computer architecture
0043: Computer hardware
etc.

I think there is a lot to learn about DVX3

Epameinondas Soufleros wrote:

The list of subjects is an old version of the Dewey decimal classification system, used by libraries. I'm not sure it is of much use to translators. Plus, it is a hassle to navigate the list and select a subject. So, you can add your own subjects. Since I don't want my subject names to be too specific, I have labels like "ICT" and "Commerce", under which fall most of my jobs. (And I don't use a "big mama"; I follow the "project TM" and "master TM" paradigm of memoQ and Trados Studio.)

Here's an extra tip:
When adding clients or subjects, you can avoid using numeric IDs, as this will cause trouble when working on more than one computers. You'll have to remember the numeric IDs by heart; there's no way of keeping the lists in sync between machines (for this, we would need a TeamServer for freelancers, something similar to Kilgray's LanguageTerminal). So, for now, you can use a string for the "ID" field:
– Instead of "01 - TechCompany Ltd", write "TechCompany - TechCompany".
– Instead of "004 - Software", write "ICT - Information and Communications Technology".

[Edited at 2016-10-27 11:34 GMT]

[Edited at 2016-10-27 11:35 GMT]

[Edited at 2016-10-27 11:40 GMT]

Epameinondas Soufleros wrote:

Yes, Michael, every term will get all the metadata of the project. Plus, every term will get as "context" the entire segment in which you were when you decided to hit on F11 and add the term to the term base. I like this feature a lot.

Now, if only we could add alternative terms like in memoQ and use RegEx or at least pipe symbols and asterisks....

Side note:
My wish is for RegEx support in term entry and term recognition. This solution would be ideal for my language, but it won't appeal to newbies (there can be a 'simple' and an 'advanced' mode, though). For example, for "mobile phone" I would like to enter this in Greek:

κινητ(ό|ού|ά|ών) τηλ[εέ]φ[ωώ]ν(ο|ου|α|ων)

which would allow for all the following to be recognized as correct translations of the term:

κινητό τηλέφωνο
κινητού τηλεφώνου
κινητά τηλέφωνα
κινητών τηλεφώνων

Till this feature is implemented, Déjà Vu will continue giving way too many false positives when checking for errors in terms and numbers.

[Edited at 2016-10-27 12:11 GMT]

Now, if only we could add alternative terms like in memoQ and use RegEx or at least pipe symbols and asterisks....

Side note:
My wish is for RegEx support in term entry and term recognition. This solution would be ideal for my language, but it won't appeal to newbies (there can be a 'simple' and an 'advanced' mode, though). For example, for "mobile phone" I would like to enter this in Greek:

κινητ(ό|ού|ά|ών) τηλ[εέ]φ[ωώ]ν(ο|ου|α|ων)

which would allow for all the following to be recognized as correct translations of the term:

κινητό τηλέφωνο
κινητού τηλεφώνου
κινητά τηλέφωνα
κινητών τηλεφώνων

Till this feature is implemented, Déjà Vu will continue giving way too many false positives when checking for errors in terms and numbers.

[Edited at 2016-10-27 12:11 GMT]

Alistair Gainey wrote:



Same for my language pair as well, although I imagine that for translators translating in the other direction it's even more inconvenient. What do you mean by "add alternative terms like in memoQ"? What is it that memoQ does?

Epameinondas Soufleros wrote:

The list of subjects is an old version of the Dewey decimal classification system, used by libraries. I'm not sure it is of much use to translators. Plus, it is a hassle to navigate the list and select a subject. So, you can add your own subjects. Since I don't want my subject names to be too specific, I have labels like "ICT" and "Commerce", under which fall most of my jobs. (And I don't use a "big mama"; I follow the "project TM" and "master TM" paradigm of memoQ and Trados Studio.)

Here's an extra tip:
When adding clients or subjects, you can avoid using numeric IDs, as this will cause trouble when working on more than one computers. You'll have to remember the numeric IDs by heart; there's no way of keeping the lists in sync between machines (for this, we would need a TeamServer for freelancers, something similar to Kilgray's LanguageTerminal). So, for now, you can use a string for the "ID" field:
– Instead of "01 - TechCompany Ltd", write "TechCompany - TechCompany".
– Instead of "004 - Software", write "ICT - Information and Communications Technology".

[Edited at 2016-10-27 11:34 GMT]

[Edited at 2016-10-27 11:35 GMT]

[Edited at 2016-10-27 11:40 GMT]