[Bioc-devel] API change for 'hexbin';
old version as 'hexbin0' package.
Martin Maechler
maechler at stat.math.ethz.ch
Fri Oct 15 19:44:09 CEST 2004
Dear Bioconductor co-developers.
Nicholas Lewin-Koh (the maintainer of 'hexbin') and I
(co-author, i.e. currently mainly tester & helper) have talked
about hexbin and the upcoming bioconductor release.
More than once we have voiced the plan to ``really fix'' hexbin
by abandoning the unfixable layout problem of hexbin graphics
{resizing: hexagons must remain symmetric hexagons; hexagon
plots as part of bigger plots, etc, etc} and
moving to *grid* graphics.
In collaboration with the grid master minder, Paul Murrell,
Nicholas has basically achieved this goal many weeks ago, also
switching from S3 to S4 classes and methods. Because time
constraints, hardware problems and the "demise" of the CVS
system, Nicholas could not commit the new changes until now.
OTOH, I have argued that bioconductor packages and users should
not program against the old ("broken by design") API but see and
start using the new one, as from Bioconductor 1.5.
For this reason, Nicholas and I have decided to put the new code
into hexbin till the feature freeze of coming Monday, and
bring it into good shape according to the "freeze release"
schedule Jeff posted yesterday.
In general, you (bio package writers) should start using the new
'hexbin' when it's available and may need a few changes in your
package examples, possibly your own functions because of the
change from base to grid graphics (from package 'graphics' to
package 'grid').
If you or other people cannot (or do not want to) do the switch
just now, we have decided to provide the old (S3, "graphics", broken)
version of hexbin as package "hexbin0".
I'm currently porting the last bits from "hexbin" to "hexbin0";
it's astonishing how many places have the package name explicitly..:-(
I hereby strongly advocate that both "hexbin" and "hexbin0"
should be put into the release of bioconductor 1.5.
Further comments are very welcome.
Martin Maechler <maechler at stat.math.ethz.ch> http://stat.ethz.ch/~maechler/
Seminar fuer Statistik, ETH-Zentrum LEO C16 Leonhardstr. 27
ETH (Federal Inst. Technology) 8092 Zurich SWITZERLAND
phone: x-41-1-632-3408 fax: ...-1228 <><
More information about the Bioc-devel
mailing list