Getting better and better.

Any chance of exchanging t-f and d-v? Then all and only the stops would have long vertical lines, and the fricatives would all be short.

Hey, good idea. I might just do that. Who are you stranger?

Consonants were selected based on readability, writ ability, and synaesthetic intuition, see here: http://en.wikipedia.org/wiki/Bouba/kiki_effect.
The glyphs are paired such that by rotating 180 degrees, one can find the unvoiced counterpart of a voiced letter or vice versa. This is not strictly true for xy, which has no counterpart and therefore looks identical both ways, and the pairs ly/ry and my/ny, which maintain symmetry but are paired for reasons other than vocalisation (namely, for being the only 2 liquids and the only 2 nasals, respectively). Pairing these is also a hangover from Srilermorna which uses the sane pairs, and helps to minimise the number of glyphs.

Vowels were chosen for minimum effort while writing, and to be distnguishable at small sizes. This is a hard goal to chase, and the shapes presented here may still change. Diphthongs were reduced in number from 14 to 4 with the aid of the semivowel system, detailed later. Full vowels exist only for transcrbing names and nonlojban text, this is to give them a distinct visual look compared to “legal” lojban, and to allow a bit more flexibility with mashing vowels about to get those foreign sounds. I have tried to maintain some similarity between the diacritic and full forms of each vowel, and also between those and their semivowel counterparts.

Semivowels

Easily the most drastic feature here, semivowels are extracted from the proper pronunciation of diphthongs with leading {i} or leading {u}, which are IPA [j] and IPA [w] respectively, according to CLL. This seemed to me to be counterintuitive to the idea of audio-visual isomorphism, and hence I had the idea to seperate the secondary pronunciations into their own glyphs, reducing {i} and {u} to a consistent pronunciation of [i] and [u].This is a minor technical distinction, but two important side-effects result, these are:

• The total elision of the comma
• Reducing the number of diphthong glyphs to just 4.

The legal dipthongs in lojban are:
au, ai, ei, io, ia, ie, ii, io, iu, ua, ue, ui, uo, and uu.
Using the consonant-like
[w] (shorthand “w”) and
[j] (shorthand “q”),
everything beginning with i and everything beginning with u can be written as consonant-vowel combos,
qa, qe, qi, qo, qu and
wa, we, wi, wo, wu.
This only leaves au, ai, ei and oi, which are not worth making trailing-[w] or -[j] glyphs for and are quite managable and learnable as units on their own.
CLL also specifies that {iy} and {uy} are legal in names. These can be written now as qY and wY.

The comma is used only in names to seperate vowels which would otherwise spell out diphthongs. Now that all diphthongs can be covered as either a full diphthong glyph or as a combination with q or w, any full vowels written next to each other can be interpreted as having a comma between them. Thus, we no longer need the comma for transcrbing words but it is included here for use at the writer's discretion, either stylistically in names or to transcribe silence.

The dot is technically “an optional reminder” to pause at a point dictacted by the grammar – it need not actually be written. Therefore, we won't. The denpabu is only useful to us as a kind of “blank consonant”, something to seat single vowels on when we want them on their own. This usage happens to coincide with the rules for denpabu anyway, which forms an elegant system where the required denpabu is incorporated into the vowel.

Don't freak out. Most punctuation here is totally optional. All except the single stress marker, that is, which is what you'd use as you would capital letters to mark nonstandard stress in nonlojban words.

All other marks are for transcribing speech dynamics, in fiction for dramatic effect, for example. None of them are meant to change the interpretation or meaning of the actual sentence.

Ba'ebu
The stress marker has three stages, which work on a relative scale. One dot loosely represents the difference between a normal unstressed lojban syllable and a normal stressed lojban syllable. Therefore you can play it by ear a little using the double and triple ba'ebu, for example to stress an entire world without ruining the stress pattern, use single ba'e on the unstressed syllables and double ba'e on the stressed syllable. Or perhaps a shorter method would be to use double on the stressed syllable and leave the rest, just to give it a bit more emphasis. The doesn't supercede the use of the actual word {ba'e}, as the ba'ebu dot is only for transcribing speech dynamics.

Stretching
The horizontal underline denotes pronouncing a syllable longer than normal. The line from a mid-word syllable should extend to underneath the next syllable, or at the end of the word it should finish before the next word. This is simply a visual aid, because technically even if you accidentally run the line too long you can still parse the sentence. Theres also a second way to write it (see diagram posted after this block) where the line runs under the rest of the letters for as long as you like, without introducing unnecessary space between characters. While this method is neater, it kind of defeats the intuitiveness of the marking because you have to look ahead to see how long the stretch is. Writer's discretion I guess.

Intonation
Tones are optional symbols again to assist in dialogue dynamics. They are useful for doing things like turning “xu” into “xu?” (of course this is not consistent across different cultures, use wisely) and trying to carry some idea of mood across in written speech, as English speakers sometimes do by using italics or other punctuation or typographical devices.
The available tones are up, down, up-down, down-up, and I'm considering including up-down-up and down-up-down. Because you never know.

Minor changes, updates and proposals.
Some glyphs have been tweaked in non-fundamental ways as well (not illustrated)

This system is now known as zbalermorna, for assembling-letter-patterns.

Perhaps diphthongs could be formed by writing both simple vowel marks, like so?

