A way to say "this is who I am, where I am, what I think is wrong, and why" ?

by Michael S. Kaplan, published on 2010/08/07 07:01 -04:00, original URI: http://blogs.msdn.com/b/michkap/archive/2010/08/07/10045212.aspx

Localization Quality.

If ever there were a loaded term, that would be it.

I mean, when one considers the differences between translation and localization, some hints about how many additional dimensions localizations adds are immediately apparent.

Ideally, localization takes the idea of perfect "translation" between a source language and a target language that an automated tool might generate, and adds all of the following additional dimensions, and more:

Now although there are ways to automate some facets of the above points, there is very little desire to avoid real human review of both what might drive that automation and on the final results thereof.

And for very good reason!

Because Microsoft, like pretty much all software companies, has made mistakes here in the past, and it is those mistakes that have driven the strong need to maximize that quality

If there is a "mistake" in a localized product, a real need to understand the cause of the mistake exists, because some types of problems might suggest other problems of the same type. And "fixing" one reported problem while ignoring the same problem elsewhere is a recipe for making quality worse by adding a terminological inconsistency on top of the reported problem. The actual fix is almost always preferred.

Now that is the view from inside the product.

Outside the product, however, the view is very different.

I mean, think of the issues that led to Microsoft creating separate localizations for European vs. Brazilian Portuguese (which I previously described in About that Portuguese localization question and About that Portuguese localization question, redux).

Most of the unanalyzed feedback would amount to bad word choices and stronger language (e.g. this sucks! and so on), most of which would ignore the fact that the real problem kin this case: that targeting Portugal with a localized product created for Brazil led to a tremendous quality issue.

Someone in Great Britain or Australia might feel the same about our English.

Someone in Morocco might feel the same about our French.

Someone in Austria might feel the same about our German.

And so on....

Perhaps they are all right, or perhaps they are not. But in all cases what the customer might find objectionable might be a "by design" issue that would only be fixable (or fixed) if that dissatisfaction led to actual usability problems that led to people not buying the localized version. Or any version....

Yet any of these potential flaws (and other problems) could all be grouped under the term Localization Quality. Since any problem in any one of these areas can lead to such a user perception.

Perhaps they understand the reasons and buy the product anyway; or perhaps they do not.

Perhaps they complain and hope that a change will be made in the future; or perhaps they just shake their heads and decide that Microsoft is a company in Redmond, WA, in the USA and that it will never change.

Even though some things can change, and do change -- if the feedback can be received and assessed, and decisions can be made that would lead to such a fix.

Thus bugs like in Does bear *** in woods^H^H^H^H^HSlovenia? are fixed prior to release -- the mere act of someone notoicing the problem and communicating it is all that is needed, sometimes.

Add to that the fact that memories last a long time. In fact, the story told in About that Portuguese localization question, redux is almost singularly unique in that people did notice the difference so quickly.

Though a huge marketing push had a lot to do with that; usually even when we make improvements in languages we seldom advertise or even mention them widely. Can you imagine marketing or commercials in a country talking how much quality used to suck for a certain language but are now much much better? :-)

Perhaps it is the process itself that is the enemy of progress here; maybe being more upfront about quality changes would make it more obvious about the nature and scope of improvements.

Now customers have new opportunities here with efforts like the Microsoft Language Portal and its downloadable terminology collection and UI translations.

Though that site is aimed more at industry then at customers, it can in theory give new insight into what might lead to something that a customer might have called simply poor localization and might instead suggest a particular set of terminology choices that aren't very good. And then the only problem would be the lack of a consistent way to get that feedback to Microsoft that would scale to the virtually unlimited number of people who might have feedback.

The problem is a difficult one, and the cost of change can be quite high (without any real guarantee of improvement that would lead to increased satisfaction overall or increased sales).

The increased sales issue is of course a factor -- if everyone is going to buy the product anyway and can live with "incorrecr" dialect/spelling choices, it is harder to sell nothing more than helping someone feel better about something they were going to buy anyway.

Though none of that makes me want to run away from the problem. It makes me think it is all the more important to try to make the situation better.

Don't you think so?

Would it make you want to download one or more of the files in your native language spoken widely in the country you live and give that feedback? Or would the scope of the problem make you want to do none of that?

Now I'm not asking you to go and do any of that right after reading this blog, since I still am not sure about the way to take in that kind of feedback, exactly.

I'm just trying to see if anyone out there would be interested in doing such a thing if there were a way to do it, and to be heard.

