![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
stephbgAm most happy to report that I got to attend this evening's ACS (Australian Computer Society) meeting featuring a presentation/workshop by Dwayne Read of Strategic Systems re Agile methods of estimating for software development projects. It's close to a year since my last one.
I've known of Dwayne for many many years, but circumstance (and mostly illness) has conspired to make me miss all of his ACS presentations thus far. It's probably 10 years or more ago that my friend Andy Dent suggested I get in contact with Dwayne as we had compatible professional interests. I got there eventually, and I hope I wasn't too fangirly :-)
The heart of the presentation was a workshop demo of the core estimation process. In brief, this method uses multiple estimators to independently assign optimistic, most likely, and pessimistic time estimates to specific tasks, and the reasons why. The numbers are crunched to come up with a single reference figure (PERT), then any differences between the estimators are considered. There's more, of course, but the very reluctant volunteer group of three had to sit out the front and estimate how long it would take to put on shoes and socks.
Tip: If one of your stakeholders is a toddler, pessimism is probably the way to go.
It was... educational. Maybe there was some performance pressure but one of the panellists made several major errors of comprehension. I was far too afraid to put myself out where failures would be examined minutely--just as I'm doing here--but once well in to the exercise I definitely wanted to charge in there to stop the pain and get things moving along again.
What I most liked about this process was the inclusion of human foibles and individual differences. OK, I also liked that my particular talent for identifying risk areas would be valued highly in such an environment. I've been fortunate that in most cases my ability to imagine doom and failure scenarios has been respected, in contrast to the places where I was merely "a downer". That was all long ago.
If I failed the courage to get up I at least asked a few questions in the course of the evening, and with the help of Mr Sticky was able to stand around and chat in a vaguely intelligent fashion afterwards (with the possible exception of the aforementioned fangirliness). And although most of the talk was about the impressive array of things that can go wrong in various software development projects--including both technical and human--I found myself missing the cut and thrust of real work. I also experienced plenty of flashbacks of horror in the process, and asked Dwayne if his audience members were ever known to burst into tears at the reopening of old war wounds. Then I heard some fresh stories from the trenches and was glad I don't have to deal with thoroughly clueless management after all.
I used to be good at this stuff. I used to be *damn* good. Grrrr!
*quiet sob*
Hopefully I can harness this newfound restlessness to do some of the work that is still left to me. I certainly feel that it's dug me out of a depression that I feel has been holding me back the last two days. That and the horrible muscular pain and spasms. Hmmm, maybe all those painkillers I took tonight had something to do with that. But they're just not sustainable; I'll pay for that short relief in days to come. Oh well, I enjoyed tonight.
We shall see what tomorrow brings.
I've known of Dwayne for many many years, but circumstance (and mostly illness) has conspired to make me miss all of his ACS presentations thus far. It's probably 10 years or more ago that my friend Andy Dent suggested I get in contact with Dwayne as we had compatible professional interests. I got there eventually, and I hope I wasn't too fangirly :-)
The heart of the presentation was a workshop demo of the core estimation process. In brief, this method uses multiple estimators to independently assign optimistic, most likely, and pessimistic time estimates to specific tasks, and the reasons why. The numbers are crunched to come up with a single reference figure (PERT), then any differences between the estimators are considered. There's more, of course, but the very reluctant volunteer group of three had to sit out the front and estimate how long it would take to put on shoes and socks.
Tip: If one of your stakeholders is a toddler, pessimism is probably the way to go.
It was... educational. Maybe there was some performance pressure but one of the panellists made several major errors of comprehension. I was far too afraid to put myself out where failures would be examined minutely--just as I'm doing here--but once well in to the exercise I definitely wanted to charge in there to stop the pain and get things moving along again.
What I most liked about this process was the inclusion of human foibles and individual differences. OK, I also liked that my particular talent for identifying risk areas would be valued highly in such an environment. I've been fortunate that in most cases my ability to imagine doom and failure scenarios has been respected, in contrast to the places where I was merely "a downer". That was all long ago.
If I failed the courage to get up I at least asked a few questions in the course of the evening, and with the help of Mr Sticky was able to stand around and chat in a vaguely intelligent fashion afterwards (with the possible exception of the aforementioned fangirliness). And although most of the talk was about the impressive array of things that can go wrong in various software development projects--including both technical and human--I found myself missing the cut and thrust of real work. I also experienced plenty of flashbacks of horror in the process, and asked Dwayne if his audience members were ever known to burst into tears at the reopening of old war wounds. Then I heard some fresh stories from the trenches and was glad I don't have to deal with thoroughly clueless management after all.
I used to be good at this stuff. I used to be *damn* good. Grrrr!
*quiet sob*
Hopefully I can harness this newfound restlessness to do some of the work that is still left to me. I certainly feel that it's dug me out of a depression that I feel has been holding me back the last two days. That and the horrible muscular pain and spasms. Hmmm, maybe all those painkillers I took tonight had something to do with that. But they're just not sustainable; I'll pay for that short relief in days to come. Oh well, I enjoyed tonight.
We shall see what tomorrow brings.
![[identity profile]](https://www.dreamwidth.org/img/silk/identity/openid.png){kind=small}
no subject
Date: 2012-02-21 09:55 pm (UTC)Does this system deal with:
Best estimate for a total project is say two years. This includes community consultation and getting legislation through parliament BEFORE the final system changes can be made (parallel development of a general brief occurs almost from day 1).
(also includes odd occurrences like: provider of low-tech solution to small requirement is taken away by PB as *shrug* inefficient, dev mgr and crew drop EVERYthing to make fix - somewhere between a week and a fortnight)
(Of course, all this occurs on top of planned version development of generally a set of changes released every two months (have to train operators, burn-out occurs if we release more frequently) and what is considered a total rewrite every say two years.)
Enter the political boss: No, I need it sooner
Harried development manager: hokay .... lets try for 18 months.
PB: No, I want by July 1 (== 9 months).
melt-down occurs.
I heard it the other side of the building ....
??
no subject
Date: 2012-02-22 05:38 am (UTC)