[ccp4bb] YSBL server
Dear all Since last announcement the number of usage of YSBL server was more than our server could cope. That is why we had to close it for few days and upgrade the sfotware on the server. Now YSBL programs server is open again and it should be able to cope (at least for a while). All jobs submitted are still available and you can take the results. After three days all old jobs will be deleted. The address of the server is: www.ysbl.york.ac.uk/YSBLPrograms/index.jsp Please contact Fei ( f...@ysbl.york.ac.uk ) if you have any problems With best regards Fei Long Alexei Vagin Garib Murhsudov P.S. The list of the programs in the current version of the server include automatic molecular replacement - BALBES, space group correction - zanuda, automatic model building - buccaneer
[ccp4bb] ccp4i, OS X 10.5.7, security enforcement
Hello All,
It seems that sh has a variable that exists and doesn't exist at the
same time (see my shell session below). Isn't such transition between
existence and non-existence reserved for quantum mechanics and eastern
philosophy? My guess is that sh has become more vigilant about
security and the exec statement. Can anyone venture a better guess as
to how the output below might happen? What's the best way around
this? But the real question is: Can the people at ccp4 look at this
and make changes to the code base before these kinds of security
features become more commonplace (for the good reason to prevent
injection attacks)?
For example: http://www.theregister.co.uk/2009/07/21/critical_ddwrt_router_vuln/
Here is my shell session:
chernev 10% cat `which ccp4i`
#!/bin/sh
# Start ccp4i interface
# \
echo CCP4I_TCLTK is $CCP4I_TCLTK
exec ${CCP4I_TCLTK}/bltwish $0 -- ${1+$@}
source [file join $env(CCP4I_TOP) bin ccp4i.tcl]
chernev 11%
chernev 12% ccp4i
CCP4I_TCLTK is /usr/local/X11/bin
Error in startup script: can't read CCP4I_TCLTK: no such variable
while executing
exec ${CCP4I_TCLTK}/bltwish $0 -- ${1+$@}
(file /usr/local/ccp4-6.1.1/bin/ccp4i line 5)
Re: [ccp4bb] ccp4i, OS X 10.5.7, security enforcement
When I do this with /bin/zsh, it does the right thing. The source command
in the bltwish shell gives me ccp4i.
When I run the commands interactively in bash or sh (which is bash on os
x), the command
exec ${CCP4I_TCLTK}/bltwish $0 -- ${1+$@}
terminates the shell session instead of turning it into a bltwish shell.
So I think your process is getting killed before you look at the cat in
the box. If it is a question of ${CCP4I_TCLTK} not getting assigned, the
error you should see looks like this:
bash: /bltwish: No such file or directory
bash: exec: /bltwish: cannot execute: No such file or directory
or
sh: /bltwish: No such file or directory
sh: exec: /bltwish: cannot execute: No such file or directory
Likewise, with quantum mechanics, the mysteries disappear if you allow
that the probability function is an objective statement about where a
particle might be located, rather than a subjective statement about our a
priori knowledge of the particle's location (i.e., you reject
wave-particle duality in favor of a particle interpretation).
On Sun, July 26, 2009 2:34 pm, James Stroud wrote:
Hello All,
It seems that sh has a variable that exists and doesn't exist at the
same time (see my shell session below). Isn't such transition between
existence and non-existence reserved for quantum mechanics and eastern
philosophy? My guess is that sh has become more vigilant about
security and the exec statement. Can anyone venture a better guess as
to how the output below might happen? What's the best way around
this? But the real question is: Can the people at ccp4 look at this
and make changes to the code base before these kinds of security
features become more commonplace (for the good reason to prevent
injection attacks)?
For example:
http://www.theregister.co.uk/2009/07/21/critical_ddwrt_router_vuln/
Here is my shell session:
chernev 10% cat `which ccp4i`
#!/bin/sh
# Start ccp4i interface
# \
echo CCP4I_TCLTK is $CCP4I_TCLTK
exec ${CCP4I_TCLTK}/bltwish $0 -- ${1+$@}
source [file join $env(CCP4I_TOP) bin ccp4i.tcl]
chernev 11%
chernev 12% ccp4i
CCP4I_TCLTK is /usr/local/X11/bin
Error in startup script: can't read CCP4I_TCLTK: no such variable
while executing
exec ${CCP4I_TCLTK}/bltwish $0 -- ${1+$@}
(file /usr/local/ccp4-6.1.1/bin/ccp4i line 5)
William G. Scott
Contact info:
http://chemistry.ucsc.edu/~wgscott/
[ccp4bb] output from DM
Dear CCP4Â’ers, This is a follow up of my previous query on density modification. I want to know if I need to use the phases (the *.mtz file) that are output from DM into subsequent cycles of Refmac or to use the original data file for refinement and model building. Thank you Kumar
Re: [ccp4bb] ccp4i, OS X 10.5.7, security enforcement
William Scott wrote:
..
