The Nineteenth Byte

flawr

2:01 PM

Heya=)

TheDoctor

38 stars on the "don't abuse stars message"

feersum

@MartinBüttner Why did you an an arbitrary base restriction to that counting question?

Martin Büttner

Because that's what we discussed in the comments.

feersum

I just finished an answer that counts in hex.

Martin Büttner

-5

A: Can numeric input/output be in unary?

Yes, I/O in any base should be acceptable by default Binary or hexadecimal, e.g. could be similarly useful for some languages. In the interests of consistency, they should also be allowed for input. Again, the challenge author can specifically disallow this if s/he wants.

-1

A: Can numeric input/output be in unary?

Yes, I/O in unary, binary, octal, decimal or hexadecimal should be acceptable by default. These conversions are generally simple, and for most languages they are trivial. For most languages and challenges, this won't make any kind of difference, but this could make a significant usability improv...

there's no consensus for allowing arbitrary bases by default

feersum

2:07 PM

@flawr well I didn't specify any numeral system, so I guess that is valid output. :p — vrwim 48 mins ago

This does not sound like he endorsed those bases and prohibited others.

Martin Büttner

I left a comment.

flawr

2:21 PM

I think most of the answers do not really solve what OP wanted to see: He said he challenges the participants to solve the problems *of having to circumvent any types such as int and long*

There are the trivial answers that are written in languages that *apparently* support arbitrary size integers. But I haven't seen any others so far.

feersum

@MartinBüttner Outputting in base 256 also does not make sense because then there isn't any possible delimiter char.

Martin Büttner

@flawr that doesn't make sense in many languages.

@feersum I considered that, but I don't think it's a big deal... since we know what sequence of numbers we're looking for it's easy to distinguish linefeed-digits from linefeed separators.

feersum

@flawr My answer is conceptually almost the same as yours, by the way...except it only needs two "digits" instead of 30-some.

flawr

@feersum Apart from the memory limitations I think OP wanted to see answers that will at least when ignoring memory limitations be able to count infinitely.

feersum

@flawr Oh, you resize the array... it would be a lot shorter not to :P

flawr

2:25 PM

@feersum But then it would not meet the specs IMHO.

See the message above.

TheDoctor

@feersum isn't base256 technically ascii encoding?

flawr

Anyway I think I have to give up the discussions as people just seem to ignore this.

feersum

@TheDoctor I don't know what you mean.

TheDoctor

more like ascii is the default base256 encoding

feersum

ASCII is 7 bits.

flawr

2:30 PM

@MartinBüttner regardin this comment I interpreted the challenge that apart from memory limitations, all the programs should be able to count infinitely, but on actual machines, the programs should at least be able to count to 2^128, don't you agree?

Martin Büttner

what does that have to do with this comment?

SuperJedi224

@flawr But java.math.BigInteger overflows at about 2^(2^34-1)-1

flawr

@MartinBüttner Well for the languages that have no specifications on the integer sizes you can always write a new interpreter now that makes it trivial!

Martin Büttner

@flawr sure.

same for any other catalogue challenge.

flawr

I do not agree on this.

As for this challenge the integer size is crucial, but it very often is not specified.

Martin Büttner

2:33 PM

doesn't matter though, because such an answer would be trivial and boring, and this is not a competition between languages... writing a new interpreter is like saying "I've added a new sub-competition and I've won it, because there's only one possible answer".

@flawr there's a consensus somewhere on meta that, for the purposes of PPCG, languages are defined by their implementations.

flawr

While for other challenges like "hello world" specs like thos do not really matter.

Martin Büttner

I really don't see the difference.

flawr

But we also had a consensus that languages versions that were created after the submission of challenges are considered invalid for that given challenge.

