Werkz stack traces are detrimental.

[Brill Pappin <br...@stabilia.com>]

I am so getting tired of the vary long and fairly useless workz tag 
stack traces when there is a problem.

I know you guys are pretty swamped but I think this one is important... 
when a non-maven user tries to build using maven and they get a stack 
trace like that, they immediately turn off even if it's a simple fix; 
they miss the message about what needs fixing because it off the screen, 
or there is a stack trace on the screen that is not relevant to whats 
happening.

I think (I beg) that before you release version 1.0, you need to do 
something about those useless (from the user point of view) stack 
traces. Ant does this fairly well by keeping the stack trace short, or 
actually telling you what the problem is.

At the very least you can chop the bottom of the stack trace off and 
only show me the relevant trace, because myself and my peers don't care 
what the internals of Maven look like... we only care that we can't 
build our own tree and we need to know why quickly and clearly.

- Brill Pappin

---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe@maven.apache.org
For additional commands, e-mail: users-help@maven.apache.org

Re: Werkz stack traces are detrimental.

["Jefferson K. French" <je...@frenches.org>]

One way to get a stack trace is to create some malformed XML. For
example, turn this:

  <j:set var="foo" value="bar"/>

into:

  <j:set var="foo" value="bar">

without the ending slash. Or embed a less-than or greater-than sign in
a string, rather than the appropriate XML entity. Both will throw a
SAXParseException with stack trace. The error message printed at the
top of the stack trace is pretty self explanatory in finding the
problem.

  Jeff
  
On Wed, 23 Jun 2004, at 13:40:11 [GMT -0400] Brill Pappin wrote:

> What version did you folks start removing those stack traces?
> It may be that the version installed here is older than the fixes.

> When I see one next I'll post it.. they are usually unexpected though 
> and I'm not sure off the top of my head how to reproduce any particular 
> trace.

> - Brill Pappin

> Brett Porter wrote:

>>Brill,
>>
>>Do you have an example of a misleading one? You shouldn't get a stack
>>trace unless you have specified -e or -X, and then by definition, you
>>asked for it :)
>>
>>If it is a fatal error (you got the blurb at the end saying how to
>>post it to the list), then we need the whole trace.
>>
>>Some plugins (linkcheck) will dump the stack even when there is no
>>problem. These need to be fixed, but its not a core issue.
>>
>>- Brett
>>
>>  
>>

> ---------------------------------------------------------------------
> To unsubscribe, e-mail: users-unsubscribe@maven.apache.org
> For additional commands, e-mail: users-help@maven.apache.org

-- 
mailto:jeff@frenches.org



---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe@maven.apache.org
For additional commands, e-mail: users-help@maven.apache.org

Re: Werkz stack traces are detrimental.

[Brill Pappin <br...@stabilia.com>]

What version did you folks start removing those stack traces?
It may be that the version installed here is older than the fixes.

When I see one next I'll post it.. they are usually unexpected though 
and I'm not sure off the top of my head how to reproduce any particular 
trace.

- Brill Pappin

Brett Porter wrote:

>Brill,
>
>Do you have an example of a misleading one? You shouldn't get a stack
>trace unless you have specified -e or -X, and then by definition, you
>asked for it :)
>
>If it is a fatal error (you got the blurb at the end saying how to
>post it to the list), then we need the whole trace.
>
>Some plugins (linkcheck) will dump the stack even when there is no
>problem. These need to be fixed, but its not a core issue.
>
>- Brett
>
>  
>

---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe@maven.apache.org
For additional commands, e-mail: users-help@maven.apache.org

Re: Werkz stack traces are detrimental.

[Brett Porter <br...@gmail.com>]

Brill,

Do you have an example of a misleading one? You shouldn't get a stack
trace unless you have specified -e or -X, and then by definition, you
asked for it :)

If it is a fatal error (you got the blurb at the end saying how to
post it to the list), then we need the whole trace.

Some plugins (linkcheck) will dump the stack even when there is no
problem. These need to be fixed, but its not a core issue.

- Brett

On Tue, 22 Jun 2004 10:09:22 -0400, Brill Pappin <br...@stabilia.com> wrote:
> 
> I am so getting tired of the vary long and fairly useless workz tag
> stack traces when there is a problem.
> 
> I know you guys are pretty swamped but I think this one is important...
> when a non-maven user tries to build using maven and they get a stack
> trace like that, they immediately turn off even if it's a simple fix;
> they miss the message about what needs fixing because it off the screen,
> or there is a stack trace on the screen that is not relevant to whats
> happening.
> 
> I think (I beg) that before you release version 1.0, you need to do
> something about those useless (from the user point of view) stack
> traces. Ant does this fairly well by keeping the stack trace short, or
> actually telling you what the problem is.
> 
> At the very least you can chop the bottom of the stack trace off and
> only show me the relevant trace, because myself and my peers don't care
> what the internals of Maven look like... we only care that we can't
> build our own tree and we need to know why quickly and clearly.
> 
> - Brill Pappin
> 
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: users-unsubscribe@maven.apache.org
> For additional commands, e-mail: users-help@maven.apache.org
> 
>

---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe@maven.apache.org
For additional commands, e-mail: users-help@maven.apache.org