| 98891f9 | Lorenzo Moneta | 04 February 2020, 08:59:02 UTC | fix comment as suggested by Philippe | 04 February 2020, 08:59:02 UTC |
| d35deff | Lorenzo Moneta | 03 February 2020, 16:40:37 UTC | Apply fixes for when RooFitMore (MathMore) is not available. This fixes ROOT-10540 and ROOT-10536 | 04 February 2020, 08:57:25 UTC |
| f00051d | Axel Naumann | 22 January 2020, 09:47:11 UTC | [core io] Mark TString(string_view) explicit (ROOT-10512): Without this, conversions from string_view to anything that TString implicitly converts to are picked up, rendering e.g. string_view == string ambiguous. | 03 February 2020, 18:08:38 UTC |
| 423c8d2 | Philippe Canal | 31 January 2020, 22:32:39 UTC | [core] Add missing op+= for string,string_view when needed | 03 February 2020, 18:08:38 UTC |
| 1dc17c0 | Stephan Hageboeck | 31 January 2020, 10:40:45 UTC | [RF] Fix crash in RooLinTransBinning due to uninitialised pointer. [ROOT-7520] The empty constructor of RooLinTransBinning was not initialising a pointer to an array. When streaming, this pointer is not streamed, keeping the dangling pointer around. Initialising it with nullptr fixes the problem. (cherry picked from commit 3a14813d23a41b8a6799e08ed11242531c4ec01f) | 03 February 2020, 08:38:34 UTC |
| b5ffc7c | Philippe Canal | 31 January 2020, 19:29:25 UTC | Merge pull request #4892 from pcanal/interpreterException-v620 Compile Interpreter.cpp with exception on. Fix script with exception with ROOT Mutex on. | 31 January 2020, 19:29:25 UTC |
| 64463d9 | Lorenzo Moneta | 31 January 2020, 15:35:34 UTC | update release notes for TMVA and add missing contributors | 31 January 2020, 16:00:41 UTC |
| 42c7e6e | Stephan Hageboeck | 29 January 2020, 07:51:45 UTC | [RF] Fix fits with multiple ranges in RooChi2Var. [ROOT-10518] Due to overwriting a function-local static, range strings for multi-range fits would be lost. The full range was fitted multiple times. This results in exactly the same coefficients as when fitting the full range, but in different errors, as the model is fitted repeatedly. (cherry picked from commit 4f53b0fd7e01a4ba40656940359eaa5567072d75) | 30 January 2020, 09:05:53 UTC |
| 0b657ca | Oksana Shadura | 22 January 2020, 18:24:15 UTC | Add release notes | 30 January 2020, 08:41:56 UTC |
| b486b8c | Chris Burr | 09 January 2020, 22:31:10 UTC | [ROOT-10472] Fix building clingutils tetss with builtin_clang=OFF for conda | 30 January 2020, 08:41:30 UTC |
| 0df0824 | Philippe Canal | 29 January 2020, 18:11:35 UTC | Update comment in patch to DiagnosticsEngine::DiagStateMap::append | 30 January 2020, 07:19:55 UTC |
| cb3e63f | Philippe Canal | 28 January 2020, 21:49:01 UTC | Add support for DiagnosticEngine state change within recursive parsing. This fixes ROOT-10504. the script: ``` namespace boost { namespace mpl { // Commenting the next line make the assert failure go away struct TTUBE {}; }} ``` reproduce the problem with 'just' ROOT. The trigger is the auto-loading of a library that has a dictionary with has forward decl string .. which all starts with: ``` static const char* fwdDeclCode = R"DICTFWDDCLS( extern int __Cling_Autoloading_Map; .... ``` The order of parsing is (with many ellipsis): ``` "<<< cling interactive line includer >>>" : line 9 : #include “standalone.C” standalone.C : line 1 : #pragma GCC diagnostic push standalone.C : line 2 : #pragma GCC diagnostic ignored "-Wuninitialized" // Inserted in Diag map standalone.C : line 3 : #pragma GCC diagnostic ignored "-Wsign-conversion" // Inserted in Diag map standalone.C : line 8 : struct TTUBE {} ; // triggers auto-loading and thus recursive parsing. "<<< cling interactive line includer >>>" : line 10 : parse dict fwd declare string input_line_9 : line 2 : #pragma clang diagnostic ignored "-Wkeyword-compat" // Inserted in Diag map input_line_9 : line 3 : #pragma clang diagnostic ignored "-Wignored-attributes" // Inserted in Diag map input_line_9 : line 4 : #pragma clang diagnostic ignored "-Wreturn-type-c-linkage" // Inserted in Diag map end of file standalone.C : line 12 : #pragma GCC diagnostic pop // Inserted in Diag map ``` The last line triggers the assert because when recording the state change, it records it as being from ``` standalone.C : line 12 "<<< cling interactive line includer >>>" : line 9 ``` but when recording the last one, it notices that the last state change that happened, indirectly, for the file '<<< cling interactive line includer >>>' happened at line 10 (because of the pragma in input_line_9 which is 'recorded' as being included by line 10), which makes that the state change for ```standalone.C:12``` happens 'ealier' than the last state change as far as the pseudo-file "<<< cling interactive line includer >>>" is concerned. For that pseudo-file, the last state change happened line 10 but the state change bbeing processed happens line 9 (where Standalone.C is being included). | 30 January 2020, 07:19:55 UTC |
| 4a3e76c | Ivana Hrivnacova | 10 January 2020, 23:09:59 UTC | Fix in TGeoMCGeometry::Mixture(Int_t& kmat, const char* name, Float_t* ...): 'wmat' values were not updated when the function is called with nlmat<0, as they should be to mimic the Geant3 behavior. | 29 January 2020, 14:08:58 UTC |
| 911140c | Oksana Shadura | 27 January 2020, 11:56:38 UTC | Use kUseGeneralPurpose to setup ZSTD as ROOT general purpose recommended algorithm (replace zlib) | 29 January 2020, 08:01:27 UTC |
| c2821d1 | Oksana Shadura | 27 January 2020, 11:55:26 UTC | Use EDefaults::kUseCompiledDefault as a default compression settings for ROOT files | 29 January 2020, 08:01:27 UTC |
| f203a97 | Philippe Canal | 24 January 2020, 18:26:51 UTC | Clarify conditional in IsDefaultVector | 27 January 2020, 15:07:58 UTC |
| 45fb919 | Philippe Canal | 21 January 2020, 17:58:53 UTC | In StreamerInfoActions always respect vector's custom-allocator Previously the handling was inconsistent resulting in mis-matches and thus memory errors. This fixes ROOT-10526. | 27 January 2020, 15:07:58 UTC |
| 5f704ed | Sergey Linev | 24 January 2020, 12:12:25 UTC | [gpad] use directly TString as file name for TPad::Print() | 27 January 2020, 07:35:30 UTC |
| fd0deb9 | Sergey Linev | 24 January 2020, 12:06:11 UTC | [gpad] fix memory leak in TPad::Print() One have to delete result of gSystem->ExpandPathName() function or just use other signature | 27 January 2020, 07:35:30 UTC |
| 7f8bc35 | Vassil Vassilev | 22 January 2020, 17:18:43 UTC | [core] Call TClass::ReadRules before the rootmap files to be read. This avoid loading the library containing the dictionary and avoids leaving the TClass::GetClass("HepMC::GenVertex") meta information in an odd state. This should bring the v6.18 behavior unintentionally broken in the refactoring commit c8cce31c0f8215897aee08d2592fe50e4dc0cfb0. This should fix ROOT-10514 completely. Patch by Philippe Canal and me! | 26 January 2020, 13:22:14 UTC |
