nutjunkie / iqmol Goto Github PK

Hi,

I'm unable to submit jobs to the default QChem sever using the lastest pre-compiled 2.15.3bs dmg for Mac; when I try to submit a job, I receive a dialogue telling me that "No queues found on server" - another dialog then pops up, which prompts me to define resource limits for a queue, but no options are given in the drop-down for Queue.

I took a look at the server configurations via the Calculation -> Edit Servers menu, and nothing there looks out-of-place to me; when I tried to swap over to HTTPS on port 443 or port 80, I get SSL handshake failed errors. Do you think this could just be a server configuration issue that I can solve locally?

This is a fresh install of IQmol 2.1.5.3bs downloaded from the official webpage and running on MacOS 11.5.1. Let me know if you need any additional information.

Hi, I am just running a little summer course where I wanted the students to run a few jobs via IQmol and the Q-Chem server.

In IQmol 2.14, I get an error message once I click submit via the Q-Chem server:
Job submission failed: Unknown error

I took a quick look at the recent commits and it seems that the problem lies in [18f457a]. I can fix the problem in IQmol if I change the server name back to QChem rather than Q-Chem. Don't know why this would make a difference but it seems to on my computer.

Cheers,
Felix

formaldehyde-eom-test.zip
Trying to learn how to setup files for QChem so I'm using IQmol to assist.
I set up an EOM-CCSD single point for formaldehyde. The SCF finishes, but the EOM doesn't go anywhere. There also does not appear to be an error printed in the output file. Attached are the files generated from IQMol.

I'd like to add the carbonyl ligand fragment (-C≡O) to IQmol, which I assume is as easy as adding a corresponding .xyz file in the fragments/Functional_Groups directory.

However, it appears a .png is also needed, and I couldn't find any documentation on how to generate a suitable one.

CMake Error at src/CMakeLists.txt:544 (qt4_wrap_cpp):
  Unknown CMake command "qt4_wrap_cpp".

Please switch to Qt5 because Qt4 is EOLed.

FreeBSD 11.2

some files in share/fragments/Molecules/Alcohols/, for example Ethelene_Glycol.xyz, Glycerol.xyz, etc. has no name inside, but oher have. Is this a bug or a fearure?

how to setting iqmol to connect on qchem local server for windows 10
my qhem installation is in C:\QChem\5.3.2

Our system is Ubuntu 18.04.
We are running Q-chem on local computer.
After I set up our sever setting in IQmol, then run the calculations, it shows permission problem.
Basically when we create a file in linux, the permission always have to be 666 (if umask is 000).
The file which was created from IQmol (".run" file) have no executable permission.
So we can never execute this script file (".run file").
Any solution for this Issue?

I want to make calculations with plain molecules, but it can't be done easily with Descartes coordinates. It would be better if I could generate z-matrix input with iqmol.

This is my proposal for iqmol
https://steemit.com/utopian-io/@upvotefree/design-logo-for-iqmol

Loading an fchk file on a vibrational frequency job does not load the frequencies. I don't know if this is an existing limitation in the fchk format or if it's just our implementation of it.

Is there a plan in place in brining IQmol to Ubuntu 22? I am having problems with installing the binary file since Qt5 is not installed/supported. Building from source can not find openbabel even when it’s installed from source, apt, and snap.

My workaround for now is to use IQmol under wine but that has its own set of problems.

Is there a way we can get IQmol running natively on Ubuntu 22?

Thanks so much in advance for your help!!

Hello, I tried setting up a TDDFT calculation in IQmol.

There is a box Advanced - Excited States - CIS. This allows to change a number of algorithmic settings. But it does not allow me to trigger the actual CIS calculation, which would need the keywords

   CIS_N_ROOTS  =  4
   CIS_SINGLETS  =  1
   CIS_TRIPLETS  =  1

So, I would still have to add these manually?

