branch: elpa/haskell-tng-mode
commit 44dc467e07859a4ab151891f3047a3ceaf4620ad
Author: Tseen She <ts33n....@gmail.com>
Commit: Tseen She <ts33n....@gmail.com>
notes on motivation, contributing and collaboration
---
README.md | 22 +++++++++++++---------
1 file changed, 13 insertions(+), 9 deletions(-)
diff --git a/README.md b/README.md
index 026f8e1..438bb4d 100644
--- a/README.md
+++ b/README.md
@@ -1,30 +1,34 @@
# Haskell Mode: The Next Generation
-This is an exploratory alternative to
[`haskell-mode`](https://github.com/haskell/haskell-mode/) that answers the
question *how would we support Haskell in GNU Emacs if we started today?*
+This is an exploratory alternative to
[`haskell-mode`](https://github.com/haskell/haskell-mode/) that answers the
question *how would I support Haskell in GNU Emacs if I started today?*
## Why?
+`haskell-mode` has been such an important tool for me that I feel the need to
explain why I am rewriting it. I have the greatest of respect for its
contributors and I have used it for many years.
+
In [Lessons from 6 Software
Rewrites](https://medium.com/@herbcaudill/lessons-from-6-software-rewrite-stories-635e4c8f7c22),
the author concludes *avoid rewrites and make incremental improvements
instead, unless you want to a) remove functionality or b) take a different
approach*.
### Remove Functionality
-`haskell-mode` is almost 30 years old and has accumulated more than 25,000
lines of code aimed at a wide variety of users from academics to industrial
software engineers. We choose to focus on the requirements of the industrial
engineer, removing features that are deeply embedded in the design of the
original codebase.
+`haskell-mode` is almost 30 years old and has accumulated more than 25,000
lines of code aimed at a wide variety of users from academics to industrial
software engineers. I choose to focus on the requirements of the industrial
engineer, removing features that are deeply embedded in the design of the
original codebase.
### Different Approach
During those past 30 years, the GNU Emacs ecosystem has evolved to provide
many features that `haskell-mode` independently implemented, such as
[`projectile`](https://github.com/bbatsov/projectile),
[`comint`](https://masteringemacs.org/article/comint-writing-command-interpreter),
[`highlight-symbol`](https://melpa.org/##/highlight-symbol),
[`pretty-symbols`](https://github.com/drothlis/pretty-symbols),
[`company`](http://company-mode.github.io),
[`yasnippet`](http://joaotavora.github.io [...]
-We choose to use idiomatic libraries to provide features, rather than building
ground-up solutions.
+I choose to use idiomatic libraries to provide features, rather than building
ground-up solutions. Many shared libraries exist to generalise solutions that
were first proposed in `haskell-mode`!
## Goal
-The goal of this friendly rewrite is to produce software that any Haskell
developer can use, understand and build upon ([Emacs
Lisp](https://www.gnu.org/software/emacs/manual/elisp.html) is fun to learn).
+I want a Haskell editing experience that lets me efficiently write commercial
Haskell, and offers me as much automation as I am happy with.
+
+I believe in helping my neighbours, so this codebase is available publicly and
I have included installation instructions. But I do not want the burden of
users.
-This can be achieved in a small codebase with zero dependencies and high test
coverage, targeting
[Haskell2010](https://www.haskell.org/onlinereport/haskell2010/).
+The goals can be achieved in a small codebase with zero dependencies and high
test coverage, targeting
[Haskell2010](https://www.haskell.org/onlinereport/haskell2010/).
-Old versions of `ghc` and extensions to the Haskell language may not be
supported, to reduce the complexity of the codebase. For example, [literate
Haskell](https://wiki.haskell.org/Literate_programming) will not be supported,
and `ghc` language extensions must be justified on a per-case basis. We are
sympathetic to language extensions that are popular in the free software and
commercial ecosystems.
+Old versions of `ghc` and extensions to the Haskell language may not be
supported, to reduce the complexity of the codebase. For example, [literate
Haskell](https://wiki.haskell.org/Literate_programming) will not be supported,
and `ghc` language extensions must be justified on a per-case basis. I am
sympathetic to language extensions that are popular in the free software and
commercial ecosystems.
-If it is possible to implement a feature using another minor mode, or command
line tool, then we would prefer not to accept the feature.
+If it is possible to implement a feature using another minor mode, or command
line tool, then I would prefer not to add the feature.
## Install
@@ -99,9 +103,9 @@ Third party Haskell tools must be installed separately (e.g.
via `cabal v2-insta
## Contributing
-Bug reports and feature requests are a source of anxiety for maintainers, and
encourage an unhealthy customer / supplier relationship between users and
contributors.
+Bug reports and feature requests are a source of anxiety for me, and encourage
an unhealthy customer / supplier relationship between users and contributors.
-Instead, and following the [anarchical spirit of
Haskell](https://www.microsoft.com/en-us/research/wp-content/uploads/2016/07/history.pdf),
we encourage discussions and debate around code contributions. Merge requests
can be raised by anybody and discussed by anybody, and do not need to be
complete. An automated test is the only way to report a bug. If the maintainers
are convinced by the technical merit and quality of a proposal, they may accept
it.
+If you wish to contribute, the best thing to do is to let me know about your
copy of this repository and we can take it from there. We may both chose to
collaborate in one place.
To run the tests, install
[`cask`](https://cask.readthedocs.io/en/latest/guide/installation.html) and type
