gcc, etc for 64 bit Suse 10.0 Linux Yast Conflicts
 

This is the conflict report I recieved from Yast and a copy of the mirror sites being used. Any suggestions on where I can find the files or libraries?

#### YaST2 conflicts list - generated 2005-11-16 12:32:47 ####

gcc-g77 3.3.5-5 conflict
Unresolved Requirements:
gcc-g77 requires gcc = 3.3.5-5
Conflict Resolution:
( ) Do Not Install gcc-g77
( ) Ignore Conflict and Risk System Inconsistencies
gcc not available
Required by:
gcc-c++ requires gcc = 4.0.2_20050901-3
gcc-c++ requires gcc = 4.0.2_20050901-3
gcc-g77 requires gcc = 3.3.5-5
libstdc++-devel requires gcc
libstdc++-devel requires gcc
Conflict Resolution:
( ) Remove the Referring Package
Do Not Install gcc-g77
( ) Ignore Conflict and Risk System Inconsistencies

#### YaST2 conflicts list END ###

Installation Sources:
[x] SUSE LINUX Version 10.0 (cd:///;devices=/dev/hdc)
[x]* dvd://// (dvd:///)
[x] SUSE LINUX Version 10.0 (ftp://mirrors.kernel.org//suse/x86_6...x10.0-GM-Extra)
[x]* dir:///home/jnistler/ (dir:///home/jnistler)
[x] SUSE LINUX Version 9.3 (ftp://mirror.mcs.anl.gov/pub/suse/i386/9.3/)
[x] contributions by Robert Schiele (http://pi3.informatik.uni-mannheim.d...iele/suse/10.0)
[x] guru (SUSE LINUX 10.0) (http://ftp.gwdg.de/pub/linux/misc/suser-guru/rpm/10.0)

What is gcc-g77? Do you need it?

gcc-g77 is a fortran compiler. I just found out that it is incorporated in gcc 3.5 and above.

Munichtexan

PS I have been in st petersburg about 3 times working with a software company. What do you do there? email: jnistler@yahoo.com

As you can see, I use Debian. In my distro there are two packages (amongst
others:)): gcc-3.3 and gcc-4.0. They are co-existing on my system without any
conflict. I am confused why does your distro do not. Some packages requires for
gcc-3.3 and some - gcc-4.0, therefore I have both gcc's installed. Which gcc
is in usage is determined by the symlink /usr/bin/gcc. It points to the actual
compiler.

It was my understanding also in regards to the symlink. The technical programs that I am using have used Matlab scripts for visualization, etc. I was originally exploring setting up Octave because it was my understanding that it has the best correlation to the Matlab script language. But with the trouble I am having in setting it up, I might as well see about setting up Mathlab instead.

One thing I still have confusion on in regards to Yast is that the 64 iso requires a DVD burn. Yast continues to ask for CD1 on //dev/hdc. I am not sure if this is a Yast set up issue, a fix required or what since this is also my first time to use Yast. I have a copy of the Suse 10.0 manual now and will read though it as I go along, but not sure why it keeps asking me for this media when I have a DVD.

I also set up a directory on my harddrive which has the necessary gccg77 and related libraries necessary. The source was accepted by Yast, but it does not appear to look at this directory. I am halfway tempted to just relearn Linux (last heavy use was prior 1995) and do everything manually.

Best regards,
Munichtexan

One acquaintance of mine use SuSe 9.3 and I saw by my own eyes how he installed
packages from the hard disk without DVD inserted. He used YaST.
I myself cannot help this question for I am using Debian, therefore no YaST.

I recognised that you have a 9.3 source on your list. Maybe you accidentally installed a package from this source and this is causing the conflict.

Thanks, gave up for now trying to use YAST. Too many problems with it or so it seems. Read through the first pages of the manual regarding Yast, but does not seem to answer the issues.

Anyway it is getting late and I have finally come up (after lots of review) where I was really trying to head with all of this work. An equivalent open source version of Matlab and a good location for download. I picked up the binary and source for SciLab. I loaded the binary (which shows my age - I know what a binary is, but rpm? :-) I WAS HOPING TO GET OFF EASY, but no way. Got to line 42 when it kicked out for not having "scilex" set up. So tomorrow, back to ./configure. Have a good night.

