Subject: [ggl] [quickbook] Algorithms quickbook sample
From: Barend Gehrels (Barend.Gehrels)
Date: 2010-02-18 15:26:13
>> - " properly by following instructions..." actually the only thing
>> necessary is to set the include folder, which is indeed written there,
>> so we can lower this a bit probably
> I've tried to update, see if it's OK now.
Yes, this is better.
Eeehm, what occurs to me now, do other incorporated boost libraries have
this section? Because if boost is installed, Boost.Geometry will be
>> - compilers, there are more now. It is also compiling on CodeWarrior (at
>> least the intersections). What is our strategy here? It probably must
>> run all unit tests. OK, then we cannot mention that one. I tested MSVC
>> 2010 express beta, but didn't do all unit tests either...
> Hmm, good question. What about maintaining two categories/lists:
> 1. Compilers confirmed as building full set of tests and examples
> 2. Compilers users mentioned but without providing further details
> what extent of BG was built with it
Seems good to me. You see it more, "compilers reported to be working" or
something like that.
>> - Boost Geometry uses Boost Build, I would mention here that this is for
>> unit tests. For users, it is really not necessary to use Boost Build
> I believe it says so:
> Boost Geometry uses Boost Build, ..., to configure, build and execute
> unit tests and example programs.
Yes, but this is not exactly true... I'm not usint Boost.Build for
example, besides occasionally. Actually unit tests are not interesting
for end uses and can probably be omitted.
The thing is, I don't want to force users to install Boost.Build etc
etc, while it is really not necessary at all. As long as they make an
example somehow and just have boost in the include path, everything works.
If they want to use Boost Build, ok, fine to me. But then it should be
stated differently, e.g. "If you want to use Boost Build, jamfiles for
building examples using Boost Build are provided".
>> - However, I would include the "Intellisense issues" from the original
>> page, because that bug really can be a irritating, reviewers mentioned
>> this and probably were glad by this hint
> I believe it is included:
OK, yes, sorry, didn't see that one. Great.
> So, you mean table with one header per row and two columns? This way?
> [ header | description ]
> [ .../*.hpp | ... ]
Yep, sounds goot to me.
-------------- next part --------------
An HTML attachment was scrubbed...
Geometry list run by mateusz at loskot.net