Hi Graeme and others,
Thank you for your replies. I was not aware of all the dependencies between
the various modules within the cctbx.
As I don't know much more about the cctbx than other developers I am not
sure what the best course of action would be. Here are some random thoughts.
Probably never gonna happen but, ideally a code freeze for 2 or 3 weeks
stopping anyone from amending or adding code to the cctbx could be spent on
disentangling modules from one another and the cctbx.
I agree that there appear to be some other projects that do not belong to
the cctbx. For instance, I think fable was conceived by Ralf as a one time
conversion tool of F77 code to C++ code. Not sure if it is used anymore.
As for pip install ability, yes that sounds like a good idea to achieve if
it is done from the lowest level first, i.e. the libtbx submodule.
Documenting how to build derived projects using the CCTBX should encourage
people not to add unrelated code into the cctbx. Although SConscripts are
familiar to many of us, to others SConscript or bootstrap are quite alien
and opaque ways of building from sources.
cheers,
Rob
-----Original Message-----
From: [email protected]
Sent: Monday, January 22, 2018 8:58 AM
To: [email protected]
Cc: [email protected]
Subject: Re: [cctbxbb] Tidy up CCTBX
Hi Rob
More thinking on this one, following from a little conversation over the
weekend at this end
If we were looking to be able to package cctbx in a more Pythonic manner -
i.e. to allow pip install - would it make sense to start at the bottom
rather than at the top? i.e. libtbx - which is pure Python with deliberately
no dependencies - would be a good place to start? The lower level code is
also more static so a safer candidate for pip installing compared with some
of the more rapid changing things…
From a diamond perspective, there are a bunch of people here who want
cctbx-without-dials-but-with-their-own-packages so being able to pip install
cctbx and it's dependencies would be pretty cool…
Another happy side effect is this would allow e.g. libtbx.bootstrap to
exist...
Cheers Graeme
On 19 Jan 2018, at 14:10, Winter, Graeme (DLSLtd,RAL,LSCI)
mailto:[email protected]> wrote:
Hi Rob
Interesting question. I have sympathy for your point of view. Some thoughts,
from a conversation here in DIALS land…
(i) dxtbx is not only there to support DIALS - cctbx image viewer, bits of
iotbx, cctbx.xfel also use it (I think)
(ii) from where I am sat this is a wider question - if I have a look in
cctbx_project I see more “non core code" than just dxtbx:
Graemes-MBP-5:cctbx_project graeme$ ls
README.md cootbx fftw3tbx omptbx sphinx
boost_adaptbx crys3d gltbx prime spotfinder
cbflib_adaptbx cudatbx iota rstbx tbxx
cctbx dox iotbx scitbx ucif
chiltbx dox.sphinx libtbx simtbx wxtbx
clipper_adaptbx dxtbx mmtbx site.pyc xfel
cma_es fable msvc9.0_include smtbx
So, in here we have several things which probably do not “belong” in the
world view you present below, for example fable, prime, simtbx, xfel etc.
(iii) If we are visiting this stuff, should we be visiting the wider build
system which I understand is one of the “blockers” for e.g. pip install
cctbx
I appreciate that this is a wider question than can we pull dxtbx out, but
is essentially derived…
Best wishes Graeme
On 18 Jan 2018, at 15:56, Dr. Robert Oeffner
mailto:[email protected]> wrote:
Hi all,
I'm wondering if we could make efforts to refactor code modules in the
cctbx_project so as making it have the minimum number of dependencies. This
would be helpful to other users of the CCTBX that wishes to develop code on
top of it. In the longer term it would also make it easier for us to deploy
CCTBX as a Python wheel installer and by implication, spread the gospel of
the wonders of the CCTBX.
What is currently complicating this is the inclusion of the dxtbx module
within the cctbx_project. It is being used by Dials. So my question is if it
would be possible to move this project into the Dials code?
If the dxtbx was just another small module it wouldn't be much of a problem.
But the dxtbx depends on the cbflib and HDF5 components. If you want to
build CCTBX from sources that's cumbersome. Why? Because CCTBX as deployed
from CCI on http://cci.lbl.gov/cctbx_build/ is only built for a few select
platforms. Building CCTBX on a different platform from those with bootstrap
compels us to also build cbflib and HDF5 from scratch. Although this is not
impossible it does increase the entry barrier for other potential users of
the CCTBX who wish to incorporate the CCTBX in their programs: If all a user
wants is to calculate allowed origins and symmetry sites for a given space
group then an alternative to CCTBX might be preferable rather than having to
massively expand their project with the seemingly unrelated HDF5 component.
In other words, as it stands now the CCTBX is not straightforward for others
to adopt.
What do other people think?
Regards,
Rob
--
Robert Oeffner, Ph.D.
Research Associate, The Read Group
Department of Haematology,
Cambridge Institute for Medical Research
University of Cambridge
Cambridge Biomedical Campus
Wellcome Trust/MRC Building
Hills Road
Cambridge CB2 0XY
www.cimr.cam.ac.uk/investigators/read/index.htmlhttp://www.cimr.cam.ac.uk/investigators/read/index.html
tel: +44(0)1223 763234
_______________________________________________
cctbxbb mailing list
[email protected]mailto:[email protected]
http://phenix-online.org/mailman/listinfo/cctbxbb
--
This e-mail and any attachments may contain confidential, copyright and or
privileged material, and are for the use of the intended addressee only. If
you are not the intended addressee or an authorised recipient of the
addressee please notify us of receipt by returning the e-mail and do not
use, copy, retain, distribute or disclose the information in or attached to
the e-mail.
Any opinions expressed within this e-mail are those of the individual and
not necessarily of Diamond Light Source Ltd.
Diamond Light Source Ltd. cannot guarantee that this e-mail or any
attachments are free from viruses and we cannot accept liability for any
damage which you may sustain as a result of software viruses which may be
transmitted in or with the message.
Diamond Light Source Limited (company no. 4375679). Registered in England
and Wales with its registered office at Diamond House, Harwell Science and
Innovation Campus, Didcot, Oxfordshire, OX11 0DE, United Kingdom
_______________________________________________
cctbxbb mailing list
[email protected]
http://phenix-online.org/mailman/listinfo/cctbxbb
---
This email has been checked for viruses by AVG.
http://www.avg.com
