Patch 1720 document lazyness of spanRL, breakRL, takeWhileRL

Title document lazyness of spanRL, breakRL, takeWhileRL
Superseder Nosy List bf, ganesh
Related Issues
Status accepted Assigned To

Created on 2018-08-28.16:32:47 by ganesh, last changed 2018-08-28.21:28:17 by ganesh.

File name Status Uploaded Type Edit Remove
document-lazyness-of-spanrl_-breakrl_-takewhilerl.dpatch ganesh, 2018-08-28.16:32:47 application/x-darcs-patch
patch-preview.txt ganesh, 2018-08-28.16:32:47 text/x-darcs-patch
unnamed ganesh, 2018-08-28.16:32:47 text/plain
See mailing list archives for discussion on individual patches.
msg20281 (view) Author: ganesh Date: 2018-08-28.16:32:47
sending this untracked patch to the tracker as I want to comment on it

1 patch for repository /home/ganesh/darcs/screened-temp:

patch 1480b4b51b250af266a764bceb5f4183517746a2
Author: Ben Franksen <ben.franksen@online.de>
Date:   Sat Aug 25 16:11:25 BST 2018
  * document lazyness of spanRL, breakRL, takeWhileRL
msg20282 (view) Author: ganesh Date: 2018-08-28.16:57:02
> +-- | Like 'takeWhile' only for 'RL's. This function is supposed to be lazy:
> +-- elements before the split point should not be touched.

Shouldn't this be elements *after* the split point? (and with the other two comments)
msg20286 (view) Author: bf Date: 2018-08-28.18:59:15
Unfortunately with RL there is an inherent ambiguity wrt the meaning of
"before" and "after". I prefer to use these terms with respect to the
application order i.e. left=earliest patch to right=latest patch. In
this sense "before" is correct. The other meaning of "before" is "what
is accessed earlier" and with RL the two notions are opposite.

Not sure what to do about that.
msg20287 (view) Author: bf Date: 2018-08-28.19:03:31
BTW for this patch I did not forget to attach a bundle, see comment
msg20260 for patch1707 (it is a follow-up on your review).
msg20288 (view) Author: bf Date: 2018-08-28.19:16:26
> The other meaning of "before" is "what
> is accessed earlier" and with RL the two notions are opposite.

Ahem. What I meant here is the order in which an RL is naturally (and
most efficiently) traversed; which is opposite to the patch application
msg20289 (view) Author: ganesh Date: 2018-08-28.21:28:11
Oh, I see, I was caught out by the fact that patch1707 was already accepted, so 
I (or rather my script) wasn't looking at it, and it also wouldn't appear in 
"Show Open" in the roundup UI. Too many things in flight for me to remember it 
all so I rely on dipping into the patch tracker to find things awaiting 

Anyway, good point about the ambiguity, I'd forgotten these were RLs. Could 
maybe expand on the comment a little bit, or just leave it as-is. It doesn't 
matter too much as anyone who cares will realise there's only one plausible 
Date User Action Args
2018-08-28 16:32:47ganeshcreate
2018-08-28 16:57:03ganeshsetstatus: needs-screening -> review-in-progress
nosy: + bf
messages: + msg20282
2018-08-28 18:59:16bfsetmessages: + msg20286
2018-08-28 19:03:31bfsetmessages: + msg20287
2018-08-28 19:16:26bfsetmessages: + msg20288
2018-08-28 21:28:12ganeshsetmessages: + msg20289
2018-08-28 21:28:17ganeshsetstatus: review-in-progress -> accepted