[Gabedit-users] (no subject)
Brought to you by:
allouchear
From: Emilio J. J. P. <ejj...@ic...> - 2008-10-26 14:25:59
|
Hello Abdul-Rahman, First at all, thank you for gabedit, it is nice and useful. > This is probably due to fact that your LOGNAME is not defined at your > remote > host computer (this variable is used by Gaussian). > Can you execute the following command at your LOCAL computer : > ssh yourRemoteHost 'echo $LOGNAME' > ssh yourRemoteHost 'ps' > and send me the error messages. Yes. The local computer (my computer) have different LOGNAME that the two remote hosts that I am using: for example, now I am using remotehost1 for run "remote process"(no batch process, remotehost1 is a old linux computer) with mopac2007. Here, there is no problem, no problem sending files, no problem with 'ps', no problem retrieving files...definetively gabedit working fantastic. Remotehost1 has logname1 (normal user privilegies, no root). In the other hand, I am using remotehost2 (a cluster of "itaniums" or some related spaceship) for run "batch process" with gaussian03. Logname2 is the logname for remotehost2, obviously logname2 has normal user privilegies. There is no problem sending files, or making chown... Only fails the last step: running the example.cmd file. I think that it is not a problem of LOGNAME because I fill the field "login" as correspond and gabedit can send the files to remotehost2. Curiously .cmd file for a calculation in hostname1 is stored in the /home/logname1 folder. The associated example.mop is stored in the working directory. /home/logname1/.../tmp. However, in the case of the gaussian03 calculation in hostname2.the .cmd file is stored in the same level (working directory) as .com file. Another difference it is the .cmd file for a gaussian calculation in hostname2 *************************** cd /home/logname2 g03-exe 1 1 NoName.com & exit ************************** where "g03-exe n p" is a bash script with the name of the queue, n: number of nodes,p: number of processors ... and other options. and it is the .cmd file for mopac calculation in hostname1 ***************************** #!/bin/sh cd /home/logname1/tmp mopac NoName.mop & exit ***************************** is it possible that the error comes from the first line: #!/bin/sh ? > You cannot change the bond types in Gabedit(presently this option is > not > implemented). > The bond types are calculated by Gabedit using the valence number of > electrons. > However : > 1) you can change the valence number of electrons of any atoms > (set/Properties > of atoms) > 2) you can activate the "not multiple bonds" by Render/Show multiple > bonds > > Noting that the bond type is not used on the quantum chemistry > calculation. The > result of your Gaussian (or other programm) is not affected by the > bond types > showed by Gabedit > Ok, I understand. I changed the "maximum bond valence" of carbon (and boron) to 6 to draw these cobaltabisdicarbollides: http://www.emiliojuarez.es/watersolubledend.html and now I see double bonds where it not correspond. However I dont find the option "not multiple bonds" by Render/Show multiple bonds. And I made the before povrays with "scale stick" option and after I modified the weidth of the cylinder in the povray file. Thank you for gabedit, sincerely Emilio José Juárez Pérez PhD Student ICMAB-CSIC Campus UAB 08193 Bellaterra (Barcelona). Spain ejj...@ic... Tel. + 34 93 580 1853 (ext. 255) Fax + 34 93 580 5729 |