It seems that the samples don't work with cocoon. So is there an example for cocoon in 
keel?

-----原始邮件-----
发件人: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED]
[EMAIL PROTECTED]
发送时间: Sunday, August 31, 2003 3:04 PM
收件人: [EMAIL PROTECTED]
主题: Keelgroup digest, Vol 1 #483 - 3 msgs


Send Keelgroup mailing list submissions to
        [EMAIL PROTECTED]

To subscribe or unsubscribe via the World Wide Web, visit
        http://lists.keelframework.com/listinfo.cgi/keelgroup-keelframework.com

or, via email, send a message with subject or body 'help' to
        [EMAIL PROTECTED]

You can reach the person managing the list at
        [EMAIL PROTECTED]

When replying, please edit your Subject line so it is more specific
than "Re: Contents of Keelgroup digest..."


Today's Topics:

   1. Re: Maven issue - 65 explicitly declared jars - Yeouch! (Pete Carapetyan)
   2. Re: Build Results, Unit and Functional Tests Online (Sasvata (Shash) Chatterjee)
   3. Re: Build Results, Unit and Functional Tests Online (Michael Nash)

--__--__--

Message: 1
Date: Sat, 30 Aug 2003 14:02:07 -0500
From: Pete Carapetyan <[EMAIL PROTECTED]>
To: [EMAIL PROTECTED]
Subject: Re: [Keelgroup] Maven issue - 65 explicitly declared jars - Yeouch!
Reply-To: [EMAIL PROTECTED]

[EMAIL PROTECTED] wrote:

> Pete Carapetyan wrote:
>
>> Much easier to plop a bunch of jars in a lib directory and be done 
>> with it !
>>
>> CONCLUSION:
>> Lots more work in the setup/transition phase. No solution is offered 
>> for this PITA.
>>
>> On the other hand, it does provide a lot more benefits, just as there 
>> are benefits to not using the wildcard in import declarations. The 
>> question is, can we get over this rather labor intensive hump?
>>
>> I have a start at placing the jars in keelframework.org/maven. Let me 
>> know if there are others that you need to place there. Adam, Shash, 
>> and Michael also have write permissions to that folder.
>
>
> Pete,
>
> I have not used Maven so this may be an over simplified approach,
> but if Maven has some kind of tag that will include an XML snippet
> inline (ex: import), it should not be difficult to write an XSL
> that will build the fileset specifying each JAR in the Maven format
> on the fly. If it were possible to write a Maven Task and place it
> in the file before the import like so:
>
>   <GenJARS dir="${MAVEN_HOME}/lib"
>            output="${MAVEN_HOME}/lib/libs.xml"
>            format="${MAVEN_HOME}/libs/default.xsl"/>
>   <import dir="{MAVEN_HOME}/lib/libs.xml">
>
> where the libs.xml is generated by the <GenJARS> task using the XSL
> format specified in default.xsl, and the <import> task pulls them
> into the Maven project file in the <fileset> format you used earlier.
> Is this a doable approach? 

There is always a lot of versioning, checking, chosing between versions, 
and ftp-ing stuff into the maven repository which takes most of the time 
the first time you do it. But still, your idea would work for the basic 
syntax, and i think I am going to try it at least to create a start on 
the xml.

Thanks for the heads up !



--__--__--

Message: 2
Date: Sat, 30 Aug 2003 17:36:08 -0500
From: "Sasvata (Shash) Chatterjee" <[EMAIL PROTECTED]>
To: [EMAIL PROTECTED]
Subject: Re: [Keelgroup] Build Results, Unit and Functional Tests Online
Reply-To: [EMAIL PROTECTED]

Mike,

Looks like the build-tests.sh script works just fine under Cygwin/Bash 
on Win-XP.

Also, I've just opted in for compile-farm access at SF, we should be 
able to setup nightly builds for Mac, Solaris, and some other Linux 
variants shortly.

I just turned on installs for Anteater and Axis to true in 
ant.properties, since those are required now.

I am having trouble with functional tests on the default deploy.  If I 
run from the script, the reports/default/anteater directory doesn't get 
created (I've attached the log file). Looks like the app-crud login is 
timing out because Tomcat is taking long to startup on my poor 500MHz 
PC, and anteater is timing out.  Where can I increae that timeout?

 If I run tomcat, wait, and then run "ant anteater" directly in 
keel-build, then the directory gets created, but the app-crud test is 
failing....because (an excerpt):

Expected to match pattern: .*Logged in as.* Got: ' <html lang="en"> 
<head> <meta http-equiv="content-type" 
content="text/html;charset=iso-8859-1"> <meta 
http-equiv="Content-Style-Type" content="text/css"> <link 
rel="stylesheet" href="/struts/navigate/menu.css" id="css_menu"> <style> 
a { color: #003366; } ul { margin-top: 1px; margin-bottom: 10px;} td { 
font-family: Tahoma, Verdana, sans-serif; font-size: 12px; } form { 
margin: 5px; } hr { color: #eeeeee; height : 1px; } .ctrl {font-family: 
Tahoma, Verdana, sans-serif; font-size: 12px

Any ideas?