Thanks,
Felix

Source code is

bool click((m_rectangle.width() < 3) && (m_rectangle.width() < 3));

Maybe

bool click((m_rectangle.width() < 3) && (m_rectangle.height() < 3));

would be better code.

Trying to compile with cmake seems to cause a lot of issues, mainly with QT. They seem to constantly mix qt4 and qt5, and it is possible that they might work with a combination of qt 4 and 5 (not good). Some other ones seem to have packages that have a different name on my qt5 such as this one from CMakeLists.txt (notice the prefix Qt is not needed anymore)

-find_package(Qt5 COMPONENTS QtCore QtGui QtNetwork QtOpenGL QtSql QtXml
+find_package(Qt5 COMPONENTS Core Gui Network OpenGL Sql Xml

On the same file, it also has "include(${QT_USE_FILE})", which is not used in QT5 anymore, even though the find_package is using qt5. Fixing these two still didn't work since most of the CMakeLists are still using qt4 and qt5 interchangeably. It might be a good idea to tell other developers to avoid using the cmake for now and stick with the qmake+make compilation, to avoid late-night tear sheds and frustration.

Hello,everyone
I use qchem to generate a .fchk file.For now I wanna know the point group directly，and know the irreduable representation of HOMO or LUMO orbitals，like Ag or Bu something like that.
I use the "symmetize structure",but couldn't see any clue here.
Please, could any one tell me what to do?

configure fails:

CMake Error at src/CMakeLists.txt:588 (add_library):
  Cannot find source file:

    Qui/QCJob.C

Did you forget to add this file?

Env: Windows 7 x64 SP 1, Domain Bound Machine.
Bug Summary: uninstaller does not completely remove IQmol 2.8.0 from machine when running with "--mode unattended" flags enabled. "C:\Program Files (x86)\IQmol-2.8.0\bin\IQmol.exe" file remains after uninstaller completes.

Remediation: Remove file & folder structure as final job in installer.

I can not build IQmol from current master because of missing header file Qui/QCJob.h

Hi, I was playing around with NBOs a bit recently. It looks like it is possible to write NBOs to the fchk file with the following input.

$rem
method          hf
basis           cc-pvtz
nbo             2
gui             2
$end

$molecule
0 1
O
H 1 0.95
H 1 0.95 2 104.5
$end

The fchk file contains the following NBO-related labels:

Alpha NBO coefficients                     R   N=         580
Beta NBO coefficients                      R   N=         580
Alpha NBO occupancies                      R   N=          10
Beta NBO occupancies                       R   N=          10
NBO Ground State(/,' Begin NBO analysis for TDA excited state ',i2,/)   I                0

But IQmol only finds the Canonical Orbitals under Surfaces and I am not able to look at the NBOs.

Cheers,
Felix

P.S. There is some more info, on ticket 2244 on jubilee

The development tree of Fedora has Boost 1.59.0, against which IQmol fails to build.

This problem was already reported in issue #8.

The public key MUST be verified before authentification of a client. Without this checking man-in-the-middle attack is possible ("ssh first time connection problem"). There are two common options to do this thing: to enforce user to check a fingerprint of the server's public key manually or to ask a trust public key server. Inserte some code in a function void SshConnection::authenticate(AuthenticationT const authentication, QString& username) from the file src/Network/SshConnection.C, for example, try this one:

