Jari Oksanen
2000-May-25 09:39 UTC
[R] problem on upgrading to RH6.2 (was problem with ts pack
plummer at iarc.fr said:> On 24-May-00 Prof Brian D Ripley wrote: > On Wed, 24 May 2000, Christian Posse wrote: > >> >> I just encountered a problem with the ts package: >> >> > > library(ts) >> Error in dyn.load(x, as.logical(local), > as.logical(now)) : >> unable to load shared library >> "/usr/ > local/lib/R/library/ts/libs/ts.so": >> /usr/local/lib/R/library/ts/ > libs/ts.so: undefined symbol: ma > > ... > > I have no idea what is going on: ma is not a symbol in the ts package, > so the problem is definitely elsewhere. We do know that this works > correctly when compiled and run on an RH6.2 system.> I''m afraid dlerror() can return garbage on RH6.2 so the "undefined > symbol: ma" message is probably not helpful.> The latest R RPM for Red Hat Linux is compiled on RH6.2 and works > fine. I suggest you try that.I have similar problems with R 1.0.1 on RH6.2 (pre-compiled rpm binaries). Some of the old packages ceased to work and when I tried to re-install them through `R INSTALL package.tar.gz'', some just fail and give similar error messages. My `ts'' works, but it is of the old original version, and I don''t dare to check if it stops working after re-installation. However, re-installed `tripack'' gives the following error message: > library(tripack) Error in dyn.load(x, as.logical(local), as.logical(now)) : unable to load shared library "/usr/lib/R/library/tripack/libs/tripack.so": /usr/lib/R/library/tripack/libs/tripack.so: undefined symbol: ? If I first gunzip the package, I get another undefined symbol, but the net effect is the unchanged: a failure to load a library. Otherwise the message is the same as for `ts'' above. Downgrading to the old binaries of the package does not help, since they fail for other reasons, in this case because they can''t find e_wsfe. This, I think, is caused by libf2g growing independent of libf2c which had this function (originally in libI77.a). cheers, jari oksanen -- Jari Oksanen - Dept Ecol Env Sci, Univ Helsinki, 15140 Lahti Ph. +358 3 89220312, Fax -- 89220289, Mobile +358 40 5136529 jari.oksanen at helsinki.fi http://www.helsinki.fi/~jhoksane -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html Send "info", "help", or "[un]subscribe" (in the "body", not the subject !) To: r-help-request at stat.math.ethz.ch _._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
Peter Dalgaard BSA
2000-May-25 10:21 UTC
[R] problem on upgrading to RH6.2 (was problem with ts pack
Jari Oksanen <jhoksane at ecology.helsinki.fi> writes:> I have similar problems with R 1.0.1 on RH6.2 (pre-compiled rpm binaries). > Some of the old packages ceased to work and when I tried to re-install them > through `R INSTALL package.tar.gz'', some just fail and give similar error > messages. My `ts'' works, but it is of the old original version, and I don''t > dare to check if it stops working after re-installation. However, re-installed > `tripack'' gives the following error message: > > > library(tripack) > Error in dyn.load(x, as.logical(local), as.logical(now)) : > unable to load shared library "/usr/lib/R/library/tripack/libs/tripack.so": > /usr/lib/R/library/tripack/libs/tripack.so: undefined symbol: ? >You''re doing better than me, I just get a segfault... It seems that something is seriously wrong with dyn.loading Fortran on RH6.2. Upgrading the C/Fortran compilers to gcc 2.95.1 or newer seems to cure it. I''ll be d**ned if I can figure out what is going wrong, but .so files created with 2.95.1 work with R compiled with egcs-1.1.2 and .so files created with egcs-1.1.2 cause segfaults when used with R compiled with 2.95.1, so the problem is likely in the shared libraries themselves. -- O__ ---- Peter Dalgaard Blegdamsvej 3 c/ /''_ --- Dept. of Biostatistics 2200 Cph. N (*) \(*) -- University of Copenhagen Denmark Ph: (+45) 35327918 ~~~~~~~~~~ - (p.dalgaard at biostat.ku.dk) FAX: (+45) 35327907 -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html Send "info", "help", or "[un]subscribe" (in the "body", not the subject !) To: r-help-request at stat.math.ethz.ch _._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