| 4a3e1b3 | Vassil Vassilev | 22 January 2020, 16:19:03 UTC | Revert "Explicitly disable autoloading when processing TClass rules." This reverts commit 8fc83bb9e2ce51b2ff171a81f818687696b2404c trying to resolve ROOT-10528 and ROOT-10514 and reintroduce the first part of ROOT-10514 where we lead the HepMC::GenVertex's dictionary. Trying to fix ROOT-10514 caused breakage all over the place. Let's remove this until we find a better fix. | 26 January 2020, 13:22:14 UTC |
| 868621a | Alja Mrak-Tadel | 23 January 2020, 19:14:48 UTC | Add Eve-7 changes in v6-20 ReleaseNotes Insert info about Eve-7 development in 3D Graphics Libraries chapter. | 24 January 2020, 11:48:17 UTC |
| d57346f | Oksana Shadura | 23 January 2020, 13:33:29 UTC | Update default compile compression settings As a part of this commit, kUseCompiledDefault was returned back to the correct value of using ZLIB as a default compression algorithm with compression level 1 (101). | 23 January 2020, 20:38:34 UTC |
| 43071b8 | Oksana Shadura | 23 January 2020, 13:31:27 UTC | Update RCompressionSettings struct with enum value of ZSTD default compression settings | 23 January 2020, 20:38:34 UTC |
| e2b30af | Chris Burr | 22 January 2020, 14:36:05 UTC | Set PYTHON_PREFER_VERSION to match PYTHON_EXECUTABLE | 23 January 2020, 18:00:37 UTC |
| 56281e0 | Chris Burr | 10 January 2020, 17:47:22 UTC | Disable SetLibraryPath for RPATH builds | 23 January 2020, 09:24:02 UTC |
| 5de2856 | Henry Fredrick Schreiner | 20 January 2020, 16:25:58 UTC | Fix travis check never failing | 23 January 2020, 09:23:37 UTC |
| baec1e7 | Henry Schreiner | 17 January 2020, 16:38:32 UTC | Fix missing standalone flag for standalone Minuit2 | 23 January 2020, 09:23:37 UTC |
| 91b72dd | Stephan Hageboeck | 21 January 2020, 13:08:59 UTC | [RF] Fix shadow warning in RooTemplateProxy. Old gcc issues a shadow warning if a local variable and a function have the same name. | 22 January 2020, 07:53:53 UTC |
| 313c9c0 | Sergey Linev | 20 January 2020, 17:07:12 UTC | [jsroot] use openui5 1.72.0 by default | 21 January 2020, 11:28:06 UTC |
| 55947b6 | Sergey Linev | 20 January 2020, 17:06:42 UTC | [jsroot] set ROOT 6.20.00 version | 21 January 2020, 11:28:06 UTC |
| 03bcc43 | Stefan Wunsch | 17 January 2020, 09:09:56 UTC | [TMVA] Add new features to release notes | 21 January 2020, 09:30:32 UTC |
| 43dfd57 | Simon Spies | 15 January 2020, 09:43:29 UTC | [hist] Fixed a possible segmentation violation that broke the test gtest-hist-hist-test-TGraphMultiErrorsTests (#4780) | 21 January 2020, 09:08:32 UTC |
| a988354 | Simon Spies | 14 January 2020, 15:46:48 UTC | [hist] Formatted the files realted to TGraphMultiErrors (#4777) * [hist] Formatted the files realted to TGraphMultiErrors accoding to ROOTs formatting rules * [hist] Corrected extremly unlikely memory leak Travis-CI warned about. Corrected extremly unlikely memory leak Travis-CI warned about in the drawing method of TGraphMultiErrors. | 21 January 2020, 09:08:32 UTC |
| e8eb836 | Stephan Hageboeck | 20 January 2020, 10:24:29 UTC | [RF] Fix reducing a dataset with multipe ranges. [ROOT-10517] When reducing a RooDataSet, selecting multiple ranges separated by ',' was ignored. (cherry picked from commit 1da7698b7cbeccccc52fd9d66f826b92fdb521b8) | 21 January 2020, 08:40:55 UTC |
