Build Figure 3: query interface, framework pipeline, and technical benchmark

[stemangiola]

The figure 3 of the manuscript

https://docs.google.com/document/d/11zGOnMKONZzAEnfzmf-wriqWIvqAcPx3lpws4MSl20w/edit?usp=sharing

will include

• The flowchart of the backend process of querying and file organisation
• The vignette of the query interface (how it looks) with an example
• The technical benchmark of H5 framework performance

[multimeric]

Should I make this figure with the assumption that we will not be using a webserver, and everything will work as it does now using a filesystem?

[stemangiola]

Should I make this figure with the assumption that we will not be using a webserver, and everything will work as it does now using a filesystem?

The assumption is that our architecture choice will work as a remote service. I don't think that we need a webserver with our architecture. For example if R points to a remote location for metadata.sqlite, and to remote location for file download, we could achieve this with any remote storage accessible with command line. Julie mentioned she knows free resource that have R API. Or a server (NECTAR that hosts the file with accessible addresses), without the need to be a webserver.

Am I missing something?