cxf-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Glen Mazza <glen.ma...@verizon.net>
Subject Re: JAXB versus xs:any
Date Mon, 24 Dec 2007 01:00:10 GMT
Am Sonntag, den 23.12.2007, 18:30 -0500 schrieb Benson Margulies:
> Glen,
> 
> So far, I have received exactly 0 replies to the several bug reports I
> have posted to JAXB. Please see defects 458, 461, and 463. The first two
> have been open for days or weeks, and have attracted no commentary
> whatsoever. If you can tell me what secret cause would cause the
> developers to respond, I'd be most grateful.
> 

I gave a comment for 458 (I'm not sure it's a bug.)  

For 461, I think it would be helpful if you dumbed down the problem a
bit more.  (I'm not sure what you're doing, so I can't duplicate the
problem.)  For one, you have too many unnecessary lines of code in your
sample download (in particular TestBean1.java), so I don't know what I'm
supposed to be looking for and what is noise.  Also, it would be better
to give sample output of what you're expecting and what you're getting.
Also, what flags, command line options, etc. are you running to generate
the output?  These things help in quickly duplicating the problem.  

For 463, I think you just want an enhancement--for it to return an error
if there is ##any instead of treating it like ##other.  In this case, a
little bit of Dale Carnegie might help, instead of labelling it as a
"DEFECT" and writing 

"Aside from the frustration of the lack of a mapping for the rest of
xs:any, it is really a trap for people that the code generator simply
spits out @'s that result in runtime for ##other instead of
complaining.", 

I'd just label it as an ENHANCEMENT request and state:  "I think it
would be better for JAXB to return an error for ##xs:any instead of
silently converting it to ##other--that could create risks downstream."
Keep it friendly.

Glen




Mime
View raw message