VMT in C++

IWETHEY v. 0.3.0 | TODO

1,095 registered users | 0 active users | 0 LpH | Statistics
Login | Create New User

Welcome to IWETHEY!

Post #175,881

by Arkadiy

9/23/04 9:56:51 AM

VMT in C++

Virtual Method Table.

For single inheritance, the simplest thing in the world: first word in object is a pointer to the table of its methods' entry points. One table per class, the pointer is set in constructor.

For multiple inheritance it gets hairy as hell.

--

... a reference to Presidente Arbusto.
-- [link|http://itre.cis.upenn.edu/~myl/languagelog/archives/001417.html|Geoffrey K. Pullum]

Lines of code in a Java method ... - (bluke) - (73) - Sept. 22, 2004, 10:23:14 AM EDT

I don't know of studies of this with OO languages - (ben_tilly) - (64) - Sept. 22, 2004, 10:48:22 AM EDT

I remember that when I read Code Complete ... - (bluke) - (60) - Sept. 22, 2004, 10:59:17 AM EDT

Hence my speculation about how and why to change that for OO -NT - (ben_tilly) - (59) - Sept. 22, 2004, 11:30:34 AM EDT

In OO code many of the decision points ... - (bluke) - (58) - Sept. 22, 2004, 11:36:37 AM EDT

Quibble - (ben_tilly) - (57) - Sept. 22, 2004, 08:04:26 PM EDT

Okay, thought it was just me - (drewk) - (46) - Sept. 22, 2004, 08:59:44 PM EDT

Not really missing anything. - (FuManChu) - (1) - Sept. 22, 2004, 09:34:24 PM EDT

VMT in C++ - (Arkadiy) - Sept. 23, 2004, 09:56:51 AM EDT

The superclass decides nothing - (ben_tilly) - (43) - Sept. 22, 2004, 09:55:16 PM EDT

Once again, it's the terminology - (drewk) - (40) - Sept. 22, 2004, 10:20:23 PM EDT

That's not multiple inheritance - (ben_tilly) - Sept. 23, 2004, 12:05:47 AM EDT

Ramble ramble ramble - (FuManChu) - Sept. 23, 2004, 02:46:35 AM EDT

Yes and No - (bluke) - (37) - Sept. 23, 2004, 03:57:24 AM EDT

I mis-read Ben - (drewk) - (36) - Sept. 23, 2004, 10:07:50 AM EDT

You misunderstand interfaces and factories ... - (bluke) - (2) - Sept. 23, 2004, 11:20:06 AM EDT

Drew is right. - (FuManChu) - (1) - Sept. 23, 2004, 12:38:44 PM EDT

My point was - (bluke) - Sept. 23, 2004, 05:00:13 PM EDT

I'm actually on the fence - (ben_tilly) - (32) - Sept. 23, 2004, 10:55:24 AM EDT

That is with implementation inheritance - (bluke) - (2) - Sept. 23, 2004, 11:00:41 AM EDT

That is both with interface and implementation inheritance - (ben_tilly) - (1) - Sept. 23, 2004, 11:08:29 AM EDT

With interfaces it is less of a concern - (bluke) - Sept. 23, 2004, 11:17:39 AM EDT

Hadn't heard of LSP - (drewk) - (28) - Sept. 23, 2004, 12:36:16 PM EDT

Huh? - (tuberculosis) - (27) - Sept. 23, 2004, 01:23:55 PM EDT

Don't get all religious on me - (drewk) - (26) - Sept. 24, 2004, 12:30:58 AM EDT

Going to disagreed on this point - (JimWeirich) - (1) - Sept. 24, 2004, 01:01:56 AM EDT

agreed - (deSitter) - Sept. 24, 2004, 04:01:49 AM EDT

Re: Don't get all religious on me - (JimWeirich) - (23) - Sept. 24, 2004, 01:11:21 AM EDT

And that's the key point - (ben_tilly) - (22) - Sept. 24, 2004, 10:34:42 AM EDT

Someday I've got to take a class - (drewk) - (21) - Sept. 24, 2004, 03:18:39 PM EDT

..when I get my inheritance -NT - (deSitter) - Sept. 24, 2004, 03:54:21 PM EDT

Yup. No generic substitute for that. -NT - (FuManChu) - Sept. 24, 2004, 04:00:08 PM EDT

Good idea in this instance. - (admin) - (14) - Sept. 24, 2004, 04:20:45 PM EDT

I object; it's part of a general pattern -NT - (FuManChu) - (13) - Sept. 24, 2004, 05:44:34 PM EDT

Ah, but I have a cunning strategy... - (admin) - (12) - Sept. 24, 2004, 06:17:48 PM EDT

Bah. Overridden. - (FuManChu) - (1) - Sept. 24, 2004, 06:26:44 PM EDT

Overloaded your capacity for understanding, eh? - (admin) - Sept. 24, 2004, 06:31:13 PM EDT

Ashton, is that you? - (Another Scott) - (7) - Sept. 24, 2004, 06:29:07 PM EDT

Better than the treasurer... - (admin) - (6) - Sept. 24, 2004, 06:33:02 PM EDT

He'll have to wait - I'm collecting garbage - (ben_tilly) - (5) - Sept. 24, 2004, 06:59:43 PM EDT

Coincidentally, there's a fellow around here named Mark... - (admin) - (4) - Sept. 24, 2004, 08:48:18 PM EDT

I'm sorry but that pointer is dangling - (ben_tilly) - (3) - Sept. 24, 2004, 09:11:32 PM EDT

Some people have too much time on their hands... - (static) - (2) - Sept. 25, 2004, 06:26:13 AM EDT

You're a real LALRf riot. - (admin) - (1) - Sept. 25, 2004, 09:35:21 AM EDT

So you sed. Personally, I think it's a Perl in the Blinker. -NT - (Another Scott) - Sept. 25, 2004, 10:12:54 AM EDT

Uuuuuhh... - (jb4) - Sept. 24, 2004, 07:27:18 PM EDT

The cun.. er, clever linguist suffers Code 715 - (Ashton) - Sept. 25, 2004, 03:08:51 AM EDT

Or an interface - (tuberculosis) - (2) - Sept. 24, 2004, 10:56:23 PM EDT

Yes, and that's what we do - (drewk) - (1) - Sept. 25, 2004, 10:04:33 PM EDT

Language doesn't need to support abstract class - (tuberculosis) - Sept. 25, 2004, 11:47:26 PM EDT

That was not what I meant - (bluke) - Sept. 26, 2004, 02:30:26 AM EDT

It's a decision, but it's abstracted out - (Arkadiy) - (1) - Sept. 23, 2004, 10:11:04 AM EDT

Yes, but I don't care how it works - (bluke) - Sept. 23, 2004, 10:39:02 AM EDT

DP's are compressed in OO - (tuberculosis) - (6) - Sept. 23, 2004, 01:09:43 PM EDT

Polymorphism relies on a lie - (tablizer) - (5) - Oct. 7, 2004, 11:07:36 PM EDT

Re: Polymorphism relies on a lie - (admin) - (2) - Oct. 7, 2004, 11:23:58 PM EDT

Mostly, it is - (tablizer) - (1) - Oct. 13, 2004, 01:10:40 PM EDT

To the naive programmer, perhaps. - (admin) - Oct. 13, 2004, 01:57:31 PM EDT

Something tells me you'd like prototype OO as a midground -NT - (FuManChu) - (1) - Oct. 8, 2004, 01:56:34 AM EDT

*No-one* can eat that much crow -NT - (pwhysall) - Oct. 8, 2004, 01:57:41 AM EDT

We had a big fight over case-vs-poly a few years ago -NT - (tablizer) - (2) - Sept. 27, 2004, 11:34:24 PM EDT

Charley Rock Tree - (ChrisR) - (1) - Sept. 27, 2004, 11:41:52 PM EDT

Ha! :-) -NT - (Another Scott) - Sept. 27, 2004, 11:45:00 PM EDT

Re: I don't know of studies of this with OO languages - (deSitter) - Sept. 22, 2004, 02:44:15 PM EDT

Interesting comment by Ron Jeffries - (bluke) - (1) - Sept. 23, 2004, 08:41:04 AM EDT

Re: Interesting comment by Ron Jeffries - (deSitter) - Sept. 23, 2004, 09:36:49 AM EDT

to minimize the damage that these kind of developers wreak - (Arkadiy) - (1) - Sept. 22, 2004, 01:04:29 PM EDT

Unfortunately I can't - (bluke) - Sept. 22, 2004, 01:17:18 PM EDT

4. -NT - (FuManChu) - (2) - Sept. 22, 2004, 02:30:58 PM EDT

In smalltalk that might be realistic - (bluke) - (1) - Sept. 23, 2004, 04:02:32 AM EDT

That was my point. Glad someone got the joke :) -NT - (FuManChu) - Sept. 23, 2004, 12:40:30 PM EDT

1/2 page - (tuberculosis) - Sept. 23, 2004, 01:18:44 PM EDT

Re: Lines of code in a Java method ... - (dshellman) - (1) - Sept. 23, 2004, 01:43:04 PM EDT

Agree - (bluke) - Sept. 26, 2004, 02:32:18 AM EDT

iwethey.org

No, it's too perilous.
302 ms