// line 241
bool is_fingerprint_issue = false;
size_t len;
int type;
LIBSSH2_KNOWNHOSTS \* knownhosts = libssh2_knownhost_init(m_session);
libssh2_knownhost_readfile(knownhosts, "known_hosts",LIBSSH2_KNOWNHOST_FILE_OPENSSH);
const char \* fingerprint = libssh2_session_hostkey(m_session, &len, &type);
struct libssh2_knownhost \* host_p;
int check = libssh2_knownhost_checkp(knownhosts, hostname, 22, fingerprint, len, LIBSSH2_KNOWNHOST_TYPE_PLAIN | LIBSSH2_KNOWNHOST_KEYENC_BASE64, &host_p);
        switch(check) {
        case LIBSSH2_KNOWNHOST_CHECK_MATCH:
        std::cout << "Server's fingerprint is found in local database\n";
        break;
        case LIBSSH2_KNOWNHOST_CHECK_MISMATCH:
        std::cout << "Keys doesn't match! Possible man-in-the-middle attack!!!\n";
        std::cout << fingerprint;
        is_fingerprint_issue = true;
        break;
        case LIBSSH2_KNOWNHOST_CHECK_NOTFOUND:
        std::cout << "No host match was found .\n";
        is_fingerprint_issue = true;
        break;
        case LIBSSH2_KNOWNHOST_CHECK_FAILURE:
        std::cout << "Something prevent the check to be made.\n";
        is_fingerprint_issue = true;
        break;
        }
libssh2_knownhost_free(knownhosts);

There is an option to reindex atoms under "Edit". I found no way to save the new indices to the .xyz file and ended up manually adjusting the .xyz file in an editor. The manual does not seem to have any documentation on that feature.

Hi,

I have a slight problem compiling IQmol 2.3.0.

First of all, makefile is broken; to fix it
IQmol/SshFileDialog.C

• IQmol/SurfaceAnimatorDialog..C
  IQmol/SurfaceAnimatorDialog.C
  in src/CMakeLists.txt.

Now, IQmol compiles but fails to link with

CMakeFiles/IQmol.dir/Data/DataFactory.C.o: In function IQmol::Data::NmrReference::NmrReference(QString const&)': DataFactory.C:(.text._ZN5IQmol4Data12NmrReferenceC2ERK7QString[_ZN5IQmol4Data12NmrReferenceC5ERK7QString]+0x23): undefined reference tovtable for IQmol::Data::NmrReference'
collect2: error: ld returned 1 exit status

When a user runs a geometry optimization, this structure correctly appears in the molecular viewer. However, if the user attempts to run a subsequent job, the correct coordinates are not translated into the new input file.

For example, let's say a user optimizes the structure in job 1. In job 2, they try to run a harmonic frequency calculation on this optimized structure. The coordinates populated in the input file for job 2 are the initial structure coordinates from job 1, not the optimized coordinates. Even if the user explicitly selects the last entry in the list of energies under the "Geometry" heading, the correct coordinates are still not used.

Hi all.

Cannot compile IQmol-2.15.0 in Fedora 36 (devel branch) against openbabel-3.1.1 with GCC-11.2.1
Please, could you render IQmol compatible to Openbabel3? Thanks.

