Agile Labs Conference Report, March 31, 2009

Last week I attended the Agile Labs conference organized by Luxoft Training Center with the support of the State University Higher School of Economics.

Initially, I was skeptical, because reports about agile-methods already became familiar, in the fall there was a lot said about this at the SECR 2008 conference, there was a specialized mini-conference AgileDays. In general, I was waiting for the repetition in the reports of the same ideas already worn out. Therefore, I planned to go and listen to a couple of reports in the morning, and then come in the evening to the promised “battle of the giants”: opponents of agile against supporters, Pankratov against Urazbayev .

But I was lurking in surprise - the conference turned out to be completely different from what I had expected. First of all, there were not so many reports about agile. Secondly, both the audience and even the speakers were very critical towards agile. And thirdly, the reports were quite fresh and interesting. That is why I stayed there for the whole day. Mostly I listened to the reports in the first section, and I’ll tell about my impressions about them.
')
The conference was opened by Askhat Urazbayev with the report “Process diseases and methods of their treatment”. I thought it was an allegorical name. Nothing of the kind - the report was purely medical! I always liked the organismic approach, so I enjoyed listening to Askhat with great pleasure.

Having carried out an analogy with medicine, he proposed a classification of process dysfunctions (diseases) into internal (infections), external (toxins) and fundamental (neuralgia), and suggested treatment methods arising from this analogy. Very interesting approach, giving good food for further reflection.

And where is agile, you ask? Yes, in general, there was nothing about agile. Well, except as an example. The proposed treatment methods are quite universal. And this is the most wonderful!

Then Slava Pankratov surprised by his report “Why I do not believe in Agile”. He attacked agile methods much weaker than I expected. Glory basically rests on two points:

1. what to do if the project is large or distributed (including the possible presence of subcontractors),
2. what to do if the customer does not work on agile (either not involved, or is involved, but starts to play against, hypertrophying the features of agile — for example, it changes requirements too often).

I feel that the attack was not too strong, because agile supporters also understand that this is not a silver bullet, that it is not always possible to apply such methods of development. And therefore fought off without much difficulty. On the one hand, there are examples of successful agile-projects, where there was a large or distributed team, methods of engagement with the customer were developed, allowing to remove many problems. On the other hand, if it seems to someone that agile "does not go" - well, do not use it, nobody forces it.

Then Alexander Ermakovich was supposed to make a story about "Testing in agile-projects." But from the very beginning, he directly stated that he is not a fan of agile, takes a neutral position, and will actually talk about testing in general. Even on the contrary, some of his theses, in my opinion, are completely incompatible with agile. Like you, for example, the following statement: test planning should be independent of development planning, and of course, testing itself, of course, should also be independent. I'm afraid that this is beyond agile. The pictures in his presentation also seemed to be drawn from RUP diagrams - artifacts, workflows, roles.

Being a supporter of the context-driven approach to testing, I generally take great care with statements that some kind of practice is good, or even better, if the context is not specified. But I also perfectly understand that people, making such statements, always implicitly imply some kind of context. So I'm just trying to restore it. However, in this case, the restored context was more like a formalized process than the work of a flexible cross-functional team.

In general, I would be suspicious of the proposal to introduce the proposed testing practices in an agile project, whereas in a more formal process they would look very nice. But it was clear that Alexander has considerable experience as a project manager, and he doesn’t care what methodology his superiors want to use; he is ready to achieve results in any conditions.

After lunch, Maxim Dorofeev gave an excellent talk about “Agile in the world of snobs”. The priests of the great standard DO-178B, the main regulatory document in the field of embedded software development for avionics, acted as snobs. That was really about agile.

Maxim showed that intelligence and wit allow you to cope even with such snobs, and playing by their rules and in their field. Snobs require you to have a project plan, and you want to maintain flexibility? No problem, we write a plan in which “Iteration 1”, “Iteration 2”, “Iteration 3” and so on are indicated as works, and to the question what is the result of each of these iterations, we answer that this will be “a working product but with not fully implemented functionality. ” Need a configuration management plan? We draw on a piece of paper or a whiteboard a repository scheme, rules for designating versions and releases, a diagram of the assembly process, scanning or photographing, put the necessary data on the resulting picture (who created the document, who signed it and when) and the hat. And so on, is the principle clear? Everything brilliant, as always, is very simple.

The next report, in which Alexander Kondakov talked about how to cross agile and CMMI, I almost completely missed. I heard about it with different variations at the last SECR conference.

Sergey Arkhipenkov’s report on “Anti-patterns of team behavior” again did not have the slightest relation to agile, but that didn’t make it any less interesting. A wonderful psychological study that perfectly complements the no less remarkable books by the same author on project management. Both books and the presentation of the report can be found at www.arkhipenkov.ru . I will not even write much - download and read, it is worth it.

Then we all went to play Lego. Under the strict guidance of Alexei Krivitsky, three teams of six people each built a toy city - houses, cars, traffic lights, trees. But built not just like that, but in full accordance with Scrum. They played Planning Poker, moved yellow pieces of paper on a whiteboard, negotiated, in general, everything is as it really is. Unfortunately, as is often the case with simulation games, the allotted time ended when the game was in full swing. Half an hour was clearly not enough, for a full result it would be necessary to allocate an hour and a half to be able not only to finish the game, but also to discuss the experience. However, it was fun, wait for the video!

Well, the curtain of the conference hosted the widely publicized "Battle of the Giants." Slava Pankratov promised to roll agile into a pancake and roll it into a tube, and Askhat Urazbayev was honored to defend flexible methods. The giants were not satisfied with martial arts, each side was doubled - Alexey Krivitsky came to the aid of Askhat, and the attack was intensified by Alexander Alexandrov. Sergey Martynenko played the role of referee.

Alas, the battle did not work. The fighters circled around each other and went without any scars or damage. From the side of the attack, the same arguments that Glory expressed in the morning, which I have already listed above, sounded mainly. And even in spite of the fact that Sergei clearly played up to the attackers, Askhat easily repulsed all the attacks aimed at agile. More precisely, he didn’t bother them at all, but simply extinguished, recognizing on all counts that agile sometimes works and sometimes does not. That, in general, is clear to almost everyone.

Perhaps the strongest blow can be considered a question thrown by Aleksandrov, as if in passing, is there really something completely unique in agile, or is it just a successful marketing move. The recognition of Askhat that this was largely true was clearly a point in favor of the attack.

Now in the English-speaking forums and blogs, more serious problems are discussed that are in the agile movement and which are more related not to the essence of the agile-methods, but to the practice of their application. I would be interested in discussing precisely these practical problems, but this just did not happen.

Well, in the end Sergey (in my opinion completely in vain) added another spoonful of tar, deciding to remind everyone that the subject of the dispute itself is very vague, no one can articulate clearly what agile is, where it starts and where it ends. As a result, the whole discussion immediately became nothing and was blown away. The giants from afar threatened each other and dispersed.

Summing up, we can say that, in general, the conference was a success, the reports were not hackneyed, the audience was lively and interested. Technical overlays are not noticed. Perhaps the only reproach towards the organizers of the conference could be the absence of a normal hot lunch. Buns are good, but not enough, the energy consumption at the conference is enormous, so I really want to eat :)

Those interested can see photos from the conference .

We are waiting for the next conference in the series - Training Labs is planned for June.