Re: Unicode certification - was RE: Dublin Conference/StandardDisclaimer
Arsa James Kass wrote: Any series of books which begins with the complete destruction of Earth is bound to be amusing, eh? Best regards, James Kass. Book 4 deals more with the creation of a new/alternative earth, James! In any case, as this is way off-topic, might I bring it back, via my earlier suggestion, as elaborated on by David Possin (below). It's perfectly acceptable for Unicode to confine itself to providing tables as touchpoints for those (its consortium members and others) actually making builds implementing principles set out in its publication. It would not require the whole consortium to get involved in the minutiae of what David describes below (a couple of boys in a backroom could do it) via a sort of Tucows site set up, giving Unicode-friendly ratings, or even broad compliance with MES/BMP/whatever, with no guarantee of performance, beyond what David has indicated. Sounds like a real time-saver, or is that a real-time saver?:-) mg David Possin wrote: It would be intereting and helpful to be able to find out if a product is Unicode-compliant before purchasing it. There are various test institutions out there that perform that work for other standards. I don't think it would be Unicode.org's responsibility to provide for the certification, to avoid membership issues, maybe it should create the certification requirements, though. I find myself wasting a lot of time figuring out if a third-party product or a certain version can handle Unicode and/or up to which version it is compliant to. I would like to be able to see a little Unicode logo on a box stamped with a release number, making it the manufacturer's responsibility to prove it. It works for operating system releases and other stuff, why not here as well? Dave = Dave Possin Globalization Consultant www.Welocalize.com -- Marion Gunn * E G T (Estab.1991) vox: +353-1-2839396 * [EMAIL PROTECTED] 27 Páirc an Fhéithlinn; Baile an Bhóthair; Contae Átha Cliath; Éire
Re: Unicode certification - was RE: Dublin Conference/Standard Disclaimer
Marion Gunn wrote, It would not require the whole consortium to get involved in the minutiae of what David describes below (a couple of boys in a backroom could do it) via a sort of Tucows site set up, giving Unicode-friendly ratings, or even broad compliance with MES/BMP/whatever, with no guarantee of performance, beyond what David has indicated. Sounds like a real time-saver, or is that a real-time saver?:-) It sounds like Hobson's choice. But, I'll agree that it's a time saver. We can also agree that the whole consortia needn't involve itself in this kind of minutiae. Indeed, since this kind of Unicode certification is beyond the realm of TUC, the consortia needn't involve itself at all. The couple of boys in the back room could do it, and possibly figure out a way to do it profitably. On the other hand, if a certification program could represent revenue for TUC*, revenue which could be used to further the cause, then who better to judge Unicode compliance? Best regards, James Kass. * ...such as use of the logo in the certificate notice... - Original Message - From: Marion Gunn [EMAIL PROTECTED] Cc: [EMAIL PROTECTED] Sent: Thursday, July 25, 2002 3:21 AM Subject: Re: Unicode certification - was RE: Dublin Conference/Standard Disclaimer Arsa James Kass wrote: Any series of books which begins with the complete destruction of Earth is bound to be amusing, eh? Best regards, James Kass. Book 4 deals more with the creation of a new/alternative earth, James! In any case, as this is way off-topic, might I bring it back, via my earlier suggestion, as elaborated on by David Possin (below). It's perfectly acceptable for Unicode to confine itself to providing tables as touchpoints for those (its consortium members and others) actually making builds implementing principles set out in its publication. It would not require the whole consortium to get involved in the minutiae of what David describes below (a couple of boys in a backroom could do it) via a sort of Tucows site set up, giving Unicode-friendly ratings, or even broad compliance with MES/BMP/whatever, with no guarantee of performance, beyond what David has indicated. Sounds like a real time-saver, or is that a real-time saver?:-) mg David Possin wrote: It would be intereting and helpful to be able to find out if a product is Unicode-compliant before purchasing it. There are various test institutions out there that perform that work for other standards. I don't think it would be Unicode.org's responsibility to provide for the certification, to avoid membership issues, maybe it should create the certification requirements, though. I find myself wasting a lot of time figuring out if a third-party product or a certain version can handle Unicode and/or up to which version it is compliant to. I would like to be able to see a little Unicode logo on a box stamped with a release number, making it the manufacturer's responsibility to prove it. It works for operating system releases and other stuff, why not here as well? Dave = Dave Possin Globalization Consultant www.Welocalize.com -- Marion Gunn * E G T (Estab.1991) vox: +353-1-2839396 * [EMAIL PROTECTED] 27 Páirc an Fhéithlinn; Baile an Bhóthair; Contae Átha Cliath; Éire