make[1]: Entering directory '/home/sagitter/rpmbuild/BUILD/IQmol-2.15.0/src/Data'
g++ -c -O2 -flto=auto -ffat-lto-objects -fexceptions -g -grecord-gcc-switches -pipe -Wall -Werror=format-security -Wp,-D_FORTIFY_SOURCE=2 -Wp,-D_GLIBCXX_ASSERTIONS -specs=/usr/lib/rpm/redhat/redhat-hardened-cc1 -fstack-protector-strong -specs=/usr/lib/rpm/redhat/redhat-annobin-cc1 -m64 -mtune=generic -fasynchronous-unwind-tables -fstack-clash-protection -fcf-protection -Wl,-z,relro -Wl,--as-needed -Wl,-z,now -specs=/usr/lib/rpm/redhat/redhat-hardened-ld -I/usr/include/openbabel3 -O2 -g -ggdb -O2 -flto=auto -ffat-lto-objects -fexceptions -g -grecord-gcc-switches -pipe -Wall -Werror=format-security -Wp,-D_FORTIFY_SOURCE=2 -Wp,-D_GLIBCXX_ASSERTIONS -specs=/usr/lib/rpm/redhat/redhat-hardened-cc1 -fstack-protector-strong -specs=/usr/lib/rpm/redhat/redhat-annobin-cc1 -m64 -mtune=generic -fasynchronous-unwind-tables -fstack-clash-protection -fcf-protection -fPIC -Wall -Wextra -D_REENTRANT -DQT_NO_DEBUG -DQT_OPENGL_LIB -DQT_PRINTSUPPORT_LIB -DQT_WIDGETS_LIB -DQT_GUI_LIB -DQT_XML_LIB -DQT_NETWORK_LIB -DQT_CORE_LIB -I. -I/usr/include/openbabel-2.0 -I../../src -I../../build -I. -I../Util -I../Parser -I../Yaml -I../OpenMesh/src -I/usr/include/qt5 -I/usr/include/qt5/QtOpenGL -I/usr/include/qt5/QtPrintSupport -I/usr/include/qt5/QtWidgets -I/usr/include/qt5/QtGui -I/usr/include/qt5/QtXml -I/usr/include/qt5/QtNetwork -I/usr/include/qt5/QtCore -I../../build -I/usr/lib64/qt5/mkspecs/linux-g++ -o ../../build/Atom.o Atom.C
In file included from Serialization.h:29,
                 from Data.h:25,
                 from DataList.h:25,
                 from Atom.h:25,
                 from Atom.C:23:
../../src/QGLViewer/vec.h: In member function 'qglviewer::Vec qglviewer::Vec::unit() const':
../../src/QGLViewer/vec.h:355:26: warning: implicitly-declared 'constexpr qglviewer::Vec::Vec(const qglviewer::Vec&)' is deprecated [-Wdeprecated-copy]
  355 |                 Vec v = *this;
      |                          ^~~~
In file included from Serialization.h:29,
                 from Data.h:25,
                 from DataList.h:25,
                 from Atom.h:25,
                 from Atom.C:23:
../../src/QGLViewer/vec.h:119:14: note: because 'qglviewer::Vec' has user-provided 'qglviewer::Vec& qglviewer::Vec::operator=(const qglviewer::Vec&)'
  119 |         Vec& operator=(const Vec& v)
      |              ^~~~~~~~
In file included from Serialization.h:29,
                 from Data.h:25,
                 from DataList.h:25,
                 from Atom.h:25,
                 from Atom.C:23:
../../src/QGLViewer/vec.h:357:24: warning: implicitly-declared 'constexpr qglviewer::Vec::Vec(const qglviewer::Vec&)' is deprecated [-Wdeprecated-copy]
  357 |                 return v;
      |                        ^
In file included from Serialization.h:29,
                 from Data.h:25,
                 from DataList.h:25,
                 from Atom.h:25,
                 from Atom.C:23:
../../src/QGLViewer/vec.h:119:14: note: because 'qglviewer::Vec' has user-provided 'qglviewer::Vec& qglviewer::Vec::operator=(const qglviewer::Vec&)'
  119 |         Vec& operator=(const Vec& v)
      |              ^~~~~~~~
In file included from DataList.h:25,
                 from Atom.h:25,
                 from Atom.C:23:
Data.h: In member function 'IQmol::Data::Base& IQmol::Data::Base::operator=(const IQmol::Data::Base&)':
Data.h:94:13: warning: this 'if' clause does not guard... [-Wmisleading-indentation]
   94 |             if (this != &that) copy(that);  return *this;
      |             ^~
Data.h:94:45: note: ...this statement, but the latter is misleadingly indented as if it were guarded by the 'if'
   94 |             if (this != &that) copy(that);  return *this;
      |                                             ^~~~~~
In file included from /usr/include/openbabel3/openbabel/mol.h:48,
                 from Atom.C:25:
/usr/include/openbabel3/openbabel/base.h: In member function 'virtual const char* OpenBabel::OBBase::GetTitle(bool) const':
/usr/include/openbabel3/openbabel/base.h:286:49: warning: unused parameter 'replaceNewlines' [-Wunused-parameter]
  286 |       virtual const char  *GetTitle(bool UNUSED(replaceNewlines) = true) const { return "";}
      |                                                 ^
/usr/include/openbabel3/openbabel/base.h:37:20: note: in definition of macro 'UNUSED'
   37 | # define UNUSED(x) x
      |                    ^
Atom.C: In static member function 'static unsigned int IQmol::Data::Atom::atomicNumber(const QString&)':
Atom.C:37:26: error: 'etab' is not a member of 'OpenBabel'
   37 |    unsigned z(OpenBabel::etab.GetAtomicNum(symbol.toLatin1().data()));
      |                          ^~~~
make[1]: *** [Makefile:686: ../../build/Atom.o] Error 1

