Fwd: Bug 89387 - improve performance for some matrix operations

[Kohei Yoshida]

On Mon, 2016-11-28 at 08:02 +0100, Jan Iversen wrote:
> Forgot reply-all
> 
> > Begin forwarded message:
> > 
> > From: Jan Iversen <jani at documentfoundation.org>
> > Subject: Re: Bug 89387 - improve performance for some matrix
> > operations
> > Date: 28 November 2016 at 07:57:07 GMT+1
> > To: Filippo giacchè <filippo.giacche at gmail.com>
> > 
> > Hi
> > 
> > > --1 has anybody some material that can help me to understand mdds
> > > library ?(some example, some pdf ……)
> > Google if your friend, first hit on a search is:
> > http://kohei.us/2012/07/20/mdds-multi_type_vector-explained/

That blog post is a little outdated now.  For multi_type_vector, I've
written a more up-to-date one over here:

http://kohei.us/files/mdds/doc/multi_type_vector.html

The bug is about the performance of Calc's internal matrix class, which
uses mdds::multi_type_matrix.  I'm afraid the documentation on that is
still largely work-in-progress, and right now we only have just some
API references:

http://kohei.us/files/mdds/doc/multi_type_matrix.html

That said, internally multi_type_matrix is a very thin wrapper around
multi_type_vector, so much of what is discussed on multi_type_vector
still applies there.

I'm also interested in some feedback on what sort of documentation,
examples, tutorials would be useful to new users of mdds as well as the
seasoned ones.  I may not be able to get something written quickly, but
knowing the areas where improvement is needed most would be a good
first step.

Thanks,

Kohei