Actually, YaST ist there to avoid problems, not causing them. I am pretty sure that your problem is caused by the 'old' YaST source. Using 9.3 packages on a 10.0 system will not work!

But doing it old-school is most fun anyway ;)

Hope you will resolve the problem following abisko00's suggestion.
Just visited www.scilab.org
Yes, it provides a gzipped binary version. It is nonsense indeed. Why do not
they post out packages??? Strange thing. Also it comes with the linux-i686
prefix. I am not sure it is compatible with AMD 64 (is it your machine?). If
you are ready to compile it from source you should download source:
http://www.scilab.org/download/stabl...1.1-src.tar.gz
But did you looked at octave and gnuplot? They both works pretty good and I am
sure you can download RPMs without any problems.

Yes, I considered Octave and gnuplot. I am working with a university on specialty code and many of the visualizations, analysis, etc. have been written in Matlab scripts. In the long run, from what I have reviewed the GUI and other aspects of SciLab should meet my criteria, but no guarantees. Plus I worked on setting up Octave and finally gave up because I did not understand what was going on. Now it is a little more understandable.

Octave and SciLab both request g77 during compiling. This is the conflict that shows up in YAST during setup of Octave.

I know that g77 (fortran compiler) is now part of gcc (after 3.5 version) and so far from my reading it indicates that I can build g77 directly from a gcc source. I have gcc and g++ set up in /usr/bin. I am now trying to figure out how to build g77. Unfortunately YAST is not set up to do it correctly, so I need information on how to do it.

I also already have a g77 rpm, I tried to get YAST to recognize a location on my disk as a source. It accepted it as a source, but does not see the g77 rpm. I can use KIIIB to burn it on to DVD, but I have not done it yet, because unlike Nero 6.0 which allows you to append onto existing data, KIIIB has issues if data is on the DVD. I hate to waste a DVD on one RPM, but if I can not figure out how otherwise, I will do it.

EDIT: Had to have 9.3 mirror in order to use YAST to set up g77. While 9.3 mirror was disabled I installed g++ and gcc so I do not know if there is a conflict installing these two items if a 9.3 and 10.0 mirror are both on Yast source.

Best regards,
MunichTexan

From octave package description:
May be problem was with gnuplot, but there is also octave-plplot package which:
I did not used it, but if Octave "mostly Matlab (R) compatible", then probably
octave-plplot more relates to Matlab (R) than gnuplot. I hope you understand me
right, I never used graphics in gnuplot from scripts. Only simple 2D of simple
functions. It is enough for my student business. I merely offer variants.
I used Octave (while I did not find Maxima) but I did not even try to compile
it. Why do you want to compile rather then download an RPM? It seems you do not
image it very clear what can happen when you compile programs instead of
installing RPMs. It is a kind of a dangerous decision. I hope to explain it in
my next post.
YaST is not the only way to install packages. If you know precisely the RPM's
file location, type 'rpm -ivh rpm_file_location' and either the package will be
installed or rpm will output alot of information about occured errors.

YGREX,
Thanks for the input. I did not find an RPM for Scilab, only a binary file setup. I installed this originally on my system and failed at line 43 for "scilex" not existing. If you happen to know of an rpm location for SciLab I would be more in happy to just install the rpm :-)

Your input on Octave is well taken. Like I said earlier, just started exploring with these different programs. I have no real preference other than the GUI and script compatibility to Matlab would be nicer in the long run. Always willing to hear other peoples experts. Not an expert by any stretch of the imagination on these programs. :-)

Munichtexan

OK: ftp://rpmfind.net/linux/powertools/6...4.1-4.i386.rpm
But it is old, the latest is 3.0-14.

OK. Looking at the rpmfind.net site for SciLab. Went ahead and set up Octave on my computer in the mean time and took a first look. A lot to it to review. But like I had understood earlier is primarily command line. Does not make it bad, but the viewing part of this work I would prefer not to spend as much time on. The optical microlithography simulations in relation to electrical design is difficult enough as far as set up