Is there anybody out there?

Wojciech Matyszkiewicz wrote:

If you're interested, this is an e-mail message I received from Etienne from Atril, in September 2015 (so quite a long time ago). I don't know if anything has changed since then, but still.
This wasn't an extremely confidential correspondence, so I hope they don't mind.
The bits of text in italics are the excerpts from my original mail to which Etienne was replying.
Here we go:

&&&&
Etienne (Atril Solutions)
Oct 1, 17:24
Hi Wojciech!
Sorry for not getting back to you earlier, the last days were pretty busy (proz events and such).
Well, make no mistake: i'm reading the proz forum every week at least and i'm taking good notes of all the requests.
We added a good amount of them that were discussed in the "Bugs and aeras to improve" in previous builds of DVX3.
By the way, thank you for your feedback which is always interesting.
I cannot share toooooo much but I can tell you that the main features that have been discussed will be implemented in a near future. We want to do it in the best possible way so it will take some more time, but you can expect big news soon.
I especially look forward to the glossary highlighting
In the pipeline
quick glossary selection (for entering terms)
In the pipeline
Autowrite for numbers and proper names
In the pipeline
Regarding the "web search" feature - I was thinking about something that could replace the Intelliwebsearch solution that I'm using now. Namely, a built-in browser where a user could pre-define search engines (Linguee, glosbe, proz.com terminology etc) and use them during the translation process. You could, for example, highlight a term or a phrase in the source text, then hit a key shortcut for DVX to produce a window (preferably auto-hiding one) to see the result in Linguee etc. It would also be great if then in the browser window you could highlight the found translation and hit a key shortcut to paste the translated text directly to your target segment.
Yes, we would like that as well (with a link tothe IATB termmbase as wel) but we have to make sure that it would not slow DVX3 down.
Also - probably it's more difficult to implement, but would it be possible to set separate font size for the active segment and the ones already translated and not translated yet? I like to have large font sizes (16 or even 18) for the active segment, but presently this must also apply to other segments and in consequence there’s little text you can see around the active segment (because all the other segments also have large fonts).
We are thinking about this for quite a while.
I’m presently trying such a plug-in with a demo of Trados 2015 and it’s frankly very useful. I can, for example make a rule to convert all English numbering (e.g. 1,000 or decimal 1.5) into continental one (1000 or 1 000 or even Polish 1 tys. and decimal 1,5 ), and all of this is fed to me via Autosuggest. A great feature, in my humble opinion.
Well you can also do this in DVX3 right now using RexExs in the search & replace module. What is missing is the automization of such things.
This, also, will come
Thank you for getting back to me.
Please understand that i cannot share too much on Proz - there is no date yet.
Have a great day,
Etienne
Kind regards,
Etienne

****


[Edited at 2017-01-22 00:13 GMT]

[Edited at 2017-01-22 15:35 GMT]

Atril_TEAM wrote:

...we really need a community manager, right?!

Atril_TEAM wrote:

...we really need a community manager, right?!

So, some news:

New website almost finished

New build being tested - with one long awaited request that has been implemented

Mind blowing new features for the next version to be unleashed in 2017 (yes, I said it)

Happy new year to you all, and do not bury us just yet.
Cheers,
Etienne

Rob Edwards wrote:



Thanks Wojciech! That is quite interesting. Hopefully much of that is covered in the final and upcoming update to DVX3. I guess web search will probably come in DVX4.

Wojciech Matyszkiewicz wrote:

Thanks, Etienne, it's really a good thing to know that you're working on the new builds.
Since our correspondence I've bought Trados, but I'm eager to put my hands on the new improved Deja Vu, as well.

So far, Trados has slightly let me down with its slowness and mediocre termbase handling, as well as, unfortunately its upLift feature, which was supposed to be a breakthrough, but turned out to be an improved version of their Autosuggest Dictionaries

There is, however, one area, where it stands above Deja Vu - namely the autosuggest feature. The suggestions are quite clever, e.g. they automatically adjust the case to your needs, and I find them really helpful.
I do hope that Atril will be able to come up with something equally intuitive.

Pavel Tsvetkov wrote:

Hi Everyone!

Just wanted to chime in a say that the next build of Déjà vu X3 will most probably include the much requested feature to mark recognized terms in the source segment.

Best regards,
Pavel Tsvetkov

Wojciech Matyszkiewicz wrote:




That's a great news. Hope they will include most of the requested ones in DVX4.

mikhailo wrote:

Just a question. How will this feature deal with flexion languages? For now filter in DVX does not support search of all wordforms.
Most of langs has a limited flexion (different sing.-pl. forms) but DVX does not support even them.

How could such defective function be very useful?

Wojciech Matyszkiewicz wrote:
Fuzzy matches is the only choice to deal with terminology in inflected languages.

Epameinondas Soufleros wrote:



This is not the only choice. You can do it like Kilgray with pipes and asterisks to separate the fixed beginning from the variable ending, or you can offer a full RegEx solution like this:

(άνθρωπος|ανθρώπου|άνθρωπο|άνθρωποι|ανθρώπων|ανθρώπους)

or like this:

(άνθρωπ(ος|ο|οι)|ανθρώπ(ου|ων|ους))


The first is what an inexperienced user would prefer, the second is for more advanced users. They both work on the basis of the same ‘mechanism’, so there's no need for the software manufacturer to select either one or the other solution.