Re: [boost] GSoC Boost.Multiprecision project

Subject: Re: [boost] GSoC Boost.Multiprecision project
From: Jyotesh Choudhari (jyoteshrc_at_[hidden])
Date: 2013-05-01 09:25:23

Hello,

> Hi Jyotesh.

>>>> Looking at the project description, I think
>>>> I have understood what needs to be implemented
>>>> in this project.

> Yes, you seem to have a good grasp of the issues.
> Good work.

>>>> I have also gone through the current
>>>> code cpp_bin_float.hpp and I have understood some
>>>> part of it. I want to know how
>>>> detailed should my proposal be? What details
>>>> should I include in my proposal?
>>>> What should be the deliverables for the project?

> Go ahead and add some details from cpp_bin_float
> if you like. You can mention that this is a potential
> solution method.

Thank you for your valuable feedback. I have written project proposal for
the first time in my life. So, I did not know what details to include. I
have made some changes to my proposal as per your suggestions.

> Other than that, I believe you have found
> the right level of detail, with a few more suggestions
> from my side below.

>> I have attached my proposal with this mail.

>> Please have a look at it and let me know your feedback.
>> Thanks and regards,

> I suggest you move the FFT algorithms to an optional
> part, add a bit more time for the elementary transcendental
> functions, and add some time for testing the various
> milestones in the project.

> After that, you should start submitting your proposal.
> As far as I know, you can modify your proposal up
> to May 03.

I have written FFT algorithm as an optional algorithm now and made some
changes to schedule. Please let me know if it is fine now.

> Keep up the good work.
> Sincerely, Chris

I was studying the code for cpp_bin_float and I have some doubts in my mind.

1. What are eval_ops_signed_type, eval_ops_unsigned_type and
eval_ops_float_type?

2. What is the use of following lines?

typedef mpl::list<eval_ops_signed_type>   signed_types;
typedef mpl::list<eval_ops_unsigned_type> unsigned_types;
typedef mpl::list<eval_ops_float_type>    float_types;

These are not used in the class elsewhere. Similar lines are present in
cpp_dec_float.hpp also and not used elsewhere in the class.

3. In the function from_float_type, I think the variable number_of_digits
should be initialized to zero before using it in while loop.

Lastly, I have seen on mailing list, that some of the students are
submitting code samples. As I am busy with my end-semester exams and R&D
project, I am not able to send any sample code. As soon as I get free, I
will send it.

Thanks and regards,
Jyotesh Choudhari