> [mailto:[log in to unmask]] On Behalf Of David J.
Peterson

> Uh...wow.  I mean, WOW.  That is incredible--and incredibly
> useful!  That is a tremendous accomplishment!  I'm adding it
to
> my webpage's list of links.

It's something I've had for many years in different formats.  I
added a little here and a little there as a came across things.


> Couple of comments:
>
> -The text for the romanized, IPA and notes columns are a bit
> on the small side.

That was intentional, so they'd fit better.  Those fields are
generally blanks anyway.

> -In the romanized column, I see some boxes instead of
characters.
> My default encoding is UTF-8, so I don't really understand why
> this should be happening.  I'm no unicode wiz by any means, so
> I don't have any suggestions.

If you're seeing boxes, it's because you don't have the right
font(s) to generate the character.  Not much I can do about
that.  Arial Unicode MS covers most of the characters.  The most
obvious ones that aren't there are Khmer and some of the
conscripts like Tengwar or Klingon though I do have those
myself.


> > -I just typed in a tough one--"be"--and I notice for your
Arabic
> romanization you have "k'n".  If I were to romanize that, I'd
> write "kaana" (IPA would be [ka:na]).  Is there a particular
reason
> you made alif an apostrophe?

Romanizations are usually based upon some (usually ISO-based)
standard I found somewhere.  They are also not entered by hand,
but actually generated from code that I wrote.  In some cases,
that code is less than perfect.  Amazing enough I managed to get
Hangul without any real problems.  Arabic for example has
complications with Alif because it can also contain diacritics
which make it something other than <a>.  Probabably the worst
case you'll see from the automatic Romanizations will be with
the Thai entries because some of the vowels occur before the
consonant.  Hanzi/Kanji are about the only ones manually
entered.


> Other than that, what an incredible resource!  Great job!

Thanks.

There is still a lot to be done with it.  I'll deal with it as
time permits.  Next, I'll be adding a page with the entries from
a single language since I already have the queries set up for
that.

I do have some conlangs in the database, but I'm currently
blocking entries with "art" in the language code.   I may allow
them later.  There are some that slipped through because they
have unique codes, including a couple of my own like QSS
(sasxsek) and QDN (Deini).  I also would like to add some
proto-language vocabulary like PIE or Protoaustronesian.