Hi all, xifan hol seems to have a variant where k is mapped to {q} and q to {Q}, and one where k is mapped to {Q} and q to {q}. The k -> {q}; q -> {Q} mapping appears to be more prevalent, but the other one is quite common, too. I can see pros and cons to both systems, although I personally prefer the k -> {q} version. For my TTS experiments, I’ve been using this particular xifan hol variant, but before I release the helper tools I’ve been using for these experiments to the public, I’d like to make sure that choosing this convention won’t be confusing or contentious, if there’s a large enough population of Klingonists that use k -> {Q}. I could always use XIFAN HOL instead, since I’ve never see. XIFAN HOL with K -> {Q} (since the main advantage of k -> {Q} is not having to explicitly convert q -> {q}.
boQwI' gives you an option of using k -> {q} or {Q}, is that something feasible on your end? I prefer to use boQwI' with k -> {Q}, since I can just use q for {q}. But I have noticed that some fonts use k for {q} and q for {Q} (or, technically, they use K for {q} and Q for {Q}, since the fonts map pIqaD to capital letters). I think I might be in the minority on this one. (Obviously, we should ask Okrand to make a ruling one way or another, just to sort it out.) On Sat, Jun 2, 2018 at 1:14 PM, Daniel Dadap <daniel@dadap.net> wrote:
Hi all,
xifan hol seems to have a variant where k is mapped to {q} and q to {Q}, and one where k is mapped to {Q} and q to {q}. The k -> {q}; q -> {Q} mapping appears to be more prevalent, but the other one is quite common, too.
I can see pros and cons to both systems, although I personally prefer the k -> {q} version. For my TTS experiments, I’ve been using this particular xifan hol variant, but before I release the helper tools I’ve been using for these experiments to the public, I’d like to make sure that choosing this convention won’t be confusing or contentious, if there’s a large enough population of Klingonists that use k -> {Q}.
I could always use XIFAN HOL instead, since I’ve never see. XIFAN HOL with K -> {Q} (since the main advantage of k -> {Q} is not having to explicitly convert q -> {q}. _______________________________________________ tlhIngan-Hol mailing list tlhIngan-Hol@lists.kli.org http://lists.kli.org/listinfo.cgi/tlhingan-hol-kli.org
On 6/2/2018 3:08 PM, nIqolay Q wrote:
boQwI' gives you an option of using k -> {q} or {Q}, is that something feasible on your end?
I prefer to use boQwI' with k -> {Q}, since I can just use q for {q}. But I have noticed that some fonts use k for {q} and q for {Q} (or, technically, they use K for {q} and Q for {Q}, since the fonts map pIqaD to capital letters). I think I might be in the minority on this one.
I prefer /k/ -> *q*. In transliteration, /k/ that isn't part of a cluster nearly always ends up as *q,* while /kr/ tends to be the transliteration of *Q. *Since native English speakers consider /k/ simpler than consonant clusters of /k,/ and since /k /is more common than /q,/ and /q /is odd to an English speaker with a following /u,/ I think /k/ is more suitable for *q.* It's just a bias for an arbitrary choice though, not a logical argument for it.
(Obviously, we should ask Okrand to make a ruling one way or another, just to sort it out.)
A ruling on what? Okrand isn't in charge of keyboard mappings or anything like that; he isn't even in charge of what Klingon LOOKS like. All he knows is what Maltz tells him, and Maltz will just say Klingon computers already have *pIqaD* on them, so there's no mapping issue. -- SuStel http://trimboli.name
On Jun 2, 2018, at 14:24, SuStel <sustel@trimboli.name> wrote: On 6/2/2018 3:08 PM, nIqolay Q wrote: boQwI' gives you an option of using k -> {q} or {Q}, is that something feasible on your end?
Right, when I ported boQwI' to iOS I made this a choice as well, but in this case, I’m using xifan hol as the internal representation of Klingon text, so I have to pick one or the other. It’s easy enough to throw up a front end to accept tlhIngan Hol (or even the opposite xifan hol), so I guess ultimately it doesn’t matter what choice I make except for people who are preparing corpora for training. I just get hung up on these types of choices when the choice is a purely personal preference, and I feel like I’m imposing my own preference on others.
