Relational-to-OOP Tax
in [OOP]
|
Prev: "OOP is Dead" article
Next: The reality of Topmind (Was: Topic-Organized Object-Oriented Programming)
From: Daniel Parker on 7 Feb 2007 11:02 On Feb 7, 1:23 am, "topmind" <topm...(a)technologist.com> wrote: > On Feb 6, 7:14 pm, "Daniel Parker" <danielapar...(a)gmail.com> wrote: > > > On Feb 5, 7:21 pm, "topmind" <topm...(a)technologist.com> wrote: > > > > Martin hyped OO ... > > > Just out of curiousity, can you point out where RCM has "hyped OO"? I > > must have missed it, in the fifteen years or so I've been following > > his writings. (Now if you'd said "hyped XP", you'd have been on > > stronger ground :-)) > > I cannot give a specific quote at the moment, but it is typical OO > book argument-by-implication. > > For example, in chapter 7 he gives a typical device-driver-like > example of a "copy" program. He then shows how bad case/if statements > are for adding new devices and how polymorphism saves the day. It > does not state that this may or may not be relavent outside of device > drivers. There is no clear disclaimer anywhere I could find. > > The *implication* is that if poly is good here, it is good > *everywhere*. > > "Argument by narrow example" you could call this phenom. > Okay, none of that qualifies as hype. Having a point of view, demonstrating it with working code, and advocating it does not qualify as hype. RCM's writing tend to be tutorial in nature, and a tutorial style doesn't really fit into the category of hype. For real OO hype, you'd have to go back quite a few years, to the days of JOOP, where people were writing articles promoting commercial interests thinly disguised as technical articles. Kind of like the stuff that's being written about SOA today. But I don't think RCM ever wrote anything like that. You might want to backtrack on this. Best regards, Daniel
From: topmind on 7 Feb 2007 11:24 Daniel Parker wrote: > On Feb 7, 1:23 am, "topmind" <topm...(a)technologist.com> wrote: > > On Feb 6, 7:14 pm, "Daniel Parker" <danielapar...(a)gmail.com> wrote: > > > > > On Feb 5, 7:21 pm, "topmind" <topm...(a)technologist.com> wrote: > > > > > > Martin hyped OO ... > > > > > Just out of curiousity, can you point out where RCM has "hyped OO"? I > > > must have missed it, in the fifteen years or so I've been following > > > his writings. (Now if you'd said "hyped XP", you'd have been on > > > stronger ground :-)) > > > > I cannot give a specific quote at the moment, but it is typical OO > > book argument-by-implication. > > > > For example, in chapter 7 he gives a typical device-driver-like > > example of a "copy" program. He then shows how bad case/if statements > > are for adding new devices and how polymorphism saves the day. It > > does not state that this may or may not be relavent outside of device > > drivers. There is no clear disclaimer anywhere I could find. > > > > The *implication* is that if poly is good here, it is good > > *everywhere*. > > > > "Argument by narrow example" you could call this phenom. > > > Okay, none of that qualifies as hype. Having a point of view, > demonstrating it with working code, and advocating it does not qualify > as hype. RCM's writing tend to be tutorial in nature, and a tutorial > style doesn't really fit into the category of hype. I don't see how a tutorial nature de-hypes something. Tutorials can be as hypee as any other form of writing. I will agree that I've seen more extreme examples of hype than Martin's, but hype is still hype. And he has in the past implied that his book had clear evidence of betterment. That is why I spent money on the damned thing. Sure, he gives one or two (well worn) change scenarios for where OO may do better, but he analyses neither the probability of those scenarios nor explores scenarios that don't favor them, such as loss of mutual exclusitivity. It is another case of excess extrapolation: Poly does better under change scenario X, so it must do better under all change scenarios. This theme runs throughout the OO world. It is either hype or bad evidence presentation. > For real OO hype, > you'd have to go back quite a few years, to the days of JOOP, where > people were writing articles promoting commercial interests thinly > disguised as technical articles. Kind of like the stuff that's being > written about SOA today. But I don't think RCM ever wrote anything > like that. You might want to backtrack on this. Again, your implication is that because there are worse offenders than Martin, that Martin should be let off the hook. You could argue that I should spend my complaints on the worse offenders, but that is not the same as demonstrating that Martin is hype-free. > > Best regards, > Daniel -T-