Martin Büttner

sure, and any default on meta can be overruled.

in fact, I think that the "languages that are newer than this challenge are disallowed" rule does more harm than good anyway.

flawr

What bothers me here is that NO ANSWER so far does adress the (in my opinion interesting) challenge of having to implement a counter that can count infinitely (if you ignore the memory constrains the hardware)

All answers that actually solve this challenge are trivial, because they use arbitrary size integers.

Nobody actually seems to care about this goal.

feersum

2:38 PM

Yes, this is typical... the greater part of challenges have poor specifications.

SuperJedi224

I've never really tested what happens when java's BigInts hit their max size.

flawr

I think it is just nonsense to have a ton of trivial or untentersting answers, to a challenge that would have been great if people actually woud care about the challenge, and not only the rep points...

Martin Büttner

@flawr I don't understand how that is different from using a built-in for prime testing.

SuperJedi224

Do they explode, or do they just wrap around?

flawr

@MartinBüttner It just means that we now have a challenge full of trivial answers. Nobody tried to actually solve the problem without relying on something that makes it trivial.

Martin Büttner

2:41 PM

It's a trivial challenge. Why would I impose an arbitrary size limit on my integers if there is no such limit in the language I'm using?

flawr

This is not what I want.

I give up this discussion, I do not want to waste my time on this anymore.

New Sandboxed Posts

0

A: Sandbox for Proposed Challenges

Rounding Fractions code-golf math rational-numbers Back in the old days of game programming, before FPUs were the norm, games predominantly used fixed-point math to represent non-integer values. Typically, the lower 8 or 12 bits of a 32-bit word are used as fractional parts, and the rest are tr...

TheDoctor

8-bit sandbox guy breaks the silence

SuperJedi224

Dimrain47 - Infernoplex

►

Martin Büttner

@flawr I can understand that, but I was actually interested in how you would implement this in a language that does have arbitrary precision types. You just complained that people used them, but if I don't impose an arbitrary limit on the type, I'm not sure what else I would do with them.

flawr

3:25 PM

@MartinBüttner The answers using arbitrary precision integers (API) are uninteresting but valid. But it bothered me that people used languages where it is not clear if they support APIs. This tells me that they do not care about the challenge and the problem of overcoming the difficulties when you actually have limited precision integers.

Martin Büttner

I see, fair enough.

feersum

When a program "dumps core" in Linux, doesn't that mean it's supposed to fill the terminal with a bunch of scary-looking hexadecimal numbers?

flawr

I want to apologize that I perhaps overreacted before, it drove me mad how people were just ignorant about this.

Martin Büttner

@feersum I don't think so. It just writes to a dump file, and even that is by default suppressed on some distributions.

@flawr No harm done.

sweerpotato

Is there an easy way to test if a program can count up to 2^128?

I feel it shouldn't be a problem for Python

Martin Büttner

3:34 PM

start at 2^128-10? :P

feersum

Does anyone have a good name idea for a language which is executed by the C++ compiler finding the type of a template?

sweerpotato

I'm more referring to "without problems and without running out of memory on a reasonable desktop PC"

Python can handle the number itself

Martin Büttner

why should there be a problem then?

feersum

It takes nearly 20 bytes to store 2^128. I'm not sure the average computer could handel that.

Martin Büttner

@feersum I'm not exactly sure what you mean.

sweerpotato

3:41 PM

hm

probably isn't a problem

Thanks :~)!

SuperJedi224

Java's BigIntegers can potentially reach 16 gigabits, which is probably about where the JVM runs out of memory anyway.

feersum

I think it's common for enterprise Java systems to use tens of gigabytes.

SuperJedi224

I think I heard somewhere that the JVM's default memory allowance was a little more than 2 gigabytes. Though, of couse, it can probably be configured to allow significantly more if you want.

feersum

-Xmx is the option for max memory

I wish Python had a default like that, so it wouldn't blow up my computer when I foolishly run a golfed program that was a little inefficient with memory.

TheDoctor

reddit.com/r/cyberpatriot/comments/3sp9se/…

SuperJedi224

3:54 PM

Transcript for