Build log: https://pastebin.com/A3XLQZFv

I want to copy the molecule (or a piece of that) and paste, so I can build a polymer structure with three or four monomers.

Steps:

1. run the application
2. build a monomer molecule
3. select the monomer (or a piece of that)
4. copy the selection and paste afterwards

actual result: I can not move the pasted monomer, it sits precisely on the original one.
expected result: the selected item should be placed somewhere near the original one on empty space.

It seems to be a bug that does not allow copying previously optimized geometries into FSM calculation. When one geometry is copied, another one becomes overwritten.

I've tried to build with Qt4, but it didn't worked (Debian Jessy, Qt 4.8.6). It seems like there are some functions used from Qt5. In order to build with Qt5 I had to change CMakeLists:

...
find_package(OpenGL REQUIRED)
find_package(Qt5 COMPONENTS Core Gui Network OpenGL Sql Xml
    REQUIRED)
...

I had to comment a lot of depricated (absense) files in src/CMakeLists, because cmake produced an error in case of having this files uncommented:

#Old/BasicServer.C
#Old/JobInfo.C
#Old/LocalConnectionThread.C
#Old/LocalHost.C
#Old/PBSServer.C
#Old/Process.C
#Old/ProcessMonitor.C
#Old/QCProcess.C
#Old/RemoteHost.C
#Old/SecureConnection.C
#Old/SecureConnectionThread.C
#Old/Server.C
#Old/ServerDialog.C
#Old/ServerListDialog.C
#Old/ServerOptionsDialog.C
#Old/ServerQueue.C
#Old/ServerQueueDialog.C
#Old/ServerRegistry.C
#Old/ServerTask.C
#Old/SGEServer.C
#Old/System.C
#Old/ProcessMonitor.ui
#Old/ServerDialog.ui
#Old/ServerListDialog.ui
#Old/ServerOptionsDialog.ui
#Old/ServerQueueDialog.ui
#Util/Matrix.h
#QUI/GeometryConstraintDialog.ui
#QUI/OptionDatabaseForm.ui
#QUI/OptionListEditor.ui
#QUI/OptionNumberEditor.ui
#QUI/QuiMainWindow.ui

There was an incorrect order of compilation: Layer before Configurator. I had to run make munually firstly in src/Configurator, then in src again.

I had to change $LIBS in src/Main/Makefile (this works on my system):

LIBS          = $(SUBLIBS) -L/usr/X11R6/lib64 /home/myself/iqmol/IQmol/src/Main/../../build/libQui.a /home/myself/iqmol/IQmol/src/Main/../../build/libLayer.a /home/myself/iqmol/IQmol/src/Main/../../build/libOld.a /home/myself/iqmol/IQmol/src/Main/../../build/libParser.a /home/myself/iqmol/IQmol/src/Main/../../build/libData.a /home/myself/iqmol/IQmol/src/Main/../../build/libConfigurator.a /home/myself/iqmol/IQmol/src/Main/../../build/libProcess.a /home/myself/iqmol/IQmol/src/Main/../../build/libNetwork.a /home/myself/iqmol/IQmol/src/Main/../../build/libYaml.a /home/myself/iqmol/IQmol/src/Main/../../build/libPlot.a /home/myself/iqmol/IQmol/src/Main/../../build/libUtil.a /usr/lib/x86_64-linux-gnu/libboost_serialization.a /usr/lib/x86_64-linux-gnu/libboost_exception.a /usr/local/lib/libOpenMeshCore.a /usr/local/lib/libOpenMeshTools.a -lz -lopenbabel -L/usr/lib/x86_64-linux-gnu -lssh2 /usr/lib/x86_64-linux-gnu/libcrypto.a -L/usr/lib/x86_64-linux-gnu -lGL -lGLU -L/usr/lib/gcc/x86_64-linux-gnu/4.9 -lgfortran -lQt5PrintSupport -lQt5OpenGL -lQt5Widgets -lQt5Sql -lQt5Network -lQt5Xml -lQt5Gui -lQt5Core -lpthread -L/usr/local/lib -lQGLViewer