| b354d0f | Stephan Hageboeck | 13 January 2020, 10:18:06 UTC | [RF] Fix filled plotting of RooCurve. [RF-8932] When a RooCurve is plotted as a filled polygon, one point has to be added at the lower left of the plot to close the polygon. A few lines of code have been cleaned up and simplified. | 20 January 2020, 16:33:59 UTC |
| b0b4ca1 | Guilherme Amadio | 17 January 2020, 10:22:49 UTC | Update release notes | 20 January 2020, 16:29:29 UTC |
| 2c26355 | Olivier Couet | 20 January 2020, 12:17:00 UTC | Fix release notes formatting [skip-ci] (#4814) | 20 January 2020, 12:17:00 UTC |
| ea5cc2f | Vassil Vassilev | 17 January 2020, 13:41:51 UTC | Explicitly disable autoloading when processing TClass rules. This should resolve ROOT-10514. | 20 January 2020, 11:10:42 UTC |
| 353b1a2 | Axel Naumann | 19 January 2020, 09:26:40 UTC | [relnotes] Fix alphabet; wording improvement (#4809) | 19 January 2020, 09:26:40 UTC |
| a5fff04 | Hadrien G | 17 January 2020, 18:17:10 UTC | Add my RHist contributions to release notes | 17 January 2020, 20:56:13 UTC |
| fa7564d | Stephan Hageboeck | 17 January 2020, 13:37:20 UTC | [RelNotes] Add RooFit release notes. | 17 January 2020, 14:29:35 UTC |
| 190e1c9 | Enric Tejedor Saavedra | 17 January 2020, 09:34:09 UTC | Add Philip to contributors | 17 January 2020, 14:28:37 UTC |
| 5208472 | Enric Tejedor Saavedra | 17 January 2020, 09:22:17 UTC | Add Massimiliano and Philip to the credits | 17 January 2020, 14:28:37 UTC |
| 35a7f3e | Enric Tejedor Saavedra | 17 January 2020, 09:13:22 UTC | [JSROOT] Add release notes: support for JupyterLab | 17 January 2020, 14:28:37 UTC |
| 5676b99 | Massimiliano Galli | 15 January 2020, 09:14:04 UTC | [Exp PyROOT/thisroot.sh] Added missing part of the path that did not allow 'out of the box' correct source Checking only for the existance of 'CMakeFiles' without including ROOTSYS in the path was causing the addition of wrong paths in case of out-of-the-box source | 17 January 2020, 14:13:47 UTC |
| 4ba9251 | Oksana Shadura | 15 January 2020, 12:22:59 UTC | Fix installation of PCMs We are excluding directories, which are accidentaly copied via unxpected behaviour of install(DIRECTORY ..) | 17 January 2020, 11:23:18 UTC |
| e183216 | Vassil Vassilev | 13 January 2020, 11:54:48 UTC | [cxxmodules] Improve ROOT startup time by not merging identifier tables Each module has a set of identifier tables which aid lookup. Based on this information clang decides if it needs a declaration to be deserialized. Namespace partitions and other C++ entities may have semantically the same identifier lookup tables across multiple modules. Since lookup is a heavily used operation in compilers clang tries to optimize it as much as possible. In case it sees more than 4 such lookup tables it merges them together into a single table aiming to keep the lookup algorithmic complexity of O(1). This logic approach has several assumptions: * The progam will use only a small superset of the modules it needs; * The program will be compiled in multiple TUs and merging of tables will not be called often; In the interpreter context where we make all module available the merging of such tables becomes CPU intense operation at runtime which produces a lot of temporary reallocations. Moreover, we have seen a lot of profiles where the merging operation dominates (by around 18%). This patch tries to make the merging far less often. On some short benchmarks ran locally we get (70-80%) runtime improvement and ~10% reduction in memory. | 17 January 2020, 05:13:58 UTC |
