Cloud cost structures - services vs containers

We now have some numbers (we'll share) and we see 1000 fold cost differences ... This alarm is sounding everywhere, we'll ignore it at our own peril. Analytical services in the cloud are now competing with our desktop and local clusters. It can be a huge leap forward, in reduced costs, persistence and scalability, or it can be a brick wall. We'll use numbers from RStudio and SAS Via to compare notes.

Note in this context the contrast with services as far as deployment - the Jupyter model and the inevitable compute service decoupling SAS is pursuing and RCloud is not far behind.

Persistent Store - CaViz

Persistent store on the Web is tied to domain origin (not domain path). Let's have a look to find out what this means for our development models. With that in mind, how should CaViz project, the entry point for upcoming interns and our contribution to Cancer Descriptive Statistics, be approached?

Google's Terra is calling at 11am

Recall the services vs containers cost structures earlier.

Have your workspaces warmed up ... this is our chance to travel back and forth to the GCP backend and your favorite analytical environments ... maybe even your current obcession with github actions ...

... more info forthcoming

ES6 Import/export (Nick)

(We never got it it last week ) Nick will lead a re-discussion of import/export. Note this is a repeat discussion of an excellent session led by Bhaumik, who already uses ES6 modules extensively in Connect. We should make it as hands-on as possible, this is one of the core bits of ES6 that is not sticking as cleanly as one might expect. Let's find out why it doesn't feel as much as "of the web" as the predecessor require. Is it just because node and the browser share the same V8 engine but are (temporarily?) diverging in code re-usability models?

___

Updates

https://episphere.github.io/quest

https://episphere.github.io/path

https://episphere.github.io/confluence

https://episphere.github.io/epibox

...