Prof. Ripley was correct suggesting a title with an RH6.2 problem and not a problem concerning the ts package. In fact, it is far worse than I thought and is similar to what Jair Oksanen and Peter Dalgaard wrote. The situation is the following: - The RPM R-base-1.0.1-2.i386.rpm from statlib works fine under RH 6.2. Loading the libraries in the base works fine too. Most contrib packages installed with this version of R with R INSTALL work except for tripack, acepack, akima, mclust (segfault!), leaps, and so on. (I stopped trying all of them.....) Strange messages of undefined symbols :> library(tripack)Error in dyn.load(x, as.logical(local), as.logical(now)) : unable to load shared library "/usr/lib/R/library/tripack/libs/tripack.s o": /usr/lib/R/library/tripack/libs/tripack.so: undefined symbol: ? The situation is far worse with the version of R built by myself after having downloaded the *.tgz file from statlib. With this version, R base works fine, but impossible to load successfully any library, even the one in the base packages (Just in case, the R-base*src.rpm provides the same source code and therefore I got the same problems). Note, my PC is pure vanilla. I just got it brand new and I installed RH6.2. Then I tried to build R and got all these problems. RH6.2 comes with <18>root at cuky/R: gcc --version egcs-2.91.66 How can I compare that with what Peter Dalgaard wrote ( I have been confused for a long time between the different brands of gcc, true gcc or egcs flavor). Peter Dalgaard wrote: You''re doing better than me, I just get a segfault... It seems that something is seriously wrong with dyn.loading Fortran on RH6.2. Upgrading the C/Fortran compilers to gcc 2.95.1 or newer seems to cure it. I''ll be d**ned if I can figure out what is going wrong, but .so files created with 2.95.1 work with R compiled with egcs-1.1.2 and .so files created with egcs-1.1.2 cause segfaults when used with R compiled with 2.95.1, so the problem is likely in the shared libraries themselves. Thanks for all answers so far. Christian ----- Original Message ----- From: "Jari Oksanen" <jhoksane at ecology.helsinki.fi> To: "r-help" <r-help at stat.math.ethz.ch> Sent: Thursday, May 25, 2000 2:39 AM Subject: Re: [R] problem on upgrading to RH6.2 (was problem with ts pack plummer at iarc.fr said:> On 24-May-00 Prof Brian D Ripley wrote: > On Wed, 24 May 2000, Christian Posse wrote: > >> >> I just encountered a problem with the ts package: >> >> > > library(ts) >> Error in dyn.load(x, as.logical(local), > as.logical(now)) : >> unable to load shared library >> "/usr/ > local/lib/R/library/ts/libs/ts.so": >> /usr/local/lib/R/library/ts/ > libs/ts.so: undefined symbol: ma > > ... > > I have no idea what is going on: ma is not a symbol in the ts package, > so the problem is definitely elsewhere. We do know that this works > correctly when compiled and run on an RH6.2 system.> I''m afraid dlerror() can return garbage on RH6.2 so the "undefined > symbol: ma" message is probably not helpful.> The latest R RPM for Red Hat Linux is compiled on RH6.2 and works > fine. I suggest you try that.I have similar problems with R 1.0.1 on RH6.2 (pre-compiled rpm binaries). Some of the old packages ceased to work and when I tried to re-install them through `R INSTALL package.tar.gz'', some just fail and give similar error messages. My `ts'' works, but it is of the old original version, and I don''t dare to check if it stops working after re-installation. However, re-installed `tripack'' gives the following error message: > library(tripack) Error in dyn.load(x, as.logical(local), as.logical(now)) : unable to load shared library "/usr/lib/R/library/tripack/libs/tripack.so": /usr/lib/R/library/tripack/libs/tripack.so: undefined symbol: ? If I first gunzip the package, I get another undefined symbol, but the net effect is the unchanged: a failure to load a library. Otherwise the message is the same as for `ts'' above. Downgrading to the old binaries of the package does not help, since they fail for other reasons, in this case because they can''t find e_wsfe. This, I think, is caused by libf2g growing independent of libf2c which had this function (originally in libI77.a). cheers, jari oksanen -- Jari Oksanen - Dept Ecol Env Sci, Univ Helsinki, 15140 Lahti Ph. +358 3 89220312, Fax -- 89220289, Mobile +358 40 5136529 jari.oksanen at helsinki.fi http://www.helsinki.fi/~jhoksane -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-. -.- r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html Send "info", "help", or "[un]subscribe" (in the "body", not the subject !) To: r-help-request at stat.math.ethz.ch _._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._. _._ -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html Send "info", "help", or "[un]subscribe" (in the "body", not the subject !) To: r-help-request at stat.math.ethz.ch _._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
