Why that is positively Ethiopic!

re: Why that is positively Ethiopic!

Igor Tandetnik — Tue, 01 Feb 2005 17:45:00 GMT

How do you represent a number with four or more consecutive zeros? Say, 1000001 or 10000000001. Do you alternate HUNDRED and TEN THOUSAND characters, with nothing between them, for each group of two zeros?

re: Why that is positively Ethiopic!

Michael Kaplan — Tue, 01 Feb 2005 17:54:00 GMT

Well, here are a few more to show the pattern:

1000001 ፻፼፩
10000000001 ፻፼፼፩
100000000000001 ፻፼፼፼፩

:-)

re: Why that is positively Ethiopic!

Ahadu — Tue, 01 Feb 2005 18:02:00 GMT

1000001 => ፻፼፩
10000000001 => ፻፼፼፩

See:

http://geez.org/Numerals/NumberSamples.html

sample numeral conversions from the sources.

re: Why that is positively Ethiopic!

Michael Kaplan — Tue, 01 Feb 2005 18:11:00 GMT

Or you can do what I did.... I took the C# source and compiled it. The code creates that NumberSamples.html file, and you can add to it whatever numbers you like. :-)

No bites yet on that algorithm?

Michael Kaplan — Tue, 01 Feb 2005 19:27:00 GMT

Darn, I figured thousands of page views would find one person who wanted to give it a shot. :-)

Explaining the Ethiopic Numbering System and a tiny lesson on Amharic

ComputerZen.com - Scott Hanselman's Weblog — Tue, 01 Feb 2005 23:45:00 GMT

re: Why that is positively Ethiopic!

Dean Harding — Tue, 01 Feb 2005 22:55:00 GMT

I never understood how they represent the answer to "5 - 5" in these numbers systems... I guess they hadn't invented subtraction when they came up with it. :p~

Besides, it's really only mathematicians who care about "0" - you don't really see it in every day life, do you?

re: Why that is positively Ethiopic!

Michael Kaplan — Tue, 01 Feb 2005 23:31:00 GMT

Well, I see zero all the time.... and so do Swedes (its on every elevator).

re: Why that is positively Ethiopic!

Dean Harding — Wed, 02 Feb 2005 00:47:00 GMT

Sure, but if we didn't have a zero to begin with, then they'd probably put "G" on there (or whatever the first letter of the Swedish word for "ground" is).

I guess my point was that if we hadn't "invented" zero so that subtraction could be properly defined, then we'd probably never need one (e.g. instead of saying "$0 deposit" in an ad for a car, you'd say "no deposit" or whatever.)

Mind you, I only really started thinking about this when I posted my first post, so maybe there *are* plenty of reasons for a zero outside of mathematics - I just can't think of one right now (that is, where you can't replace the zero by something equally meaningful...)

As an example, my local IP address is 10.0.0.45. Those two zeros could just be left blank and you'd have "10. . .45" which is equally unambiguous.

Anyway, the Ethipians, Romans and Sri Lankins seemed to get along fine without them. Maybe my point is irrelevent, I dunno, but it's interesting nonetheless.

re: Why that is positively Ethiopic!

Michael Kaplan — Wed, 02 Feb 2005 00:50:00 GMT

Definitely interesting -- the whole area fascinates me. :-)

Though it was too bad no one decided to code up the Ethopian to Arabic-Indic solution....

re: Why that is positively Ethiopic!

Marcel — Wed, 02 Feb 2005 02:09:00 GMT

Okay, all I know about this numeric system I have from this article and the linked algorithm, but don't you leave out some "power characters"?
Like in "7654321" shouldn't it be "DIGIT SEVEN; NUMBER HUNDRED; *NUMBER TEN THOUSAND;*" etc.?

re: Why that is positively Ethiopic!

Dean Harding — Wed, 02 Feb 2005 02:40:00 GMT

Just another reason I love reading your blog - there's always something that gets me thinking about things I would have never considered before. I mean, why would it have otherwise occured to me that you can get along with a character for zero anyway?

OK, you've convinced me: I'm at work now (different timezones and all) but when I get home, I'll see if I can't write a little Ethiopian to Arabic-Indic converter :)

re: Why that is positively Ethiopic!

Michael Kaplan — Wed, 02 Feb 2005 04:47:00 GMT

Well, lets see -- the number would be ፯፻፷፭፼፵፫፻፳፩.

Thats:

DIGIT SEVEN; NUMBER HUNDRED;
NUMBER SIXTY; DIGIT FIVE; NUMBER TEN THOUSAND;
NUMBER FORTY; DIGIT THREE; NUMBER HUNDRED;
NUMBER TWENTY; DIGIT ONE

So you started ok, but you forgot the two numbers between the HUNDRED and the TEN THOUSAND....

:-)

re: Why that is positively Ethiopic!

Marcel — Wed, 02 Feb 2005 10:41:00 GMT

No, I mean if you look at the linked "algorithm 4" it should be
DIGIT SEVEN; NUMBER HUNDRED; NUMBER TEN THOUSAND;
NUMBER SIXTY; DIGIT FIVE; NUMBER TEN THOUSAND;
NUMBER FORTY; DIGIT THREE; NUMBER HUNDRED;
NUMBER TWENTY; DIGIT ONE

The seven is 7*10^6 or 7*10^(2+4) after all.

re: Why that is positively Ethiopic!

