On these pages (i.e., this directory and the issues sub-directory) I plan to describe bugs and problems I encounter while working with Atril’s TM program (Translation Memory) Deja Vu X, and to suggest possible improvements.
Things I like about the program I will hardly mention, so this can cause the false impression that I am dissatisfied about it. I am not, I think it's a very useful and usable program.
See also this explanation of my possible bias.
The only Translator's Tools I ever used were Deja Vu versions 3 and X. So I can't comment on Trados, SDLX, Transit or whatever else there is on the market.
In April 2002 I bought DV3. I already knew about the program since 1999, but wasn't convinced it could be very useful for me. In those days I dedicated only part of my time to translating. One important client used their own translation memory system, with a file format that was (and is) difficult to reconcile with DV.
When trying out some German files with a fellow translator who already had DV3 in 1999, I found there were so many “rogue codes” (unnecessary codes in weird places, e.g. around umlauted characters) that the result was unworkable.
Yet, when I finally bought the program in 2002, I soon realised I should have done it earlier. The problem with the rogue codes was largely solved, by improvements in the DV3 program itself, and with the recommendation to first accept all changes in Word file in which these had been tracked.
From November 2002 I participated in beta testing DVX. On 4 May 4 2003 I expressed my amazement that DVX was brought out in a public release, at a moment when I thought it wasn't ready for it.
So I continued to use DV3, and I liked it. However, I kept making notes about possible improvements to DV3. They were only terse notes, without dedicating enough time to re-testing. Although these notes were in an unpresentable state, on 4 March 2008 I decided to put them on my site anyway.
About a month ago, I decided I’d better upgrade my hardware. Over the years, the risk of hardware failure increases, and it's better to upgrade quietly without pressure. Also, my old Windows Millennium irritated me more and more, because it quickly runs into a shortage of handles (not memory!) when many programs are open at the same time. And I wanted many: DV3, Eudora, Firefox, Acrobat, a dictionary program, sometimes Finereader for preparing a next project, etc. XP and Vista were said not to have that handle problem.
It seemed a good moment to change some software too. So I bought DVX.
First I thought I’d try DVX for one of my own web pages first, without a deadline. But it turned out it suited my rather complicated planning (moving data, installing programs, planning back-up procedures, converting translation databases, keeping up with translation work meanwhile) better to jumped in at the deep end.
So I did a real project for a real client with a real deadline as my first DVX project!
And I kept that deadline, despite having to deal with a problem concerning the file exported from DVX.
The learning curve wasn't steep at all. Of course I had to get used to some streamlined keyboard shortcuts. Wizards and the Project Explorer work a little different than in DV3. But it was all pretty clear and I could pick it all up very quickly. Of course it may have helped that I participated in beta testing, although that was many years ago.
Also, the workflow is essentially the same for DV3 and DVX.
So for new users, without experience with DV3, it is probably wise to take some time, a few days or a week, to practice with DVX, before you start with any real translation assignments.
Converting the old DV3 databases took quite some time too, because the subject codes are not fully compatible, and in my case also because I also did some cleanups that I wanted to do for some time. That may be a problem that only applies to my personal situation, so don't be alarmed by this.
This evaluation is a work in progress. I'll add issues as I discover them as soon as I find the time to make them presentable.
I'll probably also browse my old list for DV3, bit by bit, to see which old issues are no longer applicable or have meanwhile been resolved in DVX. My first informal impression is that this will often be the case: DVX is really better than DV3, and DV3 was a very good program already.
Below is an overview of menus to keep the issues accessible even if there might be many. There is a full menu listing all of them, as well as several thematic sub-menus that list only part of the issues. Hyperlinks to the same issues may occur in more than one of the sub-menus.
Issues described later will be put higher in the menus than older ones.
Each file describing an issue will have a name of the format cyymmddx.htm, where c is 2 (meaning the 21st century), yy is the year, mm and dd are the month and date, and x is a letter between ‘a’ and ‘z’, which is used to distinguish several issues that I might discover or describe all on the same day.
This naming scheme, and the order of the menus, will help returning visitors to distinguish between what they already read before, and what is new to them.
The issues sub-directory may have sub-dirs with the
same name as the description file (minus ".htm" at the end), to contain
supporting data to make the described problem repeatable.
Example: directory issues/2081130a
contains supporting files to the problem described in html file
issues/2081130a.htm.
Below are links to the menus:
| Issue grouping |
|---|
| All issues |
| Keyboard, focus, mouse |
| Search and replace |
| Spelling check |
| Screen view, presentation |
| Import and export |
| Database issues |
| Matching algorithms |
I have no business association with Atril. I won't hesitate to say it out loud when I like what I see, but I won't avoid criticism either, if there is reason for it.
The commercials on my pages are automatically selected by Google Adsense. I don't know the advertisers and I do not contact them.
Nobody's putting pressure on me to write or not to write anything at all, or to influence the way I write about any subject. And if anybody tried, I’d reject and ignore that.
© 2008 R. Harmsen