So I think your process is getting killed before you look at the cat in
the box. If it is a question of ${CCP4I_TCLTK} not getting assigned, the
error you should see looks like this:
bash: /bltwish: No such file or directory
bash: exec: /bltwish: cannot execute: No such file or directory
It seems bash treats undefined variables as the null string:
bash-3.2$ echo ${wxyz}/abc
/abc
while csh prints an error message, but not like the OP's:
oswego 185% echo ${wxyz}/abc
wxyz: Undefined variable.
apparently tcl gives the error message reported:
oswego 187% tclsh
% printf ${wxyz}/abc
can't read wxyz: no such variable
So I think the exec is successful but some ensuing tcl script is
not inheriting the environment.
Re: [ccp4bb] ccp4i, OS X 10.5.7, security enforcement
On Sun, July 26, 2009 5:38 pm, Edward A. Berry wrote:
William Scott wrote:
..
So I think your process is getting killed before you look at the cat in
the box. If it is a question of ${CCP4I_TCLTK} not getting assigned,
the
error you should see looks like this:
bash: /bltwish: No such file or directory
bash: exec: /bltwish: cannot execute: No such file or directory
It seems bash treats undefined variables as the null string:
bash-3.2$ echo ${wxyz}/abc
/abc
while csh prints an error message, but not like the OP's:
oswego 185% echo ${wxyz}/abc
wxyz: Undefined variable.
apparently tcl gives the error message reported:
oswego 187% tclsh
% printf ${wxyz}/abc
can't read wxyz: no such variable
So I think the exec is successful but some ensuing tcl script is
not inheriting the environment.
The exec is working both with bash/sh and zsh examples, but the string
source [file join $env(CCP4I_TOP) bin ccp4i.tcl] isn't getting passed
into the wish shell in any of these. The script as written doesn't work.
Maybe you need a here string. If you run the bash/sh example without the
$0 -- ${1+$@} stuff, and then paste in the source string manually
after the prompt, it does what he wants (i.e., works).
Re: [ccp4bb] ccp4i, OS X 10.5.7, security enforcement
On Sun, July 26, 2009 6:07 pm, William Scott wrote:
On Sun, July 26, 2009 5:38 pm, Edward A. Berry wrote:
William Scott wrote:
..
So I think your process is getting killed before you look at the cat in
the box. If it is a question of ${CCP4I_TCLTK} not getting assigned,
the
error you should see looks like this:
bash: /bltwish: No such file or directory
bash: exec: /bltwish: cannot execute: No such file or directory
It seems bash treats undefined variables as the null string:
bash-3.2$ echo ${wxyz}/abc
/abc
while csh prints an error message, but not like the OP's:
oswego 185% echo ${wxyz}/abc
wxyz: Undefined variable.
apparently tcl gives the error message reported:
oswego 187% tclsh
% printf ${wxyz}/abc
can't read wxyz: no such variable
So I think the exec is successful but some ensuing tcl script is
not inheriting the environment.
The exec is working both with bash/sh and zsh examples, but the string
source [file join $env(CCP4I_TOP) bin ccp4i.tcl] isn't getting passed
into the wish shell in any of these. The script as written doesn't work.
Maybe you need a here string. If you run the bash/sh example without the
$0 -- ${1+$@} stuff, and then paste in the source string manually
after the prompt, it does what he wants (i.e., works).
I think the trick is to put single quotes around the here string, eg, in
zsh, this will work:
exec ${CCP4I_TCLTK}/bltwish 'source [file join $env(CCP4I_TOP) bin
ccp4i.tcl]'
By the way, why do all this instead of just issuing the command ccp4i ?
Re: [ccp4bb] Problem with different map when ligand refinement
Dear TriNigo I would put waters in all the other difference density first (there are several obvious waters even in the section of map you sent with your email), without the substrate, and see if you see any more clues as to what's going on. Do you expect something to be added on the substrate in the reaction? Could you be looking at product? Regards Christine Dear ccp4 experts, I am trying to solve the structure of one esterase with its substrate. The structure was solved by MR to the resolution 2.2 A. In the active site, I can see clearly the different density map of the substrate (S- methyl hydroxy isobutyrate). After puting the substrate into that density map, I use Refmac to run restrain refinement at 0.1 weighting term. Even though the substrate fits to the different map very well, a big different map appears at one end of the substrate after Refmac refinement. I attach the image so that it's easy to see the problem. I tried to put water or Na (because the crystallization contains: 1M sodium citrate, 100mM Na Cacodylate, 20% glycerol) but it didn't work. I tried to use CNS instead of Refmac, however the result is worse since the R-free is increased and the different map is still there. Because I am new with CNS so I don't know which parameters I should adjust. Do you have any suggestion to solve this problem? Thank you very much for all of your helpful advice! My best regards, TriNgo, PhD student, Sungkyunkwan University, Koreaesterase_mut_S_ligand.png
