Haansoft Font Errors

Basic Fonts
-----------

C0 and C1 control characters in Basic Latin and in Latin-1 Supplement don't exist in the fonts.

The reverse solidus (92) is copied from the slash and has been reversed.

Latin letters a, e (97, 102) have wrong design in Haansoft Batang. This bug doesn't exist in Chrysanthi Unicode.

Control delete (127) looks like a not defined instead of showing actual small capital DEL in Hansoft Batang and Haansoft Dotum. This glyph doesn't exist in Chrysanthi Unicode.

Superscript 2, 3, 1 (178-179,185) aren't aligned to the superscript position.

Vulgar fractions one quarter, one half, three quarters (188-190) are nut fractions. This bug doesn't exist in Chrysanthi Unicode.

Latin kra (312) looks like an cyrillic ka instead of an small k.

Latin letters tone six (388-389) are similar to Chrysanthi Unicode.

Latin turned delta (397) is an vertically/horizontally flipped greek delta rather than a vertically flipped delta.

Latin Open E (400) isn't matching the design of Chrysanthi Unicode.

Latin letter lambda with stroke (411) has an extraneous line.

Latin letters Oi (418-419) aren't Chrysanthi Unicode.

Latin letter yr (422) is cut off in Haansoft Dotum and Batang.

Reversed esh loop (426), V hook (434), Ezh (439), Reversed ezh (440) is similar to Chrysanthi Unicode.

Latin 2 with bar (443) doesn't match Chrysanthi Unicode.

Latin small tone five (445) is smaller than the glyph in Chrysanthi Unicode. The small tone five is bigger than Chrysanthi Unicode.

Latin letters wynn (447, 503) have no serifs. However the glyph shapes are OK in Chrysanthi Unicode.

Latin letters hv (405, 502) were same as Chrysanthi Unicode.

Latin letters alpha (593-594) have a hook.

Latin letter b with hook (595) is an o with hook rather than the actual glyph in Haansoft Dotum.

Latin letter script g (609) is a g in Haansoft Dotum.

Latin letters turned h, turned k, turned m with long leg (613, 624, 670) aren't aligned to the typo descender.

Latin letter small capital I (618) have no serifs in Haansoft Dotum.

Latin letters turned m, turnd m with long leg (669-671) look like Times New Roman in Haansoft Batang.

Latin letter small capital N (628) is a reversed cyrillic small I instead of the actual glyph in Haansoft Dotum.

Latin letters r with fish hook (638-639) are similar to Gulim. However the glyph shapes are OK in Chrysanthi Unicode.

Latin letter dotless j with stroke and hook (644) is a esh with stroke instead of the actual glyph.

Latin letter t with retroflex hook (648) is a stretched t in both font families instead of retroflex hook.

Latin letter v with hook (651) is similar to Chrysanthi Unicode.

Latin stretched c (663) is aligned to the baseline of the font.

Digraphs dz, dezh, dz curl (675..677) are not connected in Haansoft Batang.

Digraphs dz, dezh, dz curl, ls, lz are not connected in Haansoft Dotum.

Spacing Modifier Letters (688..767) have bad alignment/cutoffs of letters.

Combining Diacritical Marks (768..879) have cutoffs.

Combining vertical line (781) is a apostrophe instead of showing actual glyph. However the glyph is OK in Chrysanthi Unicode.

Combining latin letters a and e (867-868) have wrong design. However the a and e glyph shapes are OK in Chrysanthi Unicode.

The Combining Diacritical Marks range was taken from Haansoft Batang, incorrectly combining it.

Greek and Coptic, Cyrillic, Cyrillic Supplement ranges don't match the designs in Haansoft Dotum.

Cyrillic, Cyrillic Supplement ranges don't match the designs in Haansoft Batang. However the designs match OK in Chrysanthi Unicode.

Armenian, Hebrew, Arabic, Syriac, Thaana, Devanagari, Bengali, Gurmukhi, Gujarati, Oriya, Tamil, Telugu, Kannada, Malayalam, Sinhala, Thai, Lao, Tibetan, Myanmar, Khmer, Tagbanwa, Buhid and Mongolian needs Opentype features to work.

