Re: [g-i] freefont vs. dejavu
On Wed, Mar 22, 2006 at 01:18:09PM +1030, Clytie Siddall wrote: Today I took a look at dejavu and found out that all the necessary glyphs needed by Vietnamese are contained in version 2.4 (I thought 2.3 was ok but I was wrong) Yay for 2.4! :) I'd like to run some more tests but here's how Vietnamese looks like when using DejavuSans (out of dejavu 2.4): http://www.webalice.it/zinosat/shots/vi_dejavu24.png looks a bit weird to me. ciao, Davide signature.asc Description: Digital signature
Dejavu and Vietnamese: unreadable (was: Re: [g-i] freefont vs. dejavu
On 23/03/2006, at 7:07 AM, Davide Viti wrote: I'd like to run some more tests but here's how Vietnamese looks like when using DejavuSans (out of dejavu 2.4): http://www.webalice.it/zinosat/shots/vi_dejavu24.png looks a bit weird to me. chokes Just a bit... it's unreadable. Most of the accents are missing or replaced by nonsense characters. This is in UTF-8?? A true Unicode font?? I don't think they can have integrated the Verajja glyphs yet. It covers Vietnamese quite adequately. I'd better write to them about this sigh... this is bad. from Clytie (vi-VN, Vietnamese free-software translation team / nhóm Việt hóa phần mềm tự do) http://groups-beta.google.com/group/vi-VN PGP.sig Description: This is a digitally signed message part
DejaVu displaying Vietnamese OK? (was: Re: [g-i] freefont vs. dejavu
On 23/03/2006, at 7:07 AM, Davide Viti wrote: http://www.webalice.it/zinosat/shots/vi_dejavu24.png looks a bit weird to me. Davide, are you sure the encoding/locale are correct? DejaVu is displaying Vietnamese readably, if not perfectly in terms of accent shape and position, on my system (Mac OSX). I've just added links and screenshots of both Freefont (26/01/06 snapshot, latest I could find) and DejaVu 2.4, to my comparative fonts page (and removed the document title-bar which was annoying you ;) ). http://www.riverland.net.au/~clytie/viet/fonts.html I have written to the DejaVu authors, giving them feedback about their font and displaying Vietnamese. I hope this is useful. :) from Clytie (vi-VN, Vietnamese free-software translation team / nhóm Việt hóa phần mềm tự do) http://groups-beta.google.com/group/vi-VN PGP.sig Description: This is a digitally signed message part
Re: [g-i] freefont vs. dejavu
On Tue, Mar 21, 2006 at 04:33:29PM +1030, Clytie Siddall wrote: Meanwhile, Freefont works for Vietnamese, it's just not as good as we would like. For example, even at 16pt, I have to guess many of the words, because the accents over the vowels blend into the line above. However, that's Freefont's issue. hopefully g-i will put more attention on font development. Today I took a look at dejavu and found out that all the necessary glyphs needed by Vietnamese are contained in version 2.4 (I thought 2.3 was ok but I was wrong) Can we use it for Vietnamese? Yes, if we don't have anything better. :) I understand your point: I still think showing our interest will inspire font developers / maintainers. As said before, I think the best option for a project like g-i would be a package like freefont or dejavu covering almost all languages without using too many bytes and with a nice visual effect. Do we have a link where I can point people, about Debian being about to release? Also, will we be doing Release Notes and generally waving the flag a bit? I'd like to know, so I can (1) translate the release notes and (2) spread the good news on Vietnamese sites. :D I'd say the wiki [1] and debian-boot. Once g-i will be more officially integrated with d-i we'll see if and how webpages will be changed. Ciao, Davide [1] http://wiki.debian.org/DebianInstaller/GUI signature.asc Description: Digital signature
Re: [g-i] freefont vs. dejavu
Do we have a link where I can point people, about Debian being about to release? Also, will we be doing Release Notes and generally waving About to release? Not exactly. The targeted release date for Debian is still the end of the year 2006. What we are closer and closer to release if the first beta version of the installer that will include a graphical version. That one is targeted for April or so signature.asc Description: Digital signature
Re: [g-i] freefont vs. dejavu
On 22/03/2006, at 3:13 AM, Christian Perrier wrote: Do we have a link where I can point people, about Debian being about to release? Also, will we be doing Release Notes and generally waving About to release? Not exactly. The targeted release date for Debian is still the end of the year 2006. Then why all the urgency to keep the Installer at 100%? I don't understand. :( What we are closer and closer to release if the first beta version of the installer that will include a graphical version. That one is targeted for April or so I am certainly learning to take Debian goal-times with a kilo of salt... ;) from Clytie (vi-VN, Vietnamese free-software translation team / nhóm Việt hóa phần mềm tự do) http://groups-beta.google.com/group/vi-VN PGP.sig Description: This is a digitally signed message part
Re: [g-i] freefont vs. dejavu
Hello again all fans of this thread (the T-shirt will be out soon) :D On 22/03/2006, at 3:52 AM, Davide Viti wrote: On Tue, Mar 21, 2006 at 04:33:29PM +1030, Clytie Siddall wrote: Meanwhile, Freefont works for Vietnamese, it's just not as good as we would like. For example, even at 16pt, I have to guess many of the words, because the accents over the vowels blend into the line above. However, that's Freefont's issue. hopefully g-i will put more attention on font development. I probably should have asked this after the first post, but ... what is g-i? GNU-install? Gnome-install? frivolousGeneral Incredulity?/ frivolous (Clytie has a terrible tendency to frivolity.) Today I took a look at dejavu and found out that all the necessary glyphs needed by Vietnamese are contained in version 2.4 (I thought 2.3 was ok but I was wrong) Yay for 2.4! :) grabs font package Can we use it for Vietnamese? Yes, if we don't have anything better. :) I understand your point: I still think showing our interest will inspire font developers / maintainers. So do I. I found the Verajja author very responsive, and I've written that detailed feedback for freefont. I can also point people to my newly-constructed comparative font page [1], which isn't definitive, but does put some examples up where people can see them. (Must find time to translate that page.) As said before, I think the best option for a project like g-i would be a package like freefont or dejavu covering almost all languages without using too many bytes and with a nice visual effect. I agree. I have said so repeatedly, during a thread about fonts for packaging, on both debian-i18n and gnome-i18n. It means most new users have a default font which will show their language readably. This is an accessibility issue, as well as an internationalization one. Do we have a link where I can point people, about Debian being about to release? Also, will we be doing Release Notes and generally waving the flag a bit? I'd like to know, so I can (1) translate the release notes and (2) spread the good news on Vietnamese sites. :D I'd say the wiki [1] and debian-boot. Once g-i will be more officially integrated with d-i we'll see if and how webpages will be changed. Thanks. I'll try to keep track of the evolving timeline. ;) [1] http://wiki.debian.org/DebianInstaller/GUI from Clytie (vi-VN, Vietnamese free-software translation team / nhóm Việt hóa phần mềm tự do) http://groups-beta.google.com/group/vi-VN [1] http://www.riverland.net.au/~clytie/viet/fonts.html PGP.sig Description: This is a digitally signed message part
Re: [g-i] freefont vs. dejavu
On Wed, Mar 22, 2006 at 01:18:09PM +1030, Clytie Siddall wrote: Hello again all fans of this thread (the T-shirt will be out soon) :D I probably should have asked this after the first post, but ... what is g-i? GNU-install? Gnome-install? frivolousGeneral Incredulity?/ frivolous (Clytie has a terrible tendency to frivolity.) g-i = Graphical Installer Translations should be kept 100%, because new installer releases are imminent. -- Tapio Lehtonen [EMAIL PROTECTED] signature.asc Description: Digital signature
Re: [g-i] freefont vs. dejavu
