Which is an argument ethically akin to car companies knowing they have a potentially fatal defect but weighing up the cost of actually fixing it and saving lives vs. the expected cost of compensation lawsuits and not fixing it if the latter is lower.

The solution, of course, is to structure the law and/or regulate the industry so that the cost of screwing people unreasonably is always substantially greater than the cost of behaving more appropriately. Passenger suffered unreasonable discomfort on any flight? Automatic 100% refund, with a presumption in favour of the passenger if your provision is significantly below the industry average (or minimum regulated standards if the industry colludes to reduce the average).

I would have a lot more sympathy if budget airlines didn't keep pulling so many obviously shady moves to try to look cheap yet acceptable quality while actually charging more that customers expected and not always offering the experience people thought they were buying. This has become so bad that we literally have new consumer protection laws taking effect in Europe around now precisely to make a bunch of the tricks that some of these airlines pull explicitly illegal.

It should have been a reasonable and simple solution to offer transparent pricing and mid-range options, but I think that ship sailed^W^Wplane departed already. Now the industry, particularly on the budget end, needs to clean up its act or face increasing levels of customer dissatisfaction at a time when people are already looking to alternatives where viable ones exist.

Yes, there's a standard way to specify autoplay for HTML5 videos. However, not all browsers will respect it. For example, Safari on iOS won't play unless the user specifically starts the video, and this was a deliberate decision on Apple's part.

In some ways, I wish more sites would do stuff like that. It would stop my mobile ISP from "helping me save bandwidth" by intercepting all the images on sites I visit and compressing them so horribly that they are often useless. I didn't ask them to do that, I don't get anywhere near my bandwidth allowance, and they aren't providing the service I pay for.

Is it? I've worked for plenty of employers who said they offered things like on-the-job training, but in practice that often meant they had a book shelf and sometimes a senior developer would give a 45 minute talk. Compared to attending multiple lectures every day for a few years... well, it's hardly a comparison at all, it's a completely different depth of study.

As I wrote elsewhere, I firmly believe that studying for a degree is not the only way to acquire this level of theoretical knowledge, but I find the idea that any developer would just pick up the same material as a normal result of day-to-day work in a typical programming job implausible. It's not about being stupid, it's about being ignorant (in the literal sense, not the derogatory one), and it takes a lot more than reading a few chapters in a book from the office library and a couple of blog posts to meet the standard.

Perhaps, if all you do is join-the-dots programming for CRUD front-ends. Personally, my clients pay me to create new data structures and algorithms that solve problems no-one ever solved before. I can't just write result=solve_my_problem(), because I'm the guy they hire to create that API.

Remember all that theory about space and time complexity, and which guarantees you can and can't achieve in distributed systems, and formally proving algorithms correct, and how compilers and virtual machines and run-time environments work? I use this stuff all the time.

Yes it is, but the quality of that experience -- what you know and how you think as a result -- matters. As has been noted many times in this industry, there is a difference between someone who has had ten years of experience and someone who has had the same year of experience ten times.

A good CS degree course is one way to gain a lot of useful experience in a relatively short time. There are others, but a job where you spend all day writing glue code that joins up someone else's library and framework APIs isn't one of them.

What do you mean "no real excuse to not help out"?

I think FOSS is great. I appreciate the work that a lot of people in the community do, and I'm happy to help out a bit myself if I can. I also think competition is healthy in the browser market, and I'm glad that Firefox is out there.

But the last time I tried to be helpful by following a Mozillian's bug reporting advice, it took me several hours to fix the damage after their instructions resulted in damage to my normal set-up. That was time I was not then billing to a client, which ultimately reduced my income by a few hundred bucks that month. Setting up a VM on my normal (Windows) PC as suggested a few posts up is not a trivial undertaking either. This is more than I'm willing to do as a favour on personal time, and it's not what my clients pay my company to do.

Mozilla Corporation is a commercial organisation. It has over 1,000 employees, it brings in millions of dollars in revenues, and it reportedly pays its CEO more than most of us here are ever likely to earn. I'm a professional, and my company is available for software development and consultancy work. If you want me to do your testing or bug fixing for you, there's a line from Goodfellas that comes to mind.

Given my own past experience with trying to help, and my position on why I won't do it myself any more, I do find it irritating when Firefox developers/fans start writing as if anyone else has some sort of moral obligation to follow onerous procedures to help out. I rarely write about this myself, but since you've pressed the issue, I thought a few facts might bring some perspective.

Obviously your second statement is over-generalised, but in any case, you're assuming that the only reason to understand different sorting algorithms is to implement them. How do you know which of a variety of algorithms to choose, or even that there's a useful question you could go ask the Internet, if you aren't even aware of the possibilities that are out there?

Sure, I could go read a bunch of papers every time I needed to choose which tool to choose, but if I had to do that then someone who actually knew their data structures and algorithms would be getting useful stuff done while I was reading. What next, don't bother learning the syntax of your programming language, you can just look up operator precedence in the on-line help of your IDE, and if you get it wrong your compiler will probably just give you a warning anyway?

