The author doesn’t seem to know what a debugger is…
Debuggers are still good at debugging serial code, but these days my code is asynchronous and distributed over many hosts. There is no concept of “stepping through code” in asynchronous systems - stepping implies that you are on a single thread, running on a single machine.
If all debuggers were good for was setting breakpoints, his headline might not be stupid.
If systems aren’t asynchronous enough for you, we’re in the process of launching the Internet of Things where we make it extremely difficult to launch a debugger on the devices where your software runs.
I’m not sure if it makes sense to describe flaws of the platform as flaws of the process.
Debuggers are well capable of attaching to remotes, it’s just a question of the remote providing the facilities.
This article is obviously terrible: inflammatory link-bait that is not merely technically naïve but technically ignorant and actively dangerous. Not that I would expect any less, but the lobste.rs crew clearly agrees: many (many!) more upvotes on the comments decrying the article that the article itself. So, serious question: why is this the top-rated lobste.rs article right now? Obviously, everyone here is disaffected by HN to some degree or another, but an HN algorithm that I actually like is that articles that receive more comments than upvotes are assumed to be generating more heat than light and drop off the front page. Worth considering something similar here?
incr flag: incorrect
Wow, what a gratuitous poo fling. All the things he outlines in the article are good tools, but telling me that my debugger is obsolete because metrics? I think not.
I don’t disagree with the need to have an appropriate measurement of how the system behaves in production, as well as a well-known set of parameters under which you know whether the system is thriving or not; I disagree that debuggers are useless because I sometimes use them and when I can, I find them delightful to use and very helpful. Granted, some environments and multi-threaded code can make this harder than it needs to be, but I don’t think such blanket statements are useful.
Why not both?
Why do people feel the need to make clickbaity, dramatic, over-generalized headlines? This article was not bad enough to be a waste of my time, but I would not have read it had the title been accurate. I’m guessing that’s what the author wanted.