From 1d861dc67f3542c691565afc4337696d0ffd0aa9 Mon Sep 17 00:00:00 2001 From: "Joachim Wuttke (h)" <j.wuttke@fz-juelich.de> Date: Sun, 4 Oct 2015 12:56:43 +0200 Subject: [PATCH] update/sort TODO --- TODO | 40 ++++++++++------------------------------ 1 file changed, 10 insertions(+), 30 deletions(-) diff --git a/TODO b/TODO index 90407bb5..7badc5b4 100644 --- a/TODO +++ b/TODO @@ -4,23 +4,16 @@ allow fit if curve has less slices than data? prevent fits for slices with m/ set; allow global fits for the remaining slices parallelize global fits -msa vs mpa sample stdv - cp: table header placement, correct 1-R^2 store status cg, cx of curve files -rework dy, dp - - == To investigate == -boost::shared_ptr -> C++11 ? boost::format -> cppformat ? == Current refactoring (everything becomes an object) == -get rid of has_err() simplfy tree_ and curve_ calls Reimplement fn key binding @@ -70,10 +63,6 @@ dialog generics - restore help on "?" (e.g. expression help for 'md') - default for plot range -How to treat blanks in user input (waiting for CLUE): -- operators "mod", "div" suggest to allow blanks in expressions -- "op* 1 .3 99" suggeststo split at blanks - == Expressions and functions == @@ -89,16 +78,14 @@ mfj, cp for curve file valgrind complains about NPloWin -> transform into a class with destructor? -resol(shift) requires resolution grid to be cut down so that - shifted resolution grid lies inside the data grid. -conv(..,shift) does not require this. -It's an inconsistency! - fs->fl: - broken for fit-with-external-program (cc ecr or similar) - y08 is not true YAML. Don't use text for data. +update gnuplot symbol explanation + + == CLEANUP == - read-plus Interface in EINEN namespace ask:: stecken @@ -125,15 +112,12 @@ allow ad-hoc addition of fit functions - plot accessible Q-E-space improve existing methods -- 'oio' wenn mehrere Files mit rank=2, automatisch in einen File packen +- 'oi' wenn mehrere Files mit rank=2, automatisch in einen File packen - 'oixy': auch 'oixyd' distribution - include wupscat, wupsbb - rewrite wupscat in bash instead of ruby -darker colors in gnuplot - - == PROJECTS == detailed balance @@ -160,6 +144,12 @@ coord name algebra == WAITING FOR CLUE == +rework dy, dp + +How to treat blanks in user input: +- operators "mod", "div" suggest to allow blanks in expressions +- "op* 1 .3 99" suggeststo split at blanks + should operations have a return value? - code for user interrupt - numeric result? @@ -185,13 +175,3 @@ script language integration ROOT integration - well-defined histogram channels - - -== The YAML problem == - -[Antti, 23oct14:] -There is another issue for your consideration: we're currently using the old yaml-cpp API. At least on my system yaml-cpp libraries for both new and old APIs cannot live together. It's not a problem while non of my installed packages requires the latest version. However, other users may have a different situation. - -We could, of course, rewrite the code to use to new API but this would be a lot of work. Luckily I have plan B. - -I have already a working implementation for cmake for building the old yaml-cpp as an external project for the project's internal use. The bundled old library is automatically included in the build process if yaml-cpp is missing or user specifies a -Duse_bundled_yaml_cpp=ON flag. We don't need to include the yaml-cpp sources themselves in the tree since cmake can download the directly from the web as needed. We could use this solution in the projects requiring yaml. -- GitLab