The c and j form is the only one which requires two pen-strokes. Perhaps you can come up with some other form that only requires one stroke?

These diphthong are a pretty good idea Anon. I had hoped to make them simpler. You also have a point about c and j, and although i'm quite fond of the shapes, you have a valid point and I will likely take you up on that.
Furthermore I'd like to remove the stems altogether. A fairly major overhaul of all the glyphs is expected. I haven't worked on this project in months, so apologies to anyone who has become used to it as it is :)

>>211

When writing in a non-Lojban text (meaning following ''' zo zoi '''), the text follows the rules of that language. I do not know if this means that it should be written strictly in only that language's alphabet, but it definately means that it must allow for that language's sounds. This means that if you intend this writing system to be able to write English's "the" it needs a voiced dental fricative. Seeing as how there are so many possible sounds in so many languages (which even vary in pronounciation from area to area), I think that you would be fighting a losing battle. Lojbanizations must be allowed int his system (for cmevla, for example), but I think that it would be better just to let other languages to use their own writing systems. This would distinguish the foreign text as foreign by aesthetics anyway.

>>223

I know that everyone says that the letter "n" represents the alveolar nasal, the velar nasal, and possibly others (such as the dental, palatal, etc.)

However, there is a strong distinction between engma and en for me. Could you possibly create an optional indicator that shows how the speaker/narrator of the text says it (possibly to show their dialect, sort of like omissions in English such as "I aint got no'in' teh hide")?

>>208

I too like the idea. If it does not mess up the 180 degree-rotation idea (which I believe it does not).

>>222

How does one separate words? Stress would either need to be marked (in text lacking any cmevla). With cmevla, both stress and the end of cmevla (marked by the denpabu) must be noted.

What about numbers? Do you plan on crafting a number system specifically designed for ease, Lojban's number/coutning system(s), mekso?

>>264
Numbers are already just one glyph as they're all CV(V). Words are marked by stress or spacing.

>>261
Indeed. Foreign languages will be written in their own alphabets.
>>262
There is no distinction between velar and alveolar nasals in lojban, but then i've included tonal marks for the sake of dramatic expression so i might keep this in mind. If theres an elegant way to do it i'll see how it turns out.
>>263
It doesnt, but I will. I've had to change a couple of letters due to writability issues, and it has ruined the 180˚ relationship for those.
>>264
Words are separated by spaces. If one has some kind of aversion to spaces, they can write in lokadin by using the stress marker.
With cmevla, one writes in non-diacritic mode, distinguishing cmevla from jbokaivla visually. Usually one would not require denpabu because they are implicit in this system, however if writing with no spaces, it may be necessary to specify them explicitly for cmevla. The denpabu is still available to use if required.
Note that when writing cmevla with spaces, the consonant gives away the end of the cmevla, and so the denpa may go back to being implied.

As for numbers, I have thought about it, but lojban is supposed to look like it is said, and writing diacritically means that single-syllable cmavo like numbers come out as one character anyway.

>>270

Thank you (for all of the responses). Speaking of here-post #262, how can one mark Roman-alphabet Lojbanic text to say that it is improper (in the case of a quotation, where zo na'i cannot be used)? Is there a way? We probably should try not to support it, but it will be there- so we have to find a manner to at least represent it (even if grudgingly). Some really simple way (similar to the comma in English, but possibly less ambiguous- one specifcally for omission/eliding, one for wrong sound, one or more that shows the span of the error, etc.)

I'm not quite sure to what you are referring here Anon. From what I understand, this is getting too far outside the scope of what lojban is supposed to be able to cope with, let alone it's writing system. Foreign text just is as it is. I dont think lojban or any lojbanic writing system should go out of its way to cope with errors in foreign text.

Oh, I played around with making symbols for numerals as well. Not gonna happen, sorry.

Wait i've misread what you said. Okay i'm not too sure there. Please provide an example.

>>274

lo'u ... le'u quotation:
http://jbotcan.org/cllc/c19/s9.html

>>274

Sorry, my curiosity got ahead of me. The ungram. quotes are perfect. I was merely asking how Lojban writes ungrammatical statements (which really has little to do with the writing system, I am afraid).

Never mind.

:)

Thank you.

I don't wish to be rude, but I really wish you wouldn't.
This is a work in progress, and as such, I'm deliberately withholding my font until its finished, so as to avoid poeple using, learning and propagating obsolete or incorrect fonts.
Because the glyphs aren't final, unofficial fonts aren't simply different typefaces for the system, they're effectively totally different systems.
In fact, in a day or so I'm about to roll out significant changes which will (regrettably) render your hard work null and void, as well as make it obsolete and therefore misleading for anyone who downloaded your file already.
For these reasons, I request that you take your font file down so we can have a controlled outcome. I do not wish this project to fork, although I admire your open source spirit ;)
I apologise for taking so long btw - been busy.
Lak

>>281

Alright, I can understand that.

Don't feel bad about nullifying my hard work; it really wasn't that hard, and I learned all kinds of new stuff about fonts and how to make them. It was fun.

.ui well thats good then. I am of the opinion that learning something can make anything worthwhile.
I'm hoping to put up the new master reference and a sheet detailing and comparing the new elements. That should be a few hours from now hopefully. Then we're one step closer to the font being freely available :)
Lak