Re: Unicode certification - was RE: Dublin Conference/StandardDisclaimer
Arsa James Kass : ... The couple of boys in the back room could do it, and possibly figure out a way to do it profitably... My thought exactly. But, to voice such equals soliciting business.:-) Still, a good idea, by all accounts. mg James Kass -- Marion Gunn * E G T (Estab.1991) vox: +353-1-2839396 * [EMAIL PROTECTED] 27 Páirc an Fhéithlinn; Baile an Bhóthair; Contae Átha Cliath; Éire
Re: Unicode certification - was RE: Dublin Conference/Standard Disclaimer
Thanks for the Fish, Marion! We could meet at Milliway's and establish the back room setup there. The compliance guidelines could then be called Unicode's Guide to the Galaxy. A 100% compliant system receives the rating '42'. Non-compliant systems are processed by the Vogons. Yes, right now my check list for Unicode compliance when contacting 3rd parties looks more like this, the higher the number the better: 0. Uni-what? 1. I know somebody who can spell Unicode. 2. I can spell Unicode. 3. Yeah, the specs say it works but we never tested it. 4. We tried it once, seemed to work. 5. We use Java, that's Unicode, right? 6. Yes, but we had to let the developer go who did it when we downsized the last time, so I am not sure about the details. 7. Yes, and it is running with different languages in Europe. 8. Yes, and it is running with different languages in Asia. 9. Yes, it is running with several languages at once. 10. Yes, and we have bidi and complex scripting too. That is about as far as I get, I can only dream of being able to get details like David Starner described for compliance. `This must be Thursday,' said [Dave] to himself, sinking low over his beer, `I never could get the hang of Thursdays.' --- Marion Gunn [EMAIL PROTECTED] wrote: Arsa James Kass wrote: Any series of books which begins with the complete destruction of Earth is bound to be amusing, eh? Best regards, James Kass. Book 4 deals more with the creation of a new/alternative earth, James! In any case, as this is way off-topic, might I bring it back, via my earlier suggestion, as elaborated on by David Possin (below). It's perfectly acceptable for Unicode to confine itself to providing tables as touchpoints for those (its consortium members and others) actually making builds implementing principles set out in its publication. It would not require the whole consortium to get involved in the minutiae of what David describes below (a couple of boys in a backroom could do it) via a sort of Tucows site set up, giving Unicode-friendly ratings, or even broad compliance with MES/BMP/whatever, with no guarantee of performance, beyond what David has indicated. Sounds like a real time-saver, or is that a real-time saver?:-) mg David Possin wrote: It would be intereting and helpful to be able to find out if a product is Unicode-compliant before purchasing it. There are various test institutions out there that perform that work for other standards. I don't think it would be Unicode.org's responsibility to provide for the certification, to avoid membership issues, maybe it should create the certification requirements, though. I find myself wasting a lot of time figuring out if a third-party product or a certain version can handle Unicode and/or up to which version it is compliant to. I would like to be able to see a little Unicode logo on a box stamped with a release number, making it the manufacturer's responsibility to prove it. It works for operating system releases and other stuff, why not here as well? Dave = Dave Possin Globalization Consultant www.Welocalize.com -- Marion Gunn * E G T (Estab.1991) vox: +353-1-2839396 * [EMAIL PROTECTED] 27 Páirc an Fhéithlinn; Baile an Bhóthair; Contae Átha Cliath; Éire = Dave Possin Globalization Consultant www.Welocalize.com http://groups.yahoo.com/group/locales/ __ Do You Yahoo!? Yahoo! Health - Feel better, live better http://health.yahoo.com
Re: Unicode certification - was RE: Dublin Conference:
