Monday, October 23, 2017

Bad programming decisions in CAT tools

Everyone knows what software bugs are: flaws in a program that make the software fail or behave in unwanted and unexpected ways. Bugs are unavoidable in something as complex as software. The most we can reasonably ask of programmers is that they try to lessen the frequency and severity of bugs by using sound programming practices, and that they correct bugs quickly, once found.

Bugs are unintentional, whereas virus, Trojans and other malware are created with malicious purpose.

But, between unintentional bugs and willful malware, there is an entire class of problems caused by intentional programming decisions: when software features work as designed, but the design itself is ill-thought-out.

I'll give two examples from SDL Trados Studio and memoQ.

In Studio, an example of flawed design is the deliberate disabling of "smart quotes" when change tracking is active. According to SDL, "This currently is by design so that no uncontrolled/automatic changes should happen when typing in review mode." However, they didn't think through the real-world consequences of their decision: now, a translator may use smart quotes during translation, but since they are disabled during review, any apostrophe or quote entered during review will be straight. After review, the text of the translation will contain a mess of straight and curly quotes and apostrophes.

SDL Studio: The translator used smart quotes, but the apostrophe used during review is a straight single quote.

Furthermore, apostrophes and quotes are tiny characters: it is entirely possible no one will notice the problem for a while. The first person to notice might very well be the customer... perhaps when he receives the final printed copy, after it is too late to correct the error.

Disabling smart quotes when change tracking is active is harmful, and the problem is made worse because it is not well documented.

For me, I have a good workaround: a short program I wrote in AutoHotkey that allows me to use two different types of smart quotes (and also straight quotes) with no tweaking of Studio's settings, no matter whether change tracking is active.

Maybe, under certain circumstances, it would be better to disable smart quotes during review, but this is a decision that should be left to the translator, not imposed by SDL.

Let's now pass to Studio's main competitor, memoQ.

Here, the flawed feature is a change introduced with version 8 of memoQ: a new behavior, touted as an ergonomic improvement, of the Shift+F3 "change case" function.

Before version 8, Shift+F3 behaved in memoQ much the same as in Word, Studio, or many other programs--it toggled through the various permutations of change case: all lowercase, ALL UPPERCASE, and Mixed Case. Now Shift+F3 opens a drop-down menu, where the user can select the case.

memoQ: An unecessary drop.down menu for a simple function.

The result is the same, but the new "feature" hinders smooth typing by shoehorning in the workflow a change no user had sought. The new behavior slows a translator used to hit Shift+F3 a couple of times, until the desired case is achieved, then press the right arrow and continue typing. Changing case now often requires at least an extra keystroke; worse, it introduces an unnecessary change in a behavior that most users had imprinted in their muscle memory. And since Shft+F3 continues working as before in other programs, the irritation caused by the change will not fade away as you form new habits.

Unlike with the Studio example, there is no workaround: the only thing you can do is return to memoQ 2015, abandoning any useful feature added in version 8.

I imagine that if Kilgray introduced this new feature, someone must have either asked for it or thought it was a brillant idea. Instead, just like SDL's disabling smart quotes in change tracking mode, it is a bad programming decision.

Special free software offer


As I mentioned before I use a short AutoHotkey program to enter smart quotes and apostrophes in Studio. The program lets me enter "smart" single and double quotes, curly apostrophes, "French"double quotes, and also, when I need them, "straight" double and single quotes.

This works for me and would work for other Italian translators as well. If you need a copy of this utility, let me know, and I'll be happy to send it to you "as is".

I can customize (for a small fee) this utility to use different sets of single and double quotes. If you are interested, please write me (you can use the contact form in this blog).

Friday, October 06, 2017

New edition of Mats Linder's Trados Studio Manual now available

Mats Linder has just released a second edition of his immensely useful Trados Studio Manual, now updated to cover Studio 2017 SP1.


Of particular interest are the extensive changes and additions to the Machine Translation section.

The new edition of the manual is free for those who purchased the first edition of the Studio 2017 manual, and available at a 50% discount for those who purchased earlier editions of the manual (those for Studio 2015, 2014, etc.)

You can get the manual from the Trados Studio Manual webpage.

Thursday, October 05, 2017

Belt and Suspenders

Some of us may have a tendency to panic, when faced with some unexpected computer error (and such incidents generally occur when little help is available, or very close to deadlines, or both), but there are steps we can take to defend ourselves from the worst effects of such mishaps.

On Sunday, my partner's computer froze: suddenly neither keyboard nor mouse responded. We tried disconnecting and reconnecting keyboard and mouse, but without result. The last thing left to try was a hard reboot. We did that, and, after restarting in safe mode and then again in normal mode, the computer seemed to be working right: all programs responded as expected.

Until, that is, my partner tried to launch Studio 2017 to continue a project she was working on. At that point the Studio splash screen briefly appeared, only to be followed by an ominous error message: "Not found".


We clicked on the Knowledge Base Community link, but to little avail: no useful help there for this particular error message, so we opened a ticket with SDL's support. Of course, since it was Sunday, the earliest we could expect to hear from support was the following day... and my partner's deadline was rapidly approaching, so we needed to find another way to continue work on her project.

