Containers: The iterator object

[jacob navia]

ng2010 a écrit :

That, of course, is false. It is only true to you, for you don't know any
better (apparently).

That, of course, is true. It is only false to you, for you don't know
any better (apparently).

Anybody can say that. If you disagree with something I write, it is not
enough to say "That is false" but you have to explain why you consider
it false.


You are unable to propose any arguments apparently.

So, before you start in on details, you better
discuss the alternative choices and why the above stated design is best.

Because in non sequential containers GetNext and GetPrevious should
traverse the sequence in some unspecified order that only guarantees
that all elements will be eventually visited. I wrote that inmy proposal
but since you did not get past the first sentence you did not read it.

HINT:

Read the propositions to the end, THEN reply. Thanks.

(Aside: If this is a user group, it surely doesn't need first-time
library builders acting like they have definitive library designs.
"neophytes" WILL get sucked into the other neophyte's personal R&D, and
surely C is "well-evolved" and not a research project in real usage (?)).

You are free to have your opinion, this is Usenet. Since (again) you do
not put any argumentation to substantiate your views, it is better to
leave them as what they are. Your personal opinion.

(Aside 2: JN, if you think C is so great and good and readily usable, why
don't you go actually USE it instead of trying to extend/modify it?

I use C every day. What is comic in your attacks is that you do not know
anything about me but somehow you think you can make statements like
"Why don't you actually USE C".

Your
actions say that C is not up to the task. And you are going on and on
about FOUNDATIONAL issues with the language.

And why should I refrain from doing that?

Ahh of course, I forgot, excuse me: Because you said so.

> I was trying to be helpful
when I said don't waste your time.)

Great! Then, you could be even more helpful and *explain* your point
ofview instead of just stating "This is wrong"...

You see? If you "explain" your views, we can discuss about them.

If you just say: "This is wrong" there is no discussion possible.

 

[Tim Rentsch]

You misunderstand completely. I have implemented already queues using a
list container.

You use the list directly, i.e. using
for (list_element=List->First; list_element != NULL;
list_element = list_element->Next ) {
// Use list_element->data here
}

You see?

You can use the components of the containers directly and iterate them
WITHOUT using any iterators at all.

Iterators are useful for MOST uses, but not for ALL uses. But you are
NOT limited to using iterators. You can use the containers directly.

I agree. It's Bad Polymorphism if you have an iterator that does
two different things (ie, what you want OR destroy itself, other
iterators and the list) depending on the specific data structure
you obtained it from. [snip]

Not "Bad Polymorphism". It may be (or may not be, depending on
specifics) bad interface design, or bad implementation choice, or
something else along those lines. The phrase "Bad Polymorphism"
just doesn't scan in this context.

 

[Phil Carmody]

I agree. It's Bad Polymorphism if you have an iterator that does
two different things (ie, what you want OR destroy itself, other
iterators and the list) depending on the specific data structure
you obtained it from. [snip]

Not "Bad Polymorphism". It may be (or may not be, depending on
specifics) bad interface design, or bad implementation choice, or
something else along those lines. The phrase "Bad Polymorphism"
just doesn't scan in this context.

"Bad polymorphism" to me just implies "violates the LSP".
His example is one of a violation of the LSP.

Phil

 

[Tim Rentsch]

I agree. It's Bad Polymorphism if you have an iterator that does
two different things (ie, what you want OR destroy itself, other
iterators and the list) depending on the specific data structure
you obtained it from. [snip]

Not "Bad Polymorphism". It may be (or may not be, depending on
specifics) bad interface design, or bad implementation choice, or
something else along those lines. The phrase "Bad Polymorphism"
just doesn't scan in this context.

"Bad polymorphism" to me just implies "violates the LSP".
His example is one of a violation of the LSP.

Yes, I understood that he intended his comment to mean
that such an implementation would violate some design
principle. But he didn't say which design principle,
and no matter which design principle was meant, "Bad
Polymorphism" is a bad name for it.

 

[ng2010]

ng2010 a écrit :

That, of course, is true. It is only false to you, for you don't know
any better (apparently).

Anybody can say that. If you disagree with something I write, it is
not enough to say "That is false" but you have to explain why you
consider it false.

You are obviously wrong on that assertion. _I_ don't have to GIVE you
anything.

You are unable to propose any arguments apparently.

Apparently you are ARGUING. I wouldn't want to get in-between anything.

Because

Well, see, go back and start again. I wasn't offering amnesty to continue
on in your immature and wasteful way. Your book report was bad. You need
to rewrite it and offer it again for regrading. There is no progression
to the next level until you meet the requirements of the preceding level.

HINT:

Read the propositions to the end, THEN reply. Thanks.

HINT:

Don't start selling snakeoil until you establish a market for it. (NOT a
recommendation to sell snakeoil!).

You are free to have your opinion, this is Usenet. Since (again) you
do not put any argumentation to substantiate your views, it is better
to leave them as what they are. Your personal opinion.

That is childishness on your part (I make some assumptions that follow
the immediately following list of your "wants").

1. This isn't a group to sell R&D projects to.
2. Your personal agenda is just that: personal.
3. If C is (still) useful, people will come in here and talk about it IN
USE, but if you FLOOD the room with your PROPOGANDA/AGENDA, ... I don't
think you want to be that.
4. Go talk in the standard C group if your intention is to "evolve" C.
Become a member of the C ISO group if that is your true intent, THEY own
C, you do not.
5. There is no "C" outside of the C standard unless you create your own
"fork"/tangent from it. Else, it is something else.

I use C every day. What is comic in your attacks is that you do not
know anything about me but somehow you think you can make statements
like "Why don't you actually USE C".

I'm going to let go that comment you made about "my attacks". Extending C
and using it are quite different. Because I know you don't use it. Your
mind is ENTIRELY occupied by "making it better", and I told you before,
stop "spinning your wheels", it's just a car. It's not alive, and it's
not your life. IT, takes your time though, and that's all you have.

And why should I refrain from doing that?

I don't think this is the group for your elementary R&D. I can tool up
and try to sell my next big stupid combobulated thing (and you're doing
it as a skunkwork) .... it's a circus. See a flea do something never
intended. Starting to sound like a crime.

Refrain from doing what? What were you doing?

 

[jacob navia]

ng2010 a écrit :

You are obviously wrong on that assertion. _I_ don't have to GIVE you
anything.

(snip)

OK. This is Usenet, you can say whatever you feel like. Personally,
since you do not propose any arguments and just dismiss everything I say
without any arguments I will stop discussing with you.