Don't do stupid things on purpose - Waterfall is one of those.
Christian Neukirchen has come up with something original for his forum, Fugi. So what's so original about it? Well, a two-dimensional layout for starters...
Cameron's serenade to Confluence coincides with a recent discussion we had in a project I'm working on. That discussion started in the aftermath of a project retrospective and involved the way we kept all of our project documentation in a wiki. Here's the short version.
We're developing a tool used by other development projects within the client's organization. We've now released 4 versions of the tool since the project's inception on my part, roughly 4 months ago. As a result, the users we go talk to have different versions of our tool installed. Usually they just upgrade to the latest release but sometimes upgrading would be too much of a hassle for them to do right there and then (the reason being that sometimes there are non-backwards compatible changes which require them to also update their working documents to match the version of the tool they're using).
So, what does this have to do with the documentation being in a wiki?
The fundamental problem is that while wikis are great for maintaining live information on a website, they are less great in publishing different versions of that information. In our case, for example, we'd like to be able to "tag" a given version of the wiki with "0.7" and then give our users a URL that points to the front page of the documentation for that specific version. Unfortunately, that's not something most wikis I'm aware of support.
This leaves us with few options. We could just clone all the pages and suffix them with a version number but that's such a horrible thought that I'll just leave it at that. We could also have the system admin clone the whole wiki directory on the server, which could be a decent solution for now if only we could perform the cloning ourselves (we're using an internal corporate wiki hosting service). We could of course just install the wiki software on our own server, which should be quite sufficient except for the backups perhaps.
The versioning issue isn't all there is, however. Even if we could work around the versioning, we'd still be dependent on online documentation and every now and then we're in a place where Internet access is a wet dream (or at least a costly one--more than 10 euros for an hour of Internet access in hotels doesn't surprise me anymore). It would be so much easier if we'd have the correct version of the documentation also distributed along with the tool.
As a result, after a four-month stint at using wiki and wiki only for documentation, we're now looking into tools that would let us work with a simple wiki-style syntax we could store in our Subversion repository and get both online and offline documentation from that single source.
We're currently looking at adopting reST. I'd love to hear your experiences with it or with similar formats/tools.
Thanks to Pete and Andy for pointing out the absolute-must-attend conference of the year,
Waterfall 2006 at Niagara Falls in April.
Be sure to check out the program. It's excellent!
I really should be in bed by now but something drew me to browsing my archive of posts from the Scrum mailing list which I haven't read in a while. Apparently, the word "swag" in "giving it a swag" stands for scientific wild a** guess. Hilarious. Or then it's just too late.
Simon Brown, a fellow JavaRancher just announced The Pragmatic Architect. And my blogroll just grew by one.
Code is clay, not marble. Or at least it should be.
My latest article titled "Ruby on Rails in the Enterprise Toolbox" has been published in the January 2006 volume of the JavaRanch Journal.
I tried to pass on my perspective about Rails in "enterprise" use without falling too much into comparing it with J2EE or .NET. In other words, I didn't want to end up with a full-blown flame war on my porch.
The article really improved massively from the first drafts thanks to Dave, David, and Allan who graciously offered to help out by reviewing the article, correcting oversights, flaws, and on the text in general.
Do check out the other articles in the issue as well--lot's of good stuff in there!
Anytime you believe automation to not be worth the trouble, I encourage you to change your viewpoint and find a way to reduce your automation cost.
Steve Jobs did his keynote yesterday in San Francisco at the MacWorld Expo. Starting at around 1:01:24 to his keynote, he spilled the beans on Apple's first Intel-based computers, the iMac and the new MacBook Pro laptop.
They say the new MacBook Pro is up to 4 times faster than the Powerbook G4. WOW!
But I'm still not an entirely happy bunny. Why? Because they only released a 15-inch MacBook Pro at this time and I was hoping for something smaller. I'm already having trouble using a 14-inch IBM in airplanes so I certainly wouldn't want to get any extra centimeters of screen height even if the package would be terrific otherwise.
Well, here's for the rumors about Apple discontinuing their 12-inch pro notebook model... Steve, I'm counting on you.
Nobody will probably argue that Apple wouldn't be a master in creating hype for its products. Add to that, and being a rather contemporary topic, sites like MacRumors.com are certainly rubbing the Mac geek population the right way by making available live updates from MacWorld San Francisco over IRC:
Of course, the worst (luckiest?) Mac geeks are not hanging out on some wacky IRC channel--they're there.
Pete Behrens talks about how organizations react when they wake up and find themselves riding a dead horse. The scary part is, of course, how many of the things Pete lists I have seen in person over the years.
The nice bunch at Dreamhost.com just quadrupled the disk space for all of their accounts and octupled the bandwidth. That's what I call a nice surprise :)
Puh-leeze, get rid of that *$@%!&* final keyword in Java! It's making my life bloody difficult at the moment. Thank you.
Alan Francis sets a great example of the right mindset, in my opinion. I've seen and still seeing way too many projects who have been running for ages without integrating the whole product. Yes, it can be difficult but it's not going to get any easier.
Via Simon Baker:
The best way to gain trust is to give trust.