Re: First release candidate for gnuplot 6

[Tatsuro M. <tma...@ya...>]

> ----- Original Message -----
> 
> From: "Ethan A Merritt" <me...@uw...>
> To: "beta" <gnu...@li...>
> Cc: "Tatsuro MATSUOKA" <tma...@ya...>
> Date: 2023/07/23 日 16:17
> Subject: Re: First release candidate for gnuplot 6
> 
> 
> On Saturday, 22 July 2023 01:06:05 PDT Tatsuro MATSUOKA via gnuplot-beta wrote:
> > 
> > BTW, 
> > --enable-stable-sort
> > seems to be missing on 6.0.rc1 but is active on the dev. branch.
> > Is this intentional?
> > 
> > Tatsuro
> 
> There are several recent features in 6.1 that I felt were not yet polished
> enough or not yet tested enough to include immediately in 6.0.rc1.
> In order of introduction (oldest first) these are
> 
> - stable sort guarantees for Z-ordering of 3D plot elements 
> 
> - z slices of pm3d surfaces (z-clipping against both min and max of slice range)
>   and two new features that depend on z slice code changes:
>     "splot with contourfill" (depends on z-slices)
>     control over fill border line properties of individual 3D objects
> 
> - colorbox gradient rendered as a pixel map rather than as separate boxes
> 
> These can be considered for inclusion in a second release candidate for 6.0
> or stay in the development version for more work.
> You have probably tested the stable sort code as much as anyone.
> Do you think it should be added to the next 6.0 release candidate?
I myself do not have a strong opinion on this matter.
It will be grateful if it is implemented in the future.

> There are also a couple of changes in 6.1 that have been marked EXPERIMENTAL
> from the time they were first added and I did not include them in 6.0 rc1.
> IMHO these should be re-thought and a better or more complete implementation
> found before inclusion in a stable release.
> 
> - "set colorbox cbtics <linetype>"  (EXPERIMENTAL)
> - "using cbticlabels(<column>)"     (EXPERIMENTAL)
> 
> On the other hand, some other features that are still marked EXPERIMENTAL
> in the documentation are included in 6.0 rc1.
> That warning should maybe be removed if no problems are reported by
> people testing the release candidate.
> 
>     Ethan
> 
> 
>