I agree with you that CS is more about theory than practice, but it's easy to construct a false dichotomy here. A lot of university-level CS courses are less beneficial than they could be even in teaching theory precisely because they get so stuck on theory that they forget to give either motivating examples or sufficient practical skills for students to explore on their own.

University-level mathematics often seems to suffer from the same affliction, presenting undergraduates with the kind of obscure, magical results that might be interesting to someone with a heavily theoretical mindset working on a PhD, but ignoring the reality that even among undergraduate mathematics students, those who have that mindset are probably a small minority. Maybe more of them would develop that kind of thinking during their studies and go on to enjoy or develop deep theoretical work, but if you confuse and/or bore them out of academia first you'll never know.

How do you spot an extrovert programmer? He's looking at your shoes while you talk.

Seriously, self-promoting incompetents are a hazard in any technical field, but in proper engineering disciplines objective assessment can make up for a lot of bull. Because software development isn't yet mature enough as an industry for that to work, it's relatively easy for a snake oil salesman to do well (almost invariably at the expense of the colleagues who are constantly clearing up his mess and ultimately the organisation they all work for).

The remarkable thing is that the executive management team are often so disconnected from the reality of their business (read: ignorant and incompetent) that they haven't even noticed their company's job ads are literally asking for things like five years of experience with Leading Programming Toolkit 2014. I'm guessing the global pool of qualified candidates by that standard is... sparsely populated.

Yes. IMHO this is what most often gets overlooked when people debate university CS/SE as a mostly-theoretical discipline as distinct from practical experience in industry.

You can study practical skills in using a certain language or library or tool, and you can become somewhat productive. But without sufficient theoretical understanding, you're just doing cookie cutter coding, and you will always have a relatively low glass ceiling on how much you can achieve.

Put more bluntly, practical skills are what you pick up to get from incompetent newbie to vaguely useful programmer in the first year or two on the job, but improving your theoretical understanding is what gets you from there to seriously useful senior developer a few years after that when you're no longer just writing simple GUI logic in C# or trivial ORM code for a Ruby on Rails web site back end.

Indeed, but someone with good theoretical understanding will pick up any given tool based on that theory fairly quickly.

Now, at no point in this post did I imply that getting a degree is either necessary or sufficient to achieve a good understanding of the theory. As far as I'm concerned, you absolutely can get there with time, effort and an open mind.

However, I think even autodidacts will find the process significantly easier if they've developed rigorous mathematical thinking and the ability to read and digest technical writing first one way or another. Also, for better or worse, the reality is that having that degree certificate will probably get you better jobs early in your career, which in turn will give you better experience and better colleagues to learn from at work.

In any case, just reading lots of casually written tutorial blog posts by people who've been playing with a tool for six months longer than you have certainly won't get you to that level of understanding alone. It's very easy to spend a lot of time doing that in a field like software development, feel like you've learned a lot and can be super-productive, and never even know how much you're missing if you've never found the right course of study or mentor or on-line learning resource to open your eyes. That, IMHO, is the biggest risk for people who haven't studied formal CS/SE one way or another, and sadly you can always find plenty of examples in the on-line forums for whatever the latest shiny technology is (currently I'd say it's front-end web development).

I'm not the one bitching and moaning about the bugs. I'm just pointing out the reality that very few people are going to go through the onerous bug reporting process that (some) Firefox developers/fans want them to, and that if they run into too many bugs in Firefox then they might choose to use another browser instead of choosing to help make Firefox better.

Unfortunately, I think your post is too close to the truth the be funny...

When all the developers move to the same platform with its non-standard implementations of everything, you get IE6.

It's not the first time they've done this "we know best" thing, unfortunately. There are cases involving HTTPS/HSTS where Firefox literally will not let you view a page it has decided is insecure, even if you explicitly want to ignore whatever the security problem is (for example, because it's a site you work on, and it's in active development and currently not fully configured).

Security warnings when encountering a likely threat = good. Overriding the user's explicit wishes = bad, always.

But there is another option for the users: they can use other software.

I sympathise with the frustrations of software developers, but the idea that any normal user (most of whom aren't going to be programmers or sysadmins themselves) is going to set up a virtual machine, reduce a bug they see down to a minimal test case, and then file a detailed bug report is crazy. It just isn't going to happen.

If a project has to keep relying on this, instead of being able to do good quality control and testing itself, the inevitable result is perpetually beta quality software, and getting left behind by other projects that are capable of doing proper quality control and testing.

This is all way off-topic by now, but my point is still the same: MojoKid's position is probably correct. There are significant costs for servers and for bandwidth for any site that scales up, and they can easily become more than it's reasonable to expect a hobbyist to pay out of their own pocket if the site becomes popular.

Of course, this is all before there is any actual content on the site! Doing the planning and research and writing and editing and presentation of original material takes about as much time and money on a web site as in any other medium.