Missing plots for new benchmarks

Missing plots for new benchmarks

[Ludovic Courtès]

Hello Neil,

There are no plots for the new benchmarks that were added recently
(‘structs.bm’ and ‘vlists.bm’).  Could you add them?

Optionally, could you add the benchmark title to the HTML code itself so
that they can be searched from the browser?

It’s really pleasant to have all these plots, especially to notice that
most of them have been going down lately.  :-)

Thank you!

Ludo’.

Re: Missing plots for new benchmarks

[Neil Jerram]

ludo@gnu.org (Ludovic Courtès) writes:

> Hello Neil,

Hi Ludo!

> There are no plots for the new benchmarks that were added recently
> (‘structs.bm’ and ‘vlists.bm’).  Could you add them?
>
> Optionally, could you add the benchmark title to the HTML code itself so
> that they can be searched from the browser?

I've made these changes now.  I also removed the hardcoded Y axis range,
as I think it was obscuring useful information (especially now that some
of the benchmarks are notably faster than when we started collecting
these results).

Please let me know if you have any comments, or can think of further
improvements...

> It’s really pleasant to have all these plots, especially to notice that
> most of them have been going down lately.  :-)

I agree!

(BTW, I really am hoping to be more available again for Guile work
soon.  Apologies all round for my general lack of contribution
recently...)

Regards,
        Neil

Re: Missing plots for new benchmarks

[Ludovic Courtès]

Hi Neil,

Neil Jerram <neil@ossau.uklinux.net> writes:

> ludo@gnu.org (Ludovic Courtès) writes:

>> There are no plots for the new benchmarks that were added recently
>> (‘structs.bm’ and ‘vlists.bm’).  Could you add them?
>>
>> Optionally, could you add the benchmark title to the HTML code itself so
>> that they can be searched from the browser?
>
> I've made these changes now.

Great, thanks!

> I also removed the hardcoded Y axis range, as I think it was obscuring
> useful information (especially now that some of the benchmarks are
> notably faster than when we started collecting these results).

Nice.

> Please let me know if you have any comments, or can think of further
> improvements...

Some of the benchmarks in vlists.bm are comparative: the idea is to see
how vlists compare to raw lists.  In these cases, it would be nice to
see the benchmarks being compared use the same Y scale.  However, it may
be inconvenient or inelegant to customize the benchmark machinery for
this.

> (BTW, I really am hoping to be more available again for Guile work
> soon.  Apologies all round for my general lack of contribution
> recently...)

Heh, that’s perfectly fine with me!

Thanks,
Ludo’.