| 65ac0ad | Oksana Shadura | 14 January 2020, 15:37:09 UTC | Update comment for NO_CXXMODULE *_rdict.pcm installation case Co-Authored-By: Vassil Vassilev <v.g.vassilev@gmail.com> | 16 January 2020, 11:47:54 UTC |
| dd0eeb9 | Oksana Shadura | 14 January 2020, 13:22:46 UTC | Restore installation of _rdict.pcms in case if NO_CXXMODULE in ROOT_GENERATE_DICTIONARY() was enabled | 16 January 2020, 11:47:54 UTC |
| 1ae29a4 | Oksana Shadura | 14 January 2020, 13:19:43 UTC | Remove malicious cleanup of CMAKE_INSTALL_PREFIX/lib/python* | 16 January 2020, 11:47:54 UTC |
| baef706 | Enric Tejedor Saavedra | 15 January 2020, 13:47:06 UTC | [RDF] Release notes for CSV data source | 16 January 2020, 08:21:43 UTC |
| deacf90 | Enric Tejedor Saavedra | 15 January 2020, 13:43:52 UTC | [PyROOT] Complete release notes | 16 January 2020, 08:21:43 UTC |
| 5f7eb06 | Massimiliano Galli | 14 January 2020, 09:59:04 UTC | [Exp PyROOT] Release notes: multi-Python versions build and installation | 15 January 2020, 12:30:07 UTC |
| b3b1c4f | Axel Naumann | 10 January 2020, 15:51:25 UTC | [relnotes] Formatting. (cherry picked from commit 00229970a4ff39e56296db3a1d25206e656872bf) | 10 January 2020, 16:08:32 UTC |
| 440c630 | Axel Naumann | 10 January 2020, 16:03:49 UTC | Update ROOT version files to v6.19/03. | 10 January 2020, 16:08:01 UTC |
| e1e2219 | Axel Naumann | 10 January 2020, 11:51:51 UTC | Update ROOT version files to v6.19/02. | 10 January 2020, 11:51:51 UTC |
| 103b9a4 | Axel Naumann | 10 January 2020, 09:29:36 UTC | [rootx] Update core team list. | 10 January 2020, 09:29:36 UTC |
| 7de93c4 | Axel Naumann | 10 January 2020, 09:29:03 UTC | [rint] Happy 2020, ROOT! | 10 January 2020, 09:29:03 UTC |
| e6a787c | Axel Naumann | 10 January 2020, 08:26:50 UTC | [relnotes] Update ETA for 6.20/00 | 10 January 2020, 08:26:50 UTC |
| b60a8ce | Enric Tejedor Saavedra | 09 January 2020, 16:00:43 UTC | [Exp PyROOT] Add patch for TROOT::Initialize in clingwrapper The following PR introduced some changes in the TCling shutdown logic: https://github.com/root-project/root/pull/4675 One of these changes affects clingwrapper.cxx in Cppyy, but the functionality it relies on (TROOT::Initialize) is not yet in the mini-ROOT of Cppyy, which is at the moment working with ROOT 6.18, so we cannot push the clingwrapper change to upstream Cppyy now. For that reason, until Cppyy updates its ROOT to 6.20, we need to keep this patch to modify our copy of Cppyy to invoke TROOT::Initialize in the constructor of ApplicationStarter. | 09 January 2020, 16:08:52 UTC |
| 8630a93 | Stephan Hageboeck | 08 January 2020, 16:54:15 UTC | [RF] Modernise loops in RooChangeTracker. | 09 January 2020, 15:29:49 UTC |
| 2db7648 | Stephan Hageboeck | 08 January 2020, 15:09:34 UTC | [RF] Enable serialisation of RooNDKeysPdf. [ROOT-10341] To enable serialisation, several members of RooNDKeysPdf had to be removed or their type had to be changed. A few members cannot be serialised and therefore have to be recomputed when necessary. Some first cleaning up was also done. | 09 January 2020, 15:29:49 UTC |
| e062269 | Stefan Wunsch | 09 January 2020, 12:25:17 UTC | [PyROOT] Protect iteration for RVec<bool> Since RVec<bool> falls back to vector<bool>, both share the same issues for iterating over the container. | 09 January 2020, 14:48:57 UTC |
