provide static build

[jw-feng]

Hi Dave,

I did struggle a bit compiling the code on CentOS 6. I found Qt5.2.1 and OEChem 2016.Feb.1 (redhat-RHEL6-g++4.4-x64) worked best. What do you think about providing a static build to make it easier for users to try out TautEnum?

Best,

JW

[DavidACosgrove]

Hi JW, I'm sorry you struggled with the compilation. Cmake makes things easier, but it is a pain tracking between different versions of Linux. I standardised on ubuntu because it seems to be a popular distro. Ironically, it was all originally developed on CentOS 6, I took a certain amount of pain switching it to ubuntu! One thing I can help with directly is to fix the dependency on Qt which isn't used by taut_enum itself at all, just the two companion viewing programs. I'll put some sort of conditional switch on it. A year or two ago Jack Delaney at OE produced a linux distro independent compilation method that he used to use at Daylight. It seemed a bit complicated for internal use at AZ where we only had 1 linux to deal with, but I'll have another look at it and see how much effort it would be to put an exe up as well. The trick he uses is to compile on the oldest linux he's interested in supporting, with the latest gcc. It does mean that you have to keep the libstdc++.so that the code was linked against in with the exe so there's a bit of extra stuff to keep track of. All the best, Dave

On Mon, Apr 17, 2017 at 6:02 PM, JW Feng notifications@github.com wrote:

Hi Dave,

I did struggle a bit compiling the code on CentOS 6. I found Qt5.2.1 and OEChem 2016.Feb.1 (redhat-RHEL6-g++4.4-x64) worked best. What do you think about providing a static build to make it easier for users to try out TautEnum?

Best,

JW

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/OpenEye-Contrib/TautEnum/issues/2, or mute the thread https://github.com/notifications/unsubscribe-auth/AIxdFux52gy9BFAO1usseWlgSqnKHvXtks5rw5stgaJpZM4M_U7d .

-- David Cosgrove Freelance computational chemistry and chemoinformatics developer http://cozchemix.co.uk

[jw-feng]

Hi Dave,

Removing the Qt dependency will be great. What about dependency on OEDepict? It requires its own license.

Best,

JW

[DavidACosgrove]

Hi JW, Yes, that too although if you don't use it, it doesn't matter if it's linked into the exe. So long as you have it to link, I guess, which might not be possible with the current oe download policy. I'll do a proper clean out when I do it. Dave

On Tue, 18 Apr 2017 at 19:16, JW Feng notifications@github.com wrote:

Hi Dave,

Removing the Qt dependency will be great. What about dependency on OEDepict? It requires its own license.

Best,

JW

— You are receiving this because you commented.

Reply to this email directly, view it on GitHub https://github.com/OpenEye-Contrib/TautEnum/issues/2#issuecomment-294934784, or mute the thread https://github.com/notifications/unsubscribe-auth/AIxdFlXv4s871SvpLRA_4YKoPF1ccODoks5rxP4WgaJpZM4M_U7d .

-- David Cosgrove Freelance computational chemistry and chemoinformatics developer http://cozchemix.co.uk

[DavidACosgrove]

I've now made compilation of mol_diff_viewer and mol_diff_viewer2 conditional, and off by default. This should remove the Qt dependency and thus make building a bit easier.

[DavidACosgrove]

If fixed the CMakeLists.txt so it works on a fairly standard Desktop Centos 6.9 installation, running in VirtualBox. You will still need to install cmake (though 'yum install cmake' is all that's required there, it works with the cmake 2.9 that that fetches) and a modern boost. Boost 1.64 compiled really quickly and doesn't need to be in a system directory, so that should be straighforward enough. I don't think I'll provide a pre-compiled exe, that seems like an extra level of support effort that I don't want to do unless pushed.