Why do you not use the alternative name tab and the name type?

So create the main name, then in the name tab of the person eidtor, create a name of type Reading Name, and a name of type Romanized name

Just like with old documents you document each name you find, so that searching on name goes smooth.
If you like sorting on Reading name, then make the reading name the main name, but be sure to use the official name for adding sources. That is, keep your system.

With such a system, you can easily do a feature request for:
1/ new predefined name types, so you do not have to make custom name types. So Reading Name and Romanized name
2/ option to have reports use names of type Reading name in output if present, or romanized name.

Such a system would keep everything nice and tidy. Adding a field 'Reading Name' to the main name editor seems like a bolted on solution to me. Good for Google in a simple interface, but not for a genealogy application.


2012/7/11 Gerrit <z0idberg@gmx.de>

Am 08.07.2012 08:47, schrieb j debert:
> I'm sorry you took it as my saying "it's too complicated" and
> unnecessary. That was certainly not the intention. I suppose I was
> thinking aloud, actually, and trying to consider everything. Many
> people on the list are not familiar with Japanese language so it is
> necessary to explain some things extensively, perhaps even
> excessively. Gramps is a recordkeeping app and accuracy is important
> so a lot more thought has to go into what is implemented and how to
> avoid mistakes that could introduce errors into the data and possibly
> make them garbage. This is not so simple and definitely not quick and
> easy.

Ok, I am sorry for my rude reply :(. I sometimes post some idea on some
mailing list, but I experienced that on some projects, I get a “oh, nice
idea, we can keep that in mind”, but on others just a “this is not
possible” comes right at the beginning. I of course know that open
source projects are just a hobby for everyone and there is nothing to
“demand”, but maybe sometimes someone is happy about a new idea. This
was just a proposed idea from my part, which I thought about when using
the program.
> Sorry I'm a terrible diplomat. I actually do like the idea. I'm sure
> can be done. I would like very much to see it implemented somehow but
> I would also like it to be at least adequate, with as little
> compromise and error as possible. And not just for Japanese.
> It would likely require enlisting the help of native Japanese
> speakers, as they are the most competent in this and can correctly
> judge how simply and how well it can be implemented.

Well, some work can be done already by people who can speak a little
Japanese. Especially because most stuff is quite straightforward
(usually, everything is written in Wikipedia).
> Actually, Google doesn't do all that good a job with Japanese although
> they are certainly better than when they started. They still get name
> readings wrong. A lot.
> By the way, in the case of 宮崎駿 it happens there is one reading for
> the family name Miyazaki and one for the given name Hayao. Too easy
> for Google. Lots of other names aren't so easy.

Yes. As I thought it would be the easiest to just let the user insert
the reading, and automatically divide between first and last name.
Google does it better and automatically, but this is not really needed
in gramps. I just used Google as an example because they actually have a
field for “reading”, which is then used for sorting the name. Basically
that’s the feature I had in mind for gramps, not the actual “it
automatically detects what kind of name it is and how it is read”. So
the user always would have to insert everything manually.
> In the case of my name, for the family name there are 14 different
> readings. For a Romaji reading, there are 4 different Kanji. For my
> given name there are 20 readings and from only the correct reading
> there are 35 Kanji. Oddly, it is much more difficult to get the
> correct Kanji from a Romaji reading. That's the simple part. Not so
> simple actually. I left some stuff out. Basically the point is that
> it's easy to make errors, especially for non-speakers of Japanese. And
> many Western systems have default font subsitutions set up to use
> Chinese Hanzi mappings which give the wrong Kanji. That's a critical
> error most Westerners would never notice. And if look-ups and research
> using Kanji fail because of system errors then the readings must be
> used and for Romaji readings it's not always so easy. Errors can
> easily be compounded and quickly become hopelessly impossible to
> correct without starting over. I know from experience.
Well, but if you insert a new person, you can just use it for yourself,
can’t you?
So for example, if you have Tanaka (family name) Ichirō (given name),
you would just insert it in gramps like this:

Family name: 田中
Family name reading: たなか (or Tanaka, depending on your preference)
Given Name: 一郎
Given name reading: いちろう (or Ichirō)

Gramps wouldn’t need to do any conversion by itself. If it is assumed
that the user can read Japanese names, his computer supports Japanese
input, then he can just insert it into the database by himself. This
would then be no different than inserting an English name, except from
the additional fields “reading”, and the different display afterwards.
But conversion itself is not necessary.

This would of course be the ame for Korean names, or Chinese ones
(Although I have to admit that I guess most Chinese would not even need
the reading fields, because they can read names without problems – It
would be nice for a non-Chinese speaker, who has a chinese ancestor,
though. So for example, if your grand-grand-mother was named 王小蘭, you
would put in the reading field Wáng Xiǎolán (I don’t know if this is a
correct Chinese name). Then you can read it, even though you may only
have a basic knowledge of Chinese, but want to have the Chinese
characters in your file).

For Chinese, it may even be nice to include all the previous names which
were in use until 100 years ago. I guess you can write these kinds of
names already with the name editor (it seems quite advanced to me), but
I guess it would be easy if, as I thought with the different
language/country templates, there would already be a preset for it. So
if I select Chinese, I would automatically have entries with the
different labels, for example for courtesy name, school name, official
name etc. This would then be easier than to do it manually with the name

Also, completely East-Asian unrelated, this reading field can be used
for  every name with a foreign script. If you have a persian ancestor,
you would put the arabic script in the ordinary field, and the
transcription in the reading field.
This is of course all under the assumption that the user can type (or at
least copy-paste) the foreign-script name into Gramps. If he can do
that, there would be no difficult technology needed, because Gramps does
not have to convert anything.

> I don't work on Gramps. I don't speak for the developers so please
> don't regard my comments as theirs. I can't say exactly how
> implementing your idea would be done. I can say that in order to get
> it right it's not likely to be as simple as it sounds and even if it's
> simple it won't stay that way because people will want enhancements.
> But I doubt anything can be done without a feature request being made
> so there's a good place to start.
> jd
Well, I could write a more detailed idea. That wouldn’t be a problem.


Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and
threat landscape has changed and how IT managers can respond. Discussions
will include endpoint security, mobile security and the latest in malware
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
Gramps-users mailing list