News for hackers. Stuff that matters.
Re:My Foreword to “The Art of Agile Development” by JohnFen on Tuesday October 26, 2021 @06:48PM (#29007386)
People often describe waterfall in that way -- but I've been in the industry for decades, and if that's what waterfall is, then I have never seen anyone do waterfall even before Agile was a thing.

I think that "waterfall" is often a straw man.

Re:My Foreword to “The Art of Agile Development” by Jtsummers on Tuesday October 26, 2021 @06:56PM (#29007449)
> I think that "waterfall" is often a straw man.

Every discussion where Waterfall comes up has someone saying this, I'm glad you've not experienced it. But it has existed, does exist, and will continue to exist (because enough project managers are idiots). One of my first jobs (I quit) was a failed project that persisted because of gov't money, but would have failed as a commercial venture (started circa 2007, failed to deliver at least through 2011, I was on board for a year and found an exit). It was 100% Waterfall, they literally delivered the wrong thing several years late and with around $1 billion spent. Great for the contractors, awful for the taxpayers.

But yes, that is what Waterfall is. It's a large scale project executed linearly with limited-to-no-feedback. Large scale is somewhat subjective, but basically if it's done in less than a few months it's probably not at a scale where your choice of development model actually matters.

Re:My Foreword to “The Art of Agile Development” by brightball on Tuesday October 26, 2021 @07:22PM (#29007681)
The issue with waterfall is that planning it properly depends entirely on accurate estimates.

Estimates are never accurate. You can’t possibly have all of the information present up front.

Re:My Foreword to “The Art of Agile Development” by dilyevsky on Tuesday October 26, 2021 @09:50PM (#29008724)
In my experience with waterfall this was solved exactly same as with making sure scrum stories can be completed in a sprint - by padding the shit out of your estimates to the point most engineers can complete their tasks working a few hours a week.
Re:My Foreword to “The Art of Agile Development” by corpMaverick on Tuesday October 26, 2021 @08:51PM (#29008375)
I think it is more correctly defined as "Big Upfront Design". This was a thing; months spent on analysis and design that filled several tomes.
Re:My Foreword to “The Art of Agile Development” by Jtsummers on Tuesday October 26, 2021 @09:00PM (#29008430)
Big Design Up Front (the more common way of ordering the words) is related to, but is not, Waterfall. Waterfall is best considered an instance of BDUF (to borrow a programming term). V-Model and (often) Iterative & Incremental (slow-motion Scrum) each have varying degrees of BDUF as well, and there are plenty of other SDLCs out there that can be categorized as BDUF-lifecycles.

Waterfall's principle failing for large scale projects is that it does not incorporate sufficient (if any) feedback or delays the feedback until too late in the cycle, and plans too far out to be reasonable (do you kwon with any real confidence what you'll have for dinner in 3 years, 2 months and a day?) V-model generally also plans out very far, but incorporates V&V as an ongoing activity (it's name is a bit of a pun, both incorporating V&V and usually being charted out as a V, vice a waterfall for Waterfall). I&I may have a big design phase, but it's usually not as fully designed as either Waterfall or the V-model because the iterations are meant to provide checkpoints for validating the design and potentially changing direction (like I wrote, slow-motion Scrum).