Do you really want to flush all these opportunities down the sink?
Gregg
Right now, yes. You haven't convinced me that the changes proposed in #2094 will make REBOL a better language.
Ladislav
Notice that there is no "right now". Once out of alpha it will not be possible any more.
Gregg
When I say "right now", I mean "based on my curreny body of knowledge". Mine is only one vote.
Ladislav
Yes, I understand what it means. I did try actually quite hard writing a lot in the ticket, discussing the things here and presenting many useful case already like:
For me these reasons are convincing. I do not think I can be much more persuasive than this.
I, of course, mentioned, that virtually every case ignoring space significance is provably a bug in either MOLD or LOAD. I mentioned correctness, reflexivity, expressivity, consistence. I think that I am done with convincing.
Gregg
I appreciate your efforts. As I said above, path improvements: YES auto-gluing blocks and strings together: NO.
Ladislav
Well, I demonstrated above that space preceding the [ character is significant. I do not need more.
I mean that the siginficance of the space preceding the [ character is not my invention. I just work with it in a way that it deserves.
Ladislav
I also must say, that it is the other way around. There is no auto-gluing proposed in #2094. What I proposed is no auto-separation.
That is, I am the one who does specify how it looks. If I write it glued, then I mean glued. If I write it separated, then I mean separated. It is actually you forbidding me to do that.
Or, I should say, forbidding Rebol to take seriously what I wrote.
Ladislav
This is the flexibility I value most.
(could be also called "freedom of expression")
Rebolek
>> trim/lines "line 1^/line 2" == "line 1 line 2"
Why is newline replaced with space? Shouldn't it be just removed?
Maxim
hum... because we can easily replace/all lines with spaces... I agree .
Gregg
I like the way it works. Line breaks *are* whitespace, unless you're dealing with hyphenates that you want to connect.
Endo
Same for R2.
trim/with "line 1^/line 2" newline ; this will give the expected result.