@ 5)
See this Typophile thread. I would personally made the stem a bit rounder, to reflect capital S character.
Definetly you need the OT substitution (also in small caps).

Thanks, Marcin. I was not aware how deeply sharp S was discussed nor it was so recently established. Time to read that whole thread.

The New Sheqel is too fancy; just keep it simple, no serif, small.

#3: Have you seen this? http://typophile.com/node/62439
#4: The rightmost one seems to better differentiate itself from a regular K or К.

#1:
Roman serifs are problematic, somewhat alien to this non-Roman glyph. Make it more familiar to the ductus of the Hebrew letters. Height? A worthy question. Consider the combination with the figures, which ought to go well. Height alignement with $ £ € etc. is not obligatory, in my view.

#2:
Drop this character. Not worth the labour. It has been encoded for some single obscure source long ago, but is used by no one.

#3:
Gravis on the right side. The same for acute.

#4 – I’d welcome some resources on Afrolatin as well !

#5:
Carefully adjust the weights of the glyph’s different parts. You’ll find lot of reference here on typophile and also at Signographie.de

Thanks a lot for all answers. I'll take your advices in account to improve New Sheqel and sharp S immediately. Let me add some considerations about the other questions:

Drachma:

You did touch an interesting issue, Andreas. Although this symbol is not currently used, don't it must be keeped for historic reasons? At last, Drachma was Greece's currency for long time. I see the same for other European national currencies whose were replaced by Euro (exception to Franc symbol, still used in African countries).

A similar situation occurs with Cruzeiro (Uni20A2), wrongly reported by Unicoed as in-use Brazilian currency. This symbol was dropped in 1967 and almost no one knows it in Brazil nowadays. Anyway, many fonts brings the symbol. The only logical reason for this is historic reference (although I believe many designers include it just because Unicode does).

.

Vietnamese:

I already knew the thread about Vietnamese diacritics, but there is no conclusion there about the grave position. An user (hdang) seems to know Vietnamese languase and says Minion solution is bad. As Minion uses the grave at right, the correct would be grave at left. But this user did not come back to confirm this.

Both options are equally strange to my Western eye. So I'd like to know how Vietnamese people actually prefer it.

.

K hook:

I agree with riccard0, K with hook at leg is more distinct from regular K. But again, how African people use it? Maybe this is the more dificult question as we have very few information about African languages even here in Typophile.

1. As noted by others, the serifs make this shape too complex. I'd make this the same height as Hebrew letters, which by long convention are most often between x-height and cap height relative to Latin. Note, however, that some modern Israeli fonts match Hebrew letters to the Latin cap height.

3. Both conventions are seen in use in Vietnam, in type and in writing. I prefer the grave on the left of the circumflex because at small sizes and low resolutions it is more distinct, so aids legibility.

4. I spent a lot of time researching African orthographies in the late '90s. The hook on the right arm is by far the most common form of this letter; in fact, I don't recall seeing the other form at all. This requires a strong hook form in order to be clearly legible and distinct from the regular K at small sizes.

Re #5 – Wait, don't automatically sub in the German cap eszett via OT. While the character has been assigned a Unicode, as far as I know it's not yet taken root in German spelling rules. Things will still take a bit of time…

John and Nina, thank you very much. All your information gave me valuable help on defining these design questions.

On the wikipedia article on the new shekel, they say that the symbol is a combination of the letters chet and shin (for shekel chadash, new shekel), so putting latin serifs on it is off. If it were to have serifs they should be Hebrew style, I would think, and probably this doesn't work anyway.

Thanks, William.

Just to give a feedback, here are the choices and improvements I made:

Vaguely related to K-hook: http://typophile.com/node/74127

To judge the shekel sign I’d need to see it in combination with figures and with Hebrew letters. In your image it appears still a bit alien to the other monetary family members.

As for the new Rupee, I notice that you intend to go for a more Latinish shaping. This aspect has been debated upon here not so long ago.

Capital Eszett: the right part is still too light. Weight distribution is fine, though. But the upper part may have some little more movement, making it less stiff. Perhaps.

my 2 cents regarding the New Shekel: (more as a Hebrew speaker, not a type designer)
As mentioned, some Hebrew faces are leveled to match Latin Cap height, some to X-height, and some in between. If your font does not include Hebrew script, then the Shekel symbol would only appear next to figures, so no reason to make it different then any other currency symbol. It will never appear next to Hebrew text because every Hebrew typeface has its own New Shekel glyph.

Regarding the serifs, everything said here was right, but I would try a version with a single serif, the one on the top left, which some Hebrew Chet letters do have. Look at this example in Fontef's Meargen, you can see both Shin and Chat at the bottom line:
http://www.flickr.com/photos/fontef/2270255289/

Thanks for more valuable inputs.

New Sheqel and Rupee:

Yotam: you touched exactly the question I did ask. As this font uses just Latin script, New Sheqel will never be together with Hebrew glyphs so it may look strange with its small height. In doubt, I did two versions: the small, traditional one, to be used with default, oldstyle and petite figures. And a larger one to be aligned with uppercases. Your idea about prtial serifs is interesting, I will do some tests.

Andreas: I believe there is no escape from a somewhat strange appearance as this came from another script. The same would occur with Drachma, old German Penny or Rial (although this one I did not include) because these are non-Latin glyphs. As you pointed, Rupee has a more Latinish shape. The alternative to Sheqel is to adopt a non-traditional design, as I made originally and as one finds in Greta Text. But this seem to produce even more strange results.

I made a sample using New Sheqel and other monetary symbols.

.
This sample is far from ideal as there are distortions due to screen resolution. In both texts (rates table and small note) I used, in order, default number set, oldstyle figures and all caps figures. Table uses tabular figures, remaining texts are proportional.

Cap Eszett:

Thank you very much, Andreas. It's a honor to receive tips from the person who proposed this glyph to Unicode! Who would know it better? I visited Signographie and also downloaded some PDFs. Unhappily, I'm not able to read more than a few words in German –although my family name would suggest the opposite– so I missed the pages without English version.

I hope this new version is a bit better. I'm trying not to make the glyph very wide:

Capital Eszett: Better! try to make the slantad bar (top right) a bit steeper. Mind the thickness of the top part.

Sheqel: I would test what happens when the glyph’s horizontal parts get stressed, instead of the vertical ones. See Hebrew.

… and as for the honor: it’s certain on the side of a German letter when it gets done in Brazil.
;-)

Are your ancestors from Freiberg/Sa.?

Andreas, thanks again for your help. I updated the PDF in Palimpsest Project thread with these new and improved glyphs. My goal is to produce a font which respects as much as possible language idiosyncrasies. Thus I'm paying attention to these characters.

My ancestors probably came from Freiberg. Unhappily, documentation about German imigrants who came to Southern Brazil between 1824 and 1890 is highly incomplete, especially about the first families –what include Freiberger's. Someday I hope to get more info on this.

Hmm. Shekel. Igor's one above has both sections with the curve on the inside stem; the ones I've seen have the second section with the curve on the outside. Is this optional?

Hi Ben, the first part—the one on the right, you read Hebrew right to left—is based on the outside of the shin, and the second part, on the left, is based on the chet. That's why the bottom right is rounded and the top left is square. It follows these letters. You can see the illustration in the wikipedia article here, which makes it clear. I don't know whether Israelis will see a bottom right square instead of round as "wrong", or whether the symbol has become sufficiently detached from the letters that you have more flexibility. Perhaps an Israeli will chime in here.

Ha, yes, I forgot Hebrew was the other way round! Think I'll go with the traditional approach for the time being. Thanks.