$epsilon = 1.0e-6 feels too big for Rat()

[Dan Kogai]

Folks,

I am only beginning to unwrap the christmas present but I immediately fell in 
love with the perl6 arithmetic system.  Not is it rich but it is also fast.

% perl6
> 340282366920938463463374607431768211297.is-prime
True
> 340282366920938460843936948965011886881.is-prime
False

And type conversions between numbers are smooth and seamless.  I couldn’t help 
loving Rat … except for one thing.

% perl6
> pi
3.14159265358979
> pi.Rat
3.141593
> pi.Rat.nude
(355 113)

That’s so 5th century!

https://en.wikipedia.org/wiki/Zu_Chongzhi

And of course,

> pi.Rat == pi
False

https://perl6advent.wordpress.com/2009/12/14/day-14-going-to-the-rats/

However, we can be more precise by simply giving 0 to $epsilon.

> pi.Rat(0).nude
(245850922 78256779)
> pi.Rat(0) == pi
True
> e.Rat(0) == e
True
> log(2).Rat(0) == log(2)
True

This I feel more natural.

http://doc.perl6.org/routine/Rat#role_Real says the default $epsilon is 1e-6.


Why so large?

Why not zero?

Dan the Perl6 Newbie




signature.asc
Description: Message signed with OpenPGP using GPGMail

Re: $epsilon = 1.0e-6 feels too big for Rat()

[Darren Duncan]

Considering that a non-fat Rat has a 64-bit denominator, I would expect 
conversions from Num to make use of that full precision by default, and not 
round off to 6 decimal places. -- Darren Duncan

Re: release?

[Brock Wilcox]

On Tue, Dec 29, 2015 at 11:39 AM, webmind  wrote:

>
> Yes, wouldn't it make sense to couple the rakudo release version to the
> language it implements?
>

Naw -- there'll be probably monthly rakudo releases but the Specification
releases should be much less frequent -- like maybe every few months (at
first) or yearly is my guess.

This might be less confusing if:
* We referred more often to rakudo instead of perl6 when we mean the
implementation (you compile with gcc, not "c"; rakudo confusingly calls
it's interpreter "perl6")
* Get more implementations! If we had like 3-4 implementations to choose
from then it might be more obvious what was going on.

Probably there would be a stronger argument for the "perl6" binary to be
either renamed to "rakudo" or to be a symlink to whatever your
current-perl6-implementation is were there an alternative implementation
... but there isn't... so ... I guess someone should do that. :)

... though there actually ARE a few others, but none nearly as complete as
Rakudo, afaik

* https://github.com/sorear/niecza - CLR
* http://fglock.github.io/Perlito/ - Perlito6 written mostly in Perl6 (lots
of other interesting Perlito stuff)
* http://perl6.org/compilers/features - comparison
* several abandoned ones (e.g. Pugs)

--Brock

Jonathan's "Perl 6 Introductory course"

[Tom Browder]

Jonathan's intro course, in pdf, here:

  https://github.com/rakudo/star/raw/master/docs/2015-spw-perl6-course.pdf

is excellent, of course.  But I really like the presentation theme and
the slide formatting!

Does anyone know what slide-making process he uses?

So far the best I have found that meets my needs is using asciidoc
input with Asciidoctor's Deck.js backend. I think I can convert from
the generated html to pdf but can't say for sure yet, but it is the
good looks of his slides that I'm primarily interested in.

Thanks.

Best regards,

-Tom

Re: release?

[Parrot Raiser]

That's how I have Perl 6 (and a number of other packages) set up; a
version-agnostic name in a $PATH place, symbolically linking to
package directory.

