Limber separators

Alexander Berntsen alexander at plaimi.net
Fri May 6 16:52:10 UTC 2016


-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA512

On 06/05/16 18:34, Iavor Diatchki wrote:
> I am not convinced by the argument that this will help make 'diffs'
>  considerably simpler: we have tools for visualizing diffs,
Most people I know read plaintext patches in emails. Others use things
like home made review tools, built-in review tools on SaaSS platforms
like GitHub, review tools in big systems like Phabricator or Gitolite,
etc. None of these help.

> and saving an extra line at the beginning/end of a long enumeration
>  seems like an unlikely source of confusion
The examples I gave were banal. More involved ones are easily
imagined. Moreover, if you view things in isolation like this then
everything is simple and clear, so I think it is an unfair argument.

> or serious merge conflicts.
I don't know what you mean by "serious", but I don't treasure any
moment I've spent on solving three-way merges from this issue.


> By the way, leading/trailing separators conflict with the syntax
> for tuple sections:
> 
> (True,) :: t -> (Bool, t) (,True) :: t -> (t, Bool)
Please read my email; I already mention this.

> I think that it wold be quite odd if leading/trailing commas meant
>  one thing in tuples and something completely different in lists.
As explained in the original email: there is already precedent for
trailing commas in import and export lists, so commas already mean
something different there. Commas are overloaded already. I don't find
this any more confusing than ad-hoc polymorphism in general.
- -- 
Alexander
alexander at plaimi.net
https://secure.plaimi.net/~alexander
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2

iQIcBAEBCgAGBQJXLMu5AAoJENQqWdRUGk8BSW0QAKgX1VIZyH+TA5b7Ac0ieZmn
8gwVeIB+rFog8TqZ/KSF4aY3kO92M35FFvFI0BV3YEaE+URW42S2r5MgtpvQBiCt
cudVW2SHZU9BAag1vU1cd8vykFwLphxZQ+/bRidKRM3+zebK6a8ON6cJTPXOrM3W
bo2s95ZjGnYkkuT+cMEXBD6V193aNJZGFWF4PnN7OBFt0u1w3Lm0ZhTH1qewQMZP
PypTlFVL6N45sWDPKqbTlrb0vES25AWNGdMjHZi5UkQ3Wv0IB8rxiEnxR4VUFe6v
8Wy/xLSnutSk+Lgz68HnoztYNrvQVwpBjRDuhpv/1QhcKedY2/daLAt9u2tBEdBs
9RpzViGI1GNexRo8obztUV4zuFHE+F8vCf4ovCoGZSUMHWM0iGjEJzPXSGvcq2Bn
tt+b0TZfH/kpUirlfkONiEHDHKANBBI2pdfsc+yqxSPfTL4wU/aTLRGbwazT26de
N5MQI5Q77kXlDjEM+2YkuD9YDWX28/LJWtUwMsAYqkrRYX+Xn7/YzpEz4M7En8SE
UWL62RXnX9BCF64Gl9GhlPnXUq8J/gTrz3iehTbKhABD+W6Ew+zpGbvMGAjz8da3
gUtOAgvKdUlvfJlE+Whr99Df3dnju43cKHeu9iTPiGbI5tkomVUMETexCO+D90BC
gsSwLod24dVNh91/7Yxk
=oHX2
-----END PGP SIGNATURE-----


More information about the Haskell-prime mailing list