Thu 24 Nov 2011 11:49:36 PM UTC, comment #12:
Concerning the size of the balls on various letters.
They are indeed different.
One interpretation is that the authors are sloppy and stupid, and have no sense of consistency...don't know how to make letters feel like they're of the same typeface.
In fact, the authors of FreeFont seem to have the same problem in the Latin range. I arranged the letters y, f, and j with their knobs aligned.
See yfj-FS.jpg
All quite different! Like they were from different typefaces! How pathetic!
And look at this!
yfj-Times.jpg
Microsoft Times New Roman! Same problem! Somebody should teach these people how to make letters!
The less completely naive interpretation of course, is that the type designers know something, and are pulling a psycho-visual trick that goes quite beyong high school geometric congruency.
There are many many books on these topics, and they're a lot of fun to read and look at. I have a few; I could send you a list.
(By the way: I am personally no artist. I couldn't have made those things the right size. But I kind of see what they're doing.)
Now we've dispensed with the "balls size" question...
Which of your complaints remain?
(file #24445, file #24446)
|
Wed 23 Nov 2011 08:32:30 AM UTC, comment #11:
Hi Alexey
OK, I think we agree on the basics, the misunderstandings are due to my crass ignorance of typographical terminology.
Of course the bottom part of ZE and KSI is different, it has to be so, but the matter of contention is the overall feeling, I mean, put the two characters together on the same line, they don't look like they belong to the same typeface!
I said there should be similarity, not identity between ZE and KSI, but now on the upper bowl ZE has a serif, while KSI has a knob; on the other hand KSI does have a serif at the intersection of the upper and lower bowl, while ZE has none. Finally, the width, the space the two characters occupy is different while even in Church Slavonic it's the same, I mean, the structure, the three-like shape. I hope we can agree on that.
Of course the swashes in KSI and ZEMLJA are different, again I said they should be uniformed, brought to a common style, just take a look at the two characters on the same line. Does it feel like the same typeface? I don't think so!
To my humble eyes, KA, ZHE and IZHITSA appear to have the same balls, while PSI and OMEGA do not. Maybe for OMEGA it's justified but for PSI I don't think so.
And while we're at balls (and since you're an expert), the ball under (Cyrillic) U is significantly larger, but I'm accustomed to having it the same as ZHE, what do you think about it?
Greetings
|
Wed 23 Nov 2011 05:37:57 AM UTC, comment #10:
Hi Alessandro,
I think any references to the Greek script are completely irrelevant here: yes, Cyrillic letters have been derived from Greek ones, but the graphical forms they acquired in poluustav can't be explained by their origin. I also had no intention to refer to the original (sharp) form of Cyrillic ZEMLYA. However if you take any correctly designed Old Slavonic or Church Slavonic font (including one used in STAROSLAVJANSKI SLOVAR'), you can see the design of ZE (the standard, "rounded" form) and XI is actually quite different, although these letters may look similar to an inexperienced eye.
The key difference between ZE and XI is that in ZE the lower bowl terminates above the baseline, while in XI it is continued with a "tail" going below. Thus there is an additional white space at the bottom part of the glyph, and to compensate this we need a heavier "knob" at the conjunction of the two bowls. In a "civilian" font this "knob" would usually have a form of a serif pointing down, and this is how the glyphs in Free Serif are designed. So reproducing the corresponding part of ZE would just be wrong here.
BTW I have no ideas what do you mean stating the swashes in KSI should be uniformed with the sharp ZEMLYA: they just cannot be the same for the obvious reason that the later character has a straight bottom, while in XI it is curved.
I also see essentially no difference between the balls in KA and IZHITSA, except that in IZHITSA it is a bit larger (which is natural if you take into account an overall shape of the glyph).
|
Tue 22 Nov 2011 01:55:04 PM UTC, comment #9:
Hi Alexey
I'm not taking any guidelines, I was just expressing the feeling that the KSI glyphs looks too Church Slavonic.
As far as history is concerned, again I'm perfectly aware that ZEMLJA primitively looked like a Greek ZE, so what? We're dealing with construction rules, not with history.
