tag:blogger.com,1999:blog-736810000699453506.post4737670039766618696..comments2023-05-11T00:49:36.314-07:00Comments on Ruminations on Computing: Why Writing Software Is HardSteve Rowehttp://www.blogger.com/profile/17905356014908630180noreply@blogger.comBlogger3125tag:blogger.com,1999:blog-736810000699453506.post-11311958434944975982007-02-03T13:35:39.000-08:002007-02-03T13:35:39.000-08:00Great post. I would like to add that, most compan...Great post. I would like to add that, most companies don't keep a historical record of how long software projects took them in the past. Then, when faced with the task of estimating a new project, there is no data to rely upon, and yet another "off the cuff" optimistic estimate is made. Without historical data, accurate future estimates cannot be made, plain and simple.tzagottanoreply@blogger.comtag:blogger.com,1999:blog-736810000699453506.post-34507383987962072712007-02-04T19:38:09.000-08:002007-02-04T19:38:09.000-08:00The approach "dive into coding without up-fro...The approach "dive into coding without up-front estimation" often lead to situation when the almost complete project stops due to technical problems (or important features are cutting off). If the project is complex, your first question is "can I complete it at all"? Then you write prototypes for features you haven't done before. If the particular feature cannot be implemented, the project cannot finish, so just don't start the project. Briefly speaking, do a good research before planning, then make the plan based on facts. Don't hope the "list of work" based on guess is doable.~AVAhttp://avapublic.blogspot.comnoreply@blogger.comtag:blogger.com,1999:blog-736810000699453506.post-10151878994682077262007-02-05T11:56:32.000-08:002007-02-05T11:56:32.000-08:00It is very interesting that we don't tend to k...It is very interesting that we don't tend to keep records. I wonder if this is just organizational laziness or if is has more to do with the fact that software organizations never really do the same work twice. Probably a bit of both.<br>@Ava-Good words of wisdom. Trying to do too much that is unknown at once is a recipe for taking forever. Doing some prototyping is always a good idea.SteveRowenoreply@blogger.com