1、,C12: “Come in Rangoon!”: Global Accessibility and Deployment,Tex Texin, General Product Manager Progress Software,“Come in Rangoon!”,“Radio Progress Calling”,“Come in Rangoon!”Even the translators did not know what it meantPossible French replacement: “Le 22 Asnires” a famous (we hope) French joke.
2、Possible Spanish replacement: “E-business? Que se ponga.” Also based on a joke from Spain. (Jose Miguel Gilas line.)But the line was not known in Latin America, so: “Adelante Timbuktu!”,I shoulda known betta! Idioms and humor do not translate.,Lessons Learned,Rangoon, Burma no longer exists. It is n
3、ow Yangon, Myanmar Avoid humor. Avoid idioms. Speak simply and directly. Test with native speakers. Test with different dialects. France vs. Fr. Canada Spain vs. Latin America,What else is there to global accessibility?,Global Accessibility and Deployment Agenda,Importance of localization Definition
4、 of terms Regional business requirements User interface requirements Character processing on the web Internet, WebSpeed, XML Translation strategies Summary,Isnt the web mostly English?,From: http:/ US,Source: ICMM V7.0, IDC 2001,36% US,Internet Commerce (B2B + B2C) ($Billions),US 36% (-13% from 2000
5、),Japan 12%,Asia Pacific 12% (+6% from 2000),West Europe34% (+13% from 2000),Rest Of World 6%,$5,026 B,Source: ICMM V7.0, IDC 2001,Users by Region in 2005 (Pie chart) Commerce By Region in 2005 (Text),People Prefer Local Language . % Preferring Native Language Sites,Source: IDC Project Atlas, 1999,S
6、ource: IDC Project Atlas, 1999,But Commerce Crosses Borders % of Home Web Spending In-Country, And Expect The Unexpected US Respondents Using Native Language,Source: IDC Project Atlas, 1999,Spanish,Chinese,Other*,English 72%,English 72%,English 52%,* Brazilian, German, French, Korean, etc.,Global Ac
7、cessibility and Deployment Agenda,Importance of localization Definition of terms Regional business requirements User interface requirements Character processing on the web Internet, WebSpeed, XML Translation strategies Summary,Definition of terms,Internationalization To design and develop an applica
8、tion: without built-in cultural assumptions that is efficient to localize Localization To tailor an application to meet the needs of a particular region, market, or culture Translation To render an application into another language,I18n,L10n,Regional business requirements,Legal requirements Privacy
9、laws Import/Export laws Advertising Language Accounting requirements Taxes, reporting, currency Support Lemon laws,Regional business requirements,Hardware and infrastructure varies Cost of phone or net services Availability and speed e.g. landline vs cellular Market leadership Japans NTT DoCoMo i-mo
10、de vs. WAP Display devices use different form factors Preference for terminal, GUI, browser or phone Computer literacy,Global Accessibility and Deployment Agenda,Importance of localization Definition of terms Regional business requirements User interface requirements Character processing on the web
11、Internet, WebSpeed, XML Translation strategies Summary,Lets start our application design with an icon for our home page,The place to return to The place to begin from Place of safety Baseballs home base Click your heels 3 timesBut, these associations are not universal,Home,Lets start our application
12、 design with an icon for our home page,French: dmarrage (start) Ger. Startseite (start page) Japan: Ho-mu pe-ji In many regions, homes do not look like this.So this icon is a poor choice, but due to wide exposure may now be recognized although still not understood.,Home,User interface design Choosin
13、g graphics and text,Avoid puns and word play (ini2reg) Avoid metaphors (home, desktop, toolkit) Avoid body parts and positions (eyes, hands, feet) Avoid sports references (homerun) Avoid animal symbols (owl, dog, cow,) Avoid religious symbols (stars, crosses) Be aware that reading directions can be
14、left to right or right to left.,What does this say?,The detergent cleans dirty shirts.,Unless you are a right-to-left reader from the Middle-East.,User interface design Avoid putting text in images,Graphics expensive to update Font difficult to create Background difficult to replace New text may int
