Hacker News Reader

Comment by azangru | Hacker News Reader

> if it fails, it is only considered evidence that you were not doing it enough.

Or not doing it properly. And I understand the suspicion, I really do; but in hindsight, if you honestly tried to review how an organisation was operating, would you sincerely be able to say that it was adhering to a certain agile methodology/framework/mindset/strategy/whatever?

I have so far not see an organisation that would be following scrum, as it is described in the scrum guide; or kanban, as it is described in the kanban guide. I have seen or heard about various organisations that use these words, but they have little resemblance to what was actually proposed. So I can't really say if agile (or any of its particular variants) work or not. I have not seen honest experiments properly run.

View on HN

> I have so far not see an organisation that would be following scrum, as it is described in the scrum guide; or kanban, as it is described in the kanban guide. I have seen or heard about various organisations that use these words, but they have little resemblance to what was actually proposed.

If that's true, wouldn't it point at the process being impossible to implement?

It is a myth. There exists a version of Agile that could be implemented, and it would be the true Agile. The pure, honest experiment that would just work, because Agile cannot fail, you can only fail to Agile.

It signals to me that the process doesn't work in reality. You are better off doing something else.

It doesn't really fail any worse than other inflexible top-down process mandates from management.

That's where it becomes "impossible to implement"—you can't impose it as a cookie-cutter solution driven and controlled by management, and get much good out of it, yet that's the usual way it manifests in the wild. But that's not so different from anything else management might push in its place.

> There exists a version of Agile that could be implemented, and it would be the true Agile. The pure, honest experiment that would just work, because Agile cannot fail, you can only fail to Agile.

"Agile" is a very vague and shapeless idea which is hard to design an experiment for; but I would settle for clean experiments with well-defined methodologies/frameworks/strategies/whatever. Specifically, for scrum or kanban. Whenever people talk about these two, they seem to misunderstand them more often than not.

>It signals to me that the process doesn't work in reality. You are better off doing something else.

Whatever you do instead, you will also cargo-cult to some degree and fail equally as badly at.

For all the "You're doing it wrong!" I've seen in industry with respect to agile, I've also felt that every team I've been part of that did some version of it, seemed to function OK. I always found the "Agile Manifesto" a completely silly nothing-burger, but always understood the core tenet of 'agile' to be "employ tighter feedback loops", which... is sort of mostly how it plays out in practice??

I've belonged to numerous teams that followed some form of agile, to varying degrees of success (or failure).

The shape of what Agile meant in each of those teams was very different from one another. It would be disingenuous to say "the ones that succeeded were truer to Agile".

If Agile can be summarized as "employ tighter feedback loops", the whole Agile thing was beyond useless. A single sentence, as useful a tenet as it may be, does not a philosophy make. And this idea was not even new by the time the Agile manifesto came out (as explained in the linked blog post).

> If Agile can be summarized as "employ tighter feedback loops", the whole Agile thing was beyond useless.

Not just that, Royce's original paper that coined the term "waterfall" in 1970[1] can be summarized as "employ tighter feedback loops" compared to top-down design (figures 2-4 in the paper).

[1] https://www.praxisframework.org/files/royce1970.pdf

Useless as it might seem, I really do think it is actually that basic. Is it useless? Hardly. Take out feedback loops and see what happens.

The tenet was not useless. But it was not bought forth by Agile. It already existed.

And if this tenet is all Agile is, then it contained zero new ideas or contributions.

It seems to have successfully popularized it though.

Reply on news.ycombinator.com