« first day (2921 days earlier)      last day (674 days later) » 

12:00 AM
Why did you need a 21G array by the way? I only needed a little over 1G.
Of memory I mean, not an array length.
@feersum I had no idea how far it would go.
@feersum Did you do pruning?
The design of my program was intended for pruning, but then I didn't actually implement it
Because I saw your result which meant it was too small to be necessary.
Well neither did I. So I may have needed the 21 GB, and for all I knew it may have not been enough
It would have been quite frustrating for the run to have to terminate due to running out of array space. Then I'd need to start it all over again with a bigger array. Now do you see why I used 21 GB?
That's the maximum I had available without closing a lot of running programs. I have 32 GB total.
Yeah sure.
Although you could have made it resizable.
It's fine.
12:04 AM
Yeah but it's such a big array that there might not be room for it to be reallocated contiguously.
I do dynamic buffer resizing in my regex engine. Doubles every time it needs to expand.
Anyway I didn't want to introduce potential bugs in my program by over-complicating it with pruning or reallocation. Since I didn't know the result ahead of time, I'd have nothing to compare against.
Trapped at step 12899744968 at position 12851850258 (max visited: 12950157910; max examined: 12951523532)
path-of-the-wildebeest.bin matches
12:34 AM
I'm curious if there are any "non-weird" moving patterns that do not get stuck
The next thing I want to do is plot it graphically... at least around the final place where it gets stuck.
Verified that the latest commit, with minimum-needed-size array, works as intended.
1 hour later…
1:38 AM
@Deadcode Sounds like you need to switch from C++ xD
@Quintec Why would I want to do that? Unless it was to something like D :)
1:59 AM
@Deadcode Oh, I guess you're thinking text animations? I was thinking something slightly more glamorous :P I heard all C++ GUI APIs are a pain
@Quintec I'm planning to write to an .SVG file.
2:43 AM
@dzaima @Mego seems like they are supported: "The final rule, Larger than Life, relies on large neighborhoods; these are supported by Golly 3.0 (up to range 500)." source. you might have to fiddle a bit though
6 hours later…
8:53 AM
Q: Compress large array of strings

m__The problem I am trying to solve is this I have have large collection of list of strings. There are more than 500 collections. Each collection contains 1 or more strings (up to 1000) Many of these strings contain common phrases The Goal - minimize the database size. Database structure Eac...

2 hours later…
11:19 AM
@ASCII-only if i'm reading that correctly, that's only for neighbor count rules though
2 hours later…
1:47 PM
@EriktheOutgolfer What do you mean? It looks fine to me
@ASCII-only Scroll up - bolded rules are supported, non-bolded are not. That page is a roadmap. While the core functionality may be in Golly for large neighborhoods, it does not support custom rules using them.
@flawr Now we can golf our posts better!
3:04 PM
@Deadcode: Huh, that was a nice surprise :D Thank you!
You're welcome BMO :) Your post was a nice surprise! I don't understand it, but it's quite impressive.
I don't understand, did you not place the bounty before I answered? I thought I'd found that challenge because of that.
Yes, I placed the bounty before. And your post is clearly the one deserving it.
(except for yours, maybe)
(which I don't understand either ^^)
:D Well maybe we could both explain it better.
I'm curious what exactly is nondeterministic in yours and how you make it deterministic.
3:11 PM
The rules which begin with s5 are introducing that
Sometimes there will be a and b_copy atoms, so it randomly picks either the first or second rule thereof
You could introduce another "state" to sequentialise it like this: Try it online!
@BMO I think what would really help is being able to see the state of the universe as it progresses. Can your interpreter do that currently?
No, you can only view the rules which are applied by using -d2 :(
How many atoms do there end up being, for example with input 144 as in your TIO link?
@Mego maybe because it shouldn't require \limit for inline support...
And is there any slight random chance that the answer could be wrong, or is your solution guaranteed deterministic?
3:22 PM
However, what I do sometimes is to just replace the right-handside with stopX and then I know where I am and can view the atoms.
@flawr lol that's been there for quite some time already
@Deadcode No idea, but not that many actually.
@Deadcode No, the only non-determinism comes from s5-rules and there the order of replacing a-/b_copy-atoms with b-atoms doesn't matter.
3:45 PM
@BMO Is there anything in particular I could explain better?
@Deadcode I think you did a pretty good job with explaining it
I'm just not too comfortable with unary-magic and have no idea how lookbehinds work :S
Lookaheads you mean? No lookbehinds in ECMA.
Here's all you need to know about lookaheads: (?=\bcreative PPCGer\b)BMO\b Ahh! Found a match :)
@Deadcode Either ^^
Going out now, will take a look later :)
Introduces a new pseudo-atom ? which will simply print the universe (can be improved I thin)
3:59 PM
4:33 PM
DNF just failed due to a 403 Forbidden to the Fedora repos and I'm a wee bit confused about that
Duke Nukem Forever?
DaNdeFied yum
Ah, package manager for Fedora can't access Fedora repos. Makes sense. (no it doesn't)
4:47 PM
Idea for an esolang: Markdown:
Idea for an esolang: Markdown:
> comment
> assignment:
[1]: 42
> addition:
plus *plus*
> subtraction:
minus _minus_
> string:
It will launch as the most frequently used programming language on SE!
5:00 PM
@Adám I like it.
Verbosity at it's finest right there.
@J.Sallé Wait, what is so verbose about it?
@Adám I meant it's a verbose language for ppcg standards, but not overly verbose like Java or something
The assignment I guess
@EriktheOutgolfer yup, but it doesn't seem that many people are aware of it yet
@Pavel It should have special rules. You can assign the same value to multiple names, but you have to use unique "assignment identifiers" to do so.
**times** and __division__ of course and ***power*** and ___root___
> while loop is the only necessary control structure:
loop body
more code

Or maybe better is to just have labels and goto:
> label:
# mylabel
> goto:
* mylabel
Ah, the mathematical operators should be outfix: *2 2 3* gives 7
5:28 PM
@Adám I like this one better
maybe jmp instead of goto:
@J.Sallé What's the difference between jmp and goto?
@Adám a whole byte!
@J.Sallé I think you've been misunderstanding all my examples ;-) Lines beginning with > are comments.
Oh yeah indeed I did
That makes much more sense now
2 hours later…
7:09 PM
@flawr lol I was there at exactly the moment that change went live
7:38 PM
@flawr And only 6-8 years later :)
8:05 PM
@trichoplax They delivered on time!
Less than 7 years, so in the optimistic half of that estimate...
8:18 PM
But we still have no design right?
in fact, we do have a newer design, just not what we wanted... :P
8:38 PM
@Mr.Xcoder Hehe, but
Q: Sum of Powers of 2

SpookyGengarThe Challenge Given an integer input x where 1 <= x <= 255, return the results of powers of two that when summed give x. Examples Given the input: 86 Your program should output: 64 16 4 2 Input: 240 Output: 128 64 32 16 Input: 1 Output: 1 Input: 64 Output: 64 Scoring T...

Q: How does the community work?

FollowerOfGodHow does this Programming Puzzles & Code Golf work? Just curious because I don't really know what this is and I want to fined out...

9:16 PM
> 4641 Segmentation fault (core dumped) jelly fu .code.tio "$@" < .input.tio
Well, "fu" too, Jelly
Ugh, up to 16 bytes already and it's still not doing what I want. Jelly's layout and my brain just don't jive.

« first day (2921 days earlier)      last day (674 days later) »