You'd know for certain that the shift towards a "naked" KSI started very early, indeed the STAROSLAVJANSKI SLOVAR' features the newer form as the standard form for ZEMLJA in its entries. But that's not relevant, I think, for we're dealing with civilian, not ecclesiastical typefaces. But that's still important for it shows that I'm not the first nor the only one to imply a commonality of glyphs (not of history, OK, we all know that). The same commonality appears in the Cyrillic version of CODEX MARIANUS edited by Jagic.
OK, I'm sorry for having coming out with the suggestion of adding a hacek to ZE, that's pure trash. But, if you read my first suggestion, I was actually proposing a mini-IZHITSA.
The balls -- OK, that's how they're called, at last I know how they're called :) -- are OK, I didn't say there should be NO balls, but they should look like what they are in say, ZHE, do you understand what I'm saying?
As for the swashes beneath, don't you think the swashes in KSI should be uniformed with U+0340/U+A641 ZEMLJA (older shape of ZE you were referring to) and U+0642/U+0643 (DZELO). This time, history is on my side :) for in the Greek alphabet the tails under ZETA and KSI are usually identical.
Greetings
|
Tue 22 Nov 2011 12:03:51 PM UTC, comment #8:
Hi Alessandro,
I have no ideas why do you think Cyrillic XI should look similar to Cyrillic ZE. Since you have access to Church Slavonic texts, you can easily make sure that the two letters have very few in common in the historical Cyrillic script. So why should they be similar in a "Civilian" type? Well, it might be reasonable to replace the ball at the upper bowl with a serif (thus making the glyph more similar to ZE and E REVERSED), but all other components are necessary to make the letter recognizable. As for the ZE + HACEK composite, it was a particularly infelicitous form, shown in some old versions of the Unicode standard, where it has been changed after a discussion in a mailing list at evertype.com. I see absolutely no reasons to return to this shape (which looks essentially unrecognizable) now.
As for IZHITSA, PSI and OMEGA, please note that Cyrillic letters normally have more features terminating with balls than you can find in the Latin alphabet. This is a distinctive feature of the Cyrillic script as a whole, not just historical letters. Compare for example Cyrillic K with its Latin counterpart. Also note that IZHITSA has a long tradition of being represented in a "Civilian" style (since it has been present in the Russian alphabet until the 1918 reform), where it always had a rounded ball at the right diagonal stem.
So once again, I think you are taking as a guideline wrong shapes from the code chart which actually should not be followed.
|
Tue 22 Nov 2011 11:07:51 AM UTC, comment #7:
Hi Steve
As I told you, I do have access to Church Slavonic texts, which feature a glyph strikingly similar to the one you or someone else included in UCS Fonts.
But finding it in the civilian script is painfully difficult, I saw it once but I can't remember where and and when.
Please mind I never said you have to redesign those letters from scratch, you know I'm an ignoramus but I thought that "composing" letters with pieces of other ones was just way easier. A ZE + caron + swash above = KSI, you did it for Serbian Cyrillic glyphs, right? I thought it should be the same, but now you're telling me I'm wrong, right?
Maybe Alexey will be of some help here.
Greetings
|
Tue 22 Nov 2011 10:58:04 AM UTC, comment #6:
Hi Alessandro,
Please provide an image illustrating what you mean.
You must have access to some hard pictures of manuscripts, or some old printed material. Just take a picture with your digital camera, scale it up so I can see what you have in mind.
Internet searches also often turn up images of old documents!
Even then, I don't know when I can do this. I have a stack of stuff right now (including other tasks for FreeFont), and as I explained before, letter design is very time-consuming.
Cheers!
|
Tue 22 Nov 2011 09:01:42 AM UTC, comment #5:
As for KSI, even a hacek or caron on top would be OK.
|
Tue 22 Nov 2011 09:00:47 AM UTC, comment #4:
Hi Steve
I thought I'd provided you with some sort of a model. The "backbone" of KSI should be ZE, with a swash above, and a mini-IZHITZA on top.
The "drops" on the right of IZHITSA and PSI and the two final "drops" of OMEGA should be modeled on the "drops" of ZHE. I'm sorry if I call them "drops", as I you know I'm an ignoramus of typography.
Maybe you could provide me with some reference work.
I hope this might help you.