Quoting Clytie Siddall ([EMAIL PROTECTED]): Then why all the urgency to keep the Installer at 100%? I don't understand. :( Because we're doing beta releases and I want these to be as fully translated as possible. However, we can't announce such releases as widely as an official release of the overall distribution. This is why I react when you talk about waving flags. I don't want to prevent you waving flags but not too big ones...:) I also want translators avoid the hurry of last minute updates which will receive less exposure if they are done just before the release. signature.asc Description: Digital signature
Re: [g-i] freefont vs. dejavu
Hi Clytie, On Sun, Mar 19, 2006 at 12:06:27PM +1030, Clytie Siddall wrote: As for the fontsize we can increase it, so it would not be a problem. Great. I'd like to have a more detailed look at the font. I took a screenshot of the same text using a fonstsize of 16 points: http://www.webalice.it/zinosat/shots/vi_16pt.png The user will be able to choose font-size for the installer, won't they? If not, I suggest a slightly larger size than normal for Vietnamese, because we have to distinguish all the accents, and some of them do blend in to each other at smaller sizes. Being realistic, I guess it'll take a while before we offer the user the chance to change properties of fonts during installation; on the other hand Frans is about to add some code which makes it possible to use different size for different languages: we will take into account your suggestion and increase default fontsize for Vietnamese (we already have to do it for Arabic / Persian) Perhaps you should file a bug for the freefont authors; I shall: I know they want to know these things. nice. I'll post some pictures of Vietnamese text using dejavu fonts as soon as the new version of the package is available so... stay tuned :) Ciao, Davide signature.asc Description: Digital signature
Re: [g-i] freefont vs. dejavu
Hellon again :) On 21/03/2006, at 9:58 AM, Davide Viti wrote: On Sun, Mar 19, 2006 at 12:06:27PM +1030, Clytie Siddall wrote: As for the fontsize we can increase it, so it would not be a problem. Great. I'd like to have a more detailed look at the font. I took a screenshot of the same text using a fonstsize of 16 points: http://www.webalice.it/zinosat/shots/vi_16pt.png Thanks, that is easier to read. The text is definitely readable, although the font does not meet some essential criteria for being a really readable Vietnamese font. I have written a detailed critique to the Freefont developers (including comparative screenshots) which I hope will be useful. Meanwhile, Freefont works for Vietnamese, it's just not as good as we would like. For example, even at 16pt, I have to guess many of the words, because the accents over the vowels blend into the line above. However, that's Freefont's issue. Can we use it for Vietnamese? Yes, if we don't have anything better. :) The user will be able to choose font-size for the installer, won't they? If not, I suggest a slightly larger size than normal for Vietnamese, because we have to distinguish all the accents, and some of them do blend in to each other at smaller sizes. Being realistic, I guess it'll take a while before we offer the user the chance to change properties of fonts during installation; on the other hand Frans is about to add some code which makes it possible to use different size for different languages: we will take into account your suggestion and increase default fontsize for Vietnamese (we already have to do it for Arabic / Persian) Thanks, that will help. I don't want outraged posts on VNOSS saying, Why didn't you package the magnifying glass?? :D I'll post some pictures of Vietnamese text using dejavu fonts as soon as the new version of the package is available so... stay tuned :) Thanks, that will be really useful. I'd like to grab a copy of the new fonts when they're available, and compare them with my existing standard comparison fonts. Hmm, I feel a quick webpage coming on... grabs text editor Gak, took much longer than I had imagined, but it's fairly detailed. http://www.riverland.net.au/~clytie/viet/fonts.html I'll have to work on the translation when I have a bit more time. I'll update it with more info on Freefont and Dejavu as it becomes available. Do we have a link where I can point people, about Debian being about to release? Also, will we be doing Release Notes and generally waving the flag a bit? I'd like to know, so I can (1) translate the release notes and (2) spread the good news on Vietnamese sites. :D from Clytie (vi-VN, Vietnamese free-software translation team / nhóm Việt hóa phần mềm tự do) http://groups-beta.google.com/group/vi-VN PGP.sig Description: This is a digitally signed message part
Re: [g-i] freefont vs. dejavu
On 17/03/2006, at 10:18 PM, Davide Viti wrote: Here's ([1]) a screenshot of some vietnamese text rendered with the new version of freefont uploaded by Christian; does it look right? This version of freefont should contain all the glyphs needed to render Vietnamese. Thanks for the screenshot, Davide. I hope it doesn't show any functional part of the installer, because it's all fragments of strings, not complete strings. ;) It's difficult to give detailed feedback on the font, because the text in the screenshot is so small. I can say that the right accents are on the right characters, so the text (when it makes sense ;) ) is readable. That's the key thing. It seems to show all our glyphs. Some of the accents are not positioned correctly when combined: should I give that detailed feedback here, or should I do that to the freefont authors? (It doesn't affect meaning, but it will have users in my community thinking, Yuk, what's wrong with this font?) Our language is a challenge to font-creators, so the freefont people have done well to get the glyphs represented readably. I have screenshots of the Verajja font compared with Lucida Grande, the general-purpose Unicode font on my system, and a font created especially for Vietnamese, if you would find those useful. Has Verajja been integrated into Dejavu yet? I have tested the Verajja font with Vietnamese, for its creator. from Clytie (vi-VN, Vietnamese free-software translation team / nhóm Việt hóa phần mềm tự do) http://groups-beta.google.com/group/vi-VN
Re: [g-i] freefont vs. dejavu
On Fri, Mar 17, 2006 at 05:14:17PM +0200, Eddy Petrişor wrote:
Macedonian and Ukrainian use the following to quote text:
U201c - Left double quotation mark
U201e - double low-9 quotation mark
mk never uses normal quote char (U0022), only minor problem
is:
94 U201c
93 U201e „
I think they know better what their quoting signs are than us :-)
agree.
I've fixed the following so the number of open / close quotes should be
right:
Are you sure you fixed them? Maybe you shouldn't be this hasty in
clasifing thngs wrong or right. Please ask the translators.
I normally ask and change things only when I'm sure what I'm doing.
Here are the changes I did (which I still consider fixes).
If those are wrong I apologize and will revert the changes.
Davide
Index: mk.po
===
--- mk.po (revisione 35468)
+++ mk.po (revisione 35481)
@@ -893,7 +893,7 @@
На некој PCMCIA хардвер му требаат специјални параметри за тој да работи и
може да предизвика компјутерот да се закочи ако не се конфигурираат. На
пример за некои Dell лаптопи потребно е тука да се внесе „exclude port 0x800-
-0x8ff„. Овие параметри ќе бидат додадени во /etc/pcmcia/config.opts. Види го
+0x8ff“. Овие параметри ќе бидат додадени во /etc/pcmcia/config.opts. Види го
инстлациониот прирачник или PCMCIA HOWTO за повеќе информации.
#. Type: string
@@ -6761,7 +6761,7 @@
#. Description
#: ../lilo-installer.templates:89
msgid Running \/sbin/lilo\ failed with error code \${ERRCODE}\.
-msgstr Извршувањето на “/sbin/lilo“ не успеа, враќајќи грешка „${ERRCODE}“.
+msgstr Извршувањето на „/sbin/lilo“ не успеа, враќајќи грешка „${ERRCODE}“.
#. Type: text
#. Description
@@ -7453,7 +7453,7 @@
msgid
Execution of preseeded command \${COMMAND}\ failed with exit code ${CODE}.
msgstr
-Извршувањето на preseeded командата “${COMMAND}“ не успеа враќајќи код на
+Извршувањето на preseeded командата „${COMMAND}“ не успеа враќајќи код на
грешка ${CODE}.
#. Type: text
@@ -7821,7 +7821,7 @@
#. Description
#: ../save-logs.templates:22
msgid The directory \${DIR}\ does not exist.
-msgstr Директориумот „${DIR} не постои.
+msgstr Директориумот „${DIR}“ не постои.
