5 ms·
Agile/XP Critique: It's not the Code, Stupid
- DanielBMarkham 19y agoInterested in any comments about this article -- I thought about submitting it to a client who is really gung-ho on XP and Agile, but I'm not sure if I'm making any sense or if it would tick them off. Feedback appreciated.
- gleb 19y agoWhat's your point? The article makes some reasonable observations, but there is no call to action. In other words, what do you suggest people do (differently)?
- projectileboy 19y agoI more or less liked it, but I was sort of confused by the sell... I do a lot of gigs where I play an agile-shmagile "coach", and much of what you wrote sounds to me like you're in violent agreement with the rational portion of the "agile" community.
- DanielBMarkham 19y agoWhat I'm seeing is people dropping one onerous system -- say traditional waterfall -- and creating a new onerous system built around favorite authors and books. I think people get the "feeling" of agile but miss out on really how it is supposed to work. I mean, I'm seeing customers quoting books back and forth to each other -- even calling in famous authors to make decisions on what they should be doing. Famous authors, bless their hearts, don't know jack squat about any particular team or project and shouldn't be making blanket announcements about what teams "ought" to do. That's the same stuff they're trying to get away from. There are no perfect answers. I come from the rational side of agile as well. This is the first time I've seen so many people so excited about giving teams freedom that they're prepared to spell out exactly what those teams should do in order to be free. If you only concentrate on delivering code you miss the other value you add. Teams can't run on manifestos and slogans.
- projectileboy 19y agoYeah, the whole cult around some of those authors is laughable; reminds of the Gilderoy Lockehart character from Harry Potter - those who can't do, write - and "teach". I did like the XP books by Kent Beck, though, because I thought they captured what for me is the most important aspect: you start with a simple process that seems reasonable, based in the values that are important to your organization. And then you adapt the process as you move along, without worrying about what some dude in a book said to do.
- DanielBMarkham 19y agoI think these guys have their heart in the right place, no doubt. I think what I see that is missing is organizational agility. There is a lot of organizational risk that just goes unanswered, and once the questions begin, these guys start sounding awful prescriptive. It's almost like they expect large organizations to run completely from a focus of one team, and one developer, instead of the entire structure. That might sell a lot of books to frustrated programmers on teams, but it kind of side-steps the whole issue of "how do we pick what to use or not use in a dynamic fashion? How dynamic can an organization be, anyway? Could you deploy to production different ways for different teams? Should project chartering be done with information that is also valuable to the team, like a business model? If you have running code and not enough people know what you are doing, do you stamp your feet and keep saying "integrated team of generalists" or do you acknowledge the fact that written communication is the only way your are going to reach some very important people?, etc." I don't see a lot of that type of thinking going on. I too do the coach thing, usually for big organizations. I also do the startup thing (which I love) with small teams. Care to guess which one I like more? (grin)
- tx 19y agoAny kind of "methodology" is aimed at improving the predictability of results expected from a group of average coders. Thus it's no surprise that "large client" appears in the first paragraph of his blog post. In reality, software engineering is more like an art, and each project calls for a somewhat unique approach, unique style. There is a lot to be said about general sense of style and beauty when it comes to programming. Imagine Leonardo "going Agile" on Mona Lisa in order to speed up the process. Of course there are always armies of copycat CRUD projects that have nothing unique about them. It only proves my point of average quality of "agile programmers" because good engineers aren't interested in CRUD work. Ruby-on-Rails is an awesome example of this: one good programmer did all necessary thinking for you. It is perfectly possible to deliver a "typical Rails" site without even caring for learning Ruby. A good methodology done right allows code monkeys with surprisingly low skill level to be productive (which is, of course, not a bad thing). Rails is agile-carved-in-stone approach that forces you to stay on track and punishes from levitating from it, I guess this is where word "rails" comes from. This drives many MBAs nuts: none of those dog-training managerial books filled with pompous common sense seem to work when applied to software: good programmers simply don't respond to "positive reinforcement" and average programmers produce average results no matter what. All those "methodologies" allow managers to get average results reliably regardless of quality of talent at their disposal.
- DanielBMarkham 19y agoThanks tx. I guess there is a cost-reward factor at work. I know RoR and the various automated DAL guys are really happy with working inside a narrow, but fully fleshed-out paradigm.