I prefer to use boQwI' with k -> {Q}, since I can just use q for {q}. But I have noticed that some fonts use k for {q} and q for {Q} (or, technically, they use K for {q} and Q for {Q}, since the fonts map pIqaD to capital letters). I think I might be in the minority on this one.
I prefer k -> q. In transliteration, k that isn't part of a cluster nearly always ends up as q, while kr tends to be the transliteration of Q. Since native English speakers consider k simpler than consonant clusters of k, and since k is more common than q, and q is odd to an English speaker with a following u, I think k is more suitable for q. It's just a bias for an arbitrary choice though, not a logical argument for it.
Yes, this is essentially the same reason my preference is for k = q.
(Obviously, we should ask Okrand to make a ruling one way or another, just to sort it out.)
A ruling on what? Okrand isn't in charge of keyboard mappings or anything like that; he isn't even in charge of what Klingon LOOKS like. All he knows is what Maltz tells him, and Maltz will just say Klingon computers already have pIqaD on them, so there's no mapping issue.
Yeah, I don’t think there needs to be an official position on this one way or another; I just wanted to get an idea as to how annoyed people who prefer k = Q would be if you had k = q forced on you.
-- SuStel http://trimboli.name _______________________________________________ tlhIngan-Hol mailing list tlhIngan-Hol@lists.kli.org http://lists.kli.org/listinfo.cgi/tlhingan-hol-kli.org
On Sat, Jun 02, 2018 at 03:24:51PM -0400, SuStel wrote:
On 6/2/2018 3:08 PM, nIqolay Q wrote:
boQwI' gives you an option of using k -> {q} or {Q}, is that something feasible on your end?
I prefer to use boQwI' with k -> {Q}, since I can just use q for {q}. But I have noticed that some fonts use k for {q} and q for {Q} (or, technically, they use K for {q} and Q for {Q}, since the fonts map pIqaD to capital letters). I think I might be in the minority on this one.
I prefer /k/ -> *q*. In transliteration, /k/ that isn't part of a cluster nearly always ends up as *q,* while /kr/ tends to be the transliteration of *Q. *Since native English speakers consider /k/ simpler than consonant clusters of /k,/ and since /k /is more common than /q,/ and /q /is odd to an English speaker with a following /u,/ I think /k/ is more suitable for *q.* It's just a bias for an arbitrary choice though, not a logical argument for it.
I lean strongly towards /q/ -> <q>, /Q/ -> <k>, due to my exposure to Swiss German (in which written* <k> often corresponds to /Q/) and Arabic (which is usually transliterated using <q> for an unaspirated uvular stop very similar to the Klingon /q/). *Swiss German has no formal orthography, and is generally written using a combination of German-based phonetic spelling and the traditional spellings of the Standard German equivalents of Swiss German words. Since Swiss German /Q/ usually corresponds to Standard German /k/, that means that <k> often ends up respresenting /Q/.
(Obviously, we should ask Okrand to make a ruling one way or another, just to sort it out.)
A ruling on what? Okrand isn't in charge of keyboard mappings or anything like that; he isn't even in charge of what Klingon LOOKS like. All he knows is what Maltz tells him, and Maltz will just say Klingon computers already have *pIqaD* on them, so there's no mapping issue.
Agreed. The xifan hol system is very clearly a Terran invention, and Mar{rc/tlz} has jurisdiction only over material that originates on Qo'noS. - SapIr
On Sat, 2 Jun 2018 at 21:08, nIqolay Q <niqolay0@gmail.com> wrote:
boQwI' gives you an option of using k -> {q} or {Q}, is that something feasible on your end?
I prefer to use boQwI' with k -> {Q}, since I can just use q for {q}. But I have noticed that some fonts use k for {q} and q for {Q} (or, technically, they use K for {q} and Q for {Q}, since the fonts map pIqaD to capital letters). I think I might be in the minority on this one.