Cherokee was copied with Linguist's Software's Cherokee Unicode font.

Arabic Supplement, N'Ko, CJK Strokes, Yijing Hexagram Symbols glyphs are blank instead of showing these actual glyphs.

Tamil digit zero (3046) is blank in Haansoft Batang and Haansoft Dotum.

Latin letter A with ring below (7680) has wrong spacing in Haansoft Dotum.

Latin letter H with cedilla (7720) has wrong spacing in Haansoft Batang. However the spacing is OK in Chrysanthi Unicode.

Latin letter a with right half ring isn't the glyph from Unicode 5.1, as the right half ring is from the base letter.

Letterlike Symbols, Number Forms, Arrows, Mathematical Operators, Miscellaneous Technical, Control Pictures, Optical Character Recognition, Geometric Shapes, Miscellaneous Symbols, Dingbats, Miscellaneous Mathematical Symbols-A, etc are monospaced.

Account of, addressed to the subject, care of, cada una (8448-8449, 8453-8454) use italic a and u instead of normal a and u.

Small script l (8467) is a capital L instead of showing actual glyph.

For all, there exists doesn't match the design.

Small roman numerals eleven, twelve (8570-8571) have an reversed x. However the roman numerals are proportional in Chrysanthi Unicode.

Square pc (13174) use the Pc instead of <square> pc.

Square Bq (13251) is a horizontal flipped p after the B.

Square mb small (13268), square Wb (13277) is a vertical flipped p after the m or after the W.

Limbu, Tai Le, New Tai Lue, Khmer Symbols, Buginese, Balinese, Sundanese, Lepcha, Ol Chiki, Phonetic Extensions, Phonetic Extensions Supplement, Combining Diacritical Marks Supplement, Miscellaneous Symbols and Arrows, Glagolitic, Latin Extended-C, Coptic, Georgian Supplement, Tifinagh, Ethiopic Extended, Cyrillic Extended-A, Supplemental Punctuation, Vai, Cyrillic Extended-B, Modifier Tone Letters, Latin Extended-D, Syloti Nagri, Phags-pa, Saurashtra, Kayah Li, Rejang, Cham, High Surrogates, High Private Use Surrogates, Low Surrogates, Variation Selectors no longer function in Haansoft Batang and Haansoft Dotum. The supported Unicode blocks are: Basic Latin, Latin-1 Supplement, Latin Extended-A, Latin Extended-B, IPA Extensions, Spacing Modifier Letters, Combining Diacritical Marks, Greek and Coptic, Cyrillic and Cyrillic Supplement, Armenian, Hebrew, Arabic, Syriac, Thaana, Bengali, Gurmukhi, Gujarati, Oriya, Tamil, Telugu, Kannada, Malayalam, Sinhala, Thai, Lao, Tibetan, Myanmar, Georgian, Hangul Jamo, Ethiopic, Cherokee, Unified Canadian Aboriginal Syllabics, Ogham, Runic, Tagalog, Hanunoo, Buhid, Tagbanwa, Khmer, Mongolian, Latin Extended Additional, Greek Extended, General Punctuation, Superscripts and Subscripts, Currency Symbols, Combining Diacritical Marks for Symbols, Letterlike Symbols, Number Forms, Arrows, Mathematical Operators, Miscellaneous Technical, Control Pictures, Optical Character Recgnition, Enclosed Alphanumerics, Box Drawing, Block Elements, Geometric Shapes, Miscellaneous Symbols, Dingbats, Miscellaneous Mathematical Symbols-A, Supplemental Arrows-A, Braille Patterns, Supplemental Arrows-B, Miscellaneous Mathematical Symbols-B, Supplemental Mathematical Operators, CJK Radicals Supplement, Kangxi Radicals, Ideographic Description Characters, CJK Symbols and Punctuation, Hiragana, Katakana, Bopomofo, Hangul Compatibility Jamo, Kanbun, Bopomofo Extended, Katakana Phonetic Extensions, Enclosed CJK Letters and Months, CJK Compatibility, CJK Unified Ideographs Extension A, CJK Compatibility Ideographs Supplement, CJK Unified Ideographs, Yi Syllables, Yi Radicals, Hangul Syllables, Private Use Area (Orthography Korean Syllables for use with Hangul), CJK Compatibility Ideographs, Alphabetic Presentation Forms, Arabic Presentation Forms-A, Variation Selectors, Combining Half Marks, Small Form Variants, Arabic Presentation Forms-B, Halfwidth and Fullwidth Forms, Specials.

