search

ropensci / RNeXML

Implementing semantically rich NeXML I/O in R
https://docs.ropensci.org/RNeXML
Other
13 stars 9 forks source link

Adds pkgdown site, appveyor, updates readme badges #223

Closed cboettig closed 5 years ago

cboettig commented 5 years ago

First steps as discussed in #222.

Still needs the larger step of refactoring the README to a minimalist description complemented by an introductory vignette.

cboettig commented 5 years ago

ah, good point, that was an old local note to myself from our previous discussion. I've dropped that now from the commit and allowed the vignette to build (dropping eval=FALSE).

I've also moved the more detailed use-focused parts of the README over to a separate quickstart vignette, and added links to all the vignettes in the README file.

hlapp commented 5 years ago

Nice! So, is the pkgdown website built automatically (by appveyor? not sure I understand in full what that part is doing 😳) or does that need to be done by hand for now?

cboettig commented 5 years ago

it's manual for now, we could automate it with tic via travis (would need to set up a deploy key and a GitHub machine account with API key I believe).

Appveyor is a Windows-based CI platform. Probably not crucial since we're not compiling code and R is otherwise pretty portable, but I've seen a few edge cases (usually around paths or dependencies) and so most rOpenSci packages try and check on both Travis and Appveyor. (Just adding it now when updating the badge list).

hlapp commented 5 years ago

Apparently taxize is now broken with regard to how it implements remote API call throttling. Very annoying.

cboettig commented 5 years ago

@hlapp ah, is that what's happening, I see. Kinda happy about this, I haven't liked the way the taxonid method works anyway and having been meaning to replace it. No need to make a separate API call for every taxon in a large tree (and deal with API limits, timeouts etc). I've been working on a project with Scott and others that caches a local name database and just does the look-ups as table joins, https://github.com/cboettig/taxald, though it's not quite ready for onboarding and CRAN yet. This'll motivate me to swap out the taxize method for this database. (Tangential but would love your input on some of the possible strategies in taxald at some point; currently, contrary to the name, it just creates a local SQL database in MonetDBLite, or can query in a in-memory cache for setup-free use. But am also exploring how best to turn each of the database dumps into RDF triples instead...)

hlapp commented 5 years ago

FYI, I created a PR for crul, which taxize uses under the hood: ropensci/crul#95