> No. I thought that the bundle included everything I needed. It does, and it is a bug. It is my fault ... I noticed this a while back, sent an email to the author of rhdf5 and promptly forgot to followup on it. The problem is that the compiler doesn't seem to pick up the current directory when searching for header files (even despite using "common.h"), and I'm not familiar enough with the way source files are compiled with R to say exactly how to fix it. I'll look into this and get it fixed. -J

> No. I thought that the bundle included everything I needed. I forgot to mention that the behaviour is system specific (I'm guessing that you're using Linux?) - so it isn't a bug in rhdf5 but rather just compiler behaviour. Its just that I don't know how to feed compiler directives in to the R CMD. -J

You should get the library package from hdf.ncsa.uiuc.edu. We don't include the source packages 'cause they're sort of large and NCSA tends to update at semi-random intervals so you'll want to get the most recent version (their backwards compatibility tends to be pretty good) Byron Ellis (bellis@hsph.harvard.edu) "Oook" - The Librarian Please finger bellis@hsph.harvard.edu for PGP keys On Mon, 8 Jul 2002, Isaac Neuhaus wrote: > No. I thought that the bundle included everything I needed. > > Isaac > > A.J. Rossini wrote: > > >>>>>>"isaac" == Isaac Neuhaus <isaac.neuhaus@bms.com> writes: > >>>>>> > > > > isaac> I got the following error while insdtalling the current bundle version > > isaac> of Bioconductor. > > > > > > > > isaac> [neuhausi@aurora Bioconductor]$ sudo R CMD INSTALL rhdf5 > > isaac> * Installing *source* package 'rhdf5' ... > > isaac> ** libs > > isaac> gcc -I/usr/local/lib/R/include -I/usr/local/include > > isaac> -D__NO_MATH_INLINES -mieee-fp -fPIC -g -O2 -c hdf5.c -o hdf5.o > > > > isaac> In file included from hdf5.c:1: > > isaac> common.h:7:18: hdf5.h: No such file or directory > > isaac> make: *** [hdf5.o] Error 1 > > isaac> ERROR: compilation failed for package 'rhdf5' > > > >Do you have the hdf5 system/development libraries on your system? > > > >best, > >-tony > > > > > > _______________________________________________ > Bioconductor mailing list > Bioconductor@stat.math.ethz.ch > http://www.stat.math.ethz.ch/mailman/listinfo/bioconductor >

because rhdf5 is an interesting package and resource, i would rather see an 'opt-in' approach at build time, so that by default the package is received but not built/installed. if a) the person gives a build option indicating that they want rhdf5 and b) a configure-level check indicates that the necessary resources are on hand, then the build proceeds with it, otherwise it skips it. On Sat, 13 Jul 2002, Robert Gentleman wrote: > Dave, > Thanks, that is a good point. We will look at removing it (and > amending the description of what users get) from the automatic > downloads. > > But we still need to make the point that the hdf5 files are not > missing. They are part of a separate application. This is the same behavior > as many other R packages (XML, all the database packages) which presume that > the application being interacted with has been installed. > It is easily downloaded and installed from: > http://hdf.ncsa.uiuc.edu/HDF5/ > > Regards, > Robert > > > On Fri, Jul 12, 2002 at 04:57:52PM -0700, David O. Nelson wrote: > > Hello all, > > > > Regarding the compile failure due to missing rhdf5 files... > > > > Might I recommend that rhdf5 not be a part of "...all Bioconductor recommneded > > (sic) packages in one bundle" ? > > > > That way potential users won't be turned off by downloading the package > > bundle, trying to install it, and being thrown by the compile failure and a > > partially installed bundle. > > > > I appreciate the convenience of one download and install, but it should be a > > bit more self-contained, eh? > > > > cheers, > > dave > > > > -- > > David O. Nelson > > Biology and Biotechnology Research Program > > Lawrence Livermore National Laboratory > > Box 808, L-441 > > Livermore, CA 94551 > > _______________________________________________ > > Bioconductor mailing list > > Bioconductor@stat.math.ethz.ch > > http://www.stat.math.ethz.ch/mailman/listinfo/bioconductor > > -- > +------------------------------------------------------------------- --------+ > | Robert Gentleman phone : (617) 632-5250 | > | Associate Professor fax: (617) 632-2444 | > | Department of Biostatistics office: M1B20 > | Harvard School of Public Health email: rgentlem@jimmy.dfci.harvard.edu | > +------------------------------------------------------------------- --------+ > _______________________________________________ > Bioconductor mailing list > Bioconductor@stat.math.ethz.ch > http://www.stat.math.ethz.ch/mailman/listinfo/bioconductor >