Re: cvs commit: jakarta-tomcat-catalina/catalina/src/share/org/apache/catalina/coreStandardContext.java
Tim Funk wrote: OK, just wanted to know if this was an 'oops' or just the way things are. I'll interpret that as the latter. ;) If you have ideas on how to improve that, post them :) Remy - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: cvs commit: jakarta-tomcat-catalina/catalina/src/share/org/apache/catalina/coreStandardContext.java
OK, just wanted to know if this was an 'oops' or just the way things are. I'll interpret that as the latter. ;) -Tim Remy Maucherat wrote: Tim Funk wrote: Sort of unrelated to the patch but noticed from line: File file = new File(configBase, name.replace('/', '_') + ".xml"); Would the following 2 names cause a confict if used at the same time? 1) name="more_cowbell"; 2) name="more/cowbell"; Well, yes. You can't avoid all conflicts, so ... Remy - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: cvs commit: jakarta-tomcat-catalina/catalina/src/share/org/apache/catalina/coreStandardContext.java
Tim Funk wrote: Sort of unrelated to the patch but noticed from line: File file = new File(configBase, name.replace('/', '_') + ".xml"); Would the following 2 names cause a confict if used at the same time? 1) name="more_cowbell"; 2) name="more/cowbell"; Well, yes. You can't avoid all conflicts, so ... Remy - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: cvs commit: jakarta-tomcat-catalina/catalina/src/share/org/apache/catalina/coreStandardContext.java
Sort of unrelated to the patch but noticed from line: File file = new File(configBase, name.replace('/', '_') + ".xml"); Would the following 2 names cause a confict if used at the same time? 1) name="more_cowbell"; 2) name="more/cowbell"; -Tim (waiting for coffee to brew before trying to think) [EMAIL PROTECTED] wrote: remm2003/07/14 03:54:11 Modified:catalina/src/share/org/apache/catalina/core StandardContext.java Log: - Strip out leading '/' for the context file name, for cosmetic reasons. Revision ChangesPath 1.70 +4 -1 jakarta-tomcat-catalina/catalina/src/share/org/apache/catalina/core/StandardContext.java Index: StandardContext.java === RCS file: /home/cvs/jakarta-tomcat-catalina/catalina/src/share/org/apache/catalina/core/StandardContext.java,v retrieving revision 1.69 retrieving revision 1.70 diff -u -r1.69 -r1.70 --- StandardContext.java 2 Jul 2003 22:00:34 - 1.69 +++ StandardContext.java 14 Jul 2003 10:54:11 - 1.70 @@ -3867,6 +3867,9 @@ if (name.equals("")) { name = "ROOT"; } +if (name.startsWith("/")) { +name = name.substring(1); +} File file = new File(configBase, name.replace('/', '_') + ".xml"); setConfigFile(file.getPath()); } - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: cvs commit: jakarta-tomcat-catalina/catalina/src/share/org/apache/catalina/coreStandardContext.java
Glenn Nielsen wrote: [EMAIL PROTECTED] wrote: remm2003/06/15 06:10:41 Modified:catalina/src/share/org/apache/catalina/core StandardContext.java Log: - Move context descriptors to $CATALINA_BASE/conf//, as proposed by Glenn. - This should make the feature secure, and I think there's no justification anymore for the deployXML flag. - Note: The manager webapp may need a few updates, which are in progress. I haven't had a chance to review the code yet. I have a question about removing the deployXML flag. In your redesign will the ability to install a {context}.xml file using an ant task or the web application manager still be available? If so, then for security the deployXML flag would still be needed. Only the location of the XML context descriptors changes. That still allows defining stuff which goes in server.xml using the manager, and is "as dangerous" (IMO) as the admin webapp. So that does give a good reason to keep the flag in :) I wasn't too sure about the change, so that's why I had left the flag in. OTOH, the feature is a little bit better now, since you can allow your users to write to the webapps folder to easily deploy their webapps, and they (hopefully) won't be able to hack the container. BTW, there's some stuff I didn't retest, in addition to the two known issues I mentioned (the WAR locking could be caused by fileupload, and is worth investigating IMO), including deploying XML descriptors from themanager. Remy - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: cvs commit: jakarta-tomcat-catalina/catalina/src/share/org/apache/catalina/coreStandardContext.java
[EMAIL PROTECTED] wrote: remm2003/06/15 06:10:41 Modified:catalina/src/share/org/apache/catalina/core StandardContext.java Log: - Move context descriptors to $CATALINA_BASE/conf//, as proposed by Glenn. - This should make the feature secure, and I think there's no justification anymore for the deployXML flag. - Note: The manager webapp may need a few updates, which are in progress. I haven't had a chance to review the code yet. I have a question about removing the deployXML flag. In your redesign will the ability to install a {context}.xml file using an ant task or the web application manager still be available? If so, then for security the deployXML flag would still be needed. Regards, Glenn - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: cvs commit: jakarta-tomcat-catalina/catalina/src/share/org/apache/catalina/coreStandardContext.java
Jean-Francois Arcand wrote: Bonne nouvelle :-) The problem right now is we ends up with 2 configuration files. If I have under webapps/ a file named test.xml that contains: that always work. But if the path is equals to: that will not work (even with my tentative patch). We will ends up with 2 configurations file (test.xml and test2.xml) and the admin tool/HostDeployer will always use webapp/test.xml According to the consensus, the context files will go in conf/[enginename]/[hostname]/ I'll try to be careful during the refactoring (which will not be done by today, sorry), and I'll try to make the deployer less dumb ;-) Remy - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: cvs commit: jakarta-tomcat-catalina/catalina/src/share/org/apache/catalina/coreStandardContext.java
Remy Maucherat wrote: [EMAIL PROTECTED] wrote: jfarcand2003/05/28 21:13:24 Modified:catalina/src/share/org/apache/catalina/core StandardContext.java Log: Revert back my latest changes since it did not fix the problem completely. Don't worry about that IMO. I'll have to rewrite that code for the deployer refactoring I'll do (today). I hope I'll be done in one day :) Bonne nouvelle :-) The problem right now is we ends up with 2 configuration files. If I have under webapps/ a file named test.xml that contains: that always work. But if the path is equals to: that will not work (even with my tentative patch). We will ends up with 2 configurations file (test.xml and test2.xml) and the admin tool/HostDeployer will always use webapp/test.xml -- Jeanfrancois Remy - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED] - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: cvs commit: jakarta-tomcat-catalina/catalina/src/share/org/apache/catalina/coreStandardContext.java
[EMAIL PROTECTED] wrote: jfarcand2003/05/28 21:13:24 Modified:catalina/src/share/org/apache/catalina/core StandardContext.java Log: Revert back my latest changes since it did not fix the problem completely. Don't worry about that IMO. I'll have to rewrite that code for the deployer refactoring I'll do (today). I hope I'll be done in one day :) Remy - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: cvs commit: jakarta-tomcat-catalina/catalina/src/share/org/apache/catalina/coreStandardContext.java
Costin Manolache wrote: [EMAIL PROTECTED] wrote: remm2003/03/12 13:32:50 Modified:catalina/src/share/org/apache/catalina/core StandardContext.java Log: - If init needs guards against being invoked twice, this also needs it. - Any explanation why, BTW ? The general idea is simple: components shouldn't assume a perfect world where everything is consistent and loaded in a certain order. In this particular case - I assume the mbean for the context couldn't find the parent to register itself - when it adds itself to the host, the host will call setParent(). On destroy - I don't think this is a problem. Well, I was getting a NPE, and the logs seemed to indicate that destroy had already been called. Remy - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]