#. Type: note
#. Description
signature.asc
Description: Digital signature
Re: [g-i] freefont vs. dejavu
On Sat, Mar 18, 2006 at 09:37:51PM +1030, Clytie Siddall wrote: On 17/03/2006, at 10:18 PM, Davide Viti wrote: Here's ([1]) a screenshot of some vietnamese text rendered with the new version of freefont uploaded by Christian; does it look right? This version of freefont should contain all the glyphs needed to render Vietnamese. Thanks for the screenshot, Davide. I hope it doesn't show any functional part of the installer, because it's all fragments of strings, not complete strings. ;) don't worry! I just select the longest lines of text from the po file so that native speakers (i.e. you) have a better idea of how glyphs look like. It's difficult to give detailed feedback on the font, because the text in the screenshot is so small. I can say that the right accents are on the right characters, so the text (when it makes sense ;) ) is readable. That's the key thing. It seems to show all our glyphs. nice: last time you pointed out it was almost unreadable because of missing glyphs. As for the fontsize we can increase it, so it would not be a problem. Some of the accents are not positioned correctly when combined: should I give that detailed feedback here, or should I do that to the freefont authors? (It doesn't affect meaning, but it will have users in my community thinking, Yuk, what's wrong with this font?) Perhaps you should file a bug for the freefont authors; consider the following though: 1) I'd like to have the new freefont as default version of freefont for g-i (we're still in the process of switching) 2) Christian is about to upload the new version of Dejavu fonts, and we might use those to render Vietnamese Our language is a challenge to font-creators, so the freefont people have done well to get the glyphs represented readably. I have screenshots of the Verajja font compared with Lucida Grande, the general-purpose Unicode font on my system, and a font created especially for Vietnamese, if you would find those useful. Has Verajja been integrated into Dejavu yet? I have tested the Verajja font with Vietnamese, for its creator. I'll let you have a screenshot of Vietnamese text rendered with Dejavu as soon as the package reaches the mirrors. Ciao, Davide signature.asc Description: Digital signature
Re: [g-i] freefont vs. dejavu
Hello again :) On 19/03/2006, at 1:10 AM, Davide Viti wrote: On Sat, Mar 18, 2006 at 09:37:51PM +1030, Clytie Siddall wrote: On 17/03/2006, at 10:18 PM, Davide Viti wrote: Here's ([1]) a screenshot of some vietnamese text rendered with the new version of freefont uploaded by Christian; does it look right? This version of freefont should contain all the glyphs needed to render Vietnamese. Thanks for the screenshot, Davide. I hope it doesn't show any functional part of the installer, because it's all fragments of strings, not complete strings. ;) don't worry! I just select the longest lines of text from the po file so that native speakers (i.e. you) have a better idea of how glyphs look like. phew I was a bit worried, there. ;) It's difficult to give detailed feedback on the font, because the text in the screenshot is so small. I can say that the right accents are on the right characters, so the text (when it makes sense ;) ) is readable. That's the key thing. It seems to show all our glyphs. nice: last time you pointed out it was almost unreadable because of missing glyphs. It is great progress. :)) Thankyou for all your work. It would be a big loss of face for Debian if we weren't able to support Vietnamese in this Installer: originally, I was told it would be out in June 2005, so I passed that on to our community, and there was great interest. Not being able to deliver then, damaged our credibility. Not being able to deliver even now, would lose us our remaining partisans. As for the fontsize we can increase it, so it would not be a problem. Great. I'd like to have a more detailed look at the font. The user will be able to choose font-size for the installer, won't they? If not, I suggest a slightly larger size than normal for Vietnamese, because we have to distinguish all the accents, and some of them do blend in to each other at smaller sizes. Some of the accents are not positioned correctly when combined: should I give that detailed feedback here, or should I do that to the freefont authors? (It doesn't affect meaning, but it will have users in my community thinking, Yuk, what's wrong with this font?) Perhaps you should file a bug for the freefont authors; I shall: I know they want to know these things. consider the following though: 1) I'd like to have the new freefont as default version of freefont for g-i (we're still in the process of switching) If we have no other choices, freefont would be fine for simply getting the information across. However, if you want to look good the users, it's better not to use a font that doesn't show the accents in the correct positions. It'd be a bit like installing Debian in English with all the dots from the is underneath them instead, and the crosses on the ts up above the t, you wouldn't have a lot of respect for that project, would you? 2) Christian is about to upload the new version of Dejavu fonts, and we might use those to render Vietnamese If they've integrated Verajja, the latest version (which has been adjusted after my feedback on, unsurprisingly, accent position), that would definitely be a better choice. I was following up Verajja- Dejavu myself (didn't know about freefont), and have had time to get that font thoroughly analyzed by several people. Our language is a challenge to font-creators, so the freefont people have done well to get the glyphs represented readably. I have screenshots of the Verajja font compared with Lucida Grande, the general-purpose Unicode font on my system, and a font created especially for Vietnamese, if you would find those useful. Has Verajja been integrated into Dejavu yet? I have tested the Verajja font with Vietnamese, for its creator. I'll let you have a screenshot of Vietnamese text rendered with Dejavu as soon as the package reaches the mirrors. Thanks, Davide. :) Good luck with everything! These are exciting times. :)) from Clytie (vi-VN, Vietnamese free-software translation team / nhóm Việt hóa phần mềm tự do) http://groups-beta.google.com/group/vi-VN PGP.sig Description: This is a digitally signed message part
Re: [g-i] freefont vs. dejavu
Hi Jens, On Fri, Mar 17, 2006 at 08:58:13AM +0100, Jens Seidel wrote: just a wishlist item: I would like to see the character (encoded in UTF-8) in a third column. It's actually already in my todo list :) ciao, Davide signature.asc Description: Digital signature
Re: [g-i] freefont vs. dejavu
On Fri, Mar 17, 2006 at 08:58:13AM +0100, Jens Seidel wrote: I would like to see the character (encoded in UTF-8) in a third column. done. I wonder if what happens to [1] (and to RTL languages in general) is a bug of the browser (I tried both Firefox and Epiphany) or is because I should take some special care when creating the text file. ciao, Davide [1] http://d-i.alioth.debian.org/spellcheck/level1/latest/nozip/ar_codes.txt signature.asc Description: Digital signature
Re: [g-i] freefont vs. dejavu
On Fri, Mar 17, 2006 at 10:45:51AM +0100, Davide Viti wrote: On Fri, Mar 17, 2006 at 08:58:13AM +0100, Jens Seidel wrote: I would like to see the character (encoded in UTF-8) in a third column. done. Thanks. I wonder if what happens to [1] (and to RTL languages in general) is a bug of the browser (I tried both Firefox and Epiphany) or is because I should take some special care when creating the text file. This is browser specific, since it looks better in konqueror and kwrite. Nevertheless the last 9 (and only these) entries look like: 3 U0652 x where x is displayed right of the closing (or opening in this language?) quote sign. Maybe it's just a negative distance which is valid. [1] http://d-i.alioth.debian.org/spellcheck/level1/latest/nozip/ar_codes.txt Jens -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: [g-i] freefont vs. dejavu
On Thu, Mar 16, 2006 at 01:54:30AM +0100, Davide Viti wrote: po file (sq.po) uses U0401 Ё twice (see [1]) U+0401: CYRILLIC CAPITAL LETTER IO I found out that there's a glyph tha looks the same in the latin range: U+00CB LATIN CAPITAL LETTER E WITH DIAERESIS I can't read Albanian, but looks to me that this should be changed. just fixed. also noticed tha U00cb is already used 15 times. regards, Davide signature.asc Description: Digital signature
Re: [g-i] freefont vs. dejavu
Hi Clytie, Here's ([1]) a screenshot of some vietnamese text rendered with the new version of freefont uploaded by Christian; does it look right? This version of freefont should contain all the glyphs needed to render Vietnamese. Ciao Davide [1] http://www.webalice.it/zinosat/shots/vi_20060317.png -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: [g-i] freefont vs. dejavu
Kazakh po file (kk.po), uses the following:
U2013 - EN dash (37 times)
U2014 - EM dash (once)
the same way as the normal - (002D HYPHEN-MINUS)
Given the way it is used (no consistency at all) I'd
change those into normal hyphens. kk is not the only
language using various flavours of the hyphen: I think also
Vietnamese uses it.