A way to say "this is who I am, where I am, what I think is wrong, and why".

Any interest out there?

Joe Clark on 8 Aug 2010 6:59 AM:

I’d just be happy to find a single Microsoft product that enfoced legitimate Canadian English when instructed to do so.

Michael S. Kaplan on 8 Aug 2010 7:33 AM:

Canadian is an especially interesting case since some people would prefer something more like the British and others would actually prefer the English....I doubt the same is true of Britain or Australia, for example....

Joe Clark on 22 Aug 2010 12:09 PM:

Canadian is exactly *not* as you have described it. You would never dare be so uninformed on, say, Brazilian Portuguese.

Canadian English spelling is *not* a choice between straight-UK and straight-US spelling. I guess I know why no Microsoft product in existence, based on testing, actually enforces Canadian spelling. You think it’s something it isn’t. It’s so ignorant an attitude it could only come from Americans.

Michael S. Kaplan on 22 Aug 2010 4:48 PM:

Joe, if you think my statements are indicative of reasons for anything, then it is *you* who lack understanding. And since different people I have talked to all over the world *including* Canada have suggested an "International English" based on British English, I can state with some authority that if nothing else there are some Canadians who disagree with you (they might find you rude as well, but I am used to that!).

Now it is unclear how much of such a request is actually people suggesting a compromise that they believe stands closer to what they want than what they currently have, yet could stand some chance of happening, and how much is genuine desire. But I think those people who take the time to describe an issue and give examples of what they mean deserve more consideration than people who simply deem me ignorant and dismiss me without doing the same.

Now additional complications on English in Canada are due to French language influence, American English influence, American technology terminology influence, and British English influence, not to mention the influence of other Englishes around the world from immigration. Given how even the people who have answered the question are quicker than most others to  both shrug/accept American English while at the same time making minor (conflicting) suggestions, I probably have more insight here than you realize (even though I believe it is nowhere near enough to launch a localized product strategy)....

But I assume you may be referring to LOCALE support -- in which case once again the most disturbing part of your response is lack of ACTIONABLE feedback.

Joe Clark on 12 Sep 2010 11:22 AM:

Michael, I’m the one who put two years’ research into double-checking the facts of Canadian English. Not you, not Thierry, not anybody at Microsoft. Extensive testing, including of MS Word, shows that not a single word processor in existence that claims to enforce Canadian English spelling does. MS Word is one of the worst performers in this regard.

I can line up all sorts of  people who offer uninformed opinions. I could find somebody who thinks the iBot sucks. What does that have to do with anything?

Canadian English is, in actual verified fact, an amalgam of U.S. and U.K. forms that nobody else uses. At some point, you’re going to have to accept you’re wrong about this. That point will come after you read my book, which you can go right ahead and pay good money for like everybody else.

Michael S. Kaplan on 12 Sep 2010 11:29 AM:


Michael S. Kaplan on 12 Sep 2010 11:58 AM:


  1. No language version of Windows "enforces" the language of its localization.
  2. I have been talking about Windows localization.
  3. I took your original, offtopic comment and answered in a way that was back on topic, by explaiing how the preferences of Canadian users for a Windows localization were interesting. Everyone I asked *assumed* no Canadian version would ever exist, but many wanted a "more British English" version. Not all, just many.
  4. I am not wrong since I have been entirely accurate within the LOCALIZATION topic I have been talking about, and localization does not enforce language. You have been talking about an entirely different topic.
  5. Declaring me wrong and insulting me under those circumstances are not the ideal way to interest me paying for your book. :-)

Joe Clark on 10 Nov 2010 11:54 AM:

Michael, ol’ pal, you and Thierry keep making mistakes when it comes to Canadian English. You, an American with an especial interest in Tamil, need to learn some humility and do actual research. Sort of what I and others have done over the years. “Research” in this case does not involve asking people who don’t have their facts straight what they think Canadian spelling really is.

Where’s the en-CA localization of Windows? Of MS Office for Mac?

If we can’t talk about anything other than localization, why did you change the subject (in your comment of 2010.08.08:0733) to actual spelling?

How is this really two separate issues? It isn’t. Microsoft products neither support nor enforce Canadian English spelling. And you have the temerity to claim you and your uninformed friends know what it is. You don’t.

Please consider a donation to keep this archive running, maintained and free of advertising.
Donate €20 or more to receive an offline copy of the whole archive including all images.

referenced by

2011/01/30 Even in India, the language is actually known as Bangla (not Bengali)

go to newer or older post, or back to index or month or day