Re: [boost] [Review] Formal Review of Proposed Boost.ChronoLibrary LAST DAY TODAY

Subject: Re: [boost] [Review] Formal Review of Proposed Boost.ChronoLibrary LAST DAY TODAY
From: vicente.botet (vicente.botet_at_[hidden])
Date: 2010-11-15 16:10:52

----- Original Message ----- 
From: "Paul A. Bristow" <pbristow_at_[hidden]>
To: <boost_at_[hidden]>
Sent: Monday, November 15, 2010 5:06 PM
Subject: Re: [boost] [Review] Formal Review of Proposed Boost.ChronoLibrary LAST DAY TODAY

>> - What is your evaluation of the design?

> Needs to provide examples showing how to use IO output (input?) Hr, Min, sec
> in languages *other than English*.
> I'm not sure the mechanism provided will work in practice?
> Especially if using locales and expecting it to work with them?
> This is such a basic need, that leaving it to the users to reinvent
> repeatedly seems unhelpful.

The example Howard has provided is quite explicit and show how we can manage with this problem with the interface provided. I'm looking for something more transparent, but I'm not aware of all the locale mechanism, so I don't know if transparency can be achieved. Anyway, I will be grateful if someone have some ideas to share.
 
> More examples are always helpful.  I didn't find all the ones referenced,
> but this may be a filing problem using sandbox.

Let me know which  references are missing. I will correct it.

> Did I miss examples of actually using Chrono with Boost.Units library?

The documentation of Boost.Ratio contains an example compining other units than the ones provided by Boost.Chrono, but not units provided by Boost.Units.
 
>> - What is your evaluation of the documentation?
> 
> Good - a few typos (sent separately).  

I have almost included all the suggestion you sent me. Your help has been considerable.

> I would have used Doxygen reference
> to reduce future maintenance of docs.
> 
> I'd like it *with Doxygen comments in the code*:  without this, reference
> docs can be hard to digest.
> And as a user (including of my own code!), I also find it very useful to
> have a standalone Doxygen.
> Again the  *with Doxygen comments in the code* makes it much more useful.
> Adding Doxygen comments in the code is easiest when writing the code, but
> perhaps a volunteer could add afterwards?

Yes, I know that  Doxygen  can help with the reference part, but up to know I've prefered to write manualy the documentation. I have always the impresion that the tools constraint more than help. I think I would give a try to doxygen, and see how it works.

>> - Did you try to use the library? With what compiler? Did you have any
> problems?
> Using MSVC 10 but without Boost.System library as included. 
> Puzzling compile problem that I didn't have time to investigate.

Yes, we need to have a coherent set of packages. Boost.Chrono couldn't be a header only library if Boost.System is not. I will try to reproduce, to see what is wrong.

>> - Do you think the library should be accepted as a Boost library?
> 
> Yes.

Thanks Paul. 
Best,
Vicente