Shash

Michael Nash wrote:

>A new top-level page is available at 
>http://66.105.113.115/vqwiki-2.3.5/jsp/Wiki?UnitTestResults, and an example set of 
>results is out on http://www.keelframework.org/testresults/jglobal/index.html.
>
>Should be fairly self-explanatory, the idea is to show as many different 
>configurations as are reasonable with
>
>1. Build results
>2. Unit Test Results
>3. Functional Test Results
>
>as well as the logs from each of these.
>
>Ideally, we'd have pages like this generated by CruiseControl instances running on a 
>number of different types of servers, verifying Keel in many different ways for every 
>CVS checkin.
>
>These results are basically the output from the build-tests.sh shell script checked 
>in earlier today (in keel-build). If anyone else can set up a test server to crank 
>out something similar, the more the merrier. :-)
>
>I should have this integrated with CruiseControl shortly, but in the meantime will 
>try to update it daily. It is currently run against HEAD from CVS.
>
>Feedback appreciated! More "green lights" should be coming along soon...
>
>Regards,
>
>Mike
>JGlobal Ltd.
>http://www.jglobal.com
>
>http://keelframework.org/documentation
>Keelgroup mailing list
>[EMAIL PROTECTED]
>http://lists.keelframework.com/listinfo.cgi/keelgroup-keelframework.com
>  
>



--__--__--

Message: 3
Date: Sun, 31 Aug 2003 07:29:33 -0600
From: Michael Nash <[EMAIL PROTECTED]>
To: [EMAIL PROTECTED]
Subject: Re: [Keelgroup] Build Results, Unit and Functional Tests Online
Organization: JGlobal
Reply-To: [EMAIL PROTECTED]

Shash:

> Looks like the build-tests.sh script works just fine under Cygwin/Bash 
> on Win-XP.

Great!

> Also, I've just opted in for compile-farm access at SF, we should be 
> able to setup nightly builds for Mac, Solaris, and some other Linux 
> variants shortly.

Excellent Idea - I'd forgotten that was possible.

> I just turned on installs for Anteater and Axis to true in 
> ant.properties, since those are required now.
> 
> I am having trouble with functional tests on the default deploy.  If I 
> run from the script, the reports/default/anteater directory doesn't get 
> created (I've attached the log file). Looks like the app-crud login is 
> timing out because Tomcat is taking long to startup on my poor 500MHz 
> PC, and anteater is timing out.  Where can I increae that timeout?

That's in build-tests.sh, there's a "sleep" while tomcat fires up - or if you mean the 
webapp itself is taking too long, I think you can add a delay in the anteater script, 
but I don't immediately recall how....

>  If I run tomcat, wait, and then run "ant anteater" directly in 
> keel-build, then the directory gets created, but the app-crud test is 
> failing....because (an excerpt):
> 
> Expected to match pattern: .*Logged in as.* Got: ' <html lang="en"> 
> <head> <meta http-equiv="content-type" 
> content="text/html;charset=iso-8859-1"> <meta 
> http-equiv="Content-Style-Type" content="text/css"> <link 
> rel="stylesheet" href="/struts/navigate/menu.css" id="css_menu"> <style> 
> a { color: #003366; } ul { margin-top: 1px; margin-bottom: 10px;} td { 
> font-family: Tahoma, Verdana, sans-serif; font-size: 12px; } form { 
> margin: 5px; } hr { color: #eeeeee; height : 1px; } .ctrl {font-family: 
> Tahoma, Verdana, sans-serif; font-size: 12px
> 
> Any ideas?

Looks like the previous test (the login) isn't logging in properly, or retaining the 
login, as that test above should have it already logged in, if I recall. Nothing in 
the log that gives a hint? Maybe the rest of the error output will have a message or 
trace - this looks like the output of the JavaScript menu, which goes on for a *long* 
while in the page :-). Are we calling verify-login-seq here? It should be just 
verify-login I think...

Mike
JGlobal Ltd.
http://www.jglobal.com



--__--__--

_______________________________________________
Keelgroup mailing list
[EMAIL PROTECTED]
http://lists.keelframework.com/listinfo.cgi/keelgroup-keelframework.com


End of Keelgroup Digest
http://keelframework.org/documentation
Keelgroup mailing list
[EMAIL PROTECTED]
http://lists.keelframework.com/listinfo.cgi/keelgroup-keelframework.com

Reply via email to