I contracted at a company where we looked into doing what you have suggested in the context of a BI team. Basically, using a wiki to submit code changes is a candidate for the Too Hard basket.
Instead, what we did was write a bunch of services that interrogated our various source code repositories (database schema, stored procs, code, cube specifications etc) and generate a page for every object we were interested in with links to related objects.
Each page will contained autogenerated documentation (formatted to look nice etc) as well as user comments. The autogenerated data was refreshed every night, but the user comments were able to persist.
Make it all searchable and you have a very usable wiki that allows users to get a handle on how things hang together.
Forget about using a wiki to submit changes to the SCC system. Nightmare !
Having read TFA, it's quite clear to me that the industry is dominated by crusty old men (COM). Whilst they are reasonably happy to tolerate Baby Boomers (let's face it, BBs have created lots of wealth for COMs), they are not particularly happy about the surly behaviour of Generations X & Y.
And why is that ? Many of us have an inherent distrust of their abuse of intellectual property.
There is exactly ONE solution to the "problem" of copyright circumvention. Namely, make *everything digital* so utterly accessible (i.e cheap & easy to download) so that it's just not worth the effort to pirate it.
Most digital "content" falls into the category of luxury (i.e it's nices to have rather then essential). Standard economic theory stresses that luxuries have a very elastic demand curve. i.e. you lower the price and sales volumes increase massively. Result: low price (i.e. $1/CD or $2.50/DVD) = huge sales and bugger all piracy.
In all my time as a professional programmer (10 yrs), I have [i]never[/i] had my own office or my own cubicle. The closest I ever got was my team of 3 sharing the same office. Which meant we could have our own stereo which got played LOUDLY. We also had clients barging in whenever they wanted.
Did that stop me grinding out efficient, elegant code ? Nope. It's all about what works for a person or team i.e. culture.
I agree with Graham when he says that it's difficult to tell a great hacker until you have worked with them - likewise it's difficult to prescribe the best working conditions. I remember reading an article about Quake I wherein id had to bring everyone into a war room so they could deliver their game.
I recently wrote an application for a mate of my brother. All I asked for was a tray of beer, but the guy was pretty happy with the app (it's a fitness and game statistics tracker for his Rugby League).
After I'd delivered the app may bro rings up and says he's got the beers plus a whole sheep carcass in his SUV and he's heading over to drop it off !
Needless to say, I had to put on a spit roast party that weekend which is A LOT more work than downing a few well earned beers;)
I case you haven't guessed, I live in New Zealand where there are about 20 sheep per human.
Slashdot Mirror
I contracted at a company where we looked into doing what you have suggested in the context of a BI team. Basically, using a wiki to submit code changes is a candidate for the Too Hard basket.
Instead, what we did was write a bunch of services that interrogated our various source code repositories (database schema, stored procs, code, cube specifications etc) and generate a page for every object we were interested in with links to related objects.
Each page will contained autogenerated documentation (formatted to look nice etc) as well as user comments. The autogenerated data was refreshed every night, but the user comments were able to persist.
Make it all searchable and you have a very usable wiki that allows users to get a handle on how things hang together.
Forget about using a wiki to submit changes to the SCC system. Nightmare !
Having read TFA, it's quite clear to me that the industry is dominated by crusty old men (COM). Whilst they are reasonably happy to tolerate Baby Boomers (let's face it, BBs have created lots of wealth for COMs), they are not particularly happy about the surly behaviour of Generations X & Y. And why is that ? Many of us have an inherent distrust of their abuse of intellectual property. There is exactly ONE solution to the "problem" of copyright circumvention. Namely, make *everything digital* so utterly accessible (i.e cheap & easy to download) so that it's just not worth the effort to pirate it. Most digital "content" falls into the category of luxury (i.e it's nices to have rather then essential). Standard economic theory stresses that luxuries have a very elastic demand curve. i.e. you lower the price and sales volumes increase massively. Result: low price (i.e. $1/CD or $2.50/DVD) = huge sales and bugger all piracy.
In all my time as a professional programmer (10 yrs), I have [i]never[/i] had my own office or my own cubicle. The closest I ever got was my team of 3 sharing the same office. Which meant we could have our own stereo which got played LOUDLY. We also had clients barging in whenever they wanted.
Did that stop me grinding out efficient, elegant code ? Nope. It's all about what works for a person or team i.e. culture.
I agree with Graham when he says that it's difficult to tell a great hacker until you have worked with them - likewise it's difficult to prescribe the best working conditions. I remember reading an article about Quake I wherein id had to bring everyone into a war room so they could deliver their game.
I recently wrote an application for a mate of my brother. All I asked for was a tray of beer, but the guy was pretty happy with the app (it's a fitness and game statistics tracker for his Rugby League). After I'd delivered the app may bro rings up and says he's got the beers plus a whole sheep carcass in his SUV and he's heading over to drop it off ! Needless to say, I had to put on a spit roast party that weekend which is A LOT more work than downing a few well earned beers ;)
I case you haven't guessed, I live in New Zealand where there are about 20 sheep per human.