yep today -> 

- got up to late see my mom already had mopped first floor and I am unhappy for not having helped house tasks for today. 

- had rested some before continuenig to task of- > yayyy first task today is getting expert in epics/monics/pullbacks to revisit some chapter 5 paragraphs. 

then second task is designing the for ever and there exists via topos as mentioned and its additional and/or traits addition(merging) mechanism.  a partial DSL task not that i start detailed DSL study yet but  this is just for exercise to get necessary enough expertise in monics/epics/pullbacks for such DSL logic quantifiers.  although there yet very ignorant of other logic systems but there exists and for every is some most basic quantifiers. eg. we would also need to check intuitionistic maths logic systems of which is of paramount importance to project of course. 

yep lets now revisit epic/monic/pullback topics to get expert to then revisit unstudied paragraphs of chapter 5 which required such expertise. 

yayyy then lets temporarily switch to partial DSL task of defining some topos based quantifier definitions and extending traits with such diagrams.

e.g. set {for every r  of R that  has trait A and B or C }  

or set {there exists r of R that  has trait A or C }   alike definitions via topos diagrams 

hmm there exists also task of how to define the topos diagrams in computer. 

should we use knowledge graph tech? (neo4j and cipher? ) 

that has had some issue of that its not sharded well that you can not index graph very well although indexing in graphs is a hard problem. 

but again imho lets not think of that task (of representation of graphs (category theory diagrams) ) design yet. most possibly initially i would have some tech of neo4j like dbs. 

also it could be some tech alike that big data wise its stored in some columnar format in parquets but then only relevant part is taken to cache which is stored/queried with neo4j alike. that main storage of course big data storage in some either parquet or so storage format but the cache (CPU cache metaphor) of knowledge graph would be in direct avail to neo4j db. hmm did not liked this. i think i would do develop some big data version of neo4j cipher alike. that it would be stored in parquet or so and queried with some query sets that is analogical to cipher and analogical to graph systems to that define the distributed graph systems there of universe of knowledge. 

hmm nevertheless this is not the constraint nor topic of current tasks. I actually came upon similar task but not fully in parsing TTLs (readable by Apache Jena but i had to modify Jena source to be able to read it via spark distributed wise) in big data wise and storing in parquet wise but had not invested necessary design to that task yet. imho these two tasks are related. We need to create a big data version of category theoretic graphs. that denormalized dbs that is fastly queriable and inference systems to fastly do inferences either.  

but this graph db part is currently unnecessary thinking currently since its not the current task set. maybe few weeks later tasks these kind of tasks. we need to first create the DSLs and inference/deduction and combinatoric wise thinking capabilities first. even ebfore that we need to also integrate NLP to DSL translation. 

hmm so task of graph db technology is not even few weeks later task, maybe 3 weeks later task earliest.