[Tellico-users] sorting and formatting

[robbystephenson]

Hey,

So with the sorting and formatting needing some work in the next release, I
thought I'd ask about best behavior.

Leaving aside bugs, in Tellico 1.3, title columns were sorting according to
typical english sorting, with the articles like "the" ignored. And it sorted
like that whether or not the auto-formatting was turned on.

Tellico 2.0 had broken formatting, and when I fixed it for 2.1, the sorting
for titles worked as before, but the sorting for names didn't follow the
formatting rules.

What makes it a little tricky is that some data sources provide names with
family name first. So it gets entered that way, whether or not auto-
formatting is turned on.

So my questions are:
1) Would you expect the title sorting to ignore articles like "the" even
when auto-formatting is turned off?
2) After fixing the name formatting, the sorting will follow whether or not
auto-formatting is on, is that what you would expect? Or should it match the
title sorting?
3) Does everything work as you expect it to for other languages?

My inclination is to make the sorting match what is displayed. That is, if
the auto-format is on, then sorting is based on the formatted values. If it
is not on, then sorting is based on whatever was entered, either directly by
the user or from the data source.

That make sense? Comments?

Thanks!
Robby
_______________________________________________
tellico-users mailing list
tellico-users@...
https://mail.kde.org/mailman/listinfo/tellico-users

[Thomas Ronayne]

tellico-users-request@... wrote:
> Hey,
Mornin'

> So my questions are:
> 1) Would you expect the title sorting to ignore articles like "the" even
> when auto-formatting is turned off?
>  
Yeah, I kind of would; my library consists of 1,150 distinct titles
(more than a few representing multi-volume sets) and there are quite a
few "The's" in there and, mostly, the-something-or-other doesn't make a
lot of sense when you're searching for a title. You know, think
old-fashioned index cards down at the library.
> 2) After fixing the name formatting, the sorting will follow whether or not
> auto-formatting is on, is that what you would expect? Or should it match the
> title sorting?
>  
Hmm. I pretty much expect that name sorting would be based on parsing
for a comma no matter what.
> 3) Does everything work as you expect it to for other languages?
>
> My inclination is to make the sorting match what is displayed. That is, if
> the auto-format is on, then sorting is based on the formatted values. If it
> is not on, then sorting is based on whatever was entered, either directly by
> the user or from the data source.
>  
I was perfectly happy with the way Tellico handled names and titles in
versions 1.x; everything made sense, looked "proper," and all was right
with the world. I'm thinking about the way I was taught to order names
and titles and subjects in school (hey, about 50 years ago, but there
your are). Seemed like a good thing then, still does, methinks.

Hope this helps some.

Oh, by the way, you're making an excellent product and is useful in so
many ways; a real pleasure to have and use. Don't beat yourself up (or
let anybody else, either).

Thomas

--
A riddle, wrapped in a mystery, inside an enigma but that's my story and I'm stickin' to it.

_______________________________________________
tellico-users mailing list
tellico-users@...
https://mail.kde.org/mailman/listinfo/tellico-users

[robbystephenson]

On Sunday 08 November 2009, Thomas Ronayne wrote:
> > 2) After fixing the name formatting, the sorting will follow whether or
> > not auto-formatting is on, is that what you would expect? Or should it
> > match the title sorting?
>
> Hmm. I pretty much expect that name sorting would be based on parsing
> for a comma no matter what.

Names are a lot harder than titles. If the value has no comma, then a
decision has to be made whether to try to determine the appropriate family
name to sort on.

> I was perfectly happy with the way Tellico handled names and titles in
> versions 1.x; everything made sense, looked "proper," and all was right
> with the world. I'm thinking about the way I was taught to order names
> and titles and subjects in school (hey, about 50 years ago, but there
> your are). Seemed like a good thing then, still does, methinks.

Tellico 1.x sorted names based on whatever was displayed. So if auto-
formatting was turned off, then no extra effort was made to determine family
name. it just sorted alphabetically by whatever was saved for the entry.

It gets tricky with other languages and  nationalities. Trust me. :)

Robby
_______________________________________________
tellico-users mailing list
tellico-users@...
https://mail.kde.org/mailman/listinfo/tellico-users

[Thomas Ronayne]

Hi, Robby,

Not trying to be a pain but I'm thinking about the way the Big Three
(let's say Amazon, LOC and ISBNdb) return information to you.

Just picking an arbitrary ISBN, 0-399-15504-X (Robert B. Parker,
/Resolution/), Amazon returns "Robert B. Parker," ISBNdb returns
"Parker, Robert Andrew" (which is actually his name) and LOC returns
"Parker, Robert B." using the Tellico internet search function; thus
there either is or is not delimiter in the string (in this case a comma
to delimit surname).

Well, OK, there's one -- now, how about a toughie: George Cordon Byron,
/Childe Harold's Pilgrimage/. George Cordon Byron, AKA Lord Byron.
Arrgghh! One begins to see the problem; do you sort it as if it were
"Byron, Lord" or "Lord, Byron?" What sort of rule do you write for
titles (his first name is not "Lord," it's George but generally nobody
knows that and, in England, he'd be called "Byron," "M' Lord" or
referred to as "George, Lord Byron." or some such) A puzzlement. The
Brits have all kinds of goofy fun things -- how, for example, does
Leicester get pronounced "Lester" I do not fathom.

Then, one goes on to the doctors (either MD or Ph.D.) and deal with
those (actually, Robert B. Parker is or has a Ph.D., but he doesn't
bother with that Dr. stuff). Those are actually easier rules to deal
with; when I have to for data base name parsing I just strip out any
pattern that matches any medical or academic title; Doctor, Dr., M.D.,
MD, MA, PHD, Ph.D., etc. -- the rule being that a title is not a
persons' name but that doesn't seem to count in England). I do the same
with military or civilian rank -- your name isn't "General," it's Your
Blasted Name, dang it! And, of course, if you want to be really picky
you can append the title to the end of the name string; e.g., George  
Smith Patton,  Jr. (also George Smith Patton III), General. Searching
for ISBN 0-395-73529-7 returns  George S. Patton Major General, /War As
I Knew It/ from Amazon, Patton, George S. (George Smith) from LOC and
Patton, George S. from ISBNdb. Rank only appears in the Amazon search;
I'd be more likely to accept the LOC rule. It's kind of the same thing
you do in genealogy: what's your name on your birth certificate.

English is bad enough -- I can't imagine (because I have zero knowledge
of other languages' sorting rules) what one would do with French,
German, Polish, Russian or who knows what.

Truly a puzzlement.

Thomas

--
A riddle, wrapped in a mystery, inside an enigma but that's my story and I'm stickin' to it.

_______________________________________________
tellico-users mailing list
tellico-users@...
https://mail.kde.org/mailman/listinfo/tellico-users

[robbystephenson]

On Wednesday 11 November 2009, Thomas Ronayne wrote:
> English is bad enough -- I can't imagine (because I have zero knowledge
> of other languages' sorting rules) what one would do with French,
> German, Polish, Russian or who knows what.

Exactly. That's why I want to make it useful for English-speakers (since
that's the style that I'm familiar with) and easy to turn off for everyone
else. :)

Robby
_______________________________________________
tellico-users mailing list
tellico-users@...
https://mail.kde.org/mailman/listinfo/tellico-users