5 ms·

I fail to understand either, but they say: > CopperSpice has been modified to remove moc, the Qt meta object compiler. Moc was removed to simplify the build pr

by fla 11y ago

I fail to understand either, but they say: > CopperSpice has been modified to remove moc, the Qt meta object compiler. Moc was removed to simplify the build process, allow CopperSpice to be a standalone library, increase run time performance, and allow developers to utilize C++11 features. So they replaced the moc functionality with some c++11 magic ?

zerr 11y ago
> So they replaced the moc functionality with some c++11 magic ? You mean they replaced moc magic with C++11?
- cjslep 11y ago
  Advanced template metaprogramming is far less common of a skill than opening a generated source file and reading source code.
  - zerr 11y ago
    Well, one of the aim of C++11 (and 14) is to make template metaprogramming much easier.
    - emsy 11y ago
      To the user it shouldn't matter wether you use a moc DSL or a template DSL. If you're a developer, you'll have to learn the innards of the moc on top of C++, which already contains templating anyway. When I first saw the signal/slot API, I wondered how they solved it and was dissappointed that they used a standalone tool for something that could've been solved with standard C++.
  - stevejones 11y ago
    C++ is a far more common skill than moc. I know C++, I can barely get my head around writing a Qt app.
    - 0xFFC 11y ago
      Exactly , Even if TMP is much harder than MOC, TMP is standard , by forcing people to learn it they are not doing something wrong or ridiculous , but by forcing people to learn something specific to framework we are doing something wrong in my opinion {then why should we have standard at first place ? every one can create its own C++ }
- fla 11y ago
  I didn't realize the c++11 standard could make it possible to get rid of moc. Defenately better this way around ;) I can't wait to see this happening upstream for Qt5.
- pjmlp 11y ago
  moc magic for signals/slots was already possible with C++98, as proven by Gtkmm already back in 2000. http://libsigc.sourceforge.net/ http://libsigc.sourceforge.net/ But it required beefy template support, which would leave out some of the compilers supported by Qt (not only MSVC).
  - roel_v 11y ago
    I think I'm not contradicting you, but I just wanted to state that I've been using libsigc with MSVC since 2002 or 2003.
    - pjmlp 11y ago
      I wanted to make a point about other compilers, as for many HNers only MSVC, gcc and clang do exist, while Qt supports many others.
pgeorgi 11y ago
> So they replaced the moc functionality with some c++11 magic ? Given that signals and slots were already implemented with templates in 2002 or before (see http://ftp.gnome.org/pub/GNOME/sources/libsigc++/1.0/ http://ftp.gnome.org/pub/GNOME/sources/libsigc++/1.0/), there's probably no need for c++11 magic.
guruz 11y ago
Indirectly related blog post: "Can Qt's moc be replaced by C++ reflection?" http://woboq.com/blog/reflection-in-cpp-and-qt-moc.html http://woboq.com/blog/reflection-in-cpp-and-qt-moc.html

by fla 11y ago

zerr 11y ago

> So they replaced the moc functionality with some c++11 magic ? You mean they replaced moc magic with C++11?

cjslep 11y ago
Advanced template metaprogramming is far less common of a skill than opening a generated source file and reading source code.
- zerr 11y ago
  Well, one of the aim of C++11 (and 14) is to make template metaprogramming much easier.
  - emsy 11y ago
    To the user it shouldn't matter wether you use a moc DSL or a template DSL. If you're a developer, you'll have to learn the innards of the moc on top of C++, which already contains templating anyway. When I first saw the signal/slot API, I wondered how they solved it and was dissappointed that they used a standalone tool for something that could've been solved with standard C++.
- stevejones 11y ago
  C++ is a far more common skill than moc. I know C++, I can barely get my head around writing a Qt app.
  - 0xFFC 11y ago
    Exactly , Even if TMP is much harder than MOC, TMP is standard , by forcing people to learn it they are not doing something wrong or ridiculous , but by forcing people to learn something specific to framework we are doing something wrong in my opinion {then why should we have standard at first place ? every one can create its own C++ }
fla 11y ago
I didn't realize the c++11 standard could make it possible to get rid of moc. Defenately better this way around ;) I can't wait to see this happening upstream for Qt5.
pjmlp 11y ago
moc magic for signals/slots was already possible with C++98, as proven by Gtkmm already back in 2000. http://libsigc.sourceforge.net/ http://libsigc.sourceforge.net/ But it required beefy template support, which would leave out some of the compilers supported by Qt (not only MSVC).
- roel_v 11y ago
  I think I'm not contradicting you, but I just wanted to state that I've been using libsigc with MSVC since 2002 or 2003.
  - pjmlp 11y ago
    I wanted to make a point about other compilers, as for many HNers only MSVC, gcc and clang do exist, while Qt supports many others.

pgeorgi 11y ago

> So they replaced the moc functionality with some c++11 magic ? Given that signals and slots were already implemented with templates in 2002 or before (see http://ftp.gnome.org/pub/GNOME/sources/libsigc++/1.0/ http://ftp.gnome.org/pub/GNOME/sources/libsigc++/1.0/), there's probably no need for c++11 magic.

guruz 11y ago

Indirectly related blog post: "Can Qt's moc be replaced by C++ reflection?" http://woboq.com/blog/reflection-in-cpp-and-qt-moc.html http://woboq.com/blog/reflection-in-cpp-and-qt-moc.html