Managers want progress reports and statistical measures to reduce the risk that that they might get fired when the project they're managing fails
That's why you don't use lines of code or some other idiotic metric to measure progress. The developer should produce a list of fairly fine grained achievements (broken down by week) that will lead to the finished product. This way the programmer determines the approach they're going to take and also the means by which progress will be measured.
Also, inevitably with larger projects there are dependencies which have to be tracked throughout the execution. Ingrid needs the foobar API, but it's not scheduled to be complete until it's too late for her, etc...
As a former engineer and two year IT manager veteran, I hope I can offer some insight into this madness. Most of it is driven by competing agendas in upper management. These people are paid to drive the company forward and whether they're motivated by personal glory or genuine interest in the fate of the company, it's a hard task. They don't always agree with each on what should be done and this results in a sense of cluelessness being conveyed to the lower levels.
I'm not excusing this environment, but I believe it is how most successful companies operate. Managers are promoted because they can drive projects and deliver. When they're all competing for a finite resource, engineering time, it often results in either overcommittments or unfinished projects.
The key to making this work is having IT management smarts and a backbone. Any IT manager with half a brain can recognize when a timeline is unreasonable. If the timeline is set in stone then make the business clearly understand what they are giving up by forcing a "hack" in terms of *money*. Don't just say it sucks to implement hacks. Explain what they are losing. If they're not losing anything than the timeline is valid and the hack solution, if it exists, is really the right solution.
I agree. I can't believe Blizzard lets a phenomenal franchise like Starcraft languish. It is hands down the best RTS ever and it's just gathering dust. They need to take a lesson from Westwood (who continues to milk the C&C teet with mixed results) and crank out a sequal.
CDK is a curses "widget set" which is marginally useful. It's easy to throw a pretty application together with it, but for large scale apps it becomes a bit of a struggle. Any dynamic sizing or placement information is left up to the application and the API is such that changing widget attributes post-initialization is a challenge.
Regardless, it's logically structured and the code is simple to dig in and hack on. I think the official site is:
http://www.vexus.ca/CDK.html
It's not event driven, however, and as my app grew it made me long for a terminal version of GTK...
Slashdot Mirror
Managers want progress reports and statistical measures to reduce the risk that that they might get fired when the project they're managing fails
That's why you don't use lines of code or some other idiotic metric to measure progress. The developer should produce a list of fairly fine grained achievements (broken down by week) that will lead to the finished product. This way the programmer determines the approach they're going to take and also the means by which progress will be measured.
Also, inevitably with larger projects there are dependencies which have to be tracked throughout the execution. Ingrid needs the foobar API, but it's not scheduled to be complete until it's too late for her, etc...
As a former engineer and two year IT manager veteran, I hope I can offer some insight into this madness. Most of it is driven by competing agendas in upper management. These people are paid to drive the company forward and whether they're motivated by personal glory or genuine interest in the fate of the company, it's a hard task. They don't always agree with each on what should be done and this results in a sense of cluelessness being conveyed to the lower levels.
I'm not excusing this environment, but I believe it is how most successful companies operate. Managers are promoted because they can drive projects and deliver. When they're all competing for a finite resource, engineering time, it often results in either overcommittments or unfinished projects.
The key to making this work is having IT management smarts and a backbone. Any IT manager with half a brain can recognize when a timeline is unreasonable. If the timeline is set in stone then make the business clearly understand what they are giving up by forcing a "hack" in terms of *money*. Don't just say it sucks to implement hacks. Explain what they are losing. If they're not losing anything than the timeline is valid and the hack solution, if it exists, is really the right solution.
I agree. I can't believe Blizzard lets a phenomenal franchise like Starcraft languish. It is hands down the best RTS ever and it's just gathering dust. They need to take a lesson from Westwood (who continues to milk the C&C teet with mixed results) and crank out a sequal.
CDK is a curses "widget set" which is marginally useful. It's easy to throw a pretty application together with it, but for large scale apps it becomes a bit of a struggle. Any dynamic sizing or placement information is left up to the application and the API is such that changing widget attributes post-initialization is a challenge.
Regardless, it's logically structured and the code is simple to dig in and hack on. I think the official site is:
http://www.vexus.ca/CDK.html
It's not event driven, however, and as my app grew it made me long for a terminal version of GTK...