Vietnamese uses the bullet (U+2022 BULLET) to
list items that in the original msgid are marked with -
the newt frontend has this bullet glyph (see [1]) and even if I
personally think the installer shouldn't use caharacters as if it was
an OO document, I can understand Clytie point.
Macedonian and Ukrainian use the following to quote text:
U201c - Left double quotation mark
U201e - double low-9 quotation mark
mk never uses normal quote char (U0022), only minor problem
is:
94 U201c “
93 U201e „
I've fixed the following so the number of open / close quotes should be right:
1) пример за некои Dell лаптопи потребно е тука да се внесе „exclude port
0x800-0x8ff„
2) Извршувањето на “/sbin/lilo“ не успеа, враќајќи грешка „${ERRCODE}“.
3) Извршувањето на preseeded командата “${COMMAND}“ не успеа враќајќи код на
4) Директориумот „${DIR} не постои.
uk uses normal quote character 8 times (always surrounding root)
and U201c U201e 97 times, which looks like consistency is kept.
One last thing:
bot es and gl use ¿ (U00bf) at the beginning of every question:
is it normal / right ?
ciao,
Davide
[1] http://www.webalice.it/zinosat/shots/vi_dot.png
signature.asc
Description: Digital signature
Re: [g-i] freefont vs. dejavu
On Friday 17 March 2006 14:37, Davide Viti wrote: One last thing: bot es and gl use ¿ (U00bf) at the beginning of every question: is it normal / right ? Yes. pgpD4eYehECPq.pgp Description: PGP signature
Re: [g-i] freefont vs. dejavu
On 3/17/06, Davide Viti [EMAIL PROTECTED] wrote: Kazakh po file (kk.po), uses the following: U2013 - EN dash (37 times) U2014 - EM dash (once) the same way as the normal - (002D HYPHEN-MINUS) Given the way it is used (no consistency at all) I'd change those into normal hyphens. kk is not the only language using various flavours of the hyphen: I think also Vietnamese uses it. Please do not fiddle in fields you don't understand. Some languages have such things as non-breaking hyphen _and_ splitting hyphen (sorry for the translation, but I can't find a better equivalent). They have diferent meaning and changing them could result in a wrongly redered text. As a side note, this is true for Romanain, but there no agreement on which unidcode points should be used. Macedonian and Ukrainian use the following to quote text: U201c - Left double quotation mark U201e - double low-9 quotation mark mk never uses normal quote char (U0022), only minor problem is: 94 U201c 93 U201e „ I think they know better what their quoting signs are than us :-) I've fixed the following so the number of open / close quotes should be right: Are you sure you fixed them? Maybe you shouldn't be this hasty in clasifing thngs wrong or right. Please ask the translators. One last thing: bot es and gl use ¿ (U00bf) at the beginning of every question: is it normal / right ? Yes -- Regards, EddyP = Imagination is more important than knowledge A.Einstein
Re: [g-i] freefont vs. dejavu
Hello everyone :) This Installer will be the very first supporting my language: it's such a huge step forward for us. There's a big conference being held by Intel in Hanoi very soon, and Javier Sola is presenting there: he's said he will talk about our progress. It's all very encouraging: government and business have the distribution capacity in Vietnam, so if we can get them on side, we will reach so many people we're not currently reaching. Our translations are used widely among students, educated people, some business people and the Net Café underground, but they're not reaching most of the people, and those people who need them most. Another important point: it's essential to have the Installer, and packages, also available offline (e.g. CDs), since Net access in many countries is unreliable at best, and completely non-existent in many cases. That is the case in Vietnam, and I heard this week from a computing engineer in Africa who has only very occasional email access and no download capacity at all. I know the Installer is available on CD, but how widely available are the CDs, physically? How about packages? I hope you don't mind me mentioning this here: I'm not sure where to mention it. On 16/03/2006, at 5:11 PM, Eddy Petrişor wrote: 4 --- Macedonian and Ukrainian use the following to quote text: U201c - Left double quotation mark U201e - double low-9 quotation mark Jutta Wrage has recently made a sumarisation of the correct quotation marks needed for each language. http://www.witch.westfalen.de/csstest/quotes/quotes.html It's very useful: we have a link to it from the Translate Wiki [1]. Thankyou, Jutta. :) Jutta, could you enlighten us? Shall we fix those, ask the translators? I think that characters like the bullet should't be used for the installer: IMO those are meant to be used in applications like Openoffice or similar. I tend to agree, but maybe we might want to hear Clytie's opinion. I use them fairly widely in PO files, because they don't add to the visual clutter of a heavily-accented language. They are discrete, so obviously (to the eye) not another character or diacritic mark. This is useful, helps to format the text. You will notice (looking at such PO files) that I use them to set points out clearly, to avoid requiring complex syntax or advanced vocabulary in the translation. A large proportion of our population is only marginally literate in this sense: the clearer the translation, with the minimum of syntactical embedding and educated vocabulary, the more likely the it is to be effective, or to be read at all. Bullets function here as a graphic aid. For Vietnamese, bullets are appropriate. They are available in the Latin-1 character-set, so everyone is likely to have them covered in default-installation fonts, for those people not automatically using a UTF-8 setting (as our users will). I think, for all Asian languages, bullets are an appropriate way to format the text. We have so many thin, straight or curved shapes and no solid, centred ones. A solid bullet makes an effective contrast, and focusses the eye. (I speak here also as a specialist on the mechanics of reading.) This is likely to apply to all languages using many diacritics. (However, individual languages may have their own formatting, or a separate meaning for this character.) I hope this is useful. :) from Clytie (vi-VN, Vietnamese free-software translation team / nhóm Việt hóa phần mềm tự do) http://groups-beta.google.com/group/vi-VN [1] http://translate.sourceforge.net/wiki/
Re: [g-i] freefont vs. dejavu
On 3/16/06, Clytie Siddall [EMAIL PROTECTED] wrote: Hello everyone :) Hi Clytie, --- Macedonian and Ukrainian use the following to quote text: U201c - Left double quotation mark U201e - double low-9 quotation mark Jutta Wrage has recently made a sumarisation of the correct quotation marks needed for each language. http://www.witch.westfalen.de/csstest/quotes/quotes.html It's very useful: we have a link to it from the Translate Wiki [1]. Thankyou, Jutta. :) Unfortunately Macedonian is missing and I can't figure out the codepoints for Ukrainian. Jutta, may I suggest adding a line after each alternative to specify the used unicode points for the quotation marks? I think this would make it a lot more clearer. The good news is that I know somebody from Macdonia who can help us ;-) with first issue. Jordanka, could you please tell us the correct quotation marks (as unicode points - see the quotation mark table[1]) for Macedonian? And/or direct us to the guy you said will be interested in contributing to the Debian Installer as a traslator for Macedonian(hint, hint, hint, you forgot that ;-) . Jutta can latter add this info to her quotation marks page[2]. [1] http://www.witch.westfalen.de/csstest/quotes/quotes-large.html [2] http://www.witch.westfalen.de/csstest/quotes/quotes.html -- Regards, EddyP = Imagination is more important than knowledge A.Einstein
Re: [g-i] freefont vs. dejavu
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Hi!
I do not read debian-boot. And I am not subscribet, so CC me please.
Am 16.03.2006 um 07:41 schrieb Eddy Petrişor:
4
---
Macedonian and Ukrainian use the following to quote text:
U201c - Left double quotation mark
U201e - double low-9 quotation mark
Jutta, could you enlighten us?
