Jonathan Carroll

The raw link is what we use in the issue, though, e.g. raw.githubusercontent.com/rweekly/image/master/YYYY-MM-DD/filename.png but for a moment I thought these files used hashes - I probably uploaded some files with...

I resized images to 600px wide with a new function in rweekly.tools this week - the email images show up only ever so slightly larger than my screen width on...

It seems this one is [my fault](https://github.com/tidyverse/forcats/commit/4ece0e736f22e5cee3cd59641449430be864a69a) - I'd blame it on being a tidyverse developer day but it's a mistake nonetheless. It doesn't appear there was a test for...

If data.frame subsetting threw a decent typed error when it couldn't find a column you could hook into that, but I don't think it does, a least yet ``` mtcars[,...

There's a higher chance *those* are properly typed errors. I'm not entirely sure how to check.

(commenting because mentioned) @noamross -- that idea became [btts](https://github.com/jonocarroll/btts) but I was uncertain about the dangers of invading users privacy with that original concept (having an identifiable ping every time...

This persists even when the outliers should be discarded, [in the examples also](https://plot.ly/ggplot2/box-plots/#outliers) ```r library(plotly) set.seed(123) df

The replacement `lapply` code is then ```r p$x$data

That's my bad - I can add that entry back into the issue shortly. The workflow for resetting the draft involves our [draft template](https://github.com/rweekly/rweekly.org/blob/gh-pages/for-editor-only-draft.txt) and doesn't preserve much at all....

The entry has been restored in the issue, but I'll leave this open for discussions about how to prevent this happening again.