We had two options: either copy the project's files and memories to her backup computer (a laptop), where we still had a copy of an earlier version of Studio, or work on the project with a different translation tool.

We chose the latter option. Getting the sdxliff file to continue work was a simple question of copying it from the SDL 2017 "Projects" folder to a different working folder, but since we couldn't launch Studio, we had to use a different tool to export the most up to date copy of the translation memory: we used Xbench to load the TM and then export it in TMX format. It was then a simple matter of creating a new project in memoQ, add to it the partially translated sdxliff file, create a new memory, and import into it the TMX file we had created in Xbench.

My partner was then able to continue her translation.

The next day, Monday, we received instructions from SDL support. We were told first to try re-installing the program; when that didn't solve the problem, we tried renaming the "projects" XML file, and then various other SDL files and folders. Nothing seemed to work, and the SDL support technician was stumped. She said she'd need to escalate the issue to a more experienced engineer, but since the second-level engineers work out of the UK office, that would have to wait until the following day.

On Tuesday, we were again on a support call, this time with the second-level engineers. They suggested various other remedies, finally succeeding in restoring Studio 2017 to life -- the culprit turned out to be an obscure Windows file (BTW: kudos to SDL's tech support -- it's well worth the money we pay for it, and they are generally patient, thorough, and professional).

In the meantime, my partner had been able to complete the translation of the project in memoQ, and she then proceeded to finalize it in the newly-repaired Studio. So, a happy ending to our short tale of technical issues.

But it got me thinking that such happy endings don't just happen: they require preparation and planning:

  • If we had relied on a single computer, and a single CAT tool, my partner would have been unable to continue working until SDL support had solved the problem;
  • If we had not installed Xbench, we would have been unable to export the data from Studio's memory;
  • If we had not paid for SDL support, our only recourse would probably have been to take the computer to a repair shop, or perhaps ask for advice in the various online forums available, and hope for the best;
  • In this case, there was no damage to the files or to the computer's hard drive, but if such damage had occurred, we would have been prepared also: we regularly back up our files both to external hard drives linked to our home network, and to online storage.

Technical problems happen, but if you plan for them, you can minimize the damage they cause.

Thursday, September 14, 2017

The Microsoft Language Portal has a new look

The Microsoft Language Portal, an indispensable resource for all translators who work with Windows software localization, has just received a new look, which brings it better in line with recent changes to other Microsoft programs, from Office to Edge and Windows itself.

The new look of the Microsoft Language Portal

Personally, I find the new look, more modern--and perhaps more attractive--but a bit less legible.

Compare the old look below:

Th old look of the Microsoft Language Portal


The addition of more color to the interface helped separate the search form from the results area. The old look also displayed more terms per page:

The old version of the Microsoft Language Portal displayed more terms per page
I haven't worked long enough with the new interface to see if there are other changes other than the obvious cosmetic ones.





Monday, September 11, 2017

GT4T - A tool for translators, instead of a tool to replace translators

Guest post by Dallas Cao, developer of GT4T


Many translators believe that machine translation (MT) is a horror story, and that using machine translation (MT) in our work only results in bad quality. Indeed, after I started advertising GT4T (Google Translate for Translators) on Facebook, the reactions I got from many translators were negative.

They are right to think that the overall quality of machine translation is bad, and that any translator who mindlessly uses machine translation puts his or her career at risk; but the quality of machine translation is improving: Google’s neural translation engine, for example, has surprised many, to the point that some agencies have started using it to replace human translators, relying afterwards on translators as post-editors--a situation that creates even greater hostility against MT among translators, who are rightfully afraid that post-editing means for them toiling at mind-numbing grunt work.

Most of us use on-line reference tools in our work; when an online reference tool gets better, it helps us more. In my opinion, MT is the most advanced technology in translation, and, therefore, it should benefit professional translators first. If we consider MT as a reference tool rather than a threat, shouldn’t we be glad when our tool gets better?

I never liked the idea of letting MT translate and translators confined to an unrewarding task of post-editing; however, we can use MT to “translate” a word, a term, a phrase, or a part of a sentence that we judge it will translate well. Sometimes MT returns nonsense, true, but most of time, when used carefully it provides a surprisingly useful translation.

I developed GT4T because I wanted a tool that could help translators (and not translation companies) make the most of Google Translate, without becoming ourselves post-editors. Copying and pasting between Google Translate and your work is not a good solution, as it takes too much time. Some TM tools already include MT, but they all submit the whole sentence to MT: you cannot choose to have MT translate only part of a sentence.

GT4T is a tool that lets you submit any portion of a sentence of your choice to MT with ease. It’s very simple: you select some text anywhere (including from inside a CAT tool), press a keyboard shortcut, and the selection is replaced by translation from MT. Simple as it is, I believe it is the correct way of using MT. As we use keyboards most of time, GT4T painlessly incorporates MT into our workflow.