"Christian Posse" <posse at talariainc.com> writes:> The situation is far worse with the version of R built by myself > after having downloaded the *.tgz file from statlib. With this > version, R base works fine, but impossible to load successfully > any library, even the one in the base packages > (Just in case, the R-base*src.rpm provides the same source code > and therefore I got the same problems). > > Note, my PC is pure vanilla. I just got it brand new and I installed > RH6.2. Then I tried to build R and got all these problems. > RH6.2 comes with > > <18>root at cuky/R: gcc --version > egcs-2.91.66 > > How can I compare that with what Peter Dalgaard wrote > ( I have been confused for a long time between the different > brands of gcc, true gcc or egcs flavor).[pd at blueberry pd]$ rpm -q egcs egcs-1.1.2-30 [pd at blueberry pd]$ gcc --version egcs-2.91.66 It''s the same one. So it''s pretty odd that you cannot build the packaces supplied with R (I assume "make check" fails?). That has worked for me all the time. I don''t think I have anything out of the ordinary in my setup, except $ more .Rconf CFLAGS="-Wall -g -O2" FFLAGS="-Wall -g -O2" -- O__ ---- Peter Dalgaard Blegdamsvej 3 c/ /''_ --- Dept. of Biostatistics 2200 Cph. N (*) \(*) -- University of Copenhagen Denmark Ph: (+45) 35327918 ~~~~~~~~~~ - (p.dalgaard at biostat.ku.dk) FAX: (+45) 35327907 -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html Send "info", "help", or "[un]subscribe" (in the "body", not the subject !) To: r-help-request at stat.math.ethz.ch _._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
I finally found a way around though the underlying problem persists. It seems that the problem is with g77. Only the packages using FORTRAN failed to be loaded correctly in either way (R from rpm or built from scratch). The solution was to build R from the source using the flag --with-f2c with ./configure. All contrib packages installed with this version of R are loaded correctly and, as far as I can superficially see, work. make check gives positive feedback. Has anyone encountered the same problem? Christian ----- Original Message ----- From: "Peter Dalgaard BSA" <p.dalgaard at biostat.ku.dk> To: "Christian Posse" <posse at talariainc.com> Cc: "r-help" <r-help at stat.math.ethz.ch> Sent: Thursday, May 25, 2000 11:20 AM Subject: Re: [R] problem on upgrading to RH6.2 (Update)> "Christian Posse" <posse at talariainc.com> writes: > > > The situation is far worse with the version of R built by myself > > after having downloaded the *.tgz file from statlib. With this > > version, R base works fine, but impossible to load successfully > > any library, even the one in the base packages > > (Just in case, the R-base*src.rpm provides the same source code > > and therefore I got the same problems). > > > > Note, my PC is pure vanilla. I just got it brand new and I installed > > RH6.2. Then I tried to build R and got all these problems. > > RH6.2 comes with > > > > <18>root at cuky/R: gcc --version > > egcs-2.91.66 > > > > How can I compare that with what Peter Dalgaard wrote > > ( I have been confused for a long time between the different > > brands of gcc, true gcc or egcs flavor). > > [pd at blueberry pd]$ rpm -q egcs > egcs-1.1.2-30 > [pd at blueberry pd]$ gcc --version > egcs-2.91.66 > > It''s the same one. So it''s pretty odd that you cannot build the > packaces supplied with R (I assume "make check" fails?). That has > worked for me all the time. I don''t think I have anything out of the > ordinary in my setup, except > > $ more .Rconf > CFLAGS="-Wall -g -O2" > FFLAGS="-Wall -g -O2" >-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html Send "info", "help", or "[un]subscribe" (in the "body", not the subject !) To: r-help-request at stat.math.ethz.ch _._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
posse at talariainc.com said:> I finally found a way around though the underlying problem persists. > It seems that the problem is with g77. Only the packages using FORTRAN > failed to be loaded correctly in either way (R from rpm or built from > scratch).> The solution was to build R from the source using the flag> --with-f2c> with ./configure.> All contrib packages installed with this version of R are loaded > correctly and, as far as I can superficially see, work. make check > gives positive feedback.I am not sure this is safe. At least I was warned when I used a similar shortcut some years ago when I had some trouble with g77 on MS-DOS (djgpp port). g77 used to apply f2c libraries in her childhood, but has grown independent since then and so the libraries forked (libf2c and libf2g). I think that there *may* be a danger of strange behaviour in some other places if we just resort to drop in f2c-libraries in g77, and at worst, this behaviour remains undetected. Not all Fortran fails. For instance, MASS (or all of VR_*) loads (and runs) beautifully, although it uses many Fortran routines. cheers, Jari Oksanen -- "Infinite variety of Nature, that is pure myth. It is not found in Nature herself. It resides in the imagination, or fancy, or cultivated blindness of the man who looks at her." Oscar Wilde -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html Send "info", "help", or "[un]subscribe" (in the "body", not the subject !) To: r-help-request at stat.math.ethz.ch _._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
Prof Brian D Ripley
2000-May-26 06:25 UTC
[R] problem on upgrading to RH6.2 (One solution)
