Home again!

Those last 50 miles or so last night were particularly stressful, because a deer managed to show up in our headlights smack dab inside the town limits of Chama, New Mexico (which probably was a plus, as I was observing the town's 45 mph speed limit). Only some well-nigh instinctive swerving reaction—accompanied by some eloquent swearing—kept the animal and our car in one piece.

We pulled into the driveway almost exactly 18 hours after departing Houston. This is not bad time, especially considering two three stops for gas and food, one stop to walk the dog, another stop to stretch our legs (and do whatever else was required), and a whole 30 minutes spent at Clines Corners wandering through their "new and improved" souvenir shop.

I tell you, I felt a little wobbly going up the front steps.

A reasonable night's sleep helped take care of the fatigue, and I've been translating urgent jobs like a maniac since about 8 am. I plan to take a short break and then get another 2,000 words in of the 40K Job (which doesn't change it's name just because the actual word count is 45K).

Oh, man was it ever nice to eat breakfast on the verandah this morning!

Cheers...

Pushing myself to push the panic button...

It took me a while, but I finally got up the guts to admit I was wrong in my choice of memoQ for the 40K Job. This is due primarily to there being about a gazillion "false tags" embedded in the source document, but more so, to the fact that the tag replacements in what memoQ provides strip away too much context, so that I need to have a printed copy of the source file next to me to understand what the source text is trying to say.

"Tags" are one way that some translation memory products deal with stuff that's either not translatable (such as graphics) or involve some kind of formatting (such as superscripts, subscripts, bold, italic, etc.). "False tags" are elements within the source document that don't really do anything, except to fire off the "insert a tag here" routine inside the translation memory product.

Through my experience in using "tag"-oriented software (e.g., Déjà Vu), I know there are ways to deal with (eliminate) false tags, and in documents where the embedded graphic or Greek letter or superscript occur a few times on a page (or even paragraph), ordinary tags are generally not a problem. But the 40K Job is filled with stuff like:

где k следует определять, как в {1}Л.2.4{2} настоящего приложения, но с заменой характеристик пояса на характеристики элемента решетки: {3} на большее из значений d или {4}, t на {5} и {6} на {7}.

which translates to

where k should be determined as in {1}J.2.4{2} of this attachment, but with the flange characteristics replaced with those of the lattice element: {3} with the larger of d or {4}, t with {5}, and {6} with {7}.

In this segment, all of the tags except the first two, which as far as I can see are "false tags," are place markers for inline graphics used that were used instead of typing out, e.g., D_b.

What makes this worse with memoQ is that the tags can't be rearranged (unfortunate, as it forces some awkward wording from time to time), and some tags don't appear at all (the software "helpfully" doesn't even show a tag if the tagged element occurs at the very beginning or very end of a segment, which on occasion just exacerbates the awkward wording).

Anyway, I've just spent about 15 minutes exporting what I've done so far back into a bilingual Word document. Here's crossing my fingers that I can do the rest in Wordfast, and do it within my deadline, which my client has extended (and which I've blown through half of already without really moving forward).

Cheers...