Documentation #1823

Updated by pospelov over 3 years ago

Consider cleaning up of our Redmine

* to make the development more agile
* to get rid from very minor issues or issues that are non-relevant anymore
* to use Redmine for support of current work, rather than as a storage of long term ideas, or minor programming suggestions

* #151 Check with ToyExperiment the influence of numerical error in chi2/residuals on fit stability
> Old issue, not important in the context of much bigger "local minima problem".

* #395 Create rpm package
> For whom? Better focus first on BornAgain inclusion on Debian scientific repository.

* #472 Support Artur and Elisabeth with simulation scripts
> ?

* #488 Fix issues encountered by Artur
> Some of issues already solved, for others is not clear what was meant.

* #504 Polarized GISANS on mesocrystals (Disch)
> Contact was lost

* #520 Fix issues encountered by JFM
> Revise and make dedicated issue, if necessary

* #525 Simulations for oxide layers with magnetization (A.Steffen)
> ?

* #532 Check distribution of mesocrystal heights in simulation
> Unclear

* #533 Check distribution of lattice parameters (both isotropic and anisotropic)
> Unclear

* #542 Windows installer: check existence of BornAgain and uninstall it if exists
> Valid request, but no sense to try solve if for current NSIS-based installer. We have to move to another type of Windows installer (e.g. qt-installer framework).

* #548 Provide script to fit mesocrystal data from Python
> We have to consider first whether to ot renew our mesocrystal mesocyrstal activity.

* #550 Simulations of multi layered sample for S. Mattauch
> ?

* #606 Restore automatic Doxygen doxygen API generation
> Should be probably dropped after migration to new statically based documentation site

* #649 Provide lattice rotation example
> Partly solved, partly mixed with other issues (new SimulationBuilder machinery)

* #650 Provide two cylinders example with two different SLD
> Who wants this? We do not have a concept of SLD in materials yet. materials.

* #652 Flat cylinders with transformation (coherent and uncoherent example)
> ?

* #707 Support Thaer Kassar: Hetero-junction thin films
> Contact was lost

* #708 Support Vicky Doan-Nguyen: In-plane lamellar packing of nanorods
> Contact was lost

* #726 Review calculation of R,T coefficients for magnetic case
> Is it still relevant?

* #728 Install BornAgain at the relevant instruments
> At Galaxi it is installed.

* #890 Formfactor restrictions in fitting and distributions
> I It think, it is already possible via RealLimits applied to parameter distribution.

* #892 Provide BornAgain support in MacPorts
> Very low priority.

* #899 GUI: create "install libBornAgainCore in Python" widget
> Not needed for Windows, for MacOS not sure that it is the a best approach.

* #929 Provide script to read offspecular data from MARIA
> Too specific. Context is lost.

* #967 Support Marvin Berlinghof: provide python example for simulation of Ruderer et al. 2012
> Have to renew acquaintance, or drop the issue.

* #936 Provide script to read GISANS data from MARIA
> Context is lost

* #939 Prepare mesocrystal simulation status report
> Too optimistic issue. First we have to re-start mesocrystal activity.

* #1029 ToF support in OffSpecSimulation
> Two years old issue by Artur.

* #1030 SLD, specular and particle density
> Is it not already solved?

* #1098 Add Python examples Interference2DLatticeSumOfRotated and TriangularRipple to Drupal
> Too specific (or have too low priority) in the view of changes in documentation

* #1139 Py: Provide Simulation class with possibility to set vector of SimulationElements from outside
> Partly coincide with other Artur's suggestions. Better to remove issue and wait the subject reappears.

* #1187 Implement 1D fitting in SpecularSimulation
> Too broad.

* #1199 Consider implementation of partial structure factor
> Unclear

* #1263 Drupal: provide site backup
> Almost obsolete

* #1282 Drupal: unique form of persistent page URLs
> Almost obsolete

* #1432 final polish of polyhedral form factor
> Is polishing still necessary?

* #1437 Release requirement: Update example scripts
> Part of web site migration. No need for an issue.

* #1441 Support Anatoly Berezkin
> Contact is lost.

* #1457 Calculate precomputed tables at compile time
> Is it still relevant? Can be done in connection with some concrete future refactoring.

* #1463 possibility to specialize|generalize physical models on the fly
> Too broad, no context.

* #1469 Package Multifit-Suite as separate library
> Should be postponed. I'm not sure that quality of our ROOT-based minimization kernel worth extra work. We might discover some better C++ library in the future (e.g. everyone all wants Baesyan minimisers, we do not have them).

* #1476 Find an appropriate path for the BornAgain executable
> Too specific while there is more global question (installation directory structure). Suggest to drop unless the future will give us extra motivation.

* #1480 Buildbot: provide tutorial how to add new build configuration: Windows
> Issue is there since a long time. Seems that nobody is actually extually interested.

* #1481 Buildbot: provide tutorial how to add new build configuration: MacOS
> Issue is there since a long time. Seems that nobody is actually extually interested.

* #1491 GUI: improve GUI fitting performance
> Still valid, but I would close to not spoil Redmine

* #1508 GUI JobView: reconsider location of activity view switch
> Seems the location is not that bad.

