"See? None of these involves requirement changing. How you can do waterfall in large projects worth millions is beyond my imagination..."

I see your point but I can't see its relationship with anything agile vs waterfall here. Nothing you talk is about your application's functional requirements so it doesn't even enter the agile realm.

"There is nothing about waterfall that requires you to make ironclad decisions at the very start"

Of course there is. Once the requirements phase is closed, it is closed. You can have iterative waterfall, but you need to wait for next iteration to alter requirements. You can have overlapping waterfall, but overlapping shouldn't make more than 10-20% of your gantt, and even then, at the cost of increasing at least on the same proportion your allowed cost/deadlines variances. In fact, the only guarantee for deliverability on waterfall comes from the fact that when a phase is closed, it's closed and you can't go back to it, which would add uncertainty to the process.

"The difference that I've seen in practice is that it's incredibly hard to implement Agile correctly"

It is, but that's the case because of the way companies are organized since Ford days: hierarchies, functional and responsability clear niches and boundaries, processes and bureaucracy, while agilism is about team playing, empowering people, planitude and results over processes and paper trails.

Agilism looks hard for too many companies the same way chess looks hard when you try to play it on a tennis field and with tennis rules.

"You are basically just saying that agile is more *fr*agile than waterfall"

It is, but then you need to add context to see what that really means.

Agile was always about empowering people over process and bureaucracy. If you are process and paper trail centered you are not doing agile. Full stop. Please, take your time to rumiate this.

Now let's imagine you really are embracing agile and let's think about your team's members (not only developers but management and even customers): is it really advantageous to empower them? Or is it like giving a loaded gun to a monkey? For one obvious thing, taken right from the agile manifesto "Customer collaboration over contract negotiation", will your customer abide to collaborate or will he want a full contract with all things strongly tied -costs, deadlines, features... by day one? Because if it's the latter, you can forget about agile right on the starting line, you see...

Now, on waterfall: of course it is more solid and efficient than agile... provided the customer (be it internal or external) perfectly understands his needs, is perfectly able to transmit them, their needs are perfectly understood and expressed formally on paper, they are agreed back by customer and you have an experienced enough team all across without a hidden agenda.

Failing anything in the list above, waterfall not only tends to quickly fall apart, but doing so in very expensive and notorious ways that agile mentality (in any of its incarnations) provide checks and balances against (but first you need to take the time to understand what is all this stuff about).

Now that I reread what I wrote, find it similar to the republic ideal: it works because its checks and balances, but obviously they only add more moving parts (thus fragility) and unefficiencies... if only everything worked as hoped.

"I've been learning to appreciate older stuff as I've gotten older"

Good. By the day you get retired you might reach master Monteverdi.

"still most music is base on repeating over and over again."

I know what you mean: Wagner -> Mahler -> Bruckner

"Of course, we're not going to do anything about the problem. Of course not."

[ironic]
Well, I used to be a denialist that thought there were nothing needed to do, because there was no problem.

But now, I'm convinced: there is a climate change and it is pushed forward by humankind.

Unfortunately, it's too late to avoid it, so I'll do nothing either.
[/ironic]

"It is called support service"

No, it isn't. Adding functionality to a new environment is not support.

"While I understand that perpetual support for older devices is not viable"

And then, wrong again. Given support for what it is, the ability to substitute broken parts and correct what was not working from the very begining should be supported basically forever, much more so for software, since software doesn't have wearing parts.

Heck, I have no problem finding parts for my 15 y.o. car but still my printer can't be the same?

"With both data collection and a restore function, Windows will just set up again from an installation image"

Yes, one that will put the user a few gigas back the times and open to vulnerabilities till upgraded -not to talk about the inability to use the computer for some few hours.

Nice.

"The blame goes where it belongs, and the consumer will buy a new printer."

Surely will. A perfectly working system stops working because Microsoft singlehandledly changes the system but still the blame is for a third party and the solution is me expending more of my hard earned money?

Ubernice.

"the ones in charge will never pay the older developers what they are worth."

That's only part of the equation. They won't pay the older developers below what they are worth and on par to the youngsters either. Olders are still harder to (mis)manage.

"Maybe 1 year experience in 20 different things... which is not a bad thing. Show you can LEARN over and over and over again "

But then, you can't offer 5 years of experience in any of those buzzwords, so you can only opt to entry positions... just to be discarded as soon as they realize your age.

It seems that nowadays basically nobody values the jack-of-all-trades approach.

"sub-meter accuracy is only really needed for smart bombs"

What kind of "smart bombs" are you talking about that, if exploded a foot away from its target, miss the mark?

"gb gun crime has doubled since they removed generalized ownership of guns."

Even if that were true:

US (ownership/100 | weapon deaths/100.000): 88,8 | 10,30 (2011)

UK (ownership/100 | weapon deaths/100.000): 6,6 | 0.25 (2010)

http://en.wikipedia.org/wiki/N...
http://en.wikipedia.org/wiki/L...

Less weapons, less deaths.

"Perhaps if it were not illegal to own the guns used to kill the Charlie Hebdo staff, the staff would be alive today as the shooters were better armed than their protectors."

Given statistics on both sides of the pond, it looks like generalized ownership of guns brings more deaths, not less.

"the more batteries installed, the fewer people paying for electrical mains infrastructure"

non sequitur. You still need to load the bateries. Doing it without support from your mains is a big if.

"but poker is about much more than just the odds.. it's about emotions, reading players expressions and tells, intimidation, and taking advantage of all that."

That's your saying.

"i don't see the point to training a computer solely based on odds and what happens 'on the table'"

Therefore the research.