15、erfere with imageSolution: Use Tooltips,User interface design Use generalization and abstraction,(Icons courtesy of The Icon Book, W. Horton),User interface design Formatted or structured data,Date, Time, Year, Calendar, Workweek Number, Precision, Size Rounding, Currency Title, Name Order, Address
16、Page Size, Measurement units,12$34Esc,User interface design Color and sound,RedDanger in the U.S.Marriage in China,WhitePurity in U.S.Funeral color in Asia,Error Beeps: A source of great embarrassment.,User interface and reports Plan for translation,Translation can be larger than original text Engli
17、sh: Speed Limit German: Geschwindigkeits-Begrenzung Translators need context and flexibility Avoid building sentences from fragments Avoid dynamic widget creation Word order may change,Dont do this: “There are ” + i + “tables in” + j “files.” Do this: Substitute(“There are &1 tables in &2 files.”, i
18、, j).Substitute(“Dans les fichiers &2, il y avait &1 tables”, i, j).,Global Accessibility and Deployment Agenda,Importance of localization Definition of terms Regional business requirements User interface requirements Character processing on the web Internet, WebSpeed, XML Translation strategies Sum
19、mary,Character handling Character sets,A collection of symbols Letters, Digits Punctuation Arithmetic Special symbols e.g. Line drawing Control codes,Character handling Code pages,Each symbol assigned a numeric value Vary by language, device, operating system = 196 in ISO8859-1 (Unix) = 142 in IBM 8
20、50 (Dos),ISO Latin-1 (ISO 8859-1) Code Page,32,96,128,160,192,224,64,Russian Windows Code Page CP1251,32,96,128,160,192,224,64,Japanese, Chinese, Korean,Languages with 255 characters Use Double-byte Character Sets (DBCS) Mixed size characters: 1 or 2 bytes ASCII compatibleHow long is a DBCS string?,
21、Text: # Chars: # Bytes:,S 1 1,S 2 2,S 3 3,L 4 4,T 4 5,L 5 6,T 5 7,S 8 11,L 6 8,T 6 9,S 7 10,Database code page,Client,Deploying across platforms/devices Real World Example: Russia,Windows 1251,DOS ibm866,Printer KOi8-R,Database Server,Dos Files,R-code ?,?,Unix ISO 8859-5,Database labeled,Client,Prog
22、ress Solution Every blue line is a conversion,-CPINTERNAL 1251,-CPSTREAM IBM866,-CPPRINT KOi8-R,Database Server,Dos Files,R-code Labeled,ISO 8859-5,-CPINTERNAL ISO 8859-5,Unicode Character Set A Worldwide, Multilingual Code Page,Example Unicode Characters,Progress and Unicode,UTF-8 is multi-byte for
23、m of Unicode e.g. is 1 character, 2 bytes, 1 column Programming is same as Double-byte Modify LENGTH, SUBSTRING, OVERLAY Specify RAW, CHARACTER, COLUMN Use -CHECKDBE to find changes,Unicode Configurations,UTF-8 AppServer,UTF-8 Progress Database Server,UTF-8 Progress Database,4GL client with any othe
24、r code page,4GL client with any code page,UTF-8 Batch client,Java or ActiveX Unicode client,SQL-92 UTF-8 Database,SQL-92 UTF-8 client,Global Accessibility and Deployment Agenda,Importance of localization Definition of terms Regional business requirements User interface requirements Character process
25、ing on the web Internet, WebSpeed, XML Translation strategies Summary,Processing of a URL,Give browser a URL http requests the page request routed to web server server processes request and serves a page http back to browser browser receives, parses and displays page,http:/,html,web server,html page
26、s to serve,http,url,But users, servers and pages can be anywhere and in any language,http:/www.progress.no,html,html,http:/www.progress.jp,Japanese,Norwegian,Windows-1252,ISO 8859-1,IBM 850,SHIFT-JIS,EUCJIS,UTF-8,What will users see?,code page GB2312 Sprechen sie Deutsch?,code page EBCDIC Parlez-vou
27、s franais?,Lets look at an example web page,The HTML,The same web page viewed in Japan,The same web page viewed in Japan,The browser has no information about the code page of the web page. It uses a default value which in this case, is very wrong and even confuses the markup (see beaut).,Solutions,T
28、ell user “set browser code page” User picks url based on code page Entity namesfor Euro Netscape 4.72 doesnt recognize File size expands easier to automate than entity names,Example with entity names and NCRs, Bienvenue au cours WebSpeed Nous esp) not Dollars ($),Or specify the charset in the HTML!,