| 3be096c | Stefan Wunsch | 07 January 2020, 16:14:58 UTC | [ROOT-10491] Don't set tp_iter for vector<bool> tp_iter is used to implement an iterator protocol for Python using the "data" member function of std::vector. However, the specialization vector<bool> is not required to have this member function, which breaks the iterator. Removing the tp_iter field does not break iterating vector<bool> in Python, it falls back to the old iterator mechanism via the get/setitem special functions. | 09 January 2020, 14:48:57 UTC |
| 2fd6114 | Stefan Wunsch | 29 December 2019, 12:06:08 UTC | [ROOT-10491] Add RDataFrame.AsNumpy test for reading boolean | 09 January 2020, 14:48:57 UTC |
| 976e833 | Sergey Linev | 09 January 2020, 13:42:45 UTC | [rbrowser] use std::multimap for drawable provider registry While nullptr entry may appear several times | 09 January 2020, 13:53:23 UTC |
| ebf3b2e | Stephan Hageboeck | 06 January 2020, 16:53:15 UTC | [RF] Improve documentation and robustness of RooStats tools. - Add error message in ToyMCSampler when trying to generate events for a simultaneous PDF that is not extended. (Lead to crash before) - Prevent crash in RooAbsTestStatistic when dataset cannot be split into categories. - Prevent nans in HypoTestInverter when scanning with logarithmic step sizes. - Improve documentation and error messages of ToyMCSampler, RooSimultaneous, RooSimSplitGenContext. | 09 January 2020, 12:31:53 UTC |
| ce8b9c3 | Stephan Hageboeck | 16 December 2019, 14:48:20 UTC | [RF] Stabilise RooJeffreysPrior. Since it is running an internal fit, RooJeffreysPrior is unstable when a parameter is close to its boundary. To fix this, the prior PDF is now cloned, and the ranges of its parameters are extended to make fits converge even at the boundary. The JeffreysPrior tutorial has been renamed and cleaned up. | 09 January 2020, 12:31:53 UTC |
| c4014ce | Stephan Hageboeck | 16 December 2019, 14:47:51 UTC | [RF] Add BuildLegend function to RooPlot. | 09 January 2020, 12:31:53 UTC |
| e5a9e9a | Stephan Hageboeck | 16 December 2019, 14:43:30 UTC | [RF] Fix formatting and documentation, rename MakeVerbose helper. - Fix the formatting of RooCache* comments and replace old loops. - Replace the 'MakeVerbose' helper by 'LocalChangeMsgLvl'. Instead of just making a section verbose, this RAII object allows to temporarily select any message level. - Fix formatting and replace old loops in RooAbsReal. | 09 January 2020, 12:31:53 UTC |
| 1a9de54 | Stephan Hageboeck | 16 December 2019, 14:38:20 UTC | [RF] Convert RooRealProxy into a template class. RooRealProxy can only take RooAbsReal* as payloads. Using the payload almost always requires casting. Therefore, the proxy was converted into a template class such that different types can be stored and retrieved without the need to cast. Typedefs for the most frequent use cases anticipated have been added: RooRealProxy (= RooTemplateProxy<RooAbsReal>), RooPdfProxy, RooLVarProxy, RooRealVarProxy. A simple test for reading a legacy RooExponential was added, which relies on correctly converting an old RooRealProxy to a RooTemplateProxy<RooAbsReal>. Removed the RooRealProxy.cxx implementation and create a legacy header that just includes RooTemplateProxy.h RooRealProxy further had functions that assume that the proxied object is a RooAbsRealLValue or a more-derived type. That's generally unsafe, so the functions have been deprecated (visible after #define R__SUGGEST_NEW_INTERFACE). Where appropriate, RooRealProxy instances should be replaced with a RooTemplateProxy and choosing the proper type for the template argument. This e.g. was done in RooPullVar. | 09 January 2020, 12:31:53 UTC |
