See r-lib/usethis#193 (comment)

 #' * Experimental: very early days, a lot of churn in search of a good API. 
 #'   Not on CRAN. Might not go anywhere. Use with care. 
 #' * Maturing: API roughed out, but finer details likely to change. Will strive 
 #'   to maintain backward compatibility, but need wider usage in order to get 
 #'   more feedback. 
 #' * Dormant: not currently working on it, but plan to come back to in the 
 #'   future. 
 #' * Stable: we're happy with the API, and the unlikely to be major changes. 
 #'   Backward incompatible changes will only be made if absolutely critical, and 
 #'   will be accompanied by a change in the major version. 
 #' * Questioning: no long convinced this is a good approach, but don't yet 
 #'   know what a better approach might be. 
 #' * Retired: known better replacement available elsewhere. Will remain 
 #'   available on CRAN. 
 #' * Archived: development complete. Archived on CRAN and on GitHub. 

Need to replace href="#" with href="{{ .RelPermalink }}"

The author of each article displays in the index but not in the article itself. Is that intentional?

CC BY-SA ?

They do this with the travis-ci docs, and it catches a number of errors.

Looks like they use https://github.com/gjtorikian/html-proofer, we could do something similar here.

e.g. google analytics

• Store metadata in yaml
• Automatically remove once done
• Only show next two

i.e. name and describe key contributors, then include a list of all contributors:

x <- gh::gh("/repos/:owner/:repo/issues", owner = "r-lib", repo = "testthat", since = "2016-04-23", state = "all", .limit = Inf)
users <- sort(unique(purrr::map_chr(x, c("user", "login"))))
length(users)
clipr::write_clip(glue::collapse(glue::glue("[\\@{users}](https://github.com/{users})"), ", ", last = ", and "))

Currently it says "There are a handful of other packages that are not in the tidyverse, but are tidyverse-adjacent. They are very useful for importing data from other sources:" and lists jsonlite, xml2, httr, rvest, and DBI under this. It's unclear to me what the difference of these from readxl and haven is since all of these packages are installed with tidyverse but are not in the core. It might be more clear to list all in the same hierarchy (as is the case in the tidyverse repo README).

Appears not to be tracking. Trying to figure out why…

And add note in readme on how to do so

We've just launched a tidyverse course (https://www.jumpingrivers.com/courses/22_r-tidyverse)

Would it be appropriate to list it at https://www.tidyverse.org/learn/?

Thanks

Maybe: All packages share a common design philosophy, grammar, and data model.

(Previously incorrectly filed at tidyverse/tidyverse#107)

And generally about how to let us know that an issue is important to you.

And don't comment on closed issues.

Idea for a blog post in the programming category.

Iterating over rows in a data frame ...
This is a common problem where there is little consensus re: best way to proceed.

Capturing a few snippets of Slack discussion to jog our memories:

• Do you really need to iterate over rows? Are you overlooking a vectorized solution? Applies especially to string manipulation. (@hadley this is related to the question of whether to attack a problem row- or column-wise)
• What about pmap()?
  • Ways to define .f (explicitly + globally, explicitly + inline + anonymously, the ~ shorthand)
  • Tension in pipelines when multiple functions are operating on the tibble (e.g. pmap() inside mutate()), competing meanings for .
  • Arguments of .f: names vs position, absorbing unused args with ...
  • Use rlang::call_inspect() to see what's going on, e.g. purrr::pmap(mtcars, rlang::call_inspect)
  • pmap() is expensive because it disaggregates the tibble. E.g. you don't want to use it naively with a social survey tibble with hundreds of columns just to work on a few variables.
• What are you trying to optimize? Performance? Readability? No dependencies? Pipe-ability?

@LucyMcGowan Let's develop this to a certain point elsewhere, but this will eventually give you the satisfaction of closing an issue when it's done!

Currently, the Contents sidebar is only populated for posts written directly in markdown, e.g.