General
-------

The fonts have too many contours and have bitmap data encoding.

When launching Font Creator, MainType or Scanahand it causes the code pages to switch to Code Page 1252, causing the Korean characters to render incorrectly. The bug will be removed when you changed a theme.

Apparantly the encoding and code page tables of Haansoft Batang and Haansoft Dotum are all properly set, which causes Windows Font Viewer to show as different previews in font preview.

The MES conformance tests used in the source is out of date. Even though the code ranges are the same, number of glyphs specified by the pages required to comply with each MES level are increased as a result of Unicode updates. This causes current fonts to fail MES tests.

The line heights of all 3 families are taller than the free Hindsight Unicode, Enigmatic Unicode and Torquemada Starved Unicode fonts for which the commercial Haansoft Unicode fonts are based. The line heights are also taller than Hindsight Unicode, which itself is taller than History Happens Std (free), Enigmatic Std (free), Hindsight Pro (free). The extra heights only serve to reduce the number of characters that can fit in a screen. Besides, the extra margins are not needed to fit the glyphs with accents (including multiple-accented Latin Extended Additional/B and Greek Extended glyphs), it only needs compressed double-height accents to fit the gap. Screen estate aside, characters such as top half integral (U+2320), bottom half integral (U+2321), summation top (U+2322), summation bottom (U+2323), are no longer connected when arranged vertically; glyphs in Box Drawing (U+2500-257F) and Block Elements (U+2580-259F) function in Courier New and other monospaced fonts like they supposed to, which can prevent equation editors and monospaced applications from working properly.

Currently the amount of symbols across font families are not the same, which is unusual since unlike letters, the stylings of symbols do not tend to change across font families. Affected regions include Combining Diacritical Marks for Symbols, Letterlike Symbols, Arrows, Mathematical Operators, Miscellaneous Technical, Optical Character Recognition, Box Drawing, Block Elements, Geometric Shapes, Miscellaneous Symbols, Dingbats, Miscellaneous Mathematical Symbols-A, Supplemental Arrows-A, Braille Patterns, Supplemental Arrows-B, Miscellaneous Mathematical Symbols-B, Supplemental Mathematical Operators, Supplemental Punctuation, Yijing Hexagram Symbols. Some of the missing symbols can be found in IPA Mona Font, Kochi Gothic/Mincho, which also include glyphs for Enclosed Alphanumerics, CJK Symbols and Punctuation, Enclosed CJK Letters and Months, CJK Compatibility.

In each font family, the development of the regular font is way far ahead of the bold and italic counterparts. Furthermore, the character set being covered between families are inconsistent. To fast track the synchronization of glyphs across the font family, UCS source should start building automated scripts that generate different font weights and styles from the base font, and glyphs that can be created by using other glyphs within the same font. Currently Haansoft Batang and Haansoft Dotum doesn't have bold, italic and bold italic versions, which causes Windows, Mac or Linux to slant the regular typestyle. A long term solution is to have source glyphs built on strokes formats like Metafont, SVG, or Type 3, which allows font developers to create glyphs of multiple styles, weights, and widths simply by writing the strokes with pen or mouse, then use algorithms to transform the strokes into font families of choice. In addition, stroke based fonts are easier to maintain even within a font family, simply because fewer control points are needed to represent a glyph. Fortunately, FontForge can handle Type 3 and SVG strokes, but the Metafont sources used by the project need to be converted to Type 3 or SVG first.

