by Michael S. Kaplan, published on 2011/04/12 07:01 -04:00, original URI: http://blogs.msdn.com/b/michkap/archive/2011/04/12/10152369.aspx
The riddle of the title is one of the primary definitions of nitpicking -- making a point that ignores the point. For the pleasure of doing so....
The other day I had someone accost me about one of my blogs and one of Raymond's blogs.
Raymond's was his Why can't AppLocale just be added to the Compatibility property sheet page? blog, which was in turn a riff on my AppLocale cannot stand mute blog.
So anyway, mine was "wrong" in a way pointed out in the comments by Ted:
Any way to hide the message? Yes, just use the __COMPAT_LAYER environment variable (set it to ApplicationLocale but be careful not to wipe out any other ones that are set, i.e. append yours to the space delimited list) and then set AppLocaleID to the locale ID e.g. 0404. Then you don't even need to run your app via AppLocale.
As the later comments indicate, he wrote up his experiments with this environment variable in a blog titled Experimenting with Microsoft Applocale.
Okay, I stand corrected. Or rather I did, right after it was pointed out. It is right there in the comments.
And the solution is pretty cool, actually.
I know for fact that the original author either didn't know about this at the time I wrote the blog, or he forgot, because he used to be my boss, and I asked him.
Or maybe he was just getting a little revenge on me for how painful I was as a direct report? :-)
Anyway, the accosting continued.
I assumed his beef with Raymond's blog was for his blog agreeing with mine.
Though it turns out that wasn't it.
Raymond's blog talked a lot about why the reboot for system locale changes was required, though he mistakenly referred to it as a logoff/logon requirement rather than a reboot requirement.
Okay, fair enough. Though his example is entirely valid and you can see the kinds of problems he mentioned by cutting/pasting between two apps when one them uses App Locale and the other doesn't. And all of his conclusions are also valid. So I don't think he was all that wrong here, really. Just replace logoff/logon cycle with reboot cycle and logoff reminder with reboot reminder (note that the warning that comes does indeed request a reboot), and everything mentioned there is then correct.
I could write the Perl script to fix that if it were the point. But it's not.
To be honest, if you logon/logoff rather than reboot, there are ALSO strange side effects, ones that are not expected. I'll tslk about those another day since today, they are not the point here, either.
Mihai pointed out the original issue in a comment with just the right amount of snark to please me:
> Why does changing the CP_ACP code page require a logoff/logon cycle?
Because it requires a reboot, and a reboot will log you off :-)
I don't think anyone was truly that confused by the whole thing, which makes the whole exercise a little silly.
The accoster did start the accosting by telling me he loved my blog. and Raymond's. Good to know (who would bother listening to an accosting by someone who said they hated them? <grin implied>).
It seems a litle like nit picking to go on about these technicalities, each of which was fixed up in the comments, already. Expalining all this now is like closing the barn door after the horses have eaten your children (to borrow a quote from a movie you probably don't know!).
And obviously we don't want that.
And the moral of the story?
Well, did you notice how Mihai made a note that was more parenthetical because he wasn't trying to be the point of the story, since he knew it wasn't?
You see, nitpickers may enjoy what they do, but they should realize there are downsides to being besides the point. And an upside to staying on message.
There's a difference between a small group of people and small as people. The former can change the world, while the latter....
Jeremy on 12 Apr 2011 4:39 PM:
Wow, what a timely post! Did you read the comments on Raymond's blog post today on a lock-free algorithm and then decide today would be good day for this post, or was that just a coincidence? The commenters over there are going off because his pseudocode refered to "oldValue" one place and "originalValue" another.
blogs.msdn.com/.../10152296.aspx
BTW, I can't post without nitpicking a couple of your typos ("right t" "lofon"), since it is on-topic today after all :)
Michael S. Kaplan on 12 Apr 2011 5:06 PM:
It was coincidence, but people often like to nitpick over there. I don't mind the timing though. :-)