* #1500 MacOS: provide vagrant test box for El Capitan
> Painful Painfull issue, and not really necessary.

* #1509 Version string should be different for executables generated from 'develop', from feature branches, or from 'master'
> Not clear, if it is that necessary.

* #1519 Linking failure of Ubuntu with Python installed via Anaconda bundle
> Should be closed. Hard to reproduce, no complains anymore.

* #1532 clarify relation of PolDWBAMagCylinders2 and StandardSimulations/StandardSimulations.h/cpp to standard Core function tests
> Nothing to do. Custom tests will at some point, become the part of standard tests. Or will not.

* #1537 parametrize Gaussians and Lorentzians by mean and fwhm
> Is this really necessary?

* #1538 AttLimits::isInRange(value) returns true for value==m_lower_limit, and false for value==m_upper_limit
> Nothing to do (or too much todo).

* #1544 require libcerf instead of providing ThirdParty/Fadeeva
> If I'm not mistaken, libcerf has to be changed first.

* #1553 Project (XML) read-in: support old versions by automatic conversion of old to new class and function names
> Seems to me, too much burden to support. Better to invest in normal back compatible project file format.

* #1554 provide script to update Python files upon name changes in BornAgain
> Not necessary in the view of python examples machinery. Easy to fix any API manually, rather than support change history. files.

* #1567 omit »FormFactor« from particle geometry class names
> Duplication of #1766

* #1570 replace parameter accessor methods getFoo() by getParameterValue("Foo")
> Minor issue which will naturally reaper reapper on the next necessary refactoring of RealParameter machinery

* #1581 get... -> create... if new pointer is returned
> No need for dedicated issue.

* #1596 handle variable instance names differently from fixed class names (=> split INamed)
> Not need for extra work here, I think. INamed class name perfectly reflect the purpose - to give names to things.

* #1598 decide on policy for incomplete ICompositeSample descendant constructors
> No issue here, I think.

* #1599 provide another suite of functional tests: simulate on different grids, and test numeric consistence of results
> Too broad issue, use case is required. Refactoring required, refactoring of functional test machinery is absolutely necessary first. necessary.

* #1622 avoid smart pointers in user API
> No need for issue

* #1626 Specular calculation gives weird results below critical angle in presence of top layer absorption
> Is it still relevant?

* #1627 python API kvector_t missing basic operator
> We have to remove, there possible usage of kvector_t in Python API. Use lists or numpy arrays instead.

* #1629 GISASSimulatin setAnalyzerProperties total_transmission default parameter generates Exception
> Is it still relevant?

* #1636 Unexpected wavelength depent intensity Roughness vs. ParticleLayout
> Is it still relevant?

* #1638 Manual: Improve / restore side bar navigation
> Is it still relevant?

* #1684 nicer layout for
> Is it still relevant?

* #1688 == Roughness ==
> No need for envelop task

* #1686 == Release procedures and pre-release actions ==
> No need for envelop task

* #1695 GUI: Implement crash handler, and restoration of state before crash
> Partly implemented (autosave), partly very difficult to implement (crash handler) in platform independent way.

* #1699 functional tests: Improve MinimizerCatalogue::toString
> Nobody ever asked for it. Given The give issue has to be, one day, the part of much bigger polishing of fitting API.

* #1701 import real-space sample configuration, e.g. from molecular dynamics (morphology file of IsGISAXS)
> No user request

* #1707 refactor LayerStrategyBuilder, DecoratedLayerComputation -> LayoutStrategyBuilder, DecoratedLayoutComputation
> Already done, in a sense

* #1713 light refactoring proposals
> Should not be part of issue tracker (I think)

* #1714 == Enhance functionality ==
> No need for envelop task

* #1715 start sample construction with SlabSample() or FilmSample()
> Nice idea to have in mind, but I'm afraid that the perception might change from sub-community to sub-community.

* #1717 GISASSimulation: setBeamParameters and setDetectorParameters have opposite order of alpha, phi
> Will affect everyone in a world. Also depends on personal preferences, some might find the ordering not that important.

* #1719 Manual: avoid the word "matrix" for embedding, ambient bulk material
> May be not for this tracker.

* #1720 Vacuum() instead of HomogeneousMaterial("Air", 0.0, 0.0)
> Issue will naturally reappear as soon as we implement normal material library. For the moment no need to bother.

* #1721 Scattered intensity should depend on beam size and sample size
> Valid reminder, but seems very big and context dependent.

* #1732 Optimization in innermost loop in RT computation
> Partly already done. improvement. Further improvements should be based on profiling.

* #1738 Rename files and classes in Multilayer to resolve ambiguity about "Matrix"
> Already done, I guess.

* #1741 Wavelength/radiatio independent way to define material parameters
> We know about it and have similar issues. Need good planing and issue-by-issue implementation. Have to wait for new collaborator.

* #1748 Consider IParameter storing the data by reference
> Why I have suggested it?

* #1767 array operations for python custom form factor
> Idea that flying in the air for a while, but let's wait for user request

* #1770 support CBF data from A. Nent
> Have to be dropped. We must agree on general solution (fabio Either with learn how to use fabio library, or data broker or something else). broker.