Unicode 5.1 supports the following ranges: Basic Latin, Latin-1 Supplement, Latin Extended-A, Latin Extended-B, IPA Extensions, Spacing Modifier Letters, Combining Diacritical Marks, Greek and Coptic, Cyrillic and Cyrillic Supplement, Armenian, Hebrew, Arabic, Syriac, Arabic Supplement, Thaana, N'Ko, Bengali, Gurmukhi, Gujarati, Oriya, Tamil, Telugu, Kannada, Malayalam, Sinhala, Thai, Lao, Tibetan, Myanmar, Georgian, Hangul Jamo, Ethiopic, Cherokee, Unified Canadian Aboriginal Syllabics, Ogham, Runic, Tagalog, Hanunoo, Buhid, Tagbanwa, Khmer, Mongolian, Limbu, Tai Le, New Tai Lue, Khmer Symbols, Buginese, Balinese, Sundanese, Lepcha, Ol Chiki, Phonetic Extensions, Phonetic Extensions Supplement, Combining Diacritical Marks Supplement, Latin Extended Additional, Greek Extended, General Punctuation, Superscripts and Subscripts, Currency Symbols, Combining Diacritical Marks for Symbols, Letterlike Symbols, Number Forms, Arrows, Mathematical Operators, Miscellaneous Technical, Control Pictures, Optical Character Recgnition, Enclosed Alphanumerics, Box Drawing, Block Elements, Geometric Shapes, Miscellaneous Symbols, Dingbats, Miscellaneous Mathematical Symbols-A, Supplemental Arrows-A, Braille Patterns, Supplemental Arrows-B, Miscellaneous Mathematical Symbols-B, Supplemental Mathematical Operators, Miscellaneous Symbols and Arrows, Glagolitic, Latin Extended-C, Coptic, Georgian Supplement, Tifinagh, Ethiopic Extended, Cyrillic Extended-A, Supplemental Punctuation, CJK Radicals Supplement, Kangxi Radicals, Ideographic Description Characters, CJK Symbols and Punctuation, Hiragana, Katakana, Bopomofo, Hangul Compatibility Jamo, Kanbun, Bopomofo Extended, CJK Strokes, Katakana Phonetic Extensions, Enclosed CJK Letters and Months, CJK Compatibility, CJK Unified Ideographs Extension A, Yijing Hexagram Symbols, CJK Compatibility Ideographs Supplement, CJK Unified Ideographs, Yi Syllables, Yi Radicals, Vai, Cyrillic Extended-B, Modifier Tone Letters, Latin Extended-D, Syloti Nagri, Phags-pa, Saurashtra, Kayah Li, Rejang, Cham, Hangul Syllables, High Surrogates, High Private Use Surrogates, Low Surrogates, Private Use Area, CJK Compatibility Ideographs, Alphabetic Presentation Forms, Arabic Presentation Forms-A, Variation Selectors, Vertical Forms, Combining Half Marks, Small Form Variants, Arabic Presentation Forms-B, Halfwidth and Fullwidth Forms, Specials.

Apparently the commercial Haansoft Font families are missing text descriptions describing the fonts' characteristics, which are necessary to build proper glyphs for a font family, especially glyphs for other languages. Here are some suggestions for the base fonts:

Haansoft Dotum: A sans-serif proportional width font with fixed medium font weight, with vertical or horizontal flat stroke terminator.

Haansoft Batang: A serif proportional width font with light horizontal and heavy vertical font weights, with flat, rounded, or pointed stroke terminator.

Documentation
-------------

The documentation incorrectly referred that the fonts are not anymore available on the Web. Last time I checked, the fonts have been hosted in SourceForge site after Microsoft had yanked them out of its own corporate pages. And thanks to file sharing networks, the core fonts, as well as the later versions of these fonts (including Windows Vista counterparts) and more, are perpetually available somewhere in the web.

I know there are legal and moral issues involved, and I do believe in software freedom, and really hate to see this and many other free software projects have become the perpetual underdogs. But the truth is, the existing arguments are not going to help developing UCS fonts, when the non-free products can be obtained very easily with almost zero chances of getting caught (in fact, in countries like Canada, it isn't even illegal to download copyright infringing materials), and the current free UCS fonts simply can't compete in quality. Also, unlike other software, the core fonts are relatively maintenance free, so there is little chance of someone calling Microsoft over the font's incompatibility (even if there are, the source of the faults are mostly anything but the fonts themselves). If that's not bad enough, there are active user groups that mix, match, and hack proprietary fonts with zero regards to EULA, which in itself has questionable legal enforceability (though in this case, there may also be issues of legality of knock-off fonts).

