Re: [boost] Can anyone take a quick read of an array type?

Subject: Re: [boost] Can anyone take a quick read of an array type?
From: Paul A. Bristow (pbristow_at_[hidden])
Date: 2013-05-10 05:38:57

> -----Original Message-----
> From: Boost [mailto:boost-bounces_at_[hidden]] On Behalf Of Daryle Walker
> Sent: Friday, May 10, 2013 2:03 AM
> To: Boost Dev-List
> Subject: Re: [boost] Can anyone take a quick read of an array type?
> 
> > From: pbristow_at_[hidden]
> > Date: Thu, 9 May 2013 09:46:54 +0100
> >
> > > -----Original Message-----
> > > From: Boost [mailto:boost-bounces_at_[hidden]] On Behalf Of
> > > Daryle Walker
> > > Sent: Thursday, May 09, 2013 9:28 AM
> > >
> > > > From: pbristow_at_[hidden]
> > > > Date: Wed, 8 May 2013 12:02:47 +0100
> > > >
> > > > > -----Original Message-----
> > > > > From: Boost [mailto:boost-bounces_at_[hidden]] On Behalf Of
> > > > > Daryle Walker
> > > > > Sent: Saturday, May 04, 2013 11:05 AM
> > > > >
> > > > > I banged out a little something at <https://github.com/CTMacUser/ArrayMD>.
> > > >
> > > > Gulp. This would seem to be quite a BIG something?
> > > >
> > > > And that I have yet to digest :-(
> > > >
> > > > You mentioned docs but I haven't found them in the link above.
> > >
> > > No, the other guy had docs. My library doesn't have any besides
> > > Doxygen comments. But those comments tend to be huge. I write the
> > > Doxygen comment after the declaration but before the definition.
> > > Doing those comments and figuring out a test take a lot longer than
> > > the actual code! I've added the ordered operators (< > <= >=) and the tuple interface (get,
tuple_size,
> tuple_element).
> > > The latter assumes increments of value_type, since users would be
> > > confused if I picked any other (I.e. sub-arrays) unit. So it should
> > > be substitutable for std::array, when given 1 dimension. (If we can
> > > find a std::array demonstration program, maybe it can be applied
> > > against array_md.) I just recently added an example. It just a quick
> > > & dirty matrix class. It demonstrates in-sync updating of iterators
> > > from two different objects (of the same shape) for addition and subtraction. The core
multiplication
> routine shows use of "apply."
> >
> > Well I'm a great fan of good Doxygen comments - and you could use
> > those with Quickbook to produce some really smart docs.
> >
> > This isn't usually possible because people don't start by writing the
> > Doxygen comments, then the code, so the prospect of going back writing
> > the comments is too much to contemplate...
> >
> > If you need help with setting this up, please ask - I've done this
> > before. Once working it is painless to write the text for introductions and tutorial in
Quickbook.
> Yeah, I've had to move Boost development to another computer, this time a Windows (8) one. This
time,
> there are too many options for development, because GCC punted to UNIX compatibility environments
> instead of direct usage within Windows. I've so far settled for a turn-key solution with
Code-Blocks
> bundled with a private version of MinGW and GCC 4.7. I've tried to build CLang, but I failed since
it's not
> turn-key. I'm obviously not using Boost.Build here right now, so I need to make either Cygwin
and/or
> MinGW usable. I got Python, Perl, and CMake, but in their Windows-GUI versions (and not
> Cygwin/MinGW CLI modes).
> I also got Visual Studio Express 2012 (for Desktop, Win8, and Web), and the November CTP, but I
have
> no idea how to use them. I didn't try since my code needs C++11 severely.
> > You can also use the snippets option to include (selected parts of)
> > the example(s) in the text, ensuring that the code shown has actually compiled and run.
> I've done that once. I'm not sure if it's been published or not.
> > I'll take a look at your example asap.

I don't think this is relevant to using Quickbook to build the docs, for which you just need to
build the Quickbook.exe, install Doxygen, write a small amount of Quickbook intro, and add the
Doxygenated source locations,  and some setup in your user jamfile.

If I get time, I'll have a go using your sources and test and example.

Paul

PS Aside, I have used NetBeans on Windows 8 to run GCC, Clang (using pre-built binaries) and VS 2012
compiler (called from NetBeans).  This allows you to switch compiler slickly.  From a brief
skirmish, I decided that I liked NetBeans more than CodeBlocks.