Negative Assert Syntax; Actual vs Documentation

Opus' documentation says:

"Additional information can be found in Microsoft's introduction to TR1 ECMAScript (the flavor of regular expressions that Opus uses by default) for more information".

The linked page says:

"A negative assert matches anything but its contents. It does not consume any characters in the target sequence. For example, "(!aa)(a*)" matches the target sequence "a" and associates capture group 1 with the subsequence "a". It does not match the target sequence "aa" or the target sequence "aaa"."

What actually works on Opus is:

(?!subexpression)

Is it an error of Microsoft's documentation or perhaps the regex flavor changed and Opus' documentation is now linking to the wrong page?

I think Microsoft have messed up their documentation as it is missing ? characters in at least one other important place:

I have a help file (chm) titled Windows Scripting Technologies. It covers JScript and VBS Script. The page for regular expression syntax seems correct.

I don't know from where I got it, but it seems Microsoft makes it available for download. Could it be distributed with Opus and linked to from Opus own documentation? Maybe it is a better option than linking to online pages?

I doubt it could unless the copyright says other people can distribute it.

As long as someone asks before and they agree, then it should be okay. I don't know how troublesome is that.

There is a boilerplate copyright notice from Microsoft on the file, dated from 2002.

Since them, they took down the links to 'Microsoft Windows Script Technologies Home Page' and 'feedback site' (links listed on the first page of the help).

Anyways, even if it can't be redistributed with Opus, might be worth saying that it exists and let users find it themselves.

I doubt it is as simple as asking with a company that size. It may involve legal sign-off which would make it not even worth their time to respond.

If we're talking about contacting Microsoft then reporting the documentation error seems a more likely path to success, and would fix it for everyone. I don't know how good they are at fixing such things these days, though.

The feedback ("is this page helpful?") link at the bottom of the documentation page suggests they may listen. I'll report it later on, but another report can't hurt if you want to as well.