Re: SaxonCS 11.0 is available

10.x has a fair bit of support life left in it: if there are issues preventing you moving, please let us know.

In principle, if we think there is demand for it, we could build 11 using IKVM, but I'd prefer to avoid that if we can. We're fighting on too many fronts.

Michael Kay
Saxonica

> On 7 Oct 2021, at 11:25, Emanuel Wlaschitz <[email protected]> wrote:
> 
> Alright, it was worth a try.
> 
> I'm guessing the plan for those stuck on .NET Framework is: stick to Saxon 10N (or perhaps even 9, which is what we do since even 10.6 still has a few kinks that need to be ironed out?)
> 
> -----Original Message-----
> From: Michael Kay <mike-JkSD5nQpfvpWk0Htik3J/[email protected]> 
> Sent: Thursday, October 7, 2021 12:05
> To: Mailing list for the SAXON XSLT and XQuery processor <[email protected]>
> Subject: Re: [saxon] SaxonCS 11.0 is available
> 
> We (Norm actually) have done a little experiment and it seems that we are indeed using covariant return types, so we get a lot of compile failures if we target netstandard2.0. I know we had the transpiler working to avoid covariant returns at one stage (it was injecting a cast into the result of every call on a virtual method with covariant overrides in the Java source) but it seems we scrapped that code.  We know it can be done because we were doing it at one stage, but it's not going to be trivial.
> 
> Michael Kay
> Saxonica
> 
> _______________________________________________
> saxon-help mailing list archived at http://saxon.markmail.org/
> [email protected]
> https://lists.sourceforge.net/lists/listinfo/saxon-help