From: Daniel Parker on 7 Feb 2007 14:20 On Feb 7, 11:24 am, "topmind" <topm...(a)technologist.com> wrote: > Daniel Parker wrote: > > For real OO hype, > > you'd have to go back quite a few years, to the days of JOOP, where > > people were writing articles promoting commercial interests thinly > > disguised as technical articles. Kind of like the stuff that's being > > written about SOA today. But I don't think RCM ever wrote anything > > like that. You might want to backtrack on this. > > Again, your implication is that because there are worse offenders than > Martin, that Martin should be let off the hook. No, that's not my implication. As far as I can tell, RCM has never written anything about OO that could reasonably be regarded as hype, ever. Feel free to have the last word (I know I won't get it.) Daniel
From: topmind on 7 Feb 2007 16:42 Daniel Parker wrote: > On Feb 7, 11:24 am, "topmind" <topm...(a)technologist.com> wrote: > > Daniel Parker wrote: > > > > For real OO hype, > > > you'd have to go back quite a few years, to the days of JOOP, where > > > people were writing articles promoting commercial interests thinly > > > disguised as technical articles. Kind of like the stuff that's being > > > written about SOA today. But I don't think RCM ever wrote anything > > > like that. You might want to backtrack on this. > > > > Again, your implication is that because there are worse offenders than > > Martin, that Martin should be let off the hook. > > No, that's not my implication. As far as I can tell, RCM has never > written anything about OO that could reasonably be regarded as hype, > ever. Feel free to have the last word (I know I won't get it.) Well, perhaps it depends on how one defines "hype". Perhaps "misleading" is a better term for the nature of Martin's OO sins. But the difference between the two is rather minor. It is also curious how Martin finds a way to work the latest software engineering buzzwords into his book titles, dispite the fact that the content and arguments do not change much. Either he is surfin' on fads to sell books, or everything in style or new happens to always be good. Not an accusation, just a watch-point. > > Daniel -T- oop.ismad.com
From: ggroups on 8 Feb 2007 04:12
On Feb 7, 9:42 pm, "topmind" <topm...(a)technologist.com> wrote: > Daniel Parker wrote: DP> For real OO hype, DP> you'd have to go back quite a few years, to the days of JOOP, where DP> people were writing articles promoting commercial interests thinly DP> disguised as technical articles. Kind of like the stuff that's being DP> written about SOA today. But I don't think RCM ever wrote anything DP> like that. You might want to backtrack on this. TM> Again, your implication is that because there are worse offenders than TM> Martin, that Martin should be let off the hook. >> No, that's not my implication. As far as I can tell, RCM has never >> written anything about OO that could reasonably be regarded as hype, >> ever. Feel free to have the last word (I know I won't get it.) > Well, perhaps it depends on how one defines "hype". 1. Well, this is what the real world seems to think the term means : http://dictionary.reference.com/browse/hyperbole http://en.wikipedia.org/wiki/Hyperbole Feel free to show us how/where Robert Martin has posted about OO in a manner consistent with the above definitions. 2. Given that you have form, IMHO the following is most appropriate : <quote> "When I use a word," Humpty Dumpty said, in a rather scornful tone, "it means just what I choose it to mean - neither more nor less." "The question is," said Alice, "whether you can make words mean so many different things." "The question is," said Humpty Dumpty, "which is to be master - that's all." </quote> If your definition of "hype" does not match the real world definitions, feel free to provide your definition of the term (preferably in a manner akin to 1) , and then show us how/where Robert Martin has posted about OO in a manner consistent with your definition(s) . If you are unable to do either 1 or 2, *retract your claims* . > Perhaps "misleading" is a better term for the nature of Martin's OO sins. > But the difference between the two is rather minor. http://dictionary.reference.com/browse/mislead There is a *BIG* difference between the two. In fact by stating the latter, you are making a far more serious claim about Robert Martin. And for this I suggest you *immediately* provide refs to where Robert Martin is misleading people, or retract the claim. > It is also curious how Martin finds a way to work the latest software > engineering buzzwords into his book titles, dispite the fact that the > content and arguments do not change much. Either he is surfin' on fads > to sell books, or everything in style or new happens to always be > good. Not an accusation, just a watch-point. All I now see is what IMHO is "hate crime" towards Robert Martin. Not the first time someone on comp.object has indulged in such behaviour ... Regards, Steven Perryman |