Everyone keeps talking about fonts but that's only relevant if you're typesetting {pIqaD}, presumably on your PC (or laptop). Many people (maybe most, in some demographics) are primarily on their mobile devices. The advantage of the "k" -> {Q} mapping is that it's effectively "optional" on soft keyboards: if you type Klingon as you normally would, you wouldn't even notice it was enabled. However, {Q} and {'} are on shifted keys, and mapping "k" and "z" to them allows a user to type them without shifting, which is a nuisance on a small device. It's also handy to be able to type "x" for {tlh} and drop "h" when its presence is unambiguous (like after "c"). With the "k" -> {Q} mapping, you can leave "xifan hol" mode on and the typical user (who knows nothing about font mappings) won't be confused when they're typing things in as-is. That's why that's the default, and swapping {q} and {Q} (i.e., "k" -> {q}, "q" -> {Q}) is an advanced option. -- De'vID
On Jun 7, 2018, at 07:16, De'vID <de.vid.jonpin@gmail.com> wrote:
On Sat, 2 Jun 2018 at 21:08, nIqolay Q <niqolay0@gmail.com> wrote: boQwI' gives you an option of using k -> {q} or {Q}, is that something feasible on your end?
I prefer to use boQwI' with k -> {Q}, since I can just use q for {q}. But I have noticed that some fonts use k for {q} and q for {Q} (or, technically, they use K for {q} and Q for {Q}, since the fonts map pIqaD to capital letters). I think I might be in the minority on this one.
Everyone keeps talking about fonts but that's only relevant if you're typesetting {pIqaD}, presumably on your PC (or laptop).
Many people (maybe most, in some demographics) are primarily on their mobile devices. The advantage of the "k" -> {Q} mapping is that it's effectively "optional" on soft keyboards: if you type Klingon as you normally would, you wouldn't even notice it was enabled. However, {Q} and {'} are on shifted keys, and mapping "k" and "z" to them allows a user to type them without shifting, which is a nuisance on a small device. It's also handy to be able to type "x" for {tlh} and drop "h" when its presence is unambiguous (like after "c"). With the "k" -> {Q} mapping, you can leave "xifan hol" mode on and the typical user (who knows nothing about font mappings) won't be confused when they're typing things in as-is. That's why that's the default, and swapping {q} and {Q} (i.e., "k" -> {q}, "q" -> {Q}) is an advanced option.
Yes, that is the thing I like about k=Q: it’s possible to implement a parser that can handle “normal” orthography (which I hate, by the way: I was trying to think of the worst thing you could possibly write in a sans-serif font the other day and came up with “lIlIllI'”: anybody have any other favorites?) or xifan hol transparently. I think for the particular project I have in mind, since the xifan hol encoding serves a practical role rather than a purely convenience based one (protecting the text from being case-mangled and avoiding problems with the qaghwI'), and it’s mandatory rather than optional, there’s really no practical advantage to either mapping. I’ll probably go with k=q, since that’s what I’ve already been using and it does seem to be more established based on the small sample size of Klingon software I’ve encountered so far.
-- De'vID _______________________________________________ tlhIngan-Hol mailing list tlhIngan-Hol@lists.kli.org http://lists.kli.org/listinfo.cgi/tlhingan-hol-kli.org
jatlh De'vID:
The advantage of the "k" -> {Q} mapping is that it's effectively
"optional" on soft keyboards: if you type Klingon as you
normally would, you wouldn't even notice it was enabled.
...
With the "k" -> {Q} mapping, you can leave "xifan hol" mode on
and the typical user (who knows nothing about font mappings)
won't be confused when they're typing things in as-is. That's
why that's the default, and swapping {q} and {Q} (i.e., "k" -> {q},
"q" -> {Q}) is an advanced option.
