[NTG-context] bottlenecks

[Hans Hagen]

On 12/17/2022 1:05 AM, Rik Kabel via ntg-context wrote:

> The laptop was hot stuff in 2017 with 16GB RAM, 1TB SSD, and a 4K touch 
> screen -- today, not so hot. Due to be replaced next year.

so a bit like the surface i have as spare (which was slower than the 
2013 precission that i recently replaced by a 2018 one)

> Here are the numbers without compact fonts enabled. Not much difference.

>     mkiv lua stats  > font engine: otf 3.131, afm 1.513, tfm 1.000, 84
>     instances, 67 shared in backend, 3 common vectors, 64 common hashes,
>     load time 1.076 seconds

which means that you don't run in compact mode

> I do have some macros that are likely to make you (and others who are 
> more adept at such things) cringe. Happy to share more of the source if 
> it is helpful.

sure

> I have also run with --profile and have attached that log as well as the 
> complete execution log (zipped together as logs.zip).
here are a few:

0.613002300 s    0.138 %     0.609968531 s    0.160 %      3674 # 
prerollreference

  0.420704300 s    0.095 %     0.420702649 s    0.110 %         2 # 
analyzeregister

0.425307400 s    0.096 %     0.425299968 s    0.111 %         9 # 
action                                   1358 
...MTX/tex/texmf-context/tex/context/base/mkxl/math-act.lmt

0.681407500 s    0.154 %     0.681338138 s    0.178 %        84 # 
otftotfm                                  611 
...MTX/tex/texmf-context/tex/context/base/mkxl/font-otl.lmt

few calls, much runtime

Hans

-----------------------------------------------------------------
                                           Hans Hagen | PRAGMA ADE
               Ridderstraat 27 | 8061 GH Hasselt | The Netherlands
        tel: 038 477 53 69 | www.pragma-ade.nl | www.pragma-pod.nl
-----------------------------------------------------------------