3 ms·
I agree that almost all of the orgs I encountered that were Doing Agile were pretty horrible. On the other hand, I've also been in orgs that were very successf
by mpweiher 7y ago
I agree that almost all of the orgs I encountered that were Doing Agile were pretty horrible.
On the other hand, I've also been in orgs that were very successful in developing software in a way that we would recognise as being agile. We just didn't make a big deal out of it. And we didn't do standups (most of the time), or 2 week sprints, or retrospectives. We didn't even pair consistently: we split up on trivial stuff, paired on hard stuff, as needed and available.
In one org we started without TDD, and even without having much in the way of automated tests at all. But when we discovered testing and TDD, boy did we ever embrace it. Not because anyone told us to (mid-to-late 90s, so no Agile Manifesto), but because the difference for us developers was night and day.
So develop software. Concentrate on what helps you develop software. In my experience, you will probably discover something recognisably agile all by yourself. So when you discover that, help yourself to what helps you. Don't make a big deal out of it.
- hef19898 7y agoSounds a lot like TPS, I assume Toyota did a lot of the stuff before any books were written. Also, my personal experience, the more companies talk about Lean and Six Sigma the less they do it.
- mpweiher 7y agoSimilar with "culture". Companies with a great culture don't talk much about it, they just get shit done without much fuss and typically have some fun doing it. Because getting shit done is fun. On the other hand, if the company all about how fantastic the culture is...run! Or at least tread very, very carefully.
- maxxxxx 7y agoTotally agree. Talking about culture or hanging up "inspirational" posters is usually a warning sign. Good organizations live their values and don't feel a need to constantly talk about them.
- HeWhoLurksLate 7y agoThat reminds me of my two favorite interviewee questions: "What's your company culture like?" - the way that the interviewer responds is a pretty good indicator of what their culture is like. "If someone does something exceptional, how do you reward them?" works in a very similar way.
- rtikulit 7y agoMaybe even step back to first principles... “Can your company even recognize when someone does something exceptional?”
- maxxxxx 7y agoWhat you are describing is exactly what Agile should be. You have a team of people who care about the craft and they slowly improve processes based on real world feedback and experience. If something doesn't work you drop it and try something else. This happens in environments where people respect each other.
- rtikulit 7y agoManagement does not want to hear that software development is a craft. They want it to be a trade, with fungible workers operating in a predictable process with fungible components. The great tragedy of Agile is that it was co-opted as the vehicle of this transformation. It was seized by management, weaponized to other purposes, and then aimed right back at us.
- bloorp 7y agoYeah, I've come to similar conclusions about Agile. It can be a great way for an already well-disciplined organization to think about the work they're doing. But many, many undisciplined organizations thought that Agile would be a catalyst for them to become disciplined. But the nitty gritty of "doing Agile" required EVEN MORE discipline than what all these organizations were already capable of exercising, so it just highlights all the frustrations everyone already has. That does seem to be the downside of Agile. It's a collection of maybe a dozen different techniques and practices. But if one of those practices falters, whether it's the TDD, or the business side still wants a certain deadline, or you don't bother to demo at the end of a sprint, then the whole house of cards falls like dominoes..... checkmate!
- asdfman123 7y agoBuild trust and respect instead of elaborate processes to compensate for the lack of them.
- JoeAltmaier 7y agoGood advice for an experienced team. Couple it with good reporting (who's job is that?) and management could be good with it. Green teams may need more structure. But the structure should not take over; it should be a tutorial and get rethought as the team matures.
- maxxxxx 7y agoYou also need to give people the chance to make mistakes and learn from them. I don’t envy a lot of the young developers who get micromanaged all the time and never have a chance to try something.
- bewo001 7y agoAgile has fallen into the hands of big consultants. Now the focus is on process, with more meetings and presentations than ever before.
- ataturk 7y agoAm tired, just dead bones tired of all of it. It's all just this or that money making scheme with guys like me caught in the middle. I despised Waterfall, fell into scrum and in the past ten or so years kinda sorta saw one project where agile basically worked right. The rest has been mostly waste. I've worked on billion dollar systems on down to just little nothings. Doesn't matter, companies are struggling to make anything work. I think there's just too much complexity and they all believe IT solutions can manage that complexity. It's failing like never before while expectations have never been higher. Message to the Fortune 500: Just get rid of anyone with the title "scrum master" they're dead weight. You were fooled, deal with it. I am trying to move my team over to Kanban right now, but we have all this reporting crap up the chain all designed around CA Rally, the worst productivity tool ever made.
- marcosdumay 7y ago> Doing Agile You mean the ones following one of those canned "agile processes"? It always amused me how people could read the Agile Manifesto and then conclude that it's best satisfied by following a canned process.
- Zenbit_UX 7y agoGot any insight as to why TDD is such a great thing? From how Ive heard it described it seems to be a way for management to impose rules on how developers can approach writing a feature and not necessarily a benefit to the devs. If I understand it properly, it's essentially write a test so it fails then fill in the feature gap. The whole write first bit is just so you don't have to retest later to verify that your test didn't hit a false positive, right? Something a careful person would check anyways...
- mpweiher 7y ago1. It's not a management-imposed thing TDD was discovered as technique by developers, for their own benefit. Management imposing it can be a bit of an anti-pattern, though it may be well-intentioned. I have certainly pushed teams to TDD, though primarily by example and by the positive experience you quickly get. Anyway, TDD has nothing to do with management imposition. It is really diametrically opposed, at most orthogonal. 2. "test didn't hit a false positive" Yes, that's a good thing, though hardly the main point. You write that a careful person does this anyway. I'd rather have a machine do that for me, and have less of a requirement on me to be careful. Computers are much better at being careful, and I want them to do the tedious stuff. I also am wary of claims of "you just need to be careful". Probably lots of bugs ahead! 3. So why is TDD so great? First, it's a design technique. Writing the test first forces you to think about how your new production code will be called. It also forces you to be clear about what the code is supposed to do, precisely, before you write the code. Writing the test tends to be significantly less difficult than writing (correct) code to solve the problem. You then write only the minimal code that is required to pass the test, which surprisingly also tends to be fairly trivial. Also, you know when you're done: when the tests are green again. So you've transformed one difficult step into two fairly trivial steps. Which sounds weirdly magical, but actually works in practice. Also, since you only wrote the minimal code that was required for the tests to pass, you know that you have good test coverage. After you've done a few, you're in a place to detect duplication and eliminate it by refactoring. This tends to be somewhat more intellectually challenging. However, you are truly refactoring, that is not changing the functionality, so you have a great safety net with the good test coverage you achieved before. So you get this small steps that are fun and easy, and you tend to not regress. Psychologically, it's also very pleasant, compared to the usual flow, which goes somewhat as follows: 1. I have a rough understanding of the problem, 2. I code a solution, which is at the limits of my understanding (because everything is up in the air) This part feels very good, "I am the master of the universe" 3. Then I should write some tests, but the incentives are a bit negative -> if the tests all pass, nothing happens, so why did I write the tests? also if nothing happens, how do I know I actually tested something? (not by "being careful") -> if the tests do show a failure, it's a distinct downer from step 2 (and so I probably subconsciously write my test so they don't fail) 4. I don't really want to refactor, because my solution from step 2 was already awesome! TDD: 1. I write a test. It fails appropriately. That makes me happy. 2. I write minimal code. Tests go green. That also makes me happy. 3. I refactor. The code now is much nicer and tests still green. More happiness. So with most normal development flows you get that initial rush and then a lot of downers. With TDD, you don't get quite the same initial excitement, but instead you get these nice regular dopamine hits without really much downside.