On 12/31/15, Philip Hazelden  wrote:
> Note that if we want scripts to be interpreter-agnostic, the perl6 binary
> needs to exist for #! purposes. So renaming it would be bad, but a simlink
> would work.
>
> On Thu, Dec 31, 2015 at 2:27 PM Brock Wilcox 
> wrote:
>
>> On Tue, Dec 29, 2015 at 11:39 AM, webmind  wrote:
>>
>>>
>>> Yes, wouldn't it make sense to couple the rakudo release version to the
>>> language it implements?
>>>
>>
>> Naw -- there'll be probably monthly rakudo releases but the Specification
>> releases should be much less frequent -- like maybe every few months (at
>> first) or yearly is my guess.
>>
>> This might be less confusing if:
>> * We referred more often to rakudo instead of perl6 when we mean the
>> implementation (you compile with gcc, not "c"; rakudo confusingly calls
>> it's interpreter "perl6")
>> * Get more implementations! If we had like 3-4 implementations to choose
>> from then it might be more obvious what was going on.
>>
>> Probably there would be a stronger argument for the "perl6" binary to be
>> either renamed to "rakudo" or to be a symlink to whatever your
>> current-perl6-implementation is were there an alternative implementation
>> ... but there isn't... so ... I guess someone should do that. :)
>>
>> ... though there actually ARE a few others, but none nearly as complete
>> as
>> Rakudo, afaik
>>
>> * https://github.com/sorear/niecza - CLR
>> * http://fglock.github.io/Perlito/ - Perlito6 written mostly in Perl6
>> (lots of other interesting Perlito stuff)
>> * http://perl6.org/compilers/features - comparison
>> * several abandoned ones (e.g. Pugs)
>>
>> --Brock
>>
>>
>

Re: release?

[webmind]

On 29/12/15 17:13, andy_b...@wiwb.uscourts.gov wrote:
> On Tue, Dec 29, 2015 at 01:57:57AM -0800, Darren Duncan wrote:
>>> On that note, are there going to be Perl 6 versions 6.x.y where {x,y} are
>> > integers?  Will 6.0.0 be the first such one? -- Darren Duncan
> 
> On Tue, Dec 29, 2015 Patrick Michaud wrote:
>> "Perl 6" is a language, not an implementation of that language.  Think
> of "Perl 6" as being like "C", "C++", "Javascript", etc., where the
> language is separate from the (many) implementations of that language.
> 
> I'm just a very ordinary perl hacker here, but Alex's point I think
> should be addressed.  Most of us (i.e ordinary, un-language
> implementation geeks) are looking to download a Perl6 and if it's rakudo
> x.y.z, fine, but make that seem like something like perl.6.tar.gz. It
> would seem that gently introducing the complete separation w/ a little
> of Perl's famous (to me) "syntatic sugar" (meta-syntatic?) to help us
> getting started.  Maybe the lower case distinction, "perl6.x.y" vs "Perl
> 6.c", would soothe both sides of the discussion.

Yes, wouldn't it make sense to couple the rakudo release version to the
language it implements?

Thanks for all your replies.


-- 
GPG Key: https://u2m.nl/data/webmind.asc
GPG Fingerprint: 0506976E 234653B4 A628EC33 E23D16EE FCF154AE
XMPP webm...@puscii.nl:  D79970A8 7EC43E29 186D86BA 590F20F6 4C7930B8
XMPP webm...@laglab.org: 11E91112 091881F7 53EF6108 63C48543 C74D035C
u2m.nl (exp: 08/04/2016) SHA256:
C2:40:67:22:25:52:29:AF:DF:50:4E:2A:6B:32:6D:BC:5B:1E:CA:7D:52:3B:4C:4A:21:5D:C8:E5:AE:7D:1A:09
Puscii (exp: 04/03/2016) SHA256:
F9:C7:B1:B7:90:6B:17:BF:84:93:93:7C:0F:B4:FD:BE:E3:C0:71:9D:83:01:ED:3A:96:FE:FC:82:9D:30:51:C9



0xFCF154AE.asc
Description: application/pgp-keys


signature.asc
Description: OpenPGP digital signature

Re: Bug or PEBCAK?

[Parrot Raiser]

Thank you.

I'm not sure how to classify this one.  It never occurred to me that a
new Linux installation would NOT include a C compiler. Installing gcc
fixed that.

