From W3C Wiki

GEO Feedback Summary

Date Page / Comment
This tool didn't tell me more about problems I meet using utf8 encoding with php/mysql on apache server.I'm going to search more specific item about my problem.Thank you anyway
i was getting mad with some strange results from SQL server to an ASP Page..the prblem was with the (in)famous acent letters (my site is in ITalian...) .......... i wasn\'t able the content-type because it was set to ISO 8591 and was locked by a template... .......... The <%Response.charset=\"utf-8\"%> really solved my problems
I just need to know how to make a proper character encoding in the <head> with the <meta> tag and I just can\'t find it here
Just make more clear, that ISO-8859-1 is the right (or the best) choice for portugese pages. Some ppl may not find it easly on the text that this enconding is the right one to portuguese language, its actually refered as 'Latin' wich may sound wierd for some. :)
Was looking for a way to resolve issues with our website and the different written variations of some of the langauges (Simplified Chinese & Tranditional Chinese to be specific).
I am having a problem when a special character like paranthese) is at the end of an english-arabic statment with DIR=RTL. Any comment in this direction?
..........javascript validations.......... ..........
The instructions for changing language in different browsers are out of date and incorrect for FireFox and Internet Explorer and Opera, others may be as well.
Target audience of current project is strictly local, so confusion over different formats globally doesn\'t apply
some links are not working and the article Character Set Considered Harmful is too outdated, is good as a basic reading, but there are many resources in your own date base better than that.
more examples are usefull for beginers
(The usefulness scale doesn't apply. I didn't want to fill it, so it's quite poor that I have to in order to place a comment :-)..........Anyway. The python solution is not quite right. The print statement adds a newline by itself, so the correct solution for a CGI script, using print would be:..........print \"Content-Type: text/html; charset=utf-8\\n\"..........It's pretty uncommon to write web applications in python that way, though.
What made this article useful was that it was turse and it linked to the type of page I was looking for. Namely: thanks.
I think somewhere here: .......... .......... or here: .......... .......... you should say that ruby is not currently supported in most browers and you have to install an extension to firefox for it to work.
Many thanks for this article...........A recommendation of a good text editor capable of handling c.q. converting between different UTF formats would be useful...........Cheers
I like this presentation but is there a link to font family and a discussion about why this character does not appear ?
3/5 is an average score, not a low one.
it's a nice site with good looking, helpfull tips are provided
I just needed to know how to specify the character set in a page of HTML. It's an attribute of the <meta> tag. .......... Where is that in here?
Too complex.
I would like to know the differents content type existing
You says that SIL Ethnologue "is the Registration Authority for ISO 639-3, that is, the folks who will maintain the code list in the future". .......... But SIL Ethnologue not list all the languages (one example: they don't code the Charrúa, an ancient language from Uruguay and a part of Argentina and Paraguay, or the Chaná, from Argentina). .......... Do they arrange that problems in the future? .......... Thank you. .......... Claudio Segovia .......... Buenos Aires .......... Argentina
I still am not clear on the difference between unicode and iso character sets.
I am currently using language negotiation on my periodic table, matching two dozen languages and differentiating Chinese Simp/Trad and American/British English based on locale. You're welcome to use the site as an example:
Too much hedging. I'd like a clear, "this is how to do it for vanilla cases". Then a disclaimer and alternatives below.
Any chance you guys could write ANY of this in English? .......... Not everybody is quite as superior as you when it comes to...Well...EVERYTHING.
The page has a potentially risky typo in the description of what happens when the UTF-8 encoded form of BOM is interpreteted as if it were ISO 8859-1 encoded. The end of the first paragraph has it right: .......... - - such as . .......... But in the fourth paragraph under the heading "answer", the guillemet has changed: .......... - - as extraneous characters 﫿.) - - .......... I actually got misled: I copied and pasted 﫿 and used it for a test, to get the octets of UTF-8 encoded BOM, then wondered what is going on...
The instructions for Firefox are incorrect. To change Languages Settings in Firefox, use: .......... Tools -> Options -> Advanced -> General -> Edit Languages
The entire topic is very informative. But I would like to point out a small mistake in one of the examples. .......... ie, in the topic of 'Composite messages that are incomplete strings'(Internationalization Issues), there is a mistake has found out in the example list, as follows: In the first example block, the word 'desiredEntries' has mistakenly used as 'requiredEntries' in the 4th and 5th example blocks. I'll be thankful to you if you check this suggestion.
