Re: Benchmark results on mdds::multi_type_vector

On Friday 13 of December 2019, Kohei Yoshida wrote:
> On 13.12.2019 05:43, Luboš Luňák wrote:
> Sure, but that's just as much of a speculation as my own interpretation.
>   To be fair, it is possible that you are right, and I am wrong.  But I
> did provide my own interpretations of those numbers based on my own
> experience and educated guesses.  I'm not claiming that I'm right, but
> I'm claiming that what I concluded in my post is my truly honest,
> hopefully reasonably researched opinions.

 Sure, I'm not questioning that. What I'm wondering is how much your educated 
guesses will match the reality of the wide range of Calc use cases, which may 
be something so hard to guess that we won't know until it gets measured for 
real.

> So, I'm a bit concerned about your use of the word "artificial" to
> describe my benchmark, because that word implies that I somehow made
> those numbers up.

 You didn't make up the numbers, but you made up the benchmark (even if based 
on an educated guess), so there's still the question of how much it will 
match various Calc usage. There is some chance that updating to this mdds 
version will regress in some areas and not provide noticeable improvements 
because of position hints use. I don't think there's any other way than to 
actually try it.

> And, while I would love to craft another benchmark test involving the
> entire Calc piece, I'm afraid I won't have enough bandwidth to do that.
> Even running this benchmark on mdds alone took me one month to do it
> end-to-end.  It would be nice to have someone else chip in and conduct
> another, more through and satisfactory benchmark test, if anybody is
> interested.

 We already do have some tests at https://perf.libreoffice.org/ , although I 
don't know if there's any other way to use those than to commit to master.

 (Also, given the last update date it's also not watched by anyone).

-- 
 Luboš Luňák
 l.lunak@collabora.com