Thread overview | ||||||||||||||||||||||||||||||||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
|
November 18, 2015 Scott Meyers wants to bring default zero-initialization to C++, mentions TDPL for precedent | ||||
---|---|---|---|---|
| ||||
He advocates for a tool like gofix, to automatically convert such features to be deprecated: http://scottmeyers.blogspot.com/2015/11/breaking-all-eggs-in-c.html Good to see C++ finally trying to deprecate more, long overdue. |
November 18, 2015 Re: Scott Meyers wants to bring default zero-initialization to C++, mentions TDPL for precedent | ||||
---|---|---|---|---|
| ||||
Posted in reply to Joakim | On Wednesday, 18 November 2015 at 15:12:27 UTC, Joakim wrote: > He advocates for a tool like gofix, to automatically convert such features to be deprecated: > > http://scottmeyers.blogspot.com/2015/11/breaking-all-eggs-in-c.html > > Good to see C++ finally trying to deprecate more, long overdue. I doubt anything will get done even in 10 years. The average C++ industrial codebase don't build with the latest compiler, or even any static analyzer, and has its own particular build system. It is already a big trouble to upgrade C++ compilers even if nothing in the language is changing. D numerous, tiny interchangeable releases are a big asset when compared to the whole compiler-tied-to-IDE-tied-to-OS-releases which is what things are in C++ land. The idea that you could bring the C++ community to use an automatic upgrade tool, or to get everyone to follow optional "Core Guidelines" is optimistic. |
November 18, 2015 Re: Scott Meyers wants to bring default zero-initialization to C++, mentions TDPL for precedent | ||||
---|---|---|---|---|
| ||||
Posted in reply to Joakim | On 11/18/15 10:12 AM, Joakim wrote:
> He advocates for a tool like gofix, to automatically convert such
> features to be deprecated:
It isn't going to happen. See the caveat at the bottom, along with his other post.
As long as C has the preprocessor, a tool like gofix is nearly impossible to create.
I'm sure Daniel has some experience with this ;)
-Steve
|
November 18, 2015 Re: Scott Meyers wants to bring default zero-initialization to C++, mentions TDPL for precedent | ||||
---|---|---|---|---|
| ||||
Posted in reply to ponce | On Wednesday, 18 November 2015 at 15:29:04 UTC, ponce wrote:
> The idea that you could bring the C++ community to use an automatic upgrade tool, or to get everyone to follow optional "Core Guidelines" is optimistic.
What community? Compilers can have pedantic compiler-switches. C++ needs what Ada has, different profiles for different scenarios. No problem in having a strict profile as the default for a compiler.
|
November 18, 2015 Re: Scott Meyers wants to bring default zero-initialization to C++, mentions TDPL for precedent | ||||
---|---|---|---|---|
| ||||
Posted in reply to Steven Schveighoffer | On Wednesday, 18 November 2015 at 15:46:38 UTC, Steven Schveighoffer wrote:
> On 11/18/15 10:12 AM, Joakim wrote:
>> He advocates for a tool like gofix, to automatically convert such
>> features to be deprecated:
>
> It isn't going to happen. See the caveat at the bottom, along with his other post.
>
> As long as C has the preprocessor, a tool like gofix is nearly impossible to create.
>
> I'm sure Daniel has some experience with this ;)
>
> -Steve
Note that we have the same problem with our unstructured templates and string mixins.
|
November 18, 2015 Re: Scott Meyers wants to bring default zero-initialization to C++, mentions TDPL for precedent | ||||
---|---|---|---|---|
| ||||
Posted in reply to Ola Fosheim Grøstad | On Wednesday, 18 November 2015 at 18:43:59 UTC, Ola Fosheim Grøstad wrote:
> On Wednesday, 18 November 2015 at 15:29:04 UTC, ponce wrote:
>> The idea that you could bring the C++ community to use an automatic upgrade tool, or to get everyone to follow optional "Core Guidelines" is optimistic.
>
> What community? Compilers can have pedantic compiler-switches. C++ needs what Ada has, different profiles for different scenarios. No problem in having a strict profile as the default for a compiler.
Oh God, please don't push for PHP-ifying C++. It's already confusing enough as it is.
For those of you who have never had the pleasure of writing PHP, lots of PHP code does completely different things depending on the compiler switches when the interpreter was compiled, the global config for the host machine, the local config files in the directory, and the warning level switches at the start of the code. So much so that it's impossible to understand what a lot of code does without consulting all of these locations to double check.
When looking at code, it should be obvious what it does without consulting outside factors.
|
November 18, 2015 Re: Scott Meyers wants to bring default zero-initialization to C++, mentions TDPL for precedent | ||||
---|---|---|---|---|
| ||||
Posted in reply to Jack Stouffer | On Wednesday, 18 November 2015 at 18:52:02 UTC, Jack Stouffer wrote: > For those of you who have never had the pleasure of writing PHP, lots of PHP code does completely different things depending on the compiler switches when the interpreter was Php is a dynamic language, that's different. > When looking at code, it should be obvious what it does without consulting outside factors. I don't think this is a problem, it probably will be better overall. When C++ adds modules they have a golden opportunity to clean up the language for those compilation units that use it. But they probably won't. Which is good for languages like Rust and D. |
November 18, 2015 Re: Scott Meyers wants to bring default zero-initialization to C++, mentions TDPL for precedent | ||||
---|---|---|---|---|
| ||||
Posted in reply to Ola Fosheim Grøstad | On Wednesday, 18 November 2015 at 19:09:48 UTC, Ola Fosheim Grøstad wrote:
> On Wednesday, 18 November 2015 at 18:52:02 UTC, Jack Stouffer wrote:
>> For those of you who have never had the pleasure of writing PHP, lots of PHP code does completely different things depending on the compiler switches when the interpreter was
>
> Php is a dynamic language, that's different.
Is it really though? In theory, yes. But in practice it's the same problem. A switch to the compiler would change the behavior of some very common and fundamental pieces of code sounds like the same problem to me.
|
November 19, 2015 Re: Scott Meyers wants to bring default zero-initialization to C++, mentions TDPL for precedent | ||||
---|---|---|---|---|
| ||||
Posted in reply to Jack Stouffer | On Wednesday, 18 November 2015 at 21:08:38 UTC, Jack Stouffer wrote:
> Is it really though? In theory, yes. But in practice it's the same problem.
Not the same problem. In dynamic languages changes are caught at runtime, not compile time.
When changes affect compile time only then you can give up upgrading the code base and use a compatibility switch.
|
November 20, 2015 Re: Scott Meyers wants to bring default zero-initialization to C++, mentions TDPL for precedent | ||||
---|---|---|---|---|
| ||||
Posted in reply to Ola Fosheim Grøstad | On Wednesday, 18 November 2015 at 19:09:48 UTC, Ola Fosheim Grøstad wrote:
> On Wednesday, 18 November 2015 at 18:52:02 UTC, Jack Stouffer wrote:
>> For those of you who have never had the pleasure of writing PHP, lots of PHP code does completely different things depending on the compiler switches when the interpreter was
>
> Php is a dynamic language, that's different.
Would you guys please stop calling PHP a language. JS and PHP are not languages, they are a fiddly feckin mess, which accounts for their high adoption rate. Sigh. Please, no gainsaying now, no "Yes, buts".
|
Copyright © 1999-2021 by the D Language Foundation