For posts written in .Rmd, headers are not added to contents. It's possible that this looks particularly weird right now because there are also no upcoming events, but I think it would also make the articles more "browseable."

n.b. Can get TOC to show up at the top of the post with the code below, but that's not the desired behaviour.

output:
  blogdown::html_page:
    toc: true

• rstudio::conf(2019)
• https://www.rstudio.com/workshops/extending-the-tidyverse/

At the end of Packages page of tidyverse.org, how about adding a "Domain Specific tidyverse Friendly Packages" (or something to that effect) section that includes list of domain/vertical-specific packages that play nice within tidyverse ecosystem? Examples:

• @rudeboybert, @ismayc, @jchunn's fivethirtyeight under "Data Journalism" and "Education"
• @walkerke's tidycensus under "Demography"
• @mdancho84 and @DavisVaughan's tidyquant under "Finance"

and eventually domain-specific packages with reverse depends/imports/suggests of tidyverse

tidyverse/tidyverse#108

They should feel more like italics than bold.

@hadley this is the section I was confused about in terms of hex placement yesterday— I didn't want to mess up the look.

Now that we have a bunch of blog posts, do we need to rethink the categories?

Most of the posts are about packages. Do we need subcategories? Like new package/major release/minor release?

https://twitter.com/martinjhnhadley/status/981494229787607040

Would be nice to have a "short"/tumble post type where you just want to link to another website with a bit of context.

Per discussion w/ @jennybc & squad, https://www.tidyverse.org/lifecycle/ should probably have a place to live that you can get to without clicking on a lifecycle button…

Possibly related to e8fcdfb (though I'm not sure how), but there's a Read More link at the bottom of the articles page that goes to an un-styled version of Jenny's workflow post at https://www.tidyverse.org/articles/1/01/.

Visual depiction:

I am skeptical that this actually captures users who only added comments (vs. opened an issue or PR). Wording of recent blog posts suggests that this is the case. If that is our the intent, we should check this. FWIW I have such a script from STAT 545 and it is not nearly this concise:

https://github.com/tidyverse/tidyverse.org/blame/a68f8ff291f4a9c73b97cf17d658043a3c0617b6/CONTRIBUTING.md#L76

Gleaned from achiever

Another possible post on workflow and setup. What are .Rprofile and .Renviron for and, more importantly, what's safe to do there vs. unsafe.

e.g. http://ggplot2.tidyverse.org

Link to pkgdown in bottom-right corner of page is broken (http://hadley.github.io/pkgdown/ gives a 404 error)

Per r-lib/usethis#236 (comment) and team discussion

#79 (comment)

A place to park photos we come across.

Currently, the RSS feed for the articles looks like this, note that the links are of the format <link>/articles/2018/03/pillar-1-2-1/</link>. These links seem to work fine in my regular RSS reader (Reeder.app, macOS), but if you import the feeds in e.g. Slack, the links look like this:
Unclickable and sad.

I've looked through the page's source and saw that you seem to use Hugo's default template, so I looked at that here, which says the RSS feed should use the .Permalink element. So I tried building the site locally but using baseurl = "https://tidyverse.org/ in config.toml, but it didn't change the RSS links. I also tried changing the relativeurls option, also without luck.

The RSS feed on my blogdown blog does have absolute/full URLs, and I tried using the same config with relativeurls and baseurl, but I can't seem to reproduce the desired behaviour.

Long story short: I would like to see full URLs in the articles RSS feed, but I can't figure out how to PR it myself :(

now happening through tidytemplate update…

• tidyverse.org
• dplyr (tidyverse/dplyr#3453)
• tidyr (pending tidyverse/tidyr#440)
• ggplot2 (see tidyverse/ggplot2#2487)
• broom
• magrittr
• purrr
• readr
• tidyverse (pending tidyverse/tidyverse#123)
• readxl
• lubridate
• reprex
• modelr
• glue
• stringr (pending tidyverse/stringr#227)
• haven
• forcats (PR tidyverse/forcats#120)
• tibble
• dbplyr
• googledrive
• style
• hms
• tidyselect
• blob
• tidytemplate
• googlesheets4

Add same to referral exclusion

• Subdomains to referral exclusion

I believe this is because the category specified in the css is "package", whereas the category specified in the articles is "packages":

I suppose it'd be easier just to change the css? I'm happy to do the 📦 PR, just wanted to run this by you first.

Maybe every month?

Using pkgdown