Re: Errors vs. exceptions

Nick,

Unless I misunderstand the `version` attribute, specifying
`version="1.1"` would enable an XForms 1.1-compatible processor (if
supported by the processor), and therefore the -exception events would
be the ones used.

If correct, then there is no particular wording to add to the spec for
this kind of backward-compatibility.

-Erik

On Wed, Sep 7, 2016 at 7:49 AM, Nick Van den Bleeken
<Nick.Van.den.Bleeken@inventivegroup.com> wrote:
> Erik,
>
>
>
> Is it an option to ‘not rename’ the events if the xforms version is
> specified and set to an older version? This tackles the backwards
> compatibility issue a bit.
>
>
>
> Regards,
>
>
>
> Nick
>
>
>
> From: <ebruchez@gmail.com> on behalf of Erik Bruchez <ebruchez@orbeon.com>
> Date: Wednesday 7 September 2016 at 07:35
> To: "public-xformsusers@w3.org" <public-xformsusers@w3.org>
> Subject: Errors vs. exceptions
> Resent-From: <public-xformsusers@w3.org>
> Resent-Date: Wednesday 7 September 2016 at 07:36
>
>
>
> All,
>
>
>
> I have a task of figuring out whether we need to rename some/all -error
> events to -exception events (or the other way around ;).
>
>
>
> In XForms 1.1, we had non-fatal -exception events, and fatal (stopping
> processing without any way of recovering) -error events (the opposite of
> Java!).
>
>
>
> In XForms 2, we no longer have events which are necessarily fatal: all
> previously-fatal events can be canceled (that is, we can cancel the default
> action which is to stop processing) and are renamed into -error.
>
>
>
> This would seem to call for calling them -error, following the XForms 1.1
> convention, and also following xforms-submit-error and xforms-output-error
> which never were fatal. Also, "error" is shorter than "exception", which is
> nice.
>
>
>
> Then there is the question of backward-compatibility. One rationale for just
> creating new -error events and removing the old -exception events is that
> there was no way to really do much with -exception events as they would stop
> processing soon after being dispatched. So I think that
> backward-compatibility is not a big issue here. We *could* consider keeping
> -exception events alongside the new -error events, but that probably
> wouldn't bring much benefit.
>
>
>
> So I think the naming -error is acceptable.
>
>
>
> One question coming to mind is whether we should consider never halting
> processing, that is making recovering from errors the default, instead of
> requiring the form author to catch all those events.
>
>
>
> -Erik

Received on Wednesday, 7 September 2016 16:07:23 UTC