Dean Harding — Wed, 02 Feb 2005 12:14:00 GMT

OK, I took up your challenge :)

You can see a screenshot of my app here: http://www.codeka.com/tmp/ethiopian.png

And you can download the C# source + binary here: http://www.codeka.com/tmp/ethiopian.zip

Simply enter an arabic-indic number in the bottom text box, click convert and it'll output an ethiopic number (the code for that is "borrowed" from that web site). If you then cut'n'paste that ethiopic number into the top text box and click that convert button, it'll convert it back to the familiar arabic-indic form. (That's the code I wrote).

I didn't write any automated tests, but you can manually test by typing a number into the arabic-indic box and clicking "Test" - this'll convert to Ethiopic then back again - you'll have to eyeball the result to make sure it's right.

It's probably not that lenient with respect to invalid ethiopic numbers, but it works OK for normalized numbers.

The code for doing the conversion is not that hard, but I won't bother explaining it here, since you should just be able to look at and see (probably better to follow it through with the debugger, it's not commented very well, heh)

re: Why that is positively Ethiopic!

Michael Kaplan — Wed, 02 Feb 2005 14:09:00 GMT

Very cool, Dean -- you took up the challenge and got the job done!

Under the circumstances, I would recommend adding a comment to the top with a copyright and a little info, adding an attribution to the other site for "algorithmic inspiration in one direction" and then posting it in a more permanent place than a "tmp" directory. Code like this ought not to be lost. :-)

re: Why that is positively Ethiopic!

Dean Harding — Wed, 02 Feb 2005 19:59:00 GMT

Well, I've been meaning to set up a blog of my own for quite a while now, so this is a good excuse to get off my butt and do it over the weekend. Once I do that, I'll fix up the code and post it in a more permanent place.

re: Why that is positively Ethiopic!

Michael Kaplan — Wed, 02 Feb 2005 20:20:00 GMT

Well, make sure it is a place that will let you host files and such.... some blogs do and some don't. :-)

This one for example does not -- but I had other sites I could use to offload content to when I have to show pictures or whatever....

re: Why that is positively Ethiopic!

Ahadu — Thu, 03 Feb 2005 04:51:00 GMT

Marcel, you did perceive a discrepancy with the Algorithm #4 correctly, it produces different results than the source codes. #4 was snared by a pitfall with the power expansions.

The geez.org article is really an algorithm #6 by the same author and supersedes anything appearing earlier. The article first appeared in Multiligual magazine in 2000 and has been used in Mozilla since 2001 or so. A prose form of the algorithm can be found in the CSS3 list module proposal. Lots of numeral algorithms there actually.

re: Why that is positively Ethiopic!

Michael Kaplan — Thu, 03 Feb 2005 05:13:00 GMT

Interesting! Too bad things are not labeled a bit more explicitly....

I'll think of slgorithm #4 like the Bohr model of the atom -- even though its not accurate, it does help explain some of the concepts. :-)

re: Why that is positively Ethiopic!

Marcel — Thu, 03 Feb 2005 14:38:00 GMT

Ah, I haven't had a look at the second article because the first one was already simple enough to understand.
I also found that it fittet the Unicode quote ("the power 100^n is indicated by n HUNDRED characters"), the only clue for the contrary now being that they're speaking of a digraph when 'n' is exactly two and not for every 2 'n's.

Ah well, it's not like I think I'll ever going to need it ;-) But interessting nonetheless.

re: Why that is positively Ethiopic!

Dean Harding — Tue, 08 Feb 2005 10:38:00 GMT

I finally set up my blog and posted a slightly newer version of my app. The main difference is a slight restructuring of the class that does the conversion, and some copyright notices :)

Have a look here: http://www.codeka.com/blogs/index.php/dean/2005/02/04/parsing_ethiopian_numbers

I'd eventually like to make it handle stuff like roman numerals and other kinds of number systems...

:-)

re: Why that is positively Ethiopic!

Michael Kaplan — Tue, 08 Feb 2005 15:55:00 GMT

Very cool. :-)

What would it mean to internationalize StrCmpLogicalW?

Sorting It All Out — Tue, 03 Oct 2006 19:40:58 GMT

If you are a regular reader of this blog then odds are you might be a little sick of hearing about StrCmpLogicalW

Rhymes with Amharic #4 (a.k.a. we're all [sub]set so turning out the lights and going to [em]bed!)

Sorting It All Out — Sun, 15 Apr 2007 04:45:39 GMT

(see also parts 1 , 2 , and 3 ) OK, we are getting close to the end of this little mini-series.... First

Behond the Table Driven Text Service, Part 14 (Don't expect too much from numbers)

Sorting it all Out — Tue, 21 Oct 2008 17:06:09 GMT

Prior posts in the series: 0 (You have to start somewhere) 1 (Starting with a dictionary simple in every