I have do not have any information about Macedonian. But for the
Ukrainian language it seems the the first alternative at http://
www.witch.westfalen.de/csstest/quotes/quotes.html#ukrainian would be
the best one.
Note: I got new info for Ukraininian lately and that part changed!
div[lang=uk] q:before { content: \00AB; }
div[lang=uk] q:after { content: \00BB; }
div[lang=uk] q q:before { content: \201E; }
div[lang=uk] q q:after { content: \201C; }
Alternative2 is used for Handwriting only normally. And two closing
quotation marks (u+201C) cannot be together using it, which makes
automatic processing quite difficult.
Some notes: neither the information from wikipedia nor that from the
standardisation pages is reliable. Both of them seem to be made by
people using computers for a long time. The very best sources are
books (especially thos using old kind of typesetting and _not_
computer typesetting).
For computer we _can_ make compromizes. But I think the best would be
to make none, if possible.
BTW: German has two different solutions, too, but U+00AB are never
used in handwriting. So handwriting rules might be another source of
confusion beside old fashioned computer typing.
Question: Do we need pages how to create quotation marks on computer
keyboards?
greetings
Jutta
- --
http://www.witch.westfalen.de
http://witch.muensterland.org
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.4.1 (Darwin)
iEYEARECAAYFAkQZJxwACgkQOgZ5N97kHkfOUQCfaqD/tYpZWhWd3JjXBp8xHmfr
kkQAnjOps4Kwrn8dVHEBV/wKBvSnV8m6
=brcu
-END PGP SIGNATURE-
Re: [g-i] freefont vs. dejavu
I'm considering to strip font files specifying the ranges to include rather than the ranges that we want to exclude, what do you think about this? For freefont exclusion (ranges we know are covered by other fonts and ranges for which we know there are no translations (yet)) is maybe more logical (probably easier to document too); for other fonts inclusion definitely. I'll leave it to you though to decide what works best. this actually simplify things, given the number of languages which is covered by freefont. If we're using freefont also to display Cyrillic, Christian just needs to comment the following line from FreeSans-strip: ranges=$ranges u0400:u052F # Cyrilic and now that I filed #357154 we have a good excuse to ask for another upload of the package; I'd like to do some tests before the new upload. removing dejavu (or freefont if the new dejavu can cover everything), would simplify and reduce overhead (basic latin glyphs, as you mentioned above, are not stripped to preserve consistency: i.e. at least u0:u7f = 128 glyphs are included in each and every ttf file including bold/oblique) Hmm. Not sure we need to include basic latin for e.g. arabic and indic fonts. Guess that depends on how different the included characters are from freefont (both style and size should be considered). I've no problem with leaving them in initially though. yes, I would start leaving those glyphs. I'd like to make a few considerations that justify the reason why I'm getting more and more convinced that we need to reduce as much as possible the number of font files: Finnish contains Moskova - +09 - Kamtšatka, Moskova+10 - Beringinmeri (Kamtšatka contains U0161 Latin Small Letter S With Caron) French (among other languages) contains São Paulo in one of its strings the risk is to have a language displayed with one particular font and one of the glyphs taken from another font file: this might look very ugly. Next step is probably creating an iso that supports the font switching from localechooser to see how that behaves. I can probably do that quite quickly when you're ready with font tarballs. K. ciao, Davide Tiscali ADSL 4 Mega Flat Naviga senza limiti con l'unica Adsl a 4 Mega di velocità a soli 19,95 € al mese! Attivala subito e hai GRATIS 2 MESI e l'ATTIVAZIONE. http://abbonati.tiscali.it/banner/middlepagetracking.html?c=webmailadslr=http://abbonati.tiscali.it/adsl/sa/4flat_tc/a=webmailz=webmailt=14
Re: [g-i] freefont vs. dejavu
On Thursday 16 March 2006 07:41, Eddy Petrişor wrote: Jutta Wrage has recently made a sumarisation of the correct quotation marks needed for each language. Eh, that assumes that these quotation marks can actually be shown in the installer (especially in the newt interface). Shouldn't we check that first? pgp8nlJ8zhlJA.pgp Description: PGP signature
Re: [g-i] freefont vs. dejavu
On Thursday 16 March 2006 11:36, Davide Viti wrote: this actually simplify things, given the number of languages which is covered by freefont. If we're using freefont also to display Cyrillic, Christian just needs to comment the following line from FreeSans-strip: Hmm. Shouldn't we talk to ru, uk, be, sr, etc. translators before switching to freefont for Cyrilic? IIRC they were the ones that asked us to use Dejavu in the first place. pgpk1hZZUGg3O.pgp Description: PGP signature
Re: [g-i] freefont vs. dejavu
On 16/03/2006, at 7:21 PM, Jutta Wrage wrote: Question: Do we need pages how to create quotation marks on computer keyboards? That's a valid point, Jutta. They are different writing environments, with different constraints. Spacing and size is more variable in handwriting, for one thing. Style of writing, both visual and syntactical, can vary a lot. It would at least be valuable to allow for separate entries for computing and handwriting. from Clytie (vi-VN, Vietnamese free-software translation team / nhóm Việt hóa phần mềm tự do) http://groups-beta.google.com/group/vi-VN PGP.sig Description: This is a digitally signed message part
Re: [g-i] freefont vs. dejavu
this actually simplify things, given the number of languages which is covered by freefont. If we're using freefont also to display Cyrillic, Christian just needs to comment the following line from FreeSans-strip: Hmm. Shouldn't we talk to ru, uk, be, sr, etc. translators before switching to freefont for Cyrilic? IIRC they were the ones that asked us to use Dejavu in the first place. yes, but we should also consider that the following glyphs are used by Cyrillic translations and are not in current (2.1-2) DejavuSans ttf files: U04a2 3 times in kk.po U04a3 1118 times in kk.po U04ae once in kk.po U04af 1059 times in kk.po U04b0 once in kk.po U04b1 422 times in kk.po U04d8 5 times in kk.po U04d9 199 times in kk.po U04e8 15 times in kk.po U04e9 734 times in kk.po U2116 38 times in bg.ko Well, considering that #349511 (new version [2.2] available) was filed 23 Jan 2006 which is almost two months ago I assume there's not much activity going on with the package. OTOH, Christian is always very fast if we need anything and he's taking care of freefont; besides that he's obviously very close to the needs of the debian installer. The best would be if both freefont and dejavu where competing for being the best font covering the largest number of languages: a project like g-i would fit perfectly for such a target. regards, Davide Tiscali ADSL 4 Mega Flat Naviga senza limiti con l'unica Adsl a 4 Mega di velocità a soli 19,95 € al mese! Attivala subito e hai GRATIS 2 MESI e l'ATTIVAZIONE. http://abbonati.tiscali.it/banner/middlepagetracking.html?c=webmailadslr=http://abbonati.tiscali.it/adsl/sa/4flat_tc/a=webmailz=webmailt=14
Re: Fw: [g-i] freefont vs. dejavu