I also had to build openmesh lib and qglviewer lib from sources, because the packages in the debian repository are old.

Please, correct config files.
I am using:

Linux version 3.16.0-4-amd64 ([email protected]) (gcc version 4.8.4 (Debian 4.8.4-1) ) gcc version 4.9.2
GNU Make 4.0
cmake 3.0.2
QMake version 3.0
Qt version 5.3.2

I work at a university and want to deploy IQmol to a virtual lab environment on Microsoft Windows RDS (remote desktop services) - the session hosts that manage this are all virtual machines without GPU's. I've noticed I can reproduce this issue on a Windows 10 VM running in VMWare. On launch we get the splash screen, however the program UI never comes up. In the event viewer application log I get a crash:

Faulting application name: IQmol.exe, version: 0.0.0.0, time stamp: 0x5eb2a8cf
Faulting module name: libstdc++-6.dll, version: 0.0.0.0, time stamp: 0x00000000
Exception code: 0x40000015
Fault offset: 0x0002085d
Faulting process id: 0x2604
Faulting application start time: 0x01d691d394c93495
Faulting application path: C:\Program Files (x86)\IQmol-2.14.1\bin\IQmol.exe
Faulting module path: C:\Program Files (x86)\IQmol-2.14.1\bin\libstdc++-6.dll
Report Id: 7d6735a3-2cd6-4739-966a-99f972de1181
Faulting package full name:
Faulting package-relative application ID:

I'm not entirely sure where else the application stores debug level logs, but can provide them if its deemed helpful.

In Main.pro, we should write version information into a header file. So if the header file is changed, main.o will update.

Cannot save eps format on Windows

As the list of jobs submitted to a queue grows, startup of IQmol gets ever slower. After a certain point (which isn't terribly many), startup can hang for 5-10 minutes. My students think IQmol is broken. :) But after experiencing it myself the last couple of days, I tried clearing the job list on a whim. Startup returned to normal. I don't know if the issue is long/dead jobs or just a long list of jobs.

Hi - compiled successfully. Note to Linux users - expects a static libopenbabel.a in the src directory (and usual Ubuntu distributions do not provide the static library) - easy to make. Some edits to common.pri and linux.pri are needed - probably good to put this in the build_linux instructions. Thank you!

I reiterate my issue here , since I am not sure on where to report the problem (I also put it into some bug-tracker site, but that is not linked to the webpage of IQmol and I am not sure whether it is maintained). Sorry for cross-posting - the website clearly asks to report issues, it just doesn't say where....
I downloaded the windows installer for IQmol from http://www.iqmol.org/downloads.html. Installation seems to have gone OK (i.e. the installer does not complain). However, when I start the IQmol executable, all that happens is that a logo shows up. Then the program stops. It is very possible that I am doing something stupid, but I would not know what.
Platform: Windows X
IQmol version: both: 2.14 and 2.13.0 as installed with the windows installer

The function boost::serialization::stl::load_collection no longer exists, neither do archive_input_seq nor no_reserve_imp

