Expose all installation options to all install functions

[jimhester]

This should make things more clear what is available, all options should
be now exposed and documented other than those to install.packages()

Mainly to address the last bit of #129 (comment)

Just a small note that upgrade isn't a documented parameter of remotes::install_github(), nor is it a documented parameter of install.packages() (which is where the documentation of install_github says the ... are passed).

[codecov-io]

Codecov Report

Merging #156 into master will decrease coverage by 2.75%.
The diff coverage is 80.51%.

@@            Coverage Diff             @@
##           master     #156      +/-   ##
==========================================
- Coverage   92.46%   89.71%   -2.76%     
==========================================
  Files          29       29              
  Lines        1712     1847     +135     
==========================================
+ Hits         1583     1657      +74     
- Misses        129      190      +61

Impacted Files	Coverage Δ
R/install-github.R	67.7% <0%> (-27.7%)	⬇️
R/install-cran.R	94.73% <100%> (+4.73%)	⬆️
R/install-gitlab.R	92.18% <100%> (+1.27%)	⬆️
R/install-remote.R	87.12% <100%> (-5.57%)	⬇️
R/install-bioc.R	90.44% <100%> (+0.67%)	⬆️
R/install-version.R	96.15% <100%> (+0.8%)	⬆️
R/install-local.R	100% <100%> (ø)	⬆️
R/install-git.R	90.72% <100%> (+0.94%)	⬆️
R/install-svn.R	93.97% <100%> (+0.73%)	⬆️
R/install-bitbucket.R	93.15% <100%> (+0.96%)	⬆️
... and 7 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update d0c05b1...63778f9. Read the comment docs.

[gaborcsardi]

Hmmm, but then of course the default values are scattered all over, and it will be hard to change them. It would be nice to have a general solution to this, e.g. roxygen could (dynamically?) generate a section for the args that can be passed down.

Hopefully we are not creating "matches multiple arguments" errors....

[jimhester]

Yes, I agree it has downsides and was a pain to add them everywhere. It would be nice to be able to have roxygen / something else forward arguments. We experimented with this a bit in tidyverse/ggplot2#1763

I deliberately removed the default arguments to the internal functions so that we would catch multiple arguments / missing arguments, and caught a few when writing this, so I think it is ok 🤞