Hello everyone Jordanka has asked me If I'm interested in contributing to the Debian installer. My name is Zoran Dimovski and I'll try to help as mush as I can. On 3/16/06, Clytie Siddall [EMAIL PROTECTED] wrote: Hello everyone :)Hi Clytie, --- Macedonian and Ukrainian use the following to quote text: U201c - Left double quotation mark U201e - double low-9 quotation mark Jutta Wrage has recently made a sumarisation of the correct quotation marks needed for each language. http://www.witch.westfalen.de/csstest/quotes/quotes.html It's very useful: we have a link to it from the Translate Wiki [1]. Thankyou, Jutta. :)Unfortunately Macedonian is missing and I can't figure out thecodepoints for Ukrainian.Jutta, may I suggest adding a line after each alternative to specifythe used unicode points for the quotation marks? I think this would make it a lot more clearer.The good news is that I know somebody from Macdonia who can help us;-) with first issue.Jordanka, could you please tell us the correct quotation marks (asunicode points - see the quotation mark table[1]) for Macedonian? The correct quotation marks for Macedonian are as mentioned abovefirst we useU201e - double low-9 quotation markand thenU201c - Left double quotation markand looks something like this „ " And/or direct us to the guy you said will be interested incontributing to the Debian Installer as a traslator for Macedonian(hint, hint, hint, you forgot that ;-) .That would be me :) so please tell me should I subscribe to the Debina-boot list or you will send me messages personaly.Best regards, Zoran. Free and Open Source Macedonia
Re: [g-i] freefont vs. dejavu
Quoting Frans Pop ([EMAIL PROTECTED]): Hmm. Shouldn't we talk to ru, uk, be, sr, etc. translators before switching to freefont for Cyrilic? IIRC they were the ones that asked us to use Dejavu in the first place. Mostly Bulgarian and Ukrainian translators or contributors (namely Ognyan Kulev, Anton Zinoviev and Eugueniy Meschcheryakov)... I don't remember advices from Russian translators, nor Belarusian and Serbian. -- signature.asc Description: Digital signature
Re: [g-i] freefont vs. dejavu
On 3/16/06, Frans Pop [EMAIL PROTECTED] wrote: On Thursday 16 March 2006 07:41, Eddy Petrişor wrote: Jutta Wrage has recently made a sumarisation of the correct quotation marks needed for each language. Eh, that assumes that these quotation marks can actually be shown in the installer (especially in the newt interface). Shouldn't we check that first? You read my mind, but I don't see why this should be a problem. -- Regards, EddyP = Imagination is more important than knowledge A.Einstein
Re: [g-i] freefont vs. dejavu
On 3/16/06, Davide Viti [EMAIL PROTECTED] wrote: this actually simplify things, given the number of languages which is covered by freefont. If we're using freefont also to display Cyrillic, Christian just needs to comment the following line from FreeSans-strip: Hmm. Shouldn't we talk to ru, uk, be, sr, etc. translators before switching to freefont for Cyrilic? IIRC they were the ones that asked us to use Dejavu in the first place. yes, but we should also consider that the following glyphs are used by Cyrillic translations and are not in current (2.1-2) DejavuSans ttf files: U04a2 3 times in kk.po U04a3 1118 times in kk.po U04ae once in kk.po U04af 1059 times in kk.po U04b0 once in kk.po U04b1 422 times in kk.po U04d8 5 times in kk.po U04d9 199 times in kk.po U04e8 15 times in kk.po U04e9 734 times in kk.po U2116 38 times in bg.ko Maybe if we collaborate with deja-vu upstream and the other BitstreamVera derivatives: http://dejavu.sourceforge.net/wiki/index.php/Bitstream_Vera_derivatives Maybe, just maybe, we'll have a decent font for cyrillic ;-) [I know my ideas seem many times a little too far from reality, but we have to have dreams, don't we?] Another hint on font coverage: http://www.fileformat.info/info/unicode/font/index.htm Well, considering that #349511 (new version [2.2] available) was filed 23 Jan 2006 which is almost two months ago I assume there's not much activity going on with the package. :( -- Regards, EddyP = Imagination is more important than knowledge A.Einstein
Re: Fw: [g-i] freefont vs. dejavu
Quoting Зоран Димовски ([EMAIL PROTECTED]): Hello everyone Jordanka has asked me If I'm interested in contributing to the Debian installer. My name is Zoran Dimovski and I'll try to help as mush as I can. Hello Зоран (I love writing cyrillic), I am the i18n coordinator for Debian Installer. If you're interested in contributing to D-I in your language, the best is probably getting first in touch with Georgi Stanojevski who's currently in charge of the Macedonian translations. Georgi did most of the current translations, but there's always work for more people. You can also contact the team at [EMAIL PROTECTED], from my records. The D-I i18n documentation is available at http://people.debian.org/~bubulle/d-i/i18n-doc and better explains the concepts we use for Debian Installer and related translation works. signature.asc Description: Digital signature
Re: [g-i] freefont vs. dejavu
Maybe if we collaborate with deja-vu upstream and the other BitstreamVera derivatives: http://dejavu.sourceforge.net/wiki/index.php/Bitstream_Vera_derivatives Maybe, just maybe, we'll have a decent font for cyrillic ;-) [I know my ideas seem many times a little too far from reality, but we have to have dreams, don't we?] Well, I also saw mentions on takiong over the maintenance of the ttf-dejavu package in the #debian-boot IRC channel. However, shouldn't all this be first discussed with Peter Cernak, the current maintainer of the ttf-dejavu font package? My feeling is that Peter is pretty responsive. Another possibility: please people join the pkg-fonts team, subscribe to the list ([EMAIL PROTECTED]) and discuss all this over there. After all, that's the point of this pkg-fonts package I launched very recently (see debian-devel): have font package maintainers work together on common issues. To Peter: all this is about DejaVu fonts covergae of some ranges, as the font is one of the possibilities we have to the graphical installer (it is requested by several Cyrillic translators, over Freefont). signature.asc Description: Digital signature
Re: [g-i] freefont vs. dejavu
On Thu, Mar 16, 2006 at 02:15:09PM +0200, Tapio Lehtonen wrote: Finnish contains Moskova - +09 - Kamtšatka, Moskova+10 - Beringinmeri (Kamtšatka contains U0161 Latin Small Letter S With Caron) Words from finnish language proper do not contain characters outside a-zåäö. But foreign words are sometimes written with special S characters. If that kind of translitteration is bad, I can write Kamtshatka, but I went to quite a bit of trouble to check the correct spelling for those names in time zones, so I prefer the current spellings. I think the current translation is just right. I just wanted to point out that if we're using various fonts to render the text of different languages, we might end up having a mixture of glyphs from different font files to just display a word. I haven't checked yet, but if for example Arabic translation contained the word Kamtšatka, the š would be picked from FreeSans.ttf, whereas the other letters would be taken from nazli.ttf and the result could be ugly. I have not really understood the problem with glyphs. Surely there is a finite number of glyphs, and the number of glyphs actually used in Debian-installer is even smaller. yes. Is it a complex process to determine what glyphs are used or does the set of glyphs change too fast? see the codepoints column in [1]. I don't know how fast does the set change (in poart because of the string freeze, in part because the computation of codepoints is a very young feature) ciao, Davide [1] http://d-i.alioth.debian.org/spellcheck/level1/index.html signature.asc Description: Digital signature
Re: [g-i] freefont vs. dejavu