/builddir/build/BUILD/IQmol-df94cf68b9d9d813e4d781c7edb9ac638724dadf/src/Data/Serialization.h: In function ‘void boost::serialization::load(Archive&, QList&, unsigned int)’:
/builddir/build/BUILD/IQmol-df94cf68b9d9d813e4d781c7edb9ac638724dadf/src/Data/Serialization.h:170:4: error: ‘load_collection’ is not a member of ‘boost::serialization::stl’
    boost::serialization::stl::load_collection< 
    ^
/builddir/build/BUILD/IQmol-df94cf68b9d9d813e4d781c7edb9ac638724dadf/src/Data/Serialization.h:171:14: error: expected primary-expression before ‘,’ token
       Archive, 
              ^
/builddir/build/BUILD/IQmol-df94cf68b9d9d813e4d781c7edb9ac638724dadf/src/Data/Serialization.h:172:15: error: expected primary-expression before ‘,’ token
       QList, 
               ^
/builddir/build/BUILD/IQmol-df94cf68b9d9d813e4d781c7edb9ac638724dadf/src/Data/Serialization.h:173:7: error: ‘archive_input_seq’ is not a member of ‘boost::serialization::stl’
       boost::serialization::stl::archive_input_seq >,
       ^
/builddir/build/BUILD/IQmol-df94cf68b9d9d813e4d781c7edb9ac638724dadf/src/Data/Serialization.h:173:59: error: expected primary-expression before ‘,’ token
       boost::serialization::stl::archive_input_seq >,
                                                           ^
/builddir/build/BUILD/IQmol-df94cf68b9d9d813e4d781c7edb9ac638724dadf/src/Data/Serialization.h:173:70: error: expected primary-expression before ‘>’ token
       boost::serialization::stl::archive_input_seq >,
                                                                      ^
/builddir/build/BUILD/IQmol-df94cf68b9d9d813e4d781c7edb9ac638724dadf/src/Data/Serialization.h:173:71: error: expected primary-expression before ‘,’ token
       boost::serialization::stl::archive_input_seq >,
                                                                       ^
/builddir/build/BUILD/IQmol-df94cf68b9d9d813e4d781c7edb9ac638724dadf/src/Data/Serialization.h:174:7: error: ‘no_reserve_imp’ is not a member of ‘boost::serialization::stl’
       boost::serialization::stl::no_reserve_imp< QList > >(ar, t);
       ^
/builddir/build/BUILD/IQmol-df94cf68b9d9d813e4d781c7edb9ac638724dadf/src/Data/Serialization.h:174:59: error: expected primary-expression before ‘>’ token
       boost::serialization::stl::no_reserve_imp< QList > >(ar, t);
                                                           ^
/builddir/build/BUILD/IQmol-df94cf68b9d9d813e4d781c7edb9ac638724dadf/src/Data/Serialization.h:174:61: error: expected primary-expression before ‘>’ token
       boost::serialization::stl::no_reserve_imp< QList > >(ar, t);
                                                             ^

hello.. this is my proposal for the web, what do you think.

https://steemit.com/utopian-io/@novianti.idr/design-new-logo-open-source-for-iqmol

Dear developers,

I cannot run I-Qmol 2.9 on Windows because of Qt platform error

Could you please fix this?

Thanks

Trying to update the Fedora packages, I get the error

MoleculeLayer.C: In member function 'void IQmol::Layer::Molecule::groupSelection()':
MoleculeLayer.C:1396:45: error: no matching function for call to 'IQmol::Layer::Group::Group(IQmol::Layer::PrimitiveList&)'
 1396 |    appendPrimitive(new Layer::Group(selected));
      |                                             ^
In file included from EfpFragmentLayer.h:25,
                 from MoleculeLayer.C:46:
GroupLayer.h:38:10: note: candidate: 'IQmol::Layer::Group::Group(const QString&)'
   38 |          Group(QString const& label = QString());
      |          ^~~~~
GroupLayer.h:38:31: note:   no known conversion for argument 1 from 'IQmol::Layer::PrimitiveList' to 'const QString&'
   38 |          Group(QString const& label = QString());
      |                ~~~~~~~~~~~~~~~^~~~~~~~~~~~~~~~~

root log https://kojipkgs.fedoraproject.org//work/tasks/4237/40314237/root.log (list of installed packages)
build log https://kojipkgs.fedoraproject.org//work/tasks/4237/40314237/build.log

