I give reference to a site that describes a class of problems that might include the one described by the op. And it is both scary, and inherent to the existing Ruby implementations. And I also mention that this description keeps me even farther away from Ruby than I was.
And I get 5 downvotes (and counting?), and no comment. I'm not new here, and I do find this puzzling.
I've learned my lesson, though - if I don't have praise for Ruby, I'll just shut up.
Maybe you might learn some kind of lesson about how to view your own comment as someone else will read it?
I.e., you seemed to first disqualify yourself as someone with expert knowledge, then offer a useless criticism of the basic technology the OP is using, then offer an article that supported your negative opinion. So after that, do you expect someone seeking expert help to click that link?
> you seemed to first disqualify yourself as someone with expert knowledge,
Right.
> then offer a useless criticism of the basic technology the OP is using,
Where did I do that? The only thing I did (other than disqualifying myself as an expert), is:
> then offer an article that supported your negative opinion.
If you actually read the article I pointed to, you might have learned something about crashing Ruby processes that leave no trace, which -- the humanity -- is what the original problem description was.
> So after that, do you expect someone seeking expert help to click that link?
Why would you (or anyone) expect the article to be "supporting my negative opinion" when the accompanying text was "perhaps this is relevant to your case?". I am not, but Joe Damato _is_ an expert in debugging undebuggable situations. Read it, you might learn something about your favorite language that you are likely to learn nowhere else.
This article helped form my negative opinion of existing Ruby implementations. Your post, among others, helped me form my opinion of the Ruby community. I totally ignored the Zedshaw bashing of it, but I'm starting to think he had it spot on.
Downvote away folks.
What I think the main problem was is that you answered someone in an emergency with a probably unrelated blog link.
The original description was "server is crashing without leaving a trace", which sounds like an obscure bug to me -- certainly of the kind described in the article I linked to. It's as related as it can be given the info they gave when they originally posted their query.
Imagine someone struggling with tuning MySQL and someone else hijacking the thread with "MySQL is slow as molasses in January, use MongoDB/Cassandra/Distributed MapReduce in Erlang!". How does it help? If you want to discuss fundamental ruby issues like this one, submit your link separately.
But that's entirely unlike what I wrote (although apparently that's how many people going through the thread read it). I was not advocating a different language/system whatever. I was pointing to an article saying "Existing Ruby implementations tend to randomly crash because of a very specific design flaw -- here's why that happens, and a possible, though not easy, way to determine if your crashes are a result of this design flaw". Given the info originally provided, it's hard to provide a more relevant link.
Personally, I don't care about Ruby, which is why I have no intention of submitting that link; and even if I did, based on the responses here, I suspect it would be downvoted as "badmouthing Ruby", rather than actually discussed. But I don't really care either way.
I was just surprised because while I have been downvoted before, it was never so harshly and without any discussion, especially since (if anyone actually followed the link they would have realized) it's possibly extremely relevant.