On Fri, 26 May 2000, Jari Oksanen wrote:> > posse at talariainc.com said: > > I finally found a way around though the underlying problem persists. > > It seems that the problem is with g77. Only the packages using FORTRAN > > failed to be loaded correctly in either way (R from rpm or built from > > scratch). > > > The solution was to build R from the source using the flag > > > --with-f2c > > > with ./configure. > > > All contrib packages installed with this version of R are loaded > > correctly and, as far as I can superficially see, work. make check > > gives positive feedback. > > I am not sure this is safe. At least I was warned when I used asimilar shortcut some years ago when I had some trouble with g77 on MS-DOS (djgpp port). g77 used to apply f2c libraries in her childhood, but has grown independent since then and so the libraries forked (libf2c and libf2g). I think that there *may* be a danger of strange behaviour in some other places if we just resort to drop in f2c-libraries in g77, and at worst, this behaviour remains undetected. I think you mean libg2c.a. That''s not what --with-f2c does though: it forces the installation routines to translate Fortran to C and then use the C compiler on them. That is, it uses f2c not just its libraries. However, I still don''t think this is a good idea. f2c + C accepts lots of code that is not valid Fortran, so in particular is a bad idea for anyone developing code. And f2c makes lots of assumptions about how the Fortran compiler should operate that are not always true of the real compiler. (Trailing underlines, passing character strings in C style, for two.) More generally, a real Fortran compiler should generate better code (and on the main platforms used for scientific computing the Fortran compilers have had enormous effort expended on them, whereas g77 seems a poor cousin to gcc).> Not all Fortran fails. For instance, MASS (or all of VR_*) loads > (and runs) beautifully, although it uses many Fortran routines.MASS does not use Fortran on R. The S-PLUS versions do, but I donated the relevant code to modreg in R. (And VR_* would not be valid Fortran symbols.) The problem here is that it appears that the Fortran system on (some?) RH6.2 installations is broken. My reaction whenever that happens on Linux (far too frequently to be comfortable) is to get the latest version, so I would suggest using gcc 2.95.2 which has been around and stable for quite a while. (BTW, g77 in egcs-1.1.2 aka 2.91.66 had problems on both Solaris and Windows with R, so I was relieved to its passing.) -- Brian D. Ripley, ripley at stats.ox.ac.uk Professor of Applied Statistics, http://www.stats.ox.ac.uk/~ripley/ University of Oxford, Tel: +44 1865 272861 (self) 1 South Parks Road, +44 1865 272860 (secr) Oxford OX1 3TG, UK Fax: +44 1865 272595 -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html Send "info", "help", or "[un]subscribe" (in the "body", not the subject !) To: r-help-request at stat.math.ethz.ch _._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
On Fri, May 26, 2000 at 07:25:50AM +0100, Prof Brian D Ripley wrote:> The problem here is that it appears that the Fortran system on (some?) > RH6.2 installations is broken. My reaction whenever that happens on Linux > (far too frequently to be comfortable) is to get the latest version, so ILinux != Red Hat Don''t trash all of Linux only because some dists don''t get development systems right. Trash those dists, and recommend one that works. -- According to the latest figures, 43% of all statistics are totally worthless. -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html Send "info", "help", or "[un]subscribe" (in the "body", not the subject !) To: r-help-request at stat.math.ethz.ch _._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
I would like first to thank Prof. Ripley for his hint. R 1.01 built from the source and all contrib packages work fine when using gcc-2.95.x Too bad RedHat stopped shipping it along ecgs since 5.2. Two questions come from all the comments: - is f2c unsafe? I was not able to find any comment in the R distribution about what I just learned from Prof Ripley and Oksanen. This would be certainly an enlightening admission. - My earlier installation of R with f2c produces no error when checking with "make check". Can I conlcude that I had a safe installation? Thanks again for any help. Christian ----- Original Message ----- From: "Prof Brian D Ripley" <ripley at stats.ox.ac.uk> To: "Jari Oksanen" <jhoksane at ecology.helsinki.fi> Cc: "Christian Posse" <posse at talariainc.com>; "r-help" <r-help at stat.math.ethz.ch> Sent: Thursday, May 25, 2000 11:25 PM Subject: Re: [R] problem on upgrading to RH6.2 (One solution)> On Fri, 26 May 2000, Jari Oksanen wrote: > > > > > posse at talariainc.com said: > > > I finally found a way around though the underlying problem persists. > > > It seems that the problem is with g77. Only the packages using FORTRAN > > > failed to be loaded correctly in either way (R from rpm or built from > > > scratch). > > > > > The solution was to build R from the source using the flag > > > > > --with-f2c > > > > > with ./configure. > > > > > All contrib packages installed with this version of R are loaded > > > correctly and, as far as I can superficially see, work. make check > > > gives positive feedback. > > > > I am not sure this is safe. At least I was warned when I used a > similar shortcut some years ago when I had some trouble with g77 > on MS-DOS (djgpp port). g77 used to apply f2c libraries in her > childhood, but has grown independent since then and so the libraries > forked (libf2c and libf2g). I think that there *may* be a danger of > strange behaviour in some other places if we just resort to drop > in f2c-libraries in g77, and at worst, this behaviour remains undetected. > > I think you mean libg2c.a. That''s not what --with-f2c does though: > it forces the installation routines to translate Fortran to C and then use > the C compiler on them. That is, it uses f2c not just its libraries. > > However, I still don''t think this is a good idea. f2c + C accepts lots of > code that is not valid Fortran, so in particular is a bad idea for anyone > developing code. And f2c makes lots of assumptions about how the Fortran > compiler should operate that are not always true of the real compiler. > (Trailing underlines, passing character strings in C style, for two.) > More generally, a real Fortran compiler should generate better code (andon> the main platforms used for scientific computing the Fortran compilershave> had enormous effort expended on them, whereas g77 seems a poor cousin to > gcc). > > > Not all Fortran fails. For instance, MASS (or all of VR_*) loads > > (and runs) beautifully, although it uses many Fortran routines. > > MASS does not use Fortran on R. The S-PLUS versions do, but I donated the > relevant code to modreg in R. (And VR_* would not be valid Fortran > symbols.) > > The problem here is that it appears that the Fortran system on (some?) > RH6.2 installations is broken. My reaction whenever that happens on Linux > (far too frequently to be comfortable) is to get the latest version, so I > would suggest using gcc 2.95.2 which has been around and stable for quitea> while. (BTW, g77 in egcs-1.1.2 aka 2.91.66 had problems on both Solaris > and Windows with R, so I was relieved to its passing.) > > -- > Brian D. Ripley, ripley at stats.ox.ac.uk > Professor of Applied Statistics, http://www.stats.ox.ac.uk/~ripley/ > University of Oxford, Tel: +44 1865 272861 (self) > 1 South Parks Road, +44 1865 272860 (secr) > Oxford OX1 3TG, UK Fax: +44 1865 272595 > > -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-> r-help mailing list -- Readhttp://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html> Send "info", "help", or "[un]subscribe" > (in the "body", not the subject !) To: r-help-request at stat.math.ethz.ch >_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._. _._ -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html Send "info", "help", or "[un]subscribe" (in the "body", not the subject !) To: r-help-request at stat.math.ethz.ch _._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
On Fri, 26 May 2000, Christian Posse wrote:> I would like first to thank Prof. Ripley for his hint. > > R 1.01 built from the source and all contrib packages > work fine when using gcc-2.95.x Too bad RedHat > stopped shipping it along ecgs since 5.2. > > Two questions come from all the comments: > > - is f2c unsafe? I was not able to find any comment > in the R distribution about what I just learned from > Prof Ripley and Oksanen. This would be certainly > an enlightening admission.I think it is slightly unsafe: see below.> - My earlier installation of R with f2c produces > no error when checking with "make check". > Can I conlcude that I had a safe installation?Probably. Unfortunately few things in make check test that you get the right answer, just that things ran. A few do, hence the reports on R-bugs of errors in example(eigen) on some machines. The problem is that there are so many machine-accuracy points that it is hard to write tests that will pass on all reasonable installations and fail on all problematic ones. (We, especially Martin Maechler, do try.) My worry is as follows. If you configure R with f2c, it uses f2c for all the contributed/local packages. We have seen examples of packages that compiled and ran under f2c even though the source code was not valid Fortran and did not do what its author intended when translated. -- Brian D. Ripley, ripley at stats.ox.ac.uk Professor of Applied Statistics, http://www.stats.ox.ac.uk/~ripley/ University of Oxford, Tel: +44 1865 272861 (self) 1 South Parks Road, +44 1865 272860 (secr) Oxford OX1 3TG, UK Fax: +44 1865 272595 -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html Send "info", "help", or "[un]subscribe" (in the "body", not the subject !) To: r-help-request at stat.math.ethz.ch _._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._