I think there are different levels of Unicode compliance we need to look at. In over 75% of the tests I am satisfied with simple compliance, I don't even expect or assume that more complex issues have been implemented or thoroughly tested. Test 1: A stream of Unicode data gets sent into the system, flows through a sequence of components, gets stored, gets retrieved, and comes back out of the system. Is the data still the same? The following tests are concerned with the different functionality of the components, tested one at a time, then combined till full functionality testing has been achieved, as if non-Unicode data had been used. (This is assuming Unicode-enablement is the objective.) This is the level of compliance I am most interested in. The component can handle Unicode data the same as it can handle legacy encodings. The vanilla test. After that we can add the chocolate sauce, the cherry, and the sprinkles of Unicode. The special Unicode compliance tests are harder to define and to perform, I agree. But in most cases these issues haven't even been implemented yet. Dave --- Tex Texin [EMAIL PROTECTED] wrote: David, Why couldn't a checklist be established for each of the functionalities that you mention, which a product could score itself against for conformance, over a state range of supported characters? Recently, I did a search for a product, and it was difficult to know which scripts were supported and whether it had the Unicode capabilities I was concerned with. It would have been nice if there was a statement of self-compliance that indicated whether or not they supported: Character ranges- broken into reasonable subgroups: Preservation of unicode characters Combining characters: normalization forms: collations: etc. I think if there were such a checklist with suitable definitions and/or conformance requirements, vendors that had done the work to support Unicode properly would be glad to declare it in their product specs or packaging. And there are probably many product developers that think they support Unicode but in fact don't and such a checklist would help make them aware of what else they need to do. And if they misadvertised or reported incorrectly, I am sure their customers would be glad to inform them of their oversight thru their support lines or by announcement to the appropriate user group lists. Sure there will be some grey areas based on particular product functionality, but it would still be a far better situation then we have today... tex = Dave Possin Globalization Consultant www.Welocalize.com http://groups.yahoo.com/group/locales/ __ Do You Yahoo!? Yahoo! Health - Feel better, live better http://health.yahoo.com
Re: Unicode certification - was RE: Dublin Conference:
At 08:07 AM 7/25/2002 -0700, David Possin wrote: After that we can add the chocolate sauce, the cherry, and the sprinkles of Unicode. The special Unicode compliance tests are harder to define and to perform, I agree. But in most cases these issues haven't even been implemented yet. But isn't the reason someone would want to quantify compliance is precisely to find out what is implemented and what is not? Barry Caplan www.i18n.com
Re: Unicode certification - was RE: Dublin Conference:
Correct, that is what I was trying to say when I added the goodies. Sorry it didn't come across that way. Let me go a bit deeper in what I mean by compliance levels. 1. Unicode support is implemented and allows for same functionality as with any other legacy encoding system. Detail: up to which Unicode release this support is implemented. 2. Additional Unicode support is implemented and and offers the following list of features beyond legacy encodings: [list of features], for example ICU is fully implemented. 3. Full Unicode support is implemented - all characters can be processed, all glyphs are available, and rendering complies to all rules for each writing system. (I hope I used the correct terms here.) I am most interested in step 1 most of the time, as it is the biggest hurdle when I perform an assessment. When or if steps 2 3 are an issue, the compliance testing gets complex on the one side, but on the other side the teams implementing them are much more knowlegable and can offer better compliance details. Tbh, I am not sure where to draw a line between 2 3, I think it is a gray zone, rarely found today. Dave --- Barry Caplan [EMAIL PROTECTED] wrote: At 08:07 AM 7/25/2002 -0700, David Possin wrote: After that we can add the chocolate sauce, the cherry, and the sprinkles of Unicode. The special Unicode compliance tests are harder to define and to perform, I agree. But in most cases these issues haven't even been implemented yet. But isn't the reason someone would want to quantify compliance is precisely to find out what is implemented and what is not? Barry Caplan www.i18n.com = Dave Possin Globalization Consultant www.Welocalize.com http://groups.yahoo.com/group/locales/ __ Do You Yahoo!? Yahoo! Health - Feel better, live better http://health.yahoo.com
Re: Unicode certification - was RE: Dublin Conference/Standard Disclaimer
On 07/25/2002 09:30:18 AM David Possin wrote: Thanks for the Fish, Marion! We could meet at Milliway's and establish the back room setup there. The compliance guidelines could then be called Unicode's Guide to the Galaxy. A 100% compliant system receives the rating '42'. Non-compliant systems are processed by the Vogons. I suspect you're going to need some fairy cake to make this happen. - Peter --- Peter Constable Non-Roman Script Initiative, SIL International 7500 W. Camp Wisdom Rd., Dallas, TX 75236, USA Tel: +1 972 708 7485 E-mail: [EMAIL PROTECTED]
Unicode certification - was RE: Dublin Conference:
