Well it would be if there weren't so many broken links to MSDN and Technet. I don't have to use those sites any more, but it always seemed like they were constantly changing the urls for stuff. Even links from product help pages or error messages were usually always broken.
Their search capability simply sucks. Unlike google, it seems knowledge from Bing doesn't seep to other parts of the company. Windows store search also sucks. It can't handle misspellings the way Youtube and Chrome store can.
It's not just the builtin search, the way the content is organized always seem be out of line with the way my brain works.
As a reference, take a look at (if you're not familiar with it already) at PostgreSQL's manual[1]. With Postgres, I don't need to search, I can (nearly always) find what I need by scanning the table of contents by eyeball. Also, I can download the whole thing as a PDF for offline reading (or printing if I wanted to).
This is an example of the standard Microsoft should aim for. And that's just the first example I can think of off the top of my hat. FreeBSD has a great manual, too, and last time I looked, the JDK also came with very good - and well-organized - documentation. I don't think it is hard to do per se, just very tedious, but Microsoft certainly has the resources to do it if management makes it a priority.
I am sorry for ranting a little here, but the state of Microsoft's documentation is very disappointing in light of the resources they have; when I was a Linux newbie and made the mistake of asking a stupid question on a discussion board, I got flamed rather hard about not having read the documentation. I then replied - my second mistake, I guess - that apparently to Unix people, "user friendly" means "comes with more documentation than you'd ever want to read". One of the flamers replied - in a very matter of fact tone - that, yes, documentation is always good, because without documentation you're screwed eventually; with good documentation, no matter how complicated and nasty a program is, at least you have a chance. It took me many years to understand the wisdom in those words, but I think I only came to really appreciate them since I became a Windows admin.
Couldn't agree more. I've always considered PostgresQL's documentation the finest of any software I've used, OSS or not. Many years' experience with all kinds of software convinces me it's nearly a law of nature: the quality of a project's documentation predicts the quality of the software itself.
However I'm pretty sure writing excellent documentation must be a very hard task, otherwise it wouldn't be such a rarity. Of course, Microsoft has the resources to do it, but the level of resources required to produce and maintain surely must be decidedly non-trivial. Reasonably, we could conclude they think it provides poor ROI.
We might well beg to differ on that point, and worth saying how much their inadequate documentation reflects the quality and utility of their products.
I think postgres' docs work quite well as a reference documentation. But in my opinion they're quite bad at introducing users to postgres/SQL/databases, including important operational tasks like backups.
In some respects that's true, though the docs do contain a lot of patient explanation at a pretty basic level.
I wouldn't exactly call it a tutorial but it did teach me a great deal about SQL when I started using the db > 15 years ago.
New standards and features have made the program more complicated over time, so periodically I need to study up on these topics. More than a few times, it's been really useful to have that "beginner level" info re: stuff that's new to me.
But like you say, not all subjects are covered that way, perhaps they haven't been considered to be basic. I bet if there are enough requests the project would improve the documentation in those areas.