A usual problem with MT is inconsistency--the MT engine translates the same term differently in different sentences. GT4T has a simple glossary feature to solve this issue. You press a keyboard shortcut to add a term to GT4T’s glossary, and that term will be pre-translated before submission to MT; thus the results suggested by MT will be consistent.

GT4T - Glossary Setup

GT4T also offers the option to use both Google Translate and Microsoft Translator at the same time. The results from both engines appear in a popup, and you can then press 1 or 2 to paste the corresponding translation.

GT4T - Alternative Translations

I expect there are still many years ahead before MT can effectively replace us. Before that happens, MT can be a great aid--a tool that can increase both the speed and the quality of our translations, if used properly. A tool for translators, instead of a tool to replace translators.

---

You can find GT4T at: https://gt4t.net/en/

Sunday, July 30, 2017

Beware: "Phras.in" is now a zombie site

Several years ago there was a site that was very useful for translators: Phras.in. It allowed you to enter two different search strings, and see how many hits each returned, and what context they had. After a while, the site disappeared. It has recently resurfaced, apparently with the same graphics and interface, but it is not its former self: it is some Chinese zombie site. If you attempt to use it, it does not return any hits, and only display random text. I hope it did not infect my computer with some malware. I would advise against trying to visit it again.

Tuesday, July 18, 2017

Deceptive advertising from Fluency?

I've just received an e-mail advertising the Fluency CAT tool.

I tried out Fluency (and paid for a couple of licenses) a few years ago and was underwhelmed, to say the least. At that point, the program was simple and easy to use, but very buggy--all the (few) projects for which I used it run into problems. All issues were quickly solved by Fluency's support, true, but there were enough problems at that time to make me decide that Fluency was not helping me, and that the program was, in fact, slowing me down.

The tool might, in the meantime, have improved greatly, but, just like many other CAT tool vendors, I see that Fluency engages in deceptive advertising: prominently displayed in their ad is a testimonial in which a satisfied translator claims that "Fluency has enabled me to double my translation speed".

I have no reason to doubt that the statement is accurate, but, the same time, I suspect it most probably is misleading: when I hear from CAT tool vendors that their particular tool doubles translation speed, they are very careful not to say what this improved productivity is measured against. A CAT tool doubling a translator's speed if that translator didn't use CAT tools before? Yes, probably, even likely. But if that doubling of speed had been achieved when compared to the use of another CAT tool Fluency would be quick to say so. Since they don't, it's fair to assume that they are comparing apples to oranges, and that the claimed speed increase would in fact be achieved by using any other decent CAT tool.

To be clear--most professional translators by now use CAT tools, hence, any productivity increased claimed by CAT tool venderos should be measured against other CAT tools, not against not using CAT tools at all.

Monday, March 06, 2017

Something from the past

From the Atril time line:
1993 – Atril develops Déjà Vu software, the first Windows-based Computer-Aided Translation Tool (CAT tool) on the market.

Yesterday I was clearing a closet where I kept some old stuff, and I found these:

A couple of original Déja Vu diskettes. They might even  still be in good working order (that is if one had a computer with Windows 3.1 on it).

There is a date written in pen on them, from '95, when I must have checked those disks for integrity. But I know I had bought DV before moving to the States… must have been back in 1993, when I was working at Logos, in Italy, and that fits right at the beginning of Atril’s time line.

So, judging from the serial number (27) still clearly visible on them, I must have been one of the very first users of CAT tools for Windows. I didn't use Déja Vu for long: in 1994 I moved to the States to work in the translation department of J.D. Edwards, a software company. Shortly after I arrived the company adopted a translation memory program, but that was IBM's TM2 (later still, J.D. Edwards changed to Trados).

But I still remember the excellent technical support we received from Emilio Benito, the late founder of Atril.

Saturday, March 04, 2017

Sometimes, the best way to mislead is to tell the truth: the case of the nonexistent 36% productivity gain

Sometimes, the best way to mislead is to tell (part of) the truth. Case in point: to much fanfare, Memsource's blog announced some time ago that translation clients can increase their productivity by 36% by using translation memory:
[...] The table above applies a predefined net rate scheme to a sample of 500+ million words. It clearly shows that Memsource’s most active users increase their productivity by an average of 36% by using translation memory. This means that if you had an average cost of 10 euro cents per word, for this volume you could save €18.6 million. Not bad!
What they imply (but cleverly don't state) is that it's only by using Memsource that you can achieve such impressive productivity gains. They are careful not to say against what they measured. My guess is that this 36% productivity gain was measured against similar translations done without the use of any CAT tool at all: If they had achieved a 36% productivity gain over what other CAT tools can do, Memsource would proudly boast of it.

However, since most professional translators already use one or more CAT tools, the productivity gains that Memsource peddles to their prospect are really not there... in fact, some professional translators complain that Memsource actually slows them down... and I can confirm that from my own experience with the tool.

I understand why the program is attractive for translation companies. I can even understand why people who translate only occasionally may find a free tool useful. For full-time professional translators, however, the slow creep of Memsource and similar online straitjackets is a big step backward.