[Num-utils] Re: Re: Re: Enhanced version of “numprocess”

[Custos-Mundi at gmx.de]

num-utils-bounces at suso.org [num-utils-bounces at suso.org] (Wed, 19 Sep 2007 14:12:32):
> Your mail to 'Num-utils' with the subject
>
>    Re: Re: Re: Enhanced version of ?numprocess?

>
> The reason it is being held:
>
>    Message body is too big: 99371 bytes with a limit of 40 KB

Argh, dammit. Forgot the “-z” switch while tarʼing everything together :/.

So I'm going to “roughly” resemble my other, unseen mail.

Suso Banderas [suso at suso.org] (Wed, 19 Sep 2007 01:03:27):
> And by the way, I found that email, it was in my inbox of all places.
> Here is the email from Bill Poser:
>
> ---------------------------------------------------------------
> Hi. In view of your interest in utilities for
> dealing with numbers, I thought  you might be
> interested in a new project of mine. I have written
> a library for converting between machine representations
> of integers and textual representations in a wide variety
> of number systems. The library comes with a command-line
> utility called numconv whose main mode of operation is
> to convert from one number system to another, e.g.
> >from Chinese to Western numbers, or Roman Numerals to
> Phoenician. See http://billposer.org/Software/libuninum.html.
>
> Bill
>
> ---------------------------------------------------------------

This project is quite promising and the “num-utils” could make good use of the capabilities these library has to offer. Unfortunately itʼs not yet in the portage tree or any portage overla
y (at least not as far I know). Are you (suso), by any chance, the maintainer of the numutils ebuild?

Suso Banderas [suso at suso.org] (Wed, 19 Sep 2007 01:03:27):
> Great.  I knew there would be some people out there that would find
> these useful.  I think what needs to be done is to make sure that we
> have a stable syntax for each command.  I mean as far as expressions for
> things like numgrep, range, etc.  Another challenge is to make fast
> algorithms for programs like random.

By “fast algorithms” you mean maybe switching to another programming language for these tasks? I just mention it because Iʼve seen “numprocess-reloaded” in action on zillions of numbers and its speed didnʼt really impress me. Though I didnʼt instantly switch to XS and started writing everything important in C. Wasnʼt that big of a deal to me (yet).

Thereʼs, however, a more important reason for this mail.                                                                                                                                    The tarball I provided carried a really erroneous version of “numprocess-reloaded” which would make the use of the C locale practicly impossible (quite bad :/). The tar ball here has this bug fixed. Furthermore I fixed some factual error in the documentation.                                                                                                                     Unfortunately I canʼt pack the git repository into the tar ball as well, because it would be quite about 180 KiB, which is too big for an attachement for this list and I have no other reasonable ways of making it available online.

Hope it works this time :).
-- 
GMX FreeMail: 1 GB Postfach, 5 E-Mail-Adressen, 10 Free SMS.
Alle Infos und kostenlose Anmeldung: http://www.gmx.net/de/go/freemail
-------------- next part --------------
A non-text attachment was scrubbed...
Name: numprocess-reloaded.tar.gz
Type: application/x-gzip
Size: 17504 bytes
Desc: not available
Url : http://mailman.suso.org/pipermail/num-utils/attachments/20070920/e62f58f0/numprocess-reloaded.tar.gz