| fdee445 | Stephan Hageboeck | 19 December 2019, 09:08:30 UTC | [RF] Prevent infinite loop in RooIntegrator1D. RooIntegrator1D could get stuck in an infinite loop if the integration range was exactly zero. Now, it immediately return zero if this happens. Further, the code was cleaned up and formatted. | 09 January 2020, 12:31:53 UTC |
| 9a540f3 | Stefan Wunsch | 07 January 2020, 14:28:31 UTC | [PyROOT exp] Disable warnings -Wmissing-field-initializers for py3.8 Warnings appear due to PEP 590 adding the tp_vectorcall field to some structs which remain uninitialized in CPyCppyy Following the conventions upstream, because the standard ensures a zero-initialization, it's safe the leave them uninitialized such as also done in the Python codebase itself. See discussion here: https://bitbucket.org/wlav/cppyy/issues/186/warnings-with-python-38-due-to-vectorcall | 09 January 2020, 10:47:16 UTC |
| 34fe0a9 | Vassil Vassilev | 09 January 2020, 07:34:10 UTC | [cxxmodules] Fix the error in std::complex on OSX. libc++ defines a macro which we need to hide from the modules system. | 09 January 2020, 10:42:30 UTC |
| 3a6172c | Olivier Couet | 09 January 2020, 10:29:38 UTC | Simplify ratio-plot examples. | 09 January 2020, 10:29:52 UTC |
| 55e6019 | Sergey Linev | 09 January 2020, 09:14:37 UTC | [ntuple] avoid windows warning | 09 January 2020, 10:17:05 UTC |
| 32b215d | Sergey Linev | 09 January 2020, 09:13:04 UTC | [win32gdk] cast char* to unsigned to use in printf Avoid warnings | 09 January 2020, 10:17:05 UTC |
| 2383ee8 | Sergey Linev | 09 January 2020, 09:10:51 UTC | [win32gdk] use explicit type casting to avoid warnings | 09 January 2020, 10:17:05 UTC |
| cd7114d | Sergey Linev | 08 January 2020, 07:19:59 UTC | [treeplayer] exclude TMPWorkerTree from dictionary on Windows Class excluded from compilation on Windows Supress warning on Windows | 09 January 2020, 10:17:05 UTC |
| 65c346d | Sergey Linev | 09 January 2020, 10:03:19 UTC | [rbrowser] use binary mode to read image files | 09 January 2020, 10:03:51 UTC |
| 4aaeaa8 | Sergey Linev | 09 January 2020, 09:03:56 UTC | [rbrowser] let configure file filters in dialog from client side | 09 January 2020, 10:03:51 UTC |
| a7cd8bb | Sergey Linev | 09 January 2020, 07:24:38 UTC | [rbrowser] implement filters in FileDialog Now one can condigure different name filters Regexp is used for file name matching, therefore more complex rules can be later applied | 09 January 2020, 10:03:51 UTC |
| edb1d39 | Sergey Linev | 08 January 2020, 13:26:28 UTC | [rbrowser] better support of windows drivers in browser | 09 January 2020, 10:03:51 UTC |
| 38d90a1 | Sergey Linev | 08 January 2020, 11:24:54 UTC | [rbrowser] use gSystem->GetVolumes to list all volumes | 09 January 2020, 10:03:51 UTC |
| 412536f | Sergey Linev | 08 January 2020, 11:32:43 UTC | [rbrowser] ignore upper/lower case on Windows for file names | 09 January 2020, 10:03:51 UTC |
| 30c18f0 | Sergey Linev | 08 January 2020, 09:55:57 UTC | [rbrowser] provide method to fill files top folder for RBrowsable Such entries used in RBrowser and RFileDialog to navigate through file system | 09 January 2020, 10:03:51 UTC |