29、Set character set in http protocol Set http charset using the HTML element META Use the IANA name for the character set,This is the preferred approach!,HTTP: ACCEPT-CHARSET ACCEPT-LANGUAGE,Internet Explorer Menu: Tools/Internet Options,HTTP: ACCEPT-CHARSET ACCEPT-LANGUAGE,Netscape Menu: Edit / Prefe
30、rences,Declaring language and locale LANG, SPAN, DIV,Example 1Bienvenue au cours WebSpeedWelcome to the WebSpeed class ,Processing of a URL,Give browser a URL Returned page offers language, locale choices Choosing determines the next page returned Declared Charset in HTML allows browser to perform c
31、onversions and correct display Accept-charset, Accept-language can be used to aid selection,http:/,html,web server,html pagesto serve,http,url,html,French, cp1252,Russian, cp1250,ISO 8859-2,Tips for Web Internationalization,User selects region & language on initial web page Dont assume language = re
32、gion Dont use country flag for language Users expect local URLs xxx.jp, , xxx.uk, xxx.de,Examples Progress local URLs,progress.co.at Austria progress.be Belgium progress.cz Czech Republic progress.de Rep. of Germany progress.dk Denmark .ec Ecuador progress.no Norway .ru Russian Fed. progress.co.cr C
33、osta Rica .gt Guatemala .mx Mexico .pk Pakistan .ro Romania progress.se Sweden progress.sk Slovakia progress.ph Philippines,Examples Progress(-)software local URLs,progress-.au Australia .au Australia progress-.ar Argentina progress-software.be Belgium progress-.br Brazil progress-software.fr France
34、 progress-.hk Hong Kong progress-software.co.jp Japan progress-software.nl Netherlands progress-software.se Sweden progress-software.ch Switzerland progress-software.co.uk United Kingdom progresssoftware.co.uk United Kingdom progress-software.es Spain progress-software.pl Poland progress-software.fi
35、 Finland,XML,XML declares the encoding in the opening declarationYou can use other encodings, but Unicode is the only one with guaranteed support XML:lang equivalent to lang in HTML What colour is it? What color is it?,Apply HTML charset via WebSpeed,Modify web-disp.p to emit the character set state
36、ment in the header of every html page Use the IANA name for the character set, not the Progress character set name Routine for converting session:cpinternal is simple.,Function Get-Mime-CodePage (available on the CD),/* Function to return the IANA name for a code page,given the Progress code page na
37、me.Example usage:DISPLAY Get-Mime-CodePage(session:cpinternal). */FUNCTION Get-MIME-CodePage RETURNS CHARACTER(INPUT progress-CodePage as CHARACTER). DEFINE VARIABLE PRO-cplist AS CHAR INIT“1250, 1251, UTF-8”. DEFINE VARIABLE MIME-cplist AS CHAR INIT“windows-1250, windows-1251, UTF-8”. DEFINE VARIAB
38、LE I AS INTEGER. I = LOOKUP(progress-CodePage,PRO-cplist). If I = 0 THEN RETURN “”. ELSE RETURN ENTRY(I,MIME-cplist). END FUNCTION.,Translation strategies,User chooses language, localeHTML labeled with IANA code pageUnicode databaseTranslations in R-code or database,Progress Service Capabilities Con
39、sulting packages for I18N, L10N,Internationalization Identify resources Provide planning and logistics Leverage Progress expertise Leverage I18N, L10N expertiseTraining Internationalization + Translation Manager For info: David McBain ,Getting You Started ,Internationalization Guide 4524 Translation
40、 Manager 4539 Visual Translator 4540 White Papers at ,Global Accessibility and Deployment Agenda,Importance of localization Definition of terms Regional business requirements User interface requirements Character processing on the web Internet, WebSpeed, XML Translation strategies Summary,Summary Yo
41、ur I18n tasklist,Localization is important on the web Have native, in-market users assess s/w Assess regional business requirements User Interface and Reports Simple, direct language is most efficient Avoid humor, word play, symbolism Generalize and abstract to create graphics Avoid putting text in
42、graphics Plan for text expansion and reformatting Avoid dynamic widget and text creation,Summary Your I18n tasklist,Code pages Label text with the correct code page Label *ML with the correct code page Use the Progress code page parameters Migrate to Unicode Use lang attribute for better browser and search engine support Translation strategy Use Translation Manager for R-Code Use database for server-side text,?,Questions,
