However, it does at least appear to be consistent, which is the most important thing, so its preferences could be considered a publisher's house style and just something you have to get used to.

Actually one other possibility strikes my mind, which is that the style of accidentals in that copy were actually what was used in Bach's time, and the current norm arose later (and was typically retroactively applied to earlier works). If that's true, then this effort could be an attempt to be "true to the original", so to speak.

Of course, this is complete and utter speculation, as I haven't read TFA or anything. :-)

It's funny, I was just playing around with a different music editor (Denemo) transcribing some written music. And that piece had a bunch of accidentals that were unnecessary and unbracketed. But it seemed to be somewhat arbitrary; some advisory accidentals were normal accidentals, some were in brackets, and some places I would have expected them based on other uses were missing.

There were advisory accidentals that crossed both barlines and octaves. Dunno if I saw cancels of cross-bar, cross-octave accidentals. :-)

I don't know if this was just sloppy editing or someone actually preferred it.

(Actually to be honest, I played cello in school orchestra for 9 years (basically stopping about a decade ago actually), I actually didn't know that accidentals didn't cross octaves until looking into related issues.)

I can't say I've used MuseScore (though I will check it out), but it has Lilypond export.

I'm not quite sure what that means though; whether you lose a lot in translation, or MuseScore pushes its defaults so that they're hard to remove, and such.

(I also understand that some of the other behavior is questionable too, e.g. strcat_s will replace the first character of the source string with '\0', which makes it hard to recover from an error.)

Or, those "security enhanced" replacements for most ANSI C functions that are actually less secure than originals (fopen_s(), etc).

OK, I understand the portability objections to these functions... but saying they're "less secure?" You're gonna have to explain that one.

Are you talking about "Eclipse when working in Java" Eclipse, or "Eclipse when working in C++" Eclipse? Because the former is pretty awesome, but my experience with the latter (admittedly several years out of date) is that it pretty much sucks.

I wouldn't be surprised if some future version of Windows doesn't even allow you to install programs that don't come from some Windows 8 version of the iTunes app store.

I suppose it's theoretically possible that I'll have to bite my tongue later, but IMO you're crazy if you actually think that. Basically MS would have to ignore the largest reason that Windows is where it is, which is backwards compatibility. Yes, there was a lot of monopolistic crap in there too, but that wouldn't have helped if they were making it hard for people to run their old software. But the degree that MS bends over backwards to ensure that old software continues to run goes to ridiculous lengths.

Despite what may be prevailing /. wisdom, MS really has been a developer-friendly company. Their main problem is changing "favored technology" like pants -- one day it's straight Win32, then MFC, then WinForms, then XAML, then Metro, and who knows what buzzwords I've missed. But the old technologies continue to be supported; hell, you can run DOS programs on Windows 7 if you've got the 32-bit version. (I think, I can't verify.) They're happy as long as people are developing on Windows.

If your prediction comes true, it'll be the death knell for Windows.

Actually crazily enough I worked for a (relatively small) company that did mixed platform development, and while there were some copies of the full VS floating around, I'm almost positive most of the day-to-day Windows development happened with the express editions. (Caveat: the build system was SCons, using the MSVC compiler.)

Or maybe that was just my day-to-day dev since I was an intern. :-)

Point being, the express editions really were a surprisingly capable tool on their own.

I haven't seen anyone use an IDE more efficiently than what I can do with Emacs and the command-line.

I use emacs for 99% of my stuff, and I have to say, while it's a great editor, I wish I had IDE-level code browsing abilities (and to a lesser extent, intellisense-style stuff). I'd kill someone for good "go to definition" support. Ctags-style stuff is a shitty substitute, at least on our code base, and I've never really been able to get the fancier stuff to work well. VS isn't perfect there either, but it's still a lot better...

There seems to be a pretty high correlation between studies showing correlations and people saying that correlation doesn't imply causation. I wonder if there's a causal relationship?