Re: Width modifier changes in Bug 29555 introduce a backwards incompatibility issue

We made the decision that this was an acceptable incompatibility, since 3.0 defined no semantics for this construct. You're right that it should probably be added to the incompatibility appendix.

Michael Kay
Saxonica


> On 8 May 2016, at 20:18, Abel Braaksma <abel.braaksma@xs4all.nl> wrote:
> 
> I wasn't present at last telcon and couldn't find this covered in the minutes, if it was discussed and dismissed as acceptable, then please disregard the following.
> 
> The proposed resolution of Bug 29555 in comment 10 (https://www.w3.org/Bugs/Public/show_bug.cgi?id=29555#c10) introduced an error case for widths with a minimum higher than the maximum (err: FOFD1340).
> 
> In FO30 there is no such error scenario, which means that format-date("[W,3-2]") was acceptable, at least it didn't throw an error. So existing XQ or XSLT with such function calls will fail now with FO31.
> 
> I don't know if it is enough an incompatibility to spend any time on, but if we accept this incompatibility then perhaps we should add it to the section "G Compatibility with Previous Versions (Non-Normative)".
> 
> (in fact, there is already a line on format-number, but not on format-date and format-dateTime).
> 
> -- Abel
> 
> 

Received on Sunday, 8 May 2016 19:53:38 UTC