I very strongly prefer <k> -> {q}, <q> -> {Q}. In my mind the {q} is a "k" sound, but the {Q} is some strange harsh sound which is appropriately represented by an unusual letter (like a capital in Okrandian transcription and a <q> in xifan hol). I much prefer to type a <k> for the "k" sound and use the <q> to represent that strange sound. I understand what De'vID is saying about <q> -> {q} allowing one to type in either Okrandian or xifan hol without any confusion over letters and I suppose I could get used to whatever, but I definitely have a strong preference (and habitualization) for the more traditional way of doing it. Jeremy (not David)
Am 02.06.2018 um 19:14 schrieb Daniel Dadap:
Hi all,
xifan hol seems to have a variant
Although there may be variants, there's two thing to remember: First, most of the fonts use the k=q q=Q variant. Second, you will certainly never get any official solution for the mapping, but I suggest you go back to to the roots, to the very first ever published Klingon font. That was the one released by the KLI in the nineties. That one also uses the k=q/q=Q variant, so that would my basic reason to follow that pattern. Most fonts that use different mappings are based on that first one, and have just changed the original one. As you mentioned before, some use lower case, some upper case. The KLI's font used uper case because the lower case is maped with a different font type. So that should be XIFAN HOL. The Klingonska Akademien font uses lower case, xifan hol, which might be easier to type, but with using caps lock makes it no difference. Another thing that might b confusing is that some fonts map the apostrophe differently (' or z), so that should also be considered. I don't know if you have checked the Klingon wiki on that, because here's quite alot about pIqaD and also about conversion. http://www.klingonwiki.net/En/PIqaD http://www.klingonwiki.net/En/Fonts http://www.klingonwiki.net/En/Perl http://www.klingonwiki.net/En/Unicode -- Lieven L. Litaer aka the "Klingon Teacher from Germany" http://www.klingonisch.de
On Jun 3, 2018, at 06:57, Lieven L. Litaer <levinius@gmx.de> wrote:
First, most of the fonts use the k=q q=Q variant.
Right, this is what I meant when I said that variant appears to be prevalent.
As you mentioned before, some use lower case, some upper case. The KLI's font used uper case because the lower case is maped with a different font type. So that should be XIFAN HOL.
Yeah, I’m really talking about more than just fonts here. In this case I’m using xifan hol as a kind of interchange format to avoid problems with case sensitivity and ' as a letter instead of punctuation.
The Klingonska Akademien font uses lower case, xifan hol, which might be easier to type, but with using caps lock makes it no difference.
It does make a difference if you’ve remapped caps lock, as many people do. I often use caps lock as compose, since it’s the only sane way to type accented characters from multiple different Latin alphabet using languages within a single keyboard layout. Otherwise, I wouldn’t have had any reservations about using XIFAN HOL, since I’ve never observed any K=Q; Q=q variant of that.
Another thing that might b confusing is that some fonts map the apostrophe differently (' or z), so that should also be considered.
Right, I’m using z, since I don’t want to spend unknown amounts of time hunting down bugs in the DNN TTS software where treating ' as a letter becomes problematic. I had already decided on using xifan hol anyway, because a cursory inspection of the source code suggested that one of the analysis passes normalizes the case of letters, and I wanted to avoid problems with q vs Q, and also words like vengHom or nenghep.
I don't know if you have checked the Klingon wiki on that, because here's quite alot about pIqaD and also about conversion.
http://www.klingonwiki.net/En/PIqaD http://www.klingonwiki.net/En/Fonts http://www.klingonwiki.net/En/Perl http://www.klingonwiki.net/En/Unicode
The wiki did come up when I searched for “xifan hol” to try to see if this had been discussed before, but I didn’t see any place where the existence of multiple variants was discussed. I can add a page dedicated to xifan hol later.
-- Lieven L. Litaer aka the "Klingon Teacher from Germany" http://www.klingonisch.de
_______________________________________________ tlhIngan-Hol mailing list tlhIngan-Hol@lists.kli.org http://lists.kli.org/listinfo.cgi/tlhingan-hol-kli.org
participants (7)
-
Daniel Dadap -
David Holt -
De'vID -
kechpaja -
Lieven L. Litaer -
nIqolay Q -
SuStel