An interesting (and pretty diverse) thread - I really should stop by more often! Anyway, I thought you just might want to hear the opinion of someone evil.
I'm stepping in with a couple of comments... (and don't have the energy to get involved in the cloud discussion, right now, so am choosing to sidestep that one
)
As a blogger I have only ever deleted comments (not including SPAM) when they have either said something illegal (e.g. "where can I get a cracked version of AutoCAD?") or insulting (usually to others, so far I've been OK with people having a go at me). But then I've been lucky enough to not have too many of either category - probably because I write my own blog, and so any comments tend to be targeted at an individual (me). When a blog has multiple (and I won't say "anonymous" - the word I'm looking for is probably "unfamiliar") authors, it's easier to say things that are blunt to the point of being disrespectful. Just to be clear - I haven't read the comments that have been removed, so this is not me passing judgment. I did want to step in to say some words in defense of whoever is managing the DevBlog's comments, though.
Whoever deleted the comments - and it could well have been Stephen, although there are lots of authors who have access the blog - surely did so to the best of their judgment. I'm almost certain that any removal of comments was done to keep the blog's tone respectful - not to stamp on dissent or criticism, per se. This isn't a bad thing, as far as I'm concerned: contributors to the blog - whether authors or commentators - should feel they can do so without risking being abused for it. Perhaps it's cultural - we Brits tend to prefer public criticism when it's stated diplomatically - but then this blog also has a broad (and diverse) audience to cater to.
That was comment number 1... onto the next:
It's more than ironic to see Kean Walmsley espousing the benefits of F# and parallel execution for developing plug-ins that leverage multiple processors, until he discovered that a major barricade to that goal was AutoCAD and its inherent lack of thread-safe code.
Well, no. I continue to espouse the benefits of F# and asynchronous programming models, I'm just a bit attention deficit.
I certainly didn't simply stop when I "discovered" an issue inside AutoCAD's architecture: I've spent quite some time and effort describing appropriate models for dealing with asynchronous calls in a way that works well with a single-threaded application (my favourite being to use an Erlang-inspired mailbox architecture).
Anyway - that's all I wanted to say. I would apologise for posting something off topic, but then this thread has covered so much ground I suspect it isn't needed.
Cheers,
Kean