At best, your existing arguments are only relevant to corporate entities, who have much less chance of escaping infringement lawsuits. But for them, cost of paying for fonts are relatively trivial, especially when they are more capable to obtain volume discounts than average end users while obtaining some, even if limited, degrees of support. On the other hand, Free UCS Outline Fonts come with no warranty. Combining with the fact that Free UCS fonts are not popular to begin with, average end users and corporate entities alike have much better chance of finding community support for the Core Fonts for the Web than Free UCS fonts ever can.

If this, or any other free software project, want to have a shot of replacing the evil proprietary software(tm) in a world of full of unrepentant freeloaders who are willing and able to get whatever software they want, virtually in anywhere, at any time, by any means necessary, it needs to compete in quality. It is an unfair fight, but this is the kind of environment this project has to compete in. Since current Free UCS Outline Fonts are not exactly in competitive shape, and last time I checked, it hasn't been updated for over a year, it will need some proper marketing to attract development resources. I am not a marketer, nor do I know of any marketer I can refer you to, but I do know a key of successful marketing is to be honest. I am not saying there are fraudulent intent involved, but claiming core fonts for the Web are unavailable on the Web when they have been hosted by SourceForge for YEARS does not sound like honesty to me. The statements need to be updated to honestly reflect the current state of the competitions, especially on technical qualities.

The other thing about marketing is, you need to send the right message to the right people. Unfortunately, that means the freedom rhetoric your home page is boasting is more of a turn off to would be users and developers than a selling point. That is not to say you should not advertise freedom, but rather, you should do it more strategically. More on that later. To attract users at the first place, you need to market pragmatic advantages over the competition. Fortunately, the saving grace comes from the vast amount of glyphs in all 2 font families. Haansoft Dotum has more glyphs than Arial 5.00 (Windows Vista Version), Arial Unicode MS 3.2 (too many chars to display), Helvetica Roman (388 chars), Helvetica World (1862 chars), Neue Helvetica (390 chars), Dotum (too many chars), Haansoft Batang currently has more glyphs and support more code pages with all unassigned code pages than even the Chrysanthi Unicode, Times New Roman 5.00 (Windows Vista version) (2790 chars), CG Times (249 chars), Times Std Roman (408 chars), Times Europa (253 chars), Batang. It is time to exploit these advantages against similar proprietary fonts for all they worth. However, keep in mind that quantity of characters are not the only benchmark of superiority, especially when the Free Font families have uneven distributions of glyphs across variants. It is just a gateway to attract developers into refining the fonts in other areas, before the temporary edge is gone.

A triple click in Opera causes Opera to select all of the sources. Oddcast's TTS translation demo has a limit of 50 characters for Asian languages and 250 characters for normal languages such as Greek and Cyrillic. There are no codes for Finnish and Hindi, causing Galician and Thai to speak as English.

An uninstallation of Safari causes Uniscribe to stop working. Can't show how to do it!

When dealing with fonts that need progress to catch up against the latest proprietary counterparts, the homepage should include progress chart, to inform users of the progress towards beating the competition, to gain the impressions this project is capable to compete against commercial foundries.

When dealing with freedom issue, remember that end users usually do not get hit by much, because they can simply get the non-free fonts with ease and virtually no consequences. This message should be targeted to corporate customers, because they are the ones who cannot get away with copyright infringements. In the project homepage, information on freedom should put into a separate section, for corporate users. However, unless there is an active community for the Free UCS scalable fonts project, this particular feature should be in the footnote section, because you can't back up these users unless the font can compete in quality against the oppositions, and you can't compete against proprietary counterparts unless there is an active development community rather than the divided efforts in current state.

I hope that helps giving the project a clearer direction for it to develop.

Syndicate content