I didn't followed the whole subject, but could you check if it works with 3.3 ( where I'm pretty sure the charset and locale had a lot of testing ) and/or with 4.1 ( as coyote-jk2 is stable and will eventually replace the ajp13 in 3.3 and the ajp connector in 4.0 ).
I expect this to be an easy. I'm also european ( 8859-2 ), and I'm pretty interested in the whole charset/internationalization/locale subject. Sorry about the whole confusion on connectors - I can only say it is a very difficult subject, with a huge amount of work and testing involved and only few people are willing to give up the comfort of java programming and debug core dumps in C. I believe mod_jk2 will be simpler enough to make future development easier, and coyote seems like a clear choice for java side. Costin On Tue, 18 Jun 2002, Sexton, George wrote: > Henri, > > I apologize. It is never fair to take out one's frustrations on another. > > I have to admit I have been pretty frustrated with connectors. We are > selling a calendaring app, and I really want people in France to see the > days of the week in French when you look at my calendar. I want an associate > in Germany to transparently see the calendar in German. I also want the > calendar to show Monday as the first day of the week. I am seriously > planning on having the application translated into the major European > languages. I really want the application to transparently run in every > ISO-8859-1 language I can. The other character sets will come, but I want > these first. > > All of these things require a working locale sub-system. > > I switched to mod_webapp because it doesn't have the locale problem. > Instead, mod_webapp is broken in that it doesn't pass all requests to tomcat > for aliased servlets. I.E. if I alias /context/foo/*.html to DisplayServlet, > it doesn't pass the request to Tomcat. So, a connector that I switched to in > order to solve one problem, has a different problem. > > Out of the two problems, I figured locales was the one that would be more > practical to resolve. > > Out in user land, there is a lot of confusion over connectors. It would be > nice if there were a "roadmap" to connectors that said "This is the future", > and "this is the past". Building connectors is insanely difficult. It took > me at least a dozen tries to get mod_jk to build correctly. Needless to say, > there should be pre-compiled connectors for the unwashed masses. > > I personally find it stunning that there are no less than 4 connectors for > Tomcat. I understand the theories of Open Source darwinism, but it would be > nice to have "one" connector that wasn't broken in some way. > > Chris Cain suggested that the connectors should be a sub-project and subject > to the same sort of discipline that the regular project is, including > voting. One thing that I can point out, is that connectors need to be > organized into teams. Your comments to the effect that "mod_jk works with > this configuration so this is not my problem" doesn't cut it in user land. > In user land, all they know is mod_jk works or it doesn't. If the sections > are maintained by different people, then there needs to be much tighter > coordination. The C connector is nothing without the Tomcat ajp listener. > The Tomcat AJP listener is nothing without the C Module. They are two halves > of a whole. > > I hope that this helps you understand more of how "users" see things. > > George Sexton > MH Software, Inc. > Voice: 303 438 9585 > http://www.mhsoftware.com > > -----Original Message----- > From: GOMEZ Henri [mailto:[EMAIL PROTECTED]] > Sent: 18 June, 2002 10:39 AM > To: Tomcat Developers List > Subject: RE: [Bug 6832] - Locale of the browser is ignored by the AJP13 > connector > > > >I find it unusual that a European would not fix such a crucial > >piece of code > >for localization of applications. > > Thanks god there is many europeans commiters on tomcat to > try to fix these problems. > > >This defect was reported in March > > And will be corrected in June by a French commiter. > BTW, ajp13 protocol forward all language, so the patch > is needed in java side. > > >It was confirmed by a second source in March. > > Yep > > >I reported it to you separately in April. > > ReYep > > >I have yet again, confirmed it's existence. To recap my > >original message > >sent directly to you in April: > > Ok > > >I am running into a problem with accept-language/getLocales() > >when I use > >mod_jk with the ajp13 connector. > > > >System Configuration: > > > >RedHat Linux 7.2 > >Apache 1.3.22 > >Tomcat 4.0.4-B1 > >mod_jk.so 4.0.4-B2 version 1.2.0 > > It works well with mod_jk 1.2.0 and tomcat 3.3.1 > > Request Information > JSP Request Method: GET > Request URI: /examples/jsp/snp/snoop.jsp > Request Protocol: HTTP/1.1 > Servlet path: /jsp/snp/snoop.jsp > Path info: null > Path translated: null > Query string: null > Content length: 0 > Content type: null > Server name: hgo1.slib.com > Server port: 80 > Remote user: null > Remote address: 172.31.1.85 > Remote host: pc0082.slib.com > Authorization scheme: null > Locale: fr_FR > Locales: fr_FR en_US en > Secure: false > Scheme: http > cipher_suite: null > key_size: null > ssl_session: null > The browser you are using is Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; > rv:1.0.0) Gecko/20020530 > > -- > To unsubscribe, e-mail: <mailto:[EMAIL PROTECTED]> > For additional commands, e-mail: <mailto:[EMAIL PROTECTED]> > > > -- > To unsubscribe, e-mail: <mailto:[EMAIL PROTECTED]> > For additional commands, e-mail: <mailto:[EMAIL PROTECTED]> > > -- To unsubscribe, e-mail: <mailto:[EMAIL PROTECTED]> For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>
