quote:Original post by SabreMan This is nonsense. How do you know what the requirements are for the next project you do? If you don''t know them, then you can''t guess what "designed correctly" means to that project. Sure, you might get to take the class and "drop it in", but that''s simply good fortune. You can''t plan on good fortune arriving too often.
You don''t know the requirements for the next project, I never said you did. Sometimes it is good fortune, I never said that when I write my classes, I make sure that they''ll be able to be used in every single project I work on from now until the day I die, and then my kids will be able to use them and pass them onto my grand kids and so on and so on. The whole point is, if you need to represent an object and it''s behaviour in a project, and someone else has already done that in another project, then why shouldn''t I use that code?
quote:Original post by SabreMan So, that''s your definition of reuse? If something is designed for multiple applications, then using it in those applications is *use*, not reuse. But, I''m repeating myself.
If it was written once, and used in different projects, then it''s being reused. What''s so hard to understand about that?
quote:Original post by SabreMan See above. You can''t know what "written correctly" means for a potential future use.
Yes you can, if my class does what it''s suppose to do, and I need something to do that in another project, and I''m able to drop the class in, then it''s written correctly. Why do you assume that your classes should be able to do everything and a bag of chips. I think you are giving your classes to broad a vision. They should be able to do one thing, and one thing well.
quote:Original post by SabreMan
Sure. But that falls into the "luck" category. It also requires you to know that the class exists. If you wrote it yourself, there''s a pretty good chance you know it exists. Otherwise, the chances are a lot lower.
Yes it''s sometimes luck. Most professional dev shops do code reviews, where you should be able to see what people are up too. If yours doesn''t, then I highly recommend you start checking out what your coworkers are up too. It can help out a lot, not only will you learn from them, you''ll be able to understand the product your working on more, and you''ll have some clue of what''s going on when q/a knocks on your door.
quote:Original post by SabreMan
Huh? A class written in a single context should represent the blueprint for how objects of that class will act in that context. Once the context changes, the class definition may no longer be correct.
That''s fine, in that case you are looking for something different.
quote:Original post by SabreMan
So, if there''s tons of proof, you''ll be able to cite.
Look at any dev shop that uses OO, we reuse our classes at my dev shop, it saves us time, it saves us money. If you want a big example, look at Windows, [sarcasm]I''m sure they completely rewrite everything from the ground up with each new release of Windows[/sarcasm]
quote:Original post by SabreMan
And people aren''t researching beyond OO just for fun, either. I suggest you reserve judgement until you have learned something about some of the new techniques on the ascendant.
People will always be researching for new and better things. They researched OO when everyone was doing procedural. mmmh I wonder why that is? Until the new techniques become available to the public, I''ll continue to use my OO because it''s the best thing we''ve got!
quote:Original post by SabreMan
Maybe it does, maybe it doesn''t. There is no empirical evidence. I''ve spent a lot of time on comp.object, and I''ve seen no proof. Not that I dislike OO. I find some of it very useful. However, I also find some of it to be marketing hype. Reuse falls into the latter category.
It does! Just because you don''t reuse, doesn''t mean it doesn''t save money and time. You speak of marketing hype, what prevents your new techniques from falling into the hype? There''s no evidence because you can''t find it in a news group? That is very naive. You should ask people working in OO dev shops, you''ll find your evidence there.
quote:Original post by SabreMan
Sure. If it''s what I need.
that''s the whole point, if you need it, use it!
quote:Original post by SabreMan No. I do exactly as much as is necessary and no more. It sounds like you need to go "beyond OO".
Beyond OO to what? A new technology that hasn''t been implimented yet? Sure boss, I wrote my stuff in this new technology. It will work eventually, you''ll see. Or should I go beyond to procedural again?