I spent some time working on collecting / stripping a nice set of fonts for the g-i. This is (was) the idea (see [1]): *** ttf-dejavu - Cyrillic scripts *** ttf-freefont - Latin - Latin-other - hi - pa_IN - bn - he - vi - el *** ttf-cjk-compact-udeb - CJK *** ttf-farsiweb (nazli) - ar / fa *** ttf-khmeros - km *** ttf-tamil-fonts (TSCu_Paranar.ttf) - ta I started from dejavu fonts and found out that the following glyphs are missing: 04a2 04a3 04ae 04af 04b0 04b1 04d8 04d9 04e8 04e9 2116 all the above glyphs are in the new ttf-freefont package; I wonder if we should consider switching to freefont whenever possible. This would bring alot of advantages (both in size and simplicity). ciao, Davide [1] http://wiki.debian.org/DebianInstaller/GUIFonts signature.asc Description: Digital signature
Re: [g-i] freefont vs. dejavu
On Thursday 16 March 2006 01:00, Davide Viti wrote: This is (was) the idea (see [1]): Looks good to me. Could you add the glyph ranges needed in the different fonts (except freefont)? Please also consider if some common codepoints, like for numbers, general punctuation and maybe accents should be included from the different fonts in addition to their base ranges. This could help to keep visual consistency within translations (only works if we implement changing the default font on selection of a different language of course). In some cases, like for CJK, we may want to keep latin characters too. For CJK this has been discussed before (and AFAIK is already implemented in the udeb). I wonder if we should consider switching to freefont whenever possible. This would bring alot of advantages (both in size and simplicity). That's basically what we've been doing so far, isn't it? Still, I think that if a different font really looks better to native speakers than freefont, we should in principle [1] use the other font [2]. [1] An exception could be if that font is much larger than alternative fonts. [2] Within reason of course. We can't let a translator for, say, Dutch decide he wants a different font. It's only acceptable for scripts that have separate ranges in the UTF table. pgpfxrnnpKrui.pgp Description: PGP signature
Re: [g-i] freefont vs. dejavu
Since we're on the subject, I've found some weirdness in the translations: 1 --- Albanian (which is listed among the latin languages) po file (sq.po) uses U0401 Ё twice (see [1]) U+0401: CYRILLIC CAPITAL LETTER IO I found out that there's a glyph tha looks the same in the latin range: U+00CB LATIN CAPITAL LETTER E WITH DIAERESIS I can't read Albanian, but looks to me that this should be changed. 2 --- Kazakh po file (kk.po), uses the following: U2013 - EN dash (37 times) U2014 - EM dash (once) the same way as the normal - (002D HYPHEN-MINUS) 3 --- Vietnamese uses the bullet (U+2022 BULLET) to list items that in the original msgid are marked with - 4 --- Macedonian and Ukrainian use the following to quote text: U201c - Left double quotation mark U201e - double low-9 quotation mark Shall we fix those, ask the translators? I think that characters like the bullet should't be used for the installer: IMO those are meant to be used in applications like Openoffice or similar. Regards, Davide [1] http://d-i.alioth.debian.org/spellcheck/level1/latest/nozip/sq_all.txt [2] http://d-i.alioth.debian.org/spellcheck/level1/latest/nozip/kk_codes.txt signature.asc Description: Digital signature
Re: [g-i] freefont vs. dejavu
On Thu, Mar 16, 2006 at 01:23:47AM +0100, Frans Pop wrote: On Thursday 16 March 2006 01:00, Davide Viti wrote: This is (was) the idea (see [1]): Looks good to me. Could you add the glyph ranges needed in the different fonts (except freefont)? I'm collecting suche ranges; I'll post them as soon as I make up my mind a little bit more. I'm considering to strip font files specifying the ranges to include rather than the ranges that we want to exclude, what do you think about this? Please also consider if some common codepoints, like for numbers, general punctuation and maybe accents should be included from the different fonts in addition to their base ranges. This could help to keep visual consistency within translations (only works if we implement changing the default font on selection of a different language of course). In some cases, like for CJK, we may want to keep latin characters too. For CJK this has been discussed before (and AFAIK is already implemented in the udeb). total agreement on both points. I wonder if we should consider switching to freefont whenever possible. This would bring alot of advantages (both in size and simplicity). That's basically what we've been doing so far, isn't it? yes. we switched to Dejavu for displaying Cyrillic and just found out 10 Cyrillic glyphs needed by the po files are missing :( Still, I think that if a different font really looks better to native speakers than freefont, we should in principle [1] use the other font [2]. [1] An exception could be if that font is much larger than alternative fonts. [2] Within reason of course. We can't let a translator for, say, Dutch decide he wants a different font. It's only acceptable for scripts that have separate ranges in the UTF table. Just to have an idea here's the situation. CJK have not been considered since it's a separate case and no stripping is needed (I know Kenshi is taking care of those) DejaVuSans-Bold.ttf: 918 glyphs [125188 bytes] DejaVuSans-Oblique.ttf: 918 glyphs [126872 bytes] DejaVuSans.ttf: 918 glyphs [141668 bytes] nazlib.ttf: 303 glyphs [68448 bytes] nazli.ttf: 303 glyphs [55941 bytes] FreeSans.ttf: 2220 glyphs [292904 bytes] FreeSansBold.ttf: 1572 glyphs [131168 bytes] FreeSansOblique.ttf: 1333 glyphs [111580 bytes] KhmerOS.ttf: 539 glyphs [265988 bytes] TSCu_Paranar.ttf: 224 glyphs [63048 bytes] Font files: 10 Total size: 1382805 bytes Unique Glyphs: 2761 Bytes per glyph: 500 removing dejavu (or freefont if the new dejavu can cover everything), would simplify and reduce overhead (basic latin glyphs, as you mentioned above, are not stripped to preserve consistency: i.e. at least u0:u7f = 128 glyphs are included in each and every ttf file including bold/oblique) ciao, Davide signature.asc Description: Digital signature
Re: [g-i] freefont vs. dejavu
On Thursday 16 March 2006 01:54, Davide Viti wrote: Since we're on the subject, I've found some weirdness in the translations: Nice work :-) 1 --- I can't read Albanian, but looks to me that this should be changed. Yes, this is a translation mistake. 2 --- U2013 - EN dash (37 times) U2014 - EM dash (once) I think we should first check how this comes out in d-i (newt). If it is displayed correctly (as -), I have no real problem with this. If it is not displayed correctly, it should definitely be changed (and maybe yet another type of check added to spellcheck scripts? :-) 3 --- Vietnamese uses the bullet (U+2022 BULLET) to list items that in the original msgid are marked with - Same as 2, though probably more likely to fail. 4 --- U201c - Left double quotation mark U201e - double low-9 quotation mark Same as 2. pgp4eDhHRonWS.pgp Description: PGP signature
Re: [g-i] freefont vs. dejavu
On Thursday 16 March 2006 02:11, Davide Viti wrote: I'm considering to strip font files specifying the ranges to include rather than the ranges that we want to exclude, what do you think about this? For freefont exclusion (ranges we know are covered by other fonts and ranges for which we know there are no translations (yet)) is maybe more logical (probably easier to document too); for other fonts inclusion definitely. I'll leave it to you though to decide what works best. removing dejavu (or freefont if the new dejavu can cover everything), would simplify and reduce overhead (basic latin glyphs, as you mentioned above, are not stripped to preserve consistency: i.e. at least u0:u7f = 128 glyphs are included in each and every ttf file including bold/oblique) Hmm. Not sure we need to include basic latin for e.g. arabic and indic fonts. Guess that depends on how different the included characters are from freefont (both style and size should be considered). I've no problem with leaving them in initially though. Next step is probably creating an iso that supports the font switching from localechooser to see how that behaves. I can probably do that quite quickly when you're ready with font tarballs. pgptfHJBorO6h.pgp Description: PGP signature
