This archive contains answers to questions sent to Unidata support through mid-2025. Note that the archive is no longer being updated. We provide the archive for reference; many of the answers presented here remain technically correct, even if somewhat outdated. For the most up-to-date information on the use of NSF Unidata software and data services, please consult the Software Documentation first.
>From: weather <address@hidden> >Organization: NMSU/NSBF >Keywords: 199905012048.OAA16960 McIDAS-X mcfc tclcomp Robert, >I know Tom appreciates an occasional success story as a change of pace. Indeed I do :-). >I got rid of Linux on wxfts.nsbf.nasa.gov, a dual-boot NT machine >that is a kind of back-up/development machine when not being used >as an NT pc. I installed Solaris 7 on it (166Mhz 80MB RAM, slow IDE >drive) and obtained gcc. How would you contrast the performance of Linux versus Solaris x86 on this platform? I have found that RedHat Linux 5.2 on my 200 Mhz Pentium PC is much more responsive than Solaris x86 on the same box. It also uses a lot less system resources. >I obtained latest McIDAS distribution and the >solx86 f2c files(and installed them) and followed instructions. Your comment "and followed instructions" is _so_ refreshing ;-) >McIDAS and XCD completed the compile, but when I went to start McGUI, >I found that mcwish had not been produced. I went back in the makelog >and found an error (bad option -R) that I had seen while perusing the >support archives. I made the change that Tom suggested in tclcomp >in the support message and did remade mcwish. The build finished >and I installed and configured it. I installed the ADDE server >and am getting data from wxmcidas (on the other side of the room) >via the ADDE server. Everything works great! I am going to put out an Addendum either today or tomorrow that includes this mod. The problem came up when I released SSEC's mcfc.sh in a previous Addendum and didn't fully test it on Solaris x86 (the build with the Sun compilers worked just fine). >Also a follow-up to a problem I had where the XCD decoders are not >producing boundary layer (30 MBAG) winds from the AVN/MRF, whereas >GEMPAK is. I believe that Don contacted SSEC about this problem. >I am not rushing for an answer, but if you think a resolution >might be far away, I can tell you that it is not imminent. >I will go ahead and produce the charts via cron jobs >using GEMPAK rather than CGI scripts and McIDAS(I have had problems with >gplt and gpend and CGI scripts), as I need to finish the web page up this >week. Sorry I didn't get to this earlier. I was trying to cleanup a bunch of stuff and then we had a tragedy in Unidata. >Thanks, Thanks for sending in a success story. It really is appreciated! Tom From: weather <address@hidden> Date: Wed, 5 May 1999 18:03:05 +0000 (GMT) Thanks Tom, I would agree that Linux running the standard window manager (FVWM2, or whatever the one is that looks similar to Windows 95) is faster than Solaris. However, when I was running KDE, the free semi-clone of CDE, I found that the two were a lot closer in terms of just using the machine. I just find Solaris more user friendly (for a new Unix user), I like CDE better than KDE or any other "free" window manager, the Solaris Answerbook has been tremendously helpful. There is a lot of info and help for Linux, but it is very hard to pull it together from all of the various sources. For us, of course, since we are a heavy GARP user, Linux is not an option because of having to purchase Motif separately. However, it would be interesting if I had another machine like wxmcidas (400Mhz PII 256MB of RAM) to load Linux on it and load it up like wxmcidas is loaded. I was sorry to hear about Glenn Davis. I did meet him once at the workshop last year, and he was very helpful and cordial as I waited around after arriving early. Robert Mullenax NSBF Meteorology