Having read the article again, I'd like to revise my feedback........... The text is informative and understandable. The "On this page" links at the top would be good in a normal (not inline) list, as they appear similar to breadcrumbs. It would also be good to have a sublist of the subtag headers. Use of Title Caps on these links (and on headers) would aid readability. The use of acronym tags would be good for terms such as RFC, IANA, IETF, and BCP (good that there\'s a definition for BCP though). It would help if the first table of examples had what is being demonstrated (script, variant etc) in a third column, and included an example of each type of subtag. Programmers will probably cut to the specs, but as it's mentioned in "Using Language Tags" #2 it would be good to have the regex for determining the matching of language tags (if feasible) ........... The main problem with the page is the use of pictures for the subtag .insidenote examples. Because .insidenote blocks are given a 24% width, the widest image (294px) requires a browser window width of 1220px to view without scrolling. I believe this information (coloured boxes around the relevant subtag part) would be better applied as styled spans directly to the text examples (eg using outline), as the images are a straight duplication of these examples. Alternatively if you want a magazine-style callout, making the examples inline-styled text in a .insidenote div would allow the site more width flexibility. .......... In "Matching language tags" it would be good to mention what browsers support the pseuo-attribute lang (ie to say IE doesn't support it before IE7, which is a pretty serious lack of support). It would also be good to have a browser support statement regarding the new language tags, even if it's just to say browsers should ignore anything they don't know. Finally it would be good to embolden or otherwise highlight this paragraph "Note that the HTML specification still recommends the use of RFC 1766 for identifying language but you should use RFC 4646 despite what the HTML specification currently says.\", and provide a direct link to the relevant section of the spec. .......... hope that was more useful .......... m_(,)_m .......... peace - oli
Your main reference for language codes is an 80KB plain text file!? How this is remotely usable to someone looking for a specific language code? .......... +1 for having \"standards developer\" I guess .......... peace - oli
This was a really helpful article. I'd like to see more about the current state-of-play regarding multiple-language content, lang="mul" and other problem cases. I'd like to see some notes or pointers to further reading concerning the state of software support for, for example Content-Language headers. This would help me devise a testing strategy to assess the impact of design decisions. (Omit xml:lang at root element? Use xml:lang="mul"? Content-Language="mul"? Omit these elements?) It's understandable that a standards body does not want to get dragged into discussing individual applications, but help with assessing deplyoment issues is really welcomed.
not enough data information was a bit unuseful
I'd add a sample with a brief explaination of the additional q preference parameter, as in: it-it,it; q=0.8,en-us; q=0.5,en; q=0.3
Good article. Could benefit from expansion, links to \"further reading\". Little discussion about the detailed breakdown of costs. Pointers to further information about topics such as parallel text, minority languages and implementation techniques could usefully be added.
In the script section, the example in the main body of text says az-Latn (because Azeri can also be written with the Arabic script) but the note on the side shows az-Cyrl, which is a valid tag because Azeri can also be written using the Cyrillic script but is inconsistent with the main body\'s example. . . . . . . . . . . . Either the body of the text should mention that the Cyrl script tag is also valid for Azeri or the example on the right should show the same example as the main text.
Maybe a note about CDATA-Sections would also match the topic, since they can be used for omitting some of the common escapings. . . . . . . . . . . I'm pretty unsure if this affect the control characters at all.
Very nicely laid-out page. Could do with a little more technical information.
The reference at the bottom: \"RFC 3066: Tags for the Identification of Languages\" cannot be reached or is expired.
Here's an addition to . . . . . . . . . . CFML. Use the cfheader-Tag . . . . . . . . . . <cfheader name=\"Content-Type\" value=\"text/html; charset=utf-8\" > . . . . . . . . . . Cheers, ym
All I want is a list of \"special\" characters that can be entered into xml. For example, how do I get a trademark symbol to appear? I tried &#8482 and ™ but they don\'t work. Isn\'t there a list somewhere?
[DRC] I think he has xml and html confused, and should be looking at entities rather than character sets
Did not give a real solution. But that is probably not a problem of the page itself, the solution does not exist. Maybe a <date> tag is really what is needed in the standard.
[DRC] On re-reading this page, I have noticed that the link to the Java/JSP code example ( is now broken.
Oh thank you for this enlightening article! I have looked for an article like this for ages!

Hi Russ. Add your feedback here.

I suggest we do it chronologically, so the most recent is nearest the top. That will make it easier to edit current stuff.

Cheers, RI