Re: [g-i] freefont vs. dejavu
[Jutta, could you please help us? - see question below] On 3/16/06, Davide Viti [EMAIL PROTECTED] wrote: Albanian (which is listed among the latin languages) po file (sq.po) uses U0401 Ё twice (see [1]) U+0401: CYRILLIC CAPITAL LETTER IO I found out that there's a glyph tha looks the same in the latin range: U+00CB LATIN CAPITAL LETTER E WITH DIAERESIS I can't read Albanian, but looks to me that this should be changed. Probably you are right 4 --- Macedonian and Ukrainian use the following to quote text: U201c - Left double quotation mark U201e - double low-9 quotation mark Jutta Wrage has recently made a sumarisation of the correct quotation marks needed for each language. Jutta, could you enlighten us? Shall we fix those, ask the translators? I think that characters like the bullet should't be used for the installer: IMO those are meant to be used in applications like Openoffice or similar. I tend to agree, but maybe we might want to hear Clytie's oppinion. -- Regards, EddyP = Imagination is more important than knowledge A.Einstein
Re: [g-i] freefont vs. dejavu
Quoting Frans Pop ([EMAIL PROTECTED]): On Thursday 16 March 2006 01:54, Davide Viti wrote: Since we're on the subject, I've found some weirdness in the translations: Nice work :-) Nothing to add to Frans comments which I fully agree with. signature.asc Description: Digital signature
Re: [g-i] freefont vs. dejavu
FYI On 3/7/06, Eddy Petrişor [EMAIL PROTECTED] wrote: On 3/7/06, Davide Viti [EMAIL PROTECTED] wrote: Tonight I've run some tests using dejavu fonts as discussed during the meeting. I created an iso [1] using an unstripped version of dejavu and without freefont; in the same directory ([2]) you can find a screenshot for each language. It turns out that dejavu does not contain Vietnames and Hebrew glyphs, meaning that we can't use it to completely substitute freefont. The latest version of freefont include all the Vietnamese glyphs listed below and the Hebrew glyphs. I was suspecting this (it says so on the dejavu home page;-), but I am glad you did the tests so we can have a clear picture. Here is the unicode coverage overview (filename: unicove.txt) of ttf-dejavu, as it is in version 2.3 (there is also a detailed map with individual unicode points in status.txt): This is the Unicode coverage file for DejaVu fonts ($Id: unicover.txt 573 2006-02-14 21:55:17Z src $) Control and similar characters are discounted from totals. Sans Serif Sans Mono U+ Basic Latin 100% (95/95) 100% (95/95) 100% (95/95) U+0080 Latin-1 Supplement 100% (96/96) 100% (96/96) 100% (96/96) U+0100 Latin Extended-A 100% (128/128) 100% (128/128) 100% (128/128) U+0180 Latin Extended-B 64% (126/194) 58% (113/194) 58% (113/194) U+0250 IPA Extensions 100% (96/96) 100% (96/96) 100% (96/96) U+02b0 Spacing Modifier Letters 51% (41/80) 51% (41/80)51% (41/80) U+0300 Combining Diacritical Marks 58% (66/112) 58% (66/112) 58% (66/112) U+0370 Greek and Coptic 88% (110/124) 88% (110/124) 88% (110/124) U+0400 Cyrillic 77% (193/248) 48% (120/248) 46% (116/248) U+0500 Cyrillic Supplement (0/16) (0/16) (0/16) U+0530 Armenian 100% (86/86) (0/86) (0/86) U+0590 Hebrew(0/86) (0/86) (0/86) U+0600 Arabic(0/235) (0/235)(0/235) U+0700 Syriac(0/77) (0/77) (0/77) U+0750 Arabic Supplement (0/30) (0/30) (0/30) U+0780 Thaana(0/50) (0/50) (0/50) U+0900 Devanagari(0/107) (0/107)(0/107) U+0980 Bengali (0/91) (0/91) (0/91) U+0a00 Gurmukhi (0/77) (0/77) (0/77) U+0a80 Gujarati (0/83) (0/83) (0/83) U+0b00 Oriya (0/81) (0/81) (0/81) U+0b80 Tamil (0/71) (0/71) (0/71) U+0c00 Telugu(0/80) (0/80) (0/80) U+0c80 Kannada (0/82) (0/82) (0/82) U+0d00 Malayalam (0/78) (0/78) (0/78) U+0d80 Sinhala (0/80) (0/80) (0/80) U+0e00 Thai (0/87) (0/87) (0/87) U+0e80 Lao (0/65) (0/65) (0/65) U+0f00 Tibetan (0/195) (0/195)(0/195) U+1000 Myanmar (0/78) (0/78) (0/78) U+10a0 Georgian (0/83) (0/83) (0/83) U+1100 Hangul Jamo (0/240) (0/240)(0/240) U+1200 Ethiopic (0/356) (0/356)(0/356) U+1380 Ethiopic Supplement (0/26) (0/26) (0/26) U+13a0 Cherokee (0/85) (0/85) (0/85) U+1400 Unified Canadian Aboriginal Syllabics (0/630) (0/630)(0/630) U+1680 Ogham (0/29) (0/29) (0/29) U+16a0 Runic (0/81) (0/81) (0/81) U+1700 Tagalog (0/20) (0/20)
Re: [g-i] freefont vs. dejavu
On 3/7/06, Davide Viti [EMAIL PROTECTED] wrote: Tonight I've run some tests using dejavu fonts as discussed during the meeting. I created an iso [1] using an unstripped version of dejavu and without freefont; in the same directory ([2]) you can find a screenshot for each language. It turns out that dejavu does not contain Vietnames and Hebrew glyphs, meaning that we can't use it to completely substitute freefont. The latest version of freefont include all the Vietnamese glyphs listed below and the Hebrew glyphs. I was suspecting this (it says so on the dejavu home page;-), but I am glad you did the tests so we can have a clear picture. Once freefont will be ok, we can decide which font should display which language: freefont will probably be used for vietnamese, hebrew, hindi, punjabi, bengali. Comments are welcome. [snip] PS: Eddy, if there was not the nasty freefont bug I would never have done this test with dejavu :)) have to admit I like the way they look like I know and I am glad you like them. [1] http://d-i.alioth.debian.org/gtk-frontend/screenshots/20060307_dejavu/mini.iso [2] http://d-i.alioth.debian.org/gtk-frontend/screenshots/20060307_dejavu/ Latin based scripts should all look better. BTW, what is the size difference? Is dejavu+freefont(for Hebrew,Vietnamese and others) freefont(for all)? -- Regards, EddyP = Imagination is more important than knowledge A.Einstein
[g-i] freefont vs. dejavu
Tonight I've run some tests using dejavu fonts as discussed during the meeting. I created an iso [1] using an unstripped version of dejavu and without freefont; in the same directory ([2]) you can find a screenshot for each language. It turns out that dejavu does not contain Vietnames and Hebrew glyphs, meaning that we can't use it to completely substitute freefont. The latest version of freefont include all the Vietnamese glyphs listed below and the Hebrew glyphs. Once freefont will be ok, we can decide which font should display which language: freefont will probably be used for vietnamese, hebrew, hindi, punjabi, bengali. Comments are welcome. 01A1 vi 01AF vi 01B0 vi 05D0 he 05D1 he 05D2 he 05D3 he 05D4 he 05D5 he 05D6 he 05D7 he 05D8 he 05D9 he 05DA he 05DB he 05DC he 05DD he 05DE he 05DF he 05E0 he 05E1 he 05E2 he 05E3 he 05E4 he 05E5 he 05E6 he 05E7 he 05E8 he 05E9 he 05EA he 1EA0 vi 1EA1 vi 1EA2 vi 1EA3 vi 1EA4 vi 1EA5 vi 1EA7 vi 1EA9 vi 1EAB vi 1EAD vi 1EAF vi 1EB1 vi 1EB3 vi 1EB5 vi 1EB7 vi 1EB9 vi 1EBB vi 1EBC vi 1EBD vi 1EBF vi 1EC1 vi 1EC3 vi 1EC5 vi 1EC7 vi 1EC9 vi 1ECA vi 1ECB vi 1ECC vi 1ECD vi 1ECE vi 1ECF vi 1ED1 vi 1ED3 vi 1ED4 vi 1ED5 vi 1ED6 vi 1ED7 vi 1ED9 vi 1EDB vi 1EDD vi 1EDE vi 1EDF vi 1EE1 vi 1EE3 vi 1EE5 vi 1EE7 vi 1EE8 vi 1EE9 vi 1EEB vi 1EED vi 1EEF vi 1EF1 vi 1EF7 vi 2116 bg 2501 vi regards, Davide PS: Eddy, if there was not the nasty freefont bug I would never have done this test with dejavu :)) have to admit I like the way they look like [1] http://d-i.alioth.debian.org/gtk-frontend/screenshots/20060307_dejavu/mini.iso [2] http://d-i.alioth.debian.org/gtk-frontend/screenshots/20060307_dejavu/ signature.asc Description: Digital signature
