Hi Llewellyn,
Great ideas - I agree with much of what you say.
We should start a bug or a Wiki page to list out all the cases for
better error messages.
I worry about having lots of silly config messages "ignore this
error". We almost need one "ignore backwards compatible errors" that
covers VelociMacro arguments, bad #includes, etc.
WILL
On 3/22/06, Llewellyn Falco <[EMAIL PROTECTED]> wrote:
> 1) I think it's a GREAT idea to explain options for avoiding an exception in
> the exception.
>
> 2)
> I agree that there are 2 paradigms, the current (work as best as you can,
> even if it is wrong) and what we are looking at as the new (fail loudly if
> it isn't right).
>
> and I understand the desire for backwards compatibility. and agree that
> there should be a way to achieve it.
> however, I would like to look this statement I bit closer.
>
> >I've personally got hundreds of templates in production and support users
> >with hundreds more. Say 5% of those templates have this error. (It's not
> >hard to make -- note that the Velocity test suite for Anakia contains this
> >error). All of a sudden I have tens of templates, scattered all over the
> >place that suddenly cause errors
>
> In my mind this is a lot of proof the log files don't really work.
> because if they did, the error's would go away. and without knowing how the
> macro's work, it's hard to tell if the breaking is important, but I tend to
> put parameters in a function because they are to be used, and therefore
> would assume it would be important at least in some situations.
>
> -------------------------------------------------------------------------------------
>
> My point is this. As the Velocity community continues to grow, I would like
> to avoid these things. I would like to make it easier to configure Velocity,
> easier to make it work, easier to debug, so future code does not suffer the
> same as the past. to this end, I once again push for the default settings
> (because new users don't usually change those for a while) to throw errors
> when you do something wrong. Error's that tell you
> 1) what went wrong
> 2) where it went wrong
> 3) what you tried.
> 4) what was possible ?
> 5) suggestions on how to fix it?
>
> for example.
> (not sure if this is still true in the latest version, as my working version
> is a couple of versions behind)
> if a template is not found in a #parse("template2") call, it is ignored.
> so
> 123#parse("unknownTemplate.vm")567
> will output -> 123567
>
> better would be it throws and error ("unknownTemplate").
>
> better still
> unknow template line 1, col 4 in yourTemplate.vm
>
> better still
> unknow template 'unknownTemplate.vm'
> line 1, col 4 in yourTemplate.vm
>
> better still
> couldn't find template 'unknownTemplate.vm'
> at line 1, col 4 in yourTemplate.vm
> from resource loaders:
> file resource load [path = c:\code]
> file resource load [path = c:\templatesss]
>
>
> best
> couldn't find template 'unknownTemplate.vm'
> at line 1, col 4 in yourTemplate.vm
> from resource loaders:
> file resource load [path = c:\code]
> file resource load [path = c:\templatesss]
>
> To ignore this error set the parameter
> 'velocity.errors.ignoreUnknownTemplate' to false
>
>
> -------------------------------------------------------------------------------------
> These type of errors really help to make debugging easy (for example, i
> would see here that i misspelled templates)
> which makes the life a bit easier for us experienced users.
> but can make a world of difference to a newbie, and that difference can be
> the difference between continuing with the product, or just being
> frustrated.
>
> also the parameter being written lines away from the checking of the
> parameter helps to ensure that it is correct (as opposed to documents
> away in the manual or code example)
>
> so anyways, that concludes my rant :-)
>
> Llewellyn.
>
>
>
>
>
>
>
>
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [EMAIL PROTECTED]
> For additional commands, e-mail: [EMAIL PROTECTED]
>
>
--
Forio Business Simulations
Will Glass-Husain
[EMAIL PROTECTED]
www.forio.com
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
