Software Time Estimation
-
Folks: What are the techniques that you use to estimate the time for software project development? Are there any scientific techniques that can help to estimate the time accurately? I know the intuitive way of breaking up a large project in to sub tasks and then estimating the time for individual sub tasks. However, it is depends on the individuals experience and gut feeling rather than any scientific approach?. Let me know your experience and any pointers (web sites, books etc). :)
-
Folks: What are the techniques that you use to estimate the time for software project development? Are there any scientific techniques that can help to estimate the time accurately? I know the intuitive way of breaking up a large project in to sub tasks and then estimating the time for individual sub tasks. However, it is depends on the individuals experience and gut feeling rather than any scientific approach?. Let me know your experience and any pointers (web sites, books etc). :)
You may find http://www.hmaster.com of some interest. Neville Franks, Author of ED for Windows http://www.getsoft.com
-
Folks: What are the techniques that you use to estimate the time for software project development? Are there any scientific techniques that can help to estimate the time accurately? I know the intuitive way of breaking up a large project in to sub tasks and then estimating the time for individual sub tasks. However, it is depends on the individuals experience and gut feeling rather than any scientific approach?. Let me know your experience and any pointers (web sites, books etc). :)
-
Folks: What are the techniques that you use to estimate the time for software project development? Are there any scientific techniques that can help to estimate the time accurately? I know the intuitive way of breaking up a large project in to sub tasks and then estimating the time for individual sub tasks. However, it is depends on the individuals experience and gut feeling rather than any scientific approach?. Let me know your experience and any pointers (web sites, books etc). :)
For most software development, you just can't do it reliably. If you are lucky enough to have a manager that understands this, then your life will be a LOT easier. Imagine yourself as an architect and builder of homes. If you are building a subdivision with only 3-4 different floor plans, once you build one of each, you will have a great idea about how long it will take. After all, you have done it before. But alas, for software it just doesn't work that way. For most of us, it is more like having a client bring their ideas of what they want to see in a house. Then you have to draw up the plans. Then get them approved. Start the building. Deal with endless change orders. No client in their right mind would ask an architect how long it would take to complete the house after 10 minutes of telling him what they wanted. But for some strange reason, this is expected in software. Of course, this doesn't answer your question. But it is a very good way of looking at the problem of giving software estimates. Tim Smith Descartes Systems Sciences, Inc.
-
Folks: What are the techniques that you use to estimate the time for software project development? Are there any scientific techniques that can help to estimate the time accurately? I know the intuitive way of breaking up a large project in to sub tasks and then estimating the time for individual sub tasks. However, it is depends on the individuals experience and gut feeling rather than any scientific approach?. Let me know your experience and any pointers (web sites, books etc). :)
Most typical way: a) Develop a gut feeling of, say, three weeks. b) Double that. c) Add another two weeks d) ... you'll miss completion time by less than two years. Same applies to building contracts, especially government-financed contracts. [ The city I live in has a very old commuter train system, which was established in the first quarter of the century. The last one, that is. Due to several political changes (World War II, the then-ensuing division of that very city) the commuter train system was partially left unused since 1961. Due to even more political changes, this train system was re-activated beginning in 1990. Now, the re-activation of an already existing piece of rail was due to governmental financing schemes more expensive than a piece of ultra-high-speed long-distance rail that was built at the same time, involving several tunnels and bridges to be built. This was due to sub-contracting every single plank, every single nail to a different local contractor, so that contractors were only sought within the city (and not on a national or even international basis). Thus the politicians and other government officials responsible got some "presents" to keep this system up ... Now it's 2001, and the contractors are still working at re-activating parts of the train system. Current hopes are that they may be finished in 2003. No, this is not a "once upon a time" story, but real life in Berlin, Germany. For those visiting, the train system is called S-Bahn. ]
-
Folks: What are the techniques that you use to estimate the time for software project development? Are there any scientific techniques that can help to estimate the time accurately? I know the intuitive way of breaking up a large project in to sub tasks and then estimating the time for individual sub tasks. However, it is depends on the individuals experience and gut feeling rather than any scientific approach?. Let me know your experience and any pointers (web sites, books etc). :)
1/3 rd design , 1/3rd planning , 2/3rds debugging and the rest working out who to blame. Regards Andrew Torrance England is not an island.