The Sigils WAR, Language Design and Coding Convention
2020-07-29 a story of why i have dollar sign in front of my variables.
Emacs Lisp Sigil War! Choose Your Side!
sigil war on reddit
sigil war on reddit. https://www.reddit.com/r/emacs/comments/4s499h/emacs_lisp_sigil_war/
voice your thoughts there.
my reasons for sigil, is Sigil for Variable Names
Sigils of the Clojure clan
Ξblood of the Φsigils
an eulogy to sigils ξ φ
(while reading this piece, you need to listen to Chopin, Funeral March 📺 )
feel pretty sad about to remove all my sigils. There are some 10 packages, thousands of sigils. Feel pretty sad, because in the process i felt something is lost. Sigils provide many linguistic advantages. For one thing, it is easy to change sigils φ ξ to the versions without, but hard to do the reverse. Precisely because the sigil'd variable provides a syntactic property that's advantageous.
ξsigil, is like a syntactic type. Sigil to syntax is types to semantics. #haskell #ocaml
i'm in the process of removing sigils on the third package, but having second thoughts. Yeah, convention didn't like it, but that's because convention isn't spiffy. 2 parts. ① There's a natural aversion, because most people have problem typing the thing, especially western people, because it's foreign and they haven't encountered input system (as if u need to type Chinese). Most emacs dev, still really want ASCII. (they are already being outdated by current crop of programing languages now, which embrace unicode chars.) ② the linguist issues: UUID, Referential Transparency, Point-Free Function Syntax, Combinatory Logic, Hygienic Macro. I outlined here Programing Style: Variable Naming: English Words Considered Harmful but most programers no unstand.
i resent wasamasa putting a damnation in there that got all upvotes to the top. (of course) Hackers are the worst enemy to programing progress, as Douglas Crockford and other observers have said.
perhaps, a comforting thought is that, if i want to do this, perhaps do it in ocaml haskell or create my own fp lang with enforced sigil! Emacs lisp, and the community, does not fit. I feel much better now.
sigil war 2
now experimenting with a new naming convention. Instead of starting with sigil @ for param and $ for local var, now use P and X prefix instead.
You see, here's the whole issue. I want sigils, cuz that truly is advantageous as a feature of programing lang syntax. . So, i started to use φ and ξ. Of course, met with lots resistence. Then, i thought, the unicode is pesky. Using ascii @ and $ is more acceptable, and familiarity, since there is already sigil of that in perl ruby php, and even clojure but with diff symbols. But in past month heavy coding, i find $ is annoying to type. So, thought about P and X. This is one more step closer to familiarity, since now it's just plain letters.
Between ξ vs $ vs X as sigil, the ξ is best, because it's unique. (best is actually have the language enforce it, as in perl ruby php, as part of syntax). Going to $, it's less unique, because for example, one cannot simply find/replace all $ to something else, since $ will appear in places that's not a sigil. The X is worst here. In order to identify them, one has to take code syntax into account, e.g. Else we have X11 Xemacs etc, no longer a matter of text processing.
looking back, people don't really like φ ξ
now, massively redo the sigil. Now, decided $ stays, but for parameters, change @name to Name
- rms Stole Emacs from Gosling
- 2020 Bozhidar Batsov RuboCop Incident
- 2019 rms Resigned from FSF
- 2019 rms at Microsoft
- 2018 rms is a Tyrant
- 2018 Linus: Respect is Earned Not Given
- 2017 rms Pushing for GPL3, Kicked Out Free Software from Free Software Platform
- 2016 Elisp Doc String Curly Quote Controversy
- 2016 rms Removes Color Emoji on Mac Emacs
- 2017 Language Server Protocol (LSP) Kills Elisp
- 2016 Ugly Redisplay Internals Hack
- 2015 rms: What's magit?
- 2013 Rants on Emacs Visual Lines by Don Hopkins, Mark Crispin
- 2013 rms Wants Emacs to be Word Processor
- 2013 How Much Donation FSF Get
- 2012 rms on Open Source
- 2012 Daniel Weinreb Died
- 2011 rms Speech Requirement
- 2007 Daniel Weinreb Rebuttal to rms's Lie
- Young rms on Software Freedom
The Emacs Cult
- 2010 Emacs Dev Inefficiency
- Emacs Dev Inefficiency, Emacs Web 2?
- 2001 Emacs and XEmacs Schism
- 2007 Emacs vs XEmacs
- 2008 Problems of Emacswiki
- 2011 Emacs vs Windows Notepad
- 2011 Emacs Undo Cult Problem
- 2010 Have You Read Emacs Manual