My students have observed that selection of atoms is sporadically unreliable. They'll click on a carbon atom at one of the molecule, for example, and a hydrogen atom several Angstroms away will become selected. We used to encounter this issue regularly, and I thought that it had been resolved (I haven't had problems with it myself), but several of them have brought it to my attention this semester. It appears to be more common among Windows users, but this observation might just be anecdotal. The students can work around it by clicking atoms in the indexed atom list, but obviously it would be preferable if they could just click on the molecule. :)

I cannot save molecule coordinates after drawing. I have clicked "File->Save as" and have chosen name (test.xyz), but the result is a messagebox with "Parser::ExtensionError()". There is no saved file, the console log is:

Attempting to save "/home/myself/test.xyz"
WARN  2015-09-10T01:11:49.348 QMsgBox.C @ 37 "Parser::ExtensionError()"

With harmonic frequency calculations, O-H stretches (or similar) can often be to the blue of 4000 cm-1. The default range in the IQmol spectrum viewer does not show these transitions. As a feature request, it would be wonderful if the spectrum auto-adjusted to encompass all transitions in the molecule. Alternatively, a default range of 0-4800ish cm-1 would likely suffice for nearly every molecule. I realize that this range is stupidly high for real spectra, but for harmonic calculations (often with less-than-perfect quantum methods), it might be needed.

Jobs on the job monitor have to be removed independently, and there's no way to prevent the job monitor from flooding with popups when booting iqmol with jobs still in the job monitor.

Great program well done. I was wondering if the ability to change atom colour would be possible to incorporate.

In the latest version of IQmol (2.15) using ssh to connect to a remote computer no seems to be functional. I can submit jobs using 2.14 with exactly the same configuration, but in 2.15, I get the error "Failed to connect to server XXX: SSH host key not found"

Under "MO Surfaces/Add Surface", it would be nice to have options for localized orbitals if Localized Alpha MO Coefficients (Boys) (or something similar) is present in a checkpoint file so that the Alpha MO coefficients section doesn't have to be manually replaced.

I can compile this code only by adding header unistd.h and replacing sleep(int) by usleep(int):

void SshTest::runDelegate()
{
   for (int i = 0; i < 10; ++i) {
      QLOG_TRACE() << "Running thread" << m_id << i;
# ifdef WIN32

```
  Sleep(2000);
```
# else

```
  //sleep(2);
  usleep(2);
```
# endif

```
  if (m_interrupt) break;
```

   }
}

In the spectrum viewer for vibrational calculations, the default range for plotting frequencies is sometimes too small. Values above 4000 cm-1 are typically neglected (see ticket #58 ). While this issue might be a mere annoyance, it becomes more of a problem when convoluting the spectrum with an artificial lineshape. In this latter case, the lineshape is only applied to transitions that appear in the displayed window. If the user first applies a lineshape and then readjusts the frequency axis, transitions that were out of range (but otherwise viewable with impulses) do not get convoluted; they simply disappear. I think the lineshape routine just needs to extend over a larger default range (say, 0-5000 cm-1?) in order to capture these high-frequency transitions.

I installed 2.15.0 for mac and managed to set up the connection with a remote server, and submit a job through Slurm (ver 19.05).

However, the job monitor fails to update the status, regardless the job is pending or running. Needless to say, it is not possible to watch the output during the calculating. Is this behavior expected? I believe it should not be difficult to retrieve the slurm status with squeue -al |grep JobID...

Hi

I tried to use the binary version on Arch Linux, with openbabel installed from the official repos of the distribution, but IQmol complains while opening due to a missing libopenbabel.so.4 library.
I agree with that complaint, since that library does not exist, in Arch the openbabel library is called libopenbabel.so.5
with openbabel version 2.4.1

Is there an easy fix?
A simple symlink does not work..