Greetings
|
Mon 21 Nov 2011 09:43:31 PM UTC, comment #3:
Hi Alessandro,
As to the style of the archaic letters, I have no idea. I didn't implement them. Some years ago I replaced the Cyrillic range on advice from Slavic speakers with glyphs from the TemporaUni font, assembled by Alexey Kryukov
http://www.thessalonica.org.ru/en/fonts.html
Now that I re-read his correspondence with me, I see he had his own doubts about some of the letter forms.
I'm therefore including him in this discussion.
Please don't refer to Gnu FreeFont as "my" font. It is a community effort which I contribute to and administer, but I neither started it nor contributed most of the letters.
I'm not going to go re-designing letters based on these sorts of complaints. The best I could do is follow some model. But even then, I have done such things in the past, and it is a hell of a lot of work, even for a few letters. If I'm to personally draw the new letters, it will not happen with the next release.
I really think the onus is on you to show me what you mean.
|
Mon 21 Nov 2011 05:40:25 PM UTC, comment #2:
No wait, I'm not complaining about UK again :) the CYRILLIC U contained in UK (the one you corrected of late) was an example of a shape appropriate for Church script, not for a modern script. I said that UK (the older one) and U didn't even look like the same font, and you cannot deny that. The U's were pretty different in shape and construction, can you deny that? The U in UK was much "rounder" and with more swashes, while the standard U was much "sharper", now sharpness is the key for civilian Cyrillic fonts, while roundness is reserved for Church fonts.
That the origins of the two letter are different, that's obvious, in Old Church Slavonic Cyrillic ZE (called ZEMLJA) did look exactly like a Greek ZETA, that's it.
That's history, but as far as we're concerned here, we deal with construction rules, don't we? And for civilian Cyrillic the "feeling" is sharp, not round. Take a look at the "sharpness" of ZE and at the "roundness" of KSI in your font, I repeat it, it doesn't look like the same font! It's like in the case of U and older UK glyphs.
It's painfully difficult to provide you with printed materials IN CIVILIAN SCRIPT, because that glyph is not in use in any Cyrillic-based alphabet since at least the mid-19th century.
I don't think you understood the issue was not really about ONE or TWO glyphs, it's about the construction rules you utilized for implementing archaic Cyrillic glyphs. I don't like that because it looks like you ported them from another font, standard and archaic glyphs don't "feel" the same, I hope you now understand the point I'm trying to make. It's like having in Latin some Blackletter glyph amid other Antiqua glyphs. Do you understand me?
P.S.
Newer UK is excluded from this issue, whatever I might think about a narrower o, the construction rules are OK.
|
Mon 21 Nov 2011 03:34:36 PM UTC, comment #1:
Sorry Alessandro,
I really don't understand what you're talking about here.
We spent a lot of effort correcting the Uk in another bug report, the correction was decided upon and committed. Now you say it's still wrong.
Please explain.
As to Ksi, I have no idea how they should look, and I didn't make the letters, but the Wikipedia article
http://en.wikipedia.org/wiki/Ksi_%28Cyrillic_letter%29
says the Ksi is based on the Greek letter Xi, whereas according to
http://en.wikipedia.org/wiki/Ze_%28Cyrillic%29
the Cyrillic Zi derives from the Greek letter Zeta.
This does not seem to agree with your claim about the history of the letters.
Perhaps you could provide some images of manuscripts or printed materials containing these letters. (Mind you I'm less impressed by how the letters are implemented in other modern fonts.)
|
Mon 21 Nov 2011 11:57:25 AM UTC, original submission:
Hi Steve
I mean, take a look at U+046F CYRILLIC SMALL LETTER KSI. In theory, the "base" of its construction should be U+0437 CYRILLIC SMALL LETTER ZE but in fact, that's not the case, neither in FreeMono, nor FreeSans, nor FreeSerif.
But the issue is far deeper than an individual character's shape. As a matter of fact, most archaic Cyrillic characters have a shape that has a "Church" flavor, round and with few angles, while they should rather have many 90° angles or the "upper point" of U+0436 CYRILLIC SMALL LETTER ZHE.
You'll excuse me for my tremendous lack of knowledge of typographical terms, but just take a look at the older shape of CYRILLIC U in CYRILLIC UK, and look at the difference between this one and the standard CYRILLIC U, you can easily spot the difference! It doesn't even look like the same font!
Greetings
|