On 12/31/15, Dominique Dumont  wrote:
> On Wednesday 30 December 2015 19:42:38 Parrot Raiser wrote:
>> probing whether your compiler thinks that it is gcc  Can't compile
>> simple gcc probe, so something is badly wrong at build/probe.pm line
>> 92.
>
> Looks like probe.pm cannot find gcc when trying to compile a small program.
>
>
> HTH
> --
>  https://github.com/dod38fr/   -o- http://search.cpan.org/~ddumont/
> http://ddumont.wordpress.com/  -o-   irc: dod at irc.debian.org
>

Re: release?

[Philip Hazelden]

Note that if we want scripts to be interpreter-agnostic, the perl6 binary
needs to exist for #! purposes. So renaming it would be bad, but a simlink
would work.

On Thu, Dec 31, 2015 at 2:27 PM Brock Wilcox 
wrote:

> On Tue, Dec 29, 2015 at 11:39 AM, webmind  wrote:
>
>>
>> Yes, wouldn't it make sense to couple the rakudo release version to the
>> language it implements?
>>
>
> Naw -- there'll be probably monthly rakudo releases but the Specification
> releases should be much less frequent -- like maybe every few months (at
> first) or yearly is my guess.
>
> This might be less confusing if:
> * We referred more often to rakudo instead of perl6 when we mean the
> implementation (you compile with gcc, not "c"; rakudo confusingly calls
> it's interpreter "perl6")
> * Get more implementations! If we had like 3-4 implementations to choose
> from then it might be more obvious what was going on.
>
> Probably there would be a stronger argument for the "perl6" binary to be
> either renamed to "rakudo" or to be a symlink to whatever your
> current-perl6-implementation is were there an alternative implementation
> ... but there isn't... so ... I guess someone should do that. :)
>
> ... though there actually ARE a few others, but none nearly as complete as
> Rakudo, afaik
>
> * https://github.com/sorear/niecza - CLR
> * http://fglock.github.io/Perlito/ - Perlito6 written mostly in Perl6
> (lots of other interesting Perlito stuff)
> * http://perl6.org/compilers/features - comparison
> * several abandoned ones (e.g. Pugs)
>
> --Brock
>
>

Re: Bug or PEBCAK?

[Dominique Dumont]

On Wednesday 30 December 2015 19:42:38 Parrot Raiser wrote:
> probing whether your compiler thinks that it is gcc  Can't compile
> simple gcc probe, so something is badly wrong at build/probe.pm line
> 92.

Looks like probe.pm cannot find gcc when trying to compile a small program. 

HTH
-- 
 https://github.com/dod38fr/   -o- http://search.cpan.org/~ddumont/
http://ddumont.wordpress.com/  -o-   irc: dod at irc.debian.org

Re: Jonathan's "Perl 6 Introductory course"

[Sitaram Chamarty]

On 31/12/15 20:43, Tom Browder wrote:
> Jonathan's intro course, in pdf, here:
> 
>   https://github.com/rakudo/star/raw/master/docs/2015-spw-perl6-course.pdf
> 
> is excellent, of course.  But I really like the presentation theme and
> the slide formatting!
> 
> Does anyone know what slide-making process he uses?
> 
> So far the best I have found that meets my needs is using asciidoc
> input with Asciidoctor's Deck.js backend. I think I can convert from
> the generated html to pdf but can't say for sure yet, but it is the
> good looks of his slides that I'm primarily interested in.

Looks like Beamer (latex+beamer).

An easy step to beamer is to use pandoc, which is what I use when I need
to make a presentation.  Though it's not as flexible as writing latex
directly, it's pretty decent because with pandoc, your input is just
markdown.  Much (much!) easier on the fingers and on the eyes :)

(Not sure how strict we are on "off-topic" stuff so I will stop here!)