It would be intereting and helpful to be able to find out if a product is Unicode-compliant before purchasing it. There are various test institutions out there that perform that work for other standards. I don't think it would be Unicode.org's responsibility to provide for the certification, to avoid membership issues, maybe it should create the certification requirements, though. I find myself wasting a lot of time figuring out if a third-party product or a certain version can handle Unicode and/or up to which version it is compliant to. I would like to be able to see a little Unicode logo on a box stamped with a release number, making it the manufacturer's responsibility to prove it. It works for operating system releases and other stuff, why not here as well? Dave --- David J. Perry [EMAIL PROTECTED] wrote: I think that it's very wise of the Unicode Consortium not to certify or officially promote any particular implementation. After all, some programmers are more skilled than others, and some implementations may not be of the quality one might wish. Or what if a member company produced a decent implementation, but the competing product by a small, non-member company was better? This could be a real mess. The Unicode web site does have a list of Unicode-enabled products (I'm not sure how complete it is), which is helpful and appropriate--but I wouldn't want to see anything beyond that. David = Dave Possin Globalization Consultant www.Welocalize.com http://groups.yahoo.com/group/locales/ __ Do You Yahoo!? Yahoo! Health - Feel better, live better http://health.yahoo.com
Re: Unicode certification - was RE: Dublin Conference:
At 11:24 AM 7/24/02 -0700, David Possin wrote: It would be intereting and helpful to be able to find out if a product is Unicode-compliant before purchasing it. The problem is too broad to be neatly solved. It's not like compliance to the Ada standard, where you can just write a bunch of test code for all compilers. You'd have to adapt the tests for each program including writing code customized for each interpreter or compiler. And after you've done this, you know that it can round-trip arbitrary Unicode and that it treats the characters as Unicode characters and not, say, Latin-1 or SJIS. You don't know whether it can handle combining characters or not, or whether or not it can handle any particular characters beyond just not messing with them. A program could pass with debilitating flaws for any real Unicode use, and still be Unicode complaint. Seems like a lot of work for little gain.
Re: Unicode certification - was RE: Dublin Conference:
David, Why couldn't a checklist be established for each of the functionalities that you mention, which a product could score itself against for conformance, over a state range of supported characters? Recently, I did a search for a product, and it was difficult to know which scripts were supported and whether it had the Unicode capabilities I was concerned with. It would have been nice if there was a statement of self-compliance that indicated whether or not they supported: Character ranges- broken into reasonable subgroups: Preservation of unicode characters Combining characters: normalization forms: collations: etc. I think if there were such a checklist with suitable definitions and/or conformance requirements, vendors that had done the work to support Unicode properly would be glad to declare it in their product specs or packaging. And there are probably many product developers that think they support Unicode but in fact don't and such a checklist would help make them aware of what else they need to do. And if they misadvertised or reported incorrectly, I am sure their customers would be glad to inform them of their oversight thru their support lines or by announcement to the appropriate user group lists. Sure there will be some grey areas based on particular product functionality, but it would still be a far better situation then we have today... tex David Starner wrote: At 11:24 AM 7/24/02 -0700, David Possin wrote: It would be intereting and helpful to be able to find out if a product is Unicode-compliant before purchasing it. The problem is too broad to be neatly solved. It's not like compliance to the Ada standard, where you can just write a bunch of test code for all compilers. You'd have to adapt the tests for each program including writing code customized for each interpreter or compiler. And after you've done this, you know that it can round-trip arbitrary Unicode and that it treats the characters as Unicode characters and not, say, Latin-1 or SJIS. You don't know whether it can handle combining characters or not, or whether or not it can handle any particular characters beyond just not messing with them. A program could pass with debilitating flaws for any real Unicode use, and still be Unicode complaint. Seems like a lot of work for little gain. -- - Tex Texin cell: +1 781 789 1898 mailto:[EMAIL PROTECTED] Xen Master http://www.i18nGuy.com XenCrafthttp://www.XenCraft.com Making e-Business Work Around the World -
