Friday, April 18, 2008

Innotech - Day 2

Only got to the morning sessions on Thursday - family emergency kept me away from some afternoon sessions that looked like they'd be good...

Agile Project Experiences
So many discussions and articles about the superiority of the agile methods lack the depth and details I am looking for - and this panel discussion did nothing to satisfy my thirst. Am I the only one feeling frustrated? Like other discussions I've had and presentations I've seen, these panelists talked about how they could do things using agile methods they couldn't do otherwise, but they didn't specify what those things are; they said their velocity increased, but they admitted that they couldn't measure this; and they said the quality of their solutions increased (in terms of bug rates), but they didn't have hard numbers. Regarding that last issue, Arlo stated that they pretty much just don't find bugs any more, but I'm more than a little skeptical about that.

My second issue is that agile proponents always compare these methodologies with the strict waterfall methodology. But I haven't seen a strict waterfall project in over 20 years. Since those early projects (at GE), every project has had constant communication with customers and other stakeholders, iterative prototyping or development, functionality prioritized and built in order of importance or risk, and/or intra-project negotiations with respect to change requests, scope changes, etc. So it seems to me the real question is one of asking about the benefits of a strict agile methodology (which is what I always hear profferred) versus the more amorphous, flexible development progression that seems common in real life.

And now my last issue: so far I haven't heard a satisfactory answer for how, using agile methods, to provide management with the information they want. At this session the panelists said management will be happy when they see the increased velocity and the list of functions completed. Is my experience with management unusual? I've never been involved with management that would accept that. In my experience management wants to be able to plan future work and the resources needed for it (i.e., project portfolio management). That means they want estimates for the length, value (e.g., ROI), and resource requirements of every project. And they want projects managed to those estimates.

Maybe my issues exist because of the space I work in - corporate, internal development. I can see how agile methodologies could work well in commercial or open source software development. For example, Microsoft provides general guidance as to the functionality they will include in their next operating system versions, but as development progresses, they drop capabilities to help reach their desired ship date. And these methods seem even more natural for open source software development. For example, I can see how Linux development is best done using agile methods.

There's so much more I could write. But I've rambled enough.

Wiki Then and Now
Ward Cunningham gave a two-part presentation. The first was a historical perspective of his invention of the wiki and his involvement in the development of agile methodologies. Fascinating and impressive.

The second half was a discussion about, the company founded by Ray King and for whom he currently works. This company troubles me. They scan domain registrations and create a wiki page for each domain. And they allow anyone to edit those pages. In his presentation Cunningham said "you've been drug out [into the public], and we're here to help you." But they're the ones who have done the "drugging"! I may rile some feathers, but I just don't see value in their site for companies:
  • To be listed, you have to have a domain. But if you have a domain, you most likely already have a web site, so you don't need AboutUs to advertise your existence.
  • AboutUs might advertise the value of allowing your customers to comment on your companies services. But if you want, you can provide this capability on your own web site. And you'd have control over the publication of these comments too.
  • People who search for your company may choose AboutUs' links rather than yours - do you see any value in that?
And companies' customers may not see much value either. AboutUs might argue that you can see comments - both positive and negative. But smart companies will lock their AboutUs pages (Cunningham said you can do that, but their web site contradicts this statement) - or have them deleted (Cunningham said they would do that for companies who request it - though the site contradicts this statement).

Given all this, I highly recommend that, at minimum, companies actively monitor their page on AboutUs.

No comments: