|
The suggested percentages are really low; it's clear that other s agree with me, given the heavy skewing towards the highest number in the responses.
Also, it's a little sad to see you shooting for 97% compliance instead of 100%.
|
|
|
|
|
If any of my existing code failed to build due to it being non standards-compliant, then that part of my code is broken!
Given the available choices, why anyone would select anything other than "More than 100 lines in every 15,000 lines" is beyond me...
IME, the fact that something "works" does not automatically mean that it is *correct*, or works *well*.
-=- James.
|
|
|
|
|
James R. Twine wrote:
Given the available choices, why anyone would select anything other than "More than 100 lines in every 15,000 lines" is beyond me...
That could be blamed to the fact that the technical expertise level among Windows developers is falling down rapidly . On average the attitude "it works so it's ok" is common and even perceived as "being productive". People don't want and often don't understand what they are doing as long as it "makes things go".
|
|
|
|
|
> That could be blamed to the fact that the technical expertise level among
> Windows developers is falling down rapidly.
I concur. I, too, have noticed entropy at work in the developer community all too often.
> On average the attitude "it works so it's ok" is common and even
> perceived as "being productive".
That only goes to show that the *wrong* people are the ones making the decisions, and/or doing the teaching.
> People don't want and often don't understand what they are doing
> as long as it "makes things go".
And the problem with that is when these people are doing something wrong, they do not know it: they lack the metacognitive ability to recognize their own incompetence. They also cannot recognize competence in others, so trying to tell these people that they are doing something wrong is difficult as well: they just do not want to hear it.
(There have been studies to this effect: http://www.apa.org/journals/psp/psp7761121.html; it is quite an interesting read!)
Thanks for the reality check: I am glad to see that there are others out there that can recognize what is going on around them.
-=- James.
|
|
|
|
|
Thanks for the link, even thought it's little lenghtly and I managed to read about 2/3 of it (and the conclusion of course). Just to not let the sad feeling persist let me just quote the Christian's current siganture:
After all, there's nothing wrong with an elite as long as I'm allowed to be part of it!! - Mike Burston Oct 23, 2001
Keep the faith! ![Smile | :)](https://codeproject.global.ssl.fastly.net/script/Forums/Images/smiley_smile.gif)
|
|
|
|
|
I think this review is very poorly phrased, as it seems to imply that users normally write nonstandard code. IMHO, users do write standard code, except for some MS-specific identifiers like __declspec and the like. Actually, users do write in a subset of the standard as MSVC++ hasn't caught up with the whole standard yet (and workarounds for standard compliance MSVC++ deficiencies are within the standard after all).
So, what is really meant by this review? Someone (Nick Hodapp?) please enlighten me.
Joaquín M López Muñoz
Telefónica, Investigación y Desarrollo
|
|
|
|
|
One of the problems with the C/C++ standards is that they are HUGE!!!!
A lot of people here probably write non-standard code out of ignorance. Since the compiler takes it, they assume it is ok.
I would nearly bet, that for most programmers, the compiler not being standard isn't that big of an issue.
HOWEVER, that still isn't an excuse for not moving the compiler forward.
Tim Smith
Descartes Systems Sciences, Inc.
|
|
|
|
|
A lot of people here probably write non-standard code out of ignorance. Since the compiler takes it, they assume it is ok.
Right off the top of my hat, these are the ways in which someone could write nonstandard code in MSVC++:- using MS-specific keywords like
declspec , - relying on behavior that diverges from the standard or it's signaled as "undefined" by the standard (vg, relying on
new throwing xalloc instead of bad_alloc , as it happens in 5.0), - using nonstandard extensions to C++ library abstract classes and template interfaces (vg, nonstandard
_Charalloc used in STL allocators by MSVC++ to cope with the lack of template member functions), - using nonstandard functions that are nevertheless included in standard headers, like
stricmp . IMHO, 2. and 3. happen in extremely rare occasions, and 1. and 2. are innocuous and certainly won't get affected by the compiler being more standard-compliant: my point is that this is a non-issue when discussing MSVC++ standard compliance.
For God's sake, standard compliance is not about extra functions and funny MS-keywords, it is about supporting all of the powerful features granted by the standard. The burden of this task is on the compiler writer, not the users of the tool.
Joaquín M López Muñoz
Telefónica, Investigación y Desarrollo
|
|
|
|
|
I'd have thought that MS compilers have had a long enough run so that people can start accepting the MS C/C++ standards as a sort of standard by itself.
Perhaps it's time other compilers started compiling MS code
Nish
Sonork ID 100.9786 voidmain
www.busterboy.org
If you don't find me on CP, I'll be at Bob's HungOut
|
|
|
|
|
Nope, not an option. Whenever MS creates standard they come up with something so yucky as VB or "managed C++", nobody will follow that. The world is much bigger than MS, they can not dictate the rules, at least not in C++ world.
|
|
|
|
|
"...they can not dictate the rules, at least not in C++ world. "
They did!
|
|
|
|
|
Stormwind wrote:
"...they can not dictate the rules, at least not in C++ world. "
They did!
No, they break the rules, but not dictate.
If MS will push tru their dumb "extensions" they will simply start to create their non-compliant C++ and thus people will use another, better C++ compiler. That in term will force MS back on track, just like in the past it did and just like now MS tries to claim to be the "most compliant". If they could didctate the rules in C++ world they would even suggest the above pool anyway...
|
|
|
|
|
I admire your faith, but you are a fool for believing that.
________________
David Wulff
http://www.davidwulff.co.uk
"I loathe people who keep dogs. They are cowards who haven't got the guts to bite people themselves" - August Strindberg
|
|
|
|
|
David Wulff wrote:
I admire your faith, but you are a fool for believing that.
Well, it's more of an observation than a faith factor. I am developer, say C++ developer, and not a "Windows developer". That means that I am able to choose the kind of tools and environment I work with. I do not depend on Microsoft all that much. Even thought right now I am doing a VC++ development I can (and I did in the past so I am not "just talking") change it to something else.
|
|
|
|
|
Oh brother.
CodeGuy
The WTL newsgroup: now 1060 members! Be a part of it. http://groups.yahoo.com/group/wtl
|
|
|
|
|
Nish, I know you're not dumb, so I can only presume this was a troll....
Christian
After all, there's nothing wrong with an elite as long as I'm allowed to be part of it!! - Mike Burston Oct 23, 2001
Sonork ID 100.10002:MeanManOzI live in Bob's HungOut now
|
|
|
|
|
Christian, Christian, Christian
Yup, it was a teaser-post. Jus to get people's reactions.
a bit like posting "do you really think god created this planet" or "hey, did you see that new IE version!"
Nish
Sonork ID 100.9786 voidmain
www.busterboy.org
If you don't find me on CP, I'll be at Bob's HungOut
|
|
|
|
|
Nish [BusterBoy] wrote:
Yup, it was a teaser-post. Jus to get people's reactions.
I thought so. Did one of those myself earlier today - fun isn't it ?
Christian
After all, there's nothing wrong with an elite as long as I'm allowed to be part of it!! - Mike Burston Oct 23, 2001
Sonork ID 100.10002:MeanManOzI live in Bob's HungOut now
|
|
|
|
|
Christian Graus wrote:
Did one of those myself earlier today - fun isn't it ?
Yeah. Sort of
Nish
Sonork ID 100.9786 voidmain
www.busterboy.org
If you don't find me on CP, I'll be at Bob's HungOut
|
|
|
|
|
why don't just the VC compiler team allow this:
#define I_WANT_TO_DEVELOP_IN_WINDOWS_AND_NOTHING_ELSE
// VC6 compliant code
#undef I_WANT_TO_DEVELOP_IN_WINDOWS_AND_NOTHING_ELSE
// ANSI C++ code
Surely, adding these macros in our old code isn't much trouble.
"Sometimes the solutions of our problems are the problems themselves."
|
|
|
|
|
There should have been an option: "I already write standard-compliant code. I would not need to rewrite."
For several years, I have been writing in a subset of the standard that will compile on MSVC. I don't want to lock my code in unnecessarily to Windows, should I decide some day that another OS serves me better. I don't lose anything by being standards-compliant (libraries and APIs are another matter, but all sensible developers separate the OS-glue from the core).
Thus, I don't anticipate that 100% C++ compliance by VC++ would require me to rewrite any of my code.
|
|
|
|
|
I would tend to think the same. However, we are writing a cross platform app at the moment, and we've only done the Windows part. I maintain it won't be cross platform until it has compiled under Mac and *nix. Until that happens, our cross platformness, like your and my standard compliant code, are just a theory. Therefore it makes sense to answer according to the degree to which we'd be happy to change our code if the theory is proven wrong.
Christian
After all, there's nothing wrong with an elite as long as I'm allowed to be part of it!! - Mike Burston Oct 23, 2001
Sonork ID 100.10002:MeanManOzI live in Bob's HungOut now
|
|
|
|
|
How many lines of MFC sources are *you* willing to change for better standard compliance?
|
|
|
|
|
A very good point!
However, it seems that MFC is not developed anylonger, and they are focusing on Managed C++, they will probably not change a bit of MFC.
Best regards,
Alexandru Savescu
|
|
|
|
|
Alexandru Savescu wrote:
they will probably not change a bit of MFC.
MFC has been updated to MFC 7.0, and they have probably made a few internal tweaks and changes as well as adding all the new stuff. You have to remember that Microsoft use MFC extensively (i.e. Visio), so they aren't going to stop updating it anytime soon. Whether we get to see these updates is another story...
________________
David Wulff
http://www.davidwulff.co.uk
"I loathe people who keep dogs. They are cowards who haven't got the guts to bite people themselves" - August Strindberg
|
|
|
|