| 2f0e997 | Sergey Linev | 07 January 2020, 17:17:38 UTC | [rbrowser] correctly initialize RFileDialog on Windows Require special set of existing drive letters. For now limit by working dir. | 09 January 2020, 10:03:51 UTC |
| e4bb25b | Sergey Linev | 08 January 2020, 11:24:54 UTC | [rbrowser] use gSystem->GetVolumes to list all volumes | 09 January 2020, 10:03:51 UTC |
| c280d65 | Sergey Linev | 08 January 2020, 09:55:57 UTC | [rbrowser] provide method to fill files top folder for RBrowsable Such entries used in RBrowser and RFileDialog to navigate through file system | 09 January 2020, 10:03:51 UTC |
| d3c12cd | Sergey Linev | 08 January 2020, 08:41:16 UTC | [rbrowser] fix bug in file dialog initialization | 09 January 2020, 10:03:51 UTC |
| d3a3fcb | Sergey Linev | 08 January 2020, 08:26:08 UTC | [http] handle symbolic links on Windows when sending file | 09 January 2020, 10:03:51 UTC |
| e42570d | Sergey Linev | 07 January 2020, 17:17:38 UTC | [rbrowser] correctly initialize RFileDialog on Windows Require special set of existing drive letters. For now limit by working dir. | 09 January 2020, 10:03:51 UTC |
| a62024f | Sergey Linev | 07 January 2020, 15:37:16 UTC | [rbrowser] handle situation when initial message appears before dialog initialized | 09 January 2020, 10:03:51 UTC |
| 47f2b4e | Sergey Linev | 06 January 2020, 16:10:15 UTC | [rbrowser] introduce filetype selection in RFileDialog | 09 January 2020, 10:03:51 UTC |
| 9eb4491 | Sergey Linev | 06 January 2020, 15:03:23 UTC | [rbrowser] add Browsable namespace to LinkDef | 09 January 2020, 10:03:51 UTC |
| 1d2e767 | Massimiliano Galli | 08 January 2020, 14:14:40 UTC | [Exp PyROOT] Remove directory wrongly created in installation directory For reasons that we need to clarify, when cmake install is run, all the directories inside {CMAKE_LIBRARY_OUTPUT_DIRECTORY} are copied inside ${CMAKE_INSTALL_LIBDIR}. This affects also lib/pythonX.Y, which in the case of installation directory should not exist at all. Here we manually remove them after installation. This commit can be removed once we figure out why this happens. | 09 January 2020, 10:02:45 UTC |
| 52d8ff8 | Massimiliano Galli | 08 January 2020, 08:58:20 UTC | [Exp PyROOT/thisroot.sh] EXP_PYROOT variable passed by command line that covers the case of instalallation N.B.: will go away once experimental will become the default | 09 January 2020, 10:02:45 UTC |
| 996762c | Massimiliano Galli | 17 December 2019, 22:52:00 UTC | [Exp PyROOT] Create the eggs only in a dummy way for simplicity | 09 January 2020, 10:02:45 UTC |
| 6ddc793 | Massimiliano Galli | 07 January 2020, 15:45:38 UTC | [Exp PyROOT/thisroot.sh] Conditional silencing of ZSH nomatch option Due to the way in commit 4700 we changed how clean_environment works (we get rid of $pyroot_dir and all the cases for lib/python*), we got back to the ZSH problem solved in commit 4693. In order to solve it once and for all, we proceed like the following: at the beginning of the clean_environment function we check whether we are using ZSH; if so, we check if nonomatch is enabled; if it is already enabled (behavior that we want), we don't do anything; if it's not enabled, we enable it and set a variable that reminds us to disable at the end. | 09 January 2020, 10:02:45 UTC |