Re: Jonathan's "Perl 6 Introductory course"

[Tom Browder]

On Thu, Dec 31, 2015 at 9:26 AM, Sitaram Chamarty  wrote:
> On 31/12/15 20:43, Tom Browder wrote:
>>   https://github.com/rakudo/star/raw/master/docs/2015-spw-perl6-course.pdf
...
>> Does anyone know what slide-making process he uses?
...
> Looks like Beamer (latex+beamer).
>
> An easy step to beamer is to use pandoc, which is what I use when I need
> to make a presentation.  Though it's not as flexible as writing latex
> directly, it's pretty decent because with pandoc, your input is just
> markdown.  Much (much!) easier on the fingers and on the eyes :)

Great info, Sitaram, thanks so much!

...Oops, I see several paths leading there.  Any hints on your work flow?

Best regards,

-Tom

Re: $epsilon = 1.0e-6 feels too big for Rat()

[James Ellis Osborne III]

Considering that the Chinese 5th Century is 1849, I would expect
http://babel.hathitrust.org/cgi/pt?id=mdp.39015046627348 to present
less zuo.

-jas

On 31 December 2015 at 03:48, Darren Duncan  wrote:
> Considering that a non-fat Rat has a 64-bit denominator, I would expect
> conversions from Num to make use of that full precision by default, and not
> round off to 6 decimal places. -- Darren Duncan
>

Re: Jonathan's "Perl 6 Introductory course"

[Tom Browder]

Thanks Sitaram and Moritz!

-Tom

Re: Jonathan's "Perl 6 Introductory course"

[Sitaram Chamarty]

On 31/12/15 21:26, Tom Browder wrote:
> On Thu, Dec 31, 2015 at 9:26 AM, Sitaram Chamarty  wrote:
>> On 31/12/15 20:43, Tom Browder wrote:
>>>   https://github.com/rakudo/star/raw/master/docs/2015-spw-perl6-course.pdf
> ...
>>> Does anyone know what slide-making process he uses?
> ...
>> Looks like Beamer (latex+beamer).
>>
>> An easy step to beamer is to use pandoc, which is what I use when I need
>> to make a presentation.  Though it's not as flexible as writing latex
>> directly, it's pretty decent because with pandoc, your input is just
>> markdown.  Much (much!) easier on the fingers and on the eyes :)
> 
> Great info, Sitaram, thanks so much!
> 
> ...Oops, I see several paths leading there.  Any hints on your work flow?

The basic stuff is fairly simple, and the pandoc site has decent help.
It's basically markdown, except a simple (non-standard-markdown) syntax
for the title and the author.  Then you run a command like `pandoc -i
foo.mkd -o foo.pdf -t beamer -V theme:Darmstadt [...]`.

Note that pandoc can't do things like vertical split slides, etc.
(AFAIK), and I've never really tried to figure out how to change the
colors, fonts, etc., either.

The nice thing about pandoc is that it can write several different
formats; see [1] for an example that uses the "HTML Slidy" format.  (The
pictures are done by graphviz, not pandoc, as you will probably guess
when you see it.)

[1]: http://gitolite.com/gcs.html

Re: Jonathan's "Perl 6 Introductory course"

[Moritz Lenz]

On 12/31/2015 04:26 PM, Sitaram Chamarty wrote:
> On 31/12/15 20:43, Tom Browder wrote:
>> Jonathan's intro course, in pdf, here:
>> 
>>   https://github.com/rakudo/star/raw/master/docs/2015-spw-perl6-course.pdf
>> 
>> is excellent, of course.  But I really like the presentation theme and
>> the slide formatting!
>> 
>> Does anyone know what slide-making process he uses?
>> 
>> So far the best I have found that meets my needs is using asciidoc
>> input with Asciidoctor's Deck.js backend. I think I can convert from
>> the generated html to pdf but can't say for sure yet, but it is the
>> good looks of his slides that I'm primarily interested in.
> 
> Looks like Beamer (latex+beamer).

The meta data of the PDF agrees, it says "LaTeX with Beamer class
version 3.10" in the "creator" field.

Cheers,
Moritz