Hi folks,





Thanks to all who helped, it looks like the Squeak/Spur64 will be faster than the pharo. It is currently at (120 Million elements parsed in ~33 minutes).



woot!





The next step is an architecture plan/ brainstorming.





What I am thinking is to take the components of the SAX Parse and wrap them in a Squeak object containing the contents and meta-information.



I then want to make those objects available to other instances of Squeak hosting Xtreams-Parsing where the contents can be parsed and converted to different formats: xHTML to start.



If http://www.squeaksource.com/AMQP/ AMQP can handle "objects" then I can use STOMP (which I used before, about 10 years ago for a commercial app involving squeak)



so...SAX(Wrap)->Wrap->AMQP(Competing consumers model) https://www.enterpriseintegrationpatterns.com/patterns/messaging/CompetingConsumers.html)  ->*  "many squeaks(Xtreams-Parsing).



And, per the "75 Gb discussion" store those things in multiple nodes according to some internal logic.



Sound reasonable?



cheers,



tty
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.squeakfoundation.org/pipermail/squeak-dev/attachments/20211021/0c85d6cb/attachment.html>