- 02 May, 2021 1 commit
-
-
Albert Astals Cid authored
I think it's a bug, but now sure if of libstdc++ or the spec, have opened https://gcc.gnu.org/bugzilla/show_bug.cgi?id=100384 for the gcc people to comment on
-
- 13 Jan, 2019 1 commit
-
-
Laurent Montel authored
-
- 24 Feb, 2013 3 commits
-
-
David Narváez authored
-
David Narváez authored
Wrong interpretation of the braces This reverts commit c12194d5.
-
David Narváez authored
-
- 21 Sep, 2006 1 commit
-
-
Thomas Häber authored
svn path=/trunk/KDE/kdeedu/kig/; revision=587176
-
- 07 Dec, 2005 1 commit
-
-
Maurizio Paolini authored
svn path=/trunk/KDE/kdeedu/kig/; revision=486409
-
- 05 Oct, 2005 1 commit
-
-
Dirk Mueller authored
svn path=/trunk/KDE/kdeedu/kig/; revision=467420
-
- 16 Jul, 2005 1 commit
-
-
Dirk Mueller authored
svn path=/trunk/KDE/kdeedu/kig/; revision=435269
-
- 28 Feb, 2005 1 commit
-
-
Maurizio Paolini authored
- infrared vision (possibility to operate with hidden objects) - optimal "calcpath" algorithm svn path=/trunk/kdeedu/kig/; revision=393911
-
- 14 Feb, 2005 2 commits
-
-
Maurizio Paolini authored
misunderstood its semantics (my fault!). However, due to the freeze, the new code is presently disabled if favour to the old one; the new algorithm can be activated by "define"ing the "NEWCALCPATH" symbol in kig/misc/calcpaths.cc. svn path=/trunk/kdeedu/kig/; revision=389363
-
Maurizio Paolini authored
is worse than what it is meant to fix I am temporarily reverting to old behaviour till properly fixing the problem :-( svn path=/trunk/kdeedu/kig/; revision=389349
-
- 07 Feb, 2005 1 commit
-
-
Maurizio Paolini authored
had exponential complexity for some examples. svn path=/trunk/kdeedu/kig/; revision=387217
-
- 14 Dec, 2004 1 commit
-
-
Pino Toscano authored
* More tooltips and "what's this" to some dialog. * No more file name in every file. * More control on macro name: when loading from file, if a macro has no name, we'll assign it a bogus one (like "Unnamed macro #id"). Furthermore, we ensure that the user don't set an empty name for a macro in the Edit Types Dialog. * Other misc changes. svn path=/trunk/kdeedu/kig/; revision=370664
-
- 18 May, 2004 1 commit
-
-
Dominique Devriese authored
* Add a --convert-to-native command line option, which converts the given file to the native Kig file format ( without a GUI ). * Split up KigDocument into KigPart and KigDocument svn path=/trunk/kdeedu/kig/; revision=312389
-
- 02 Apr, 2004 1 commit
-
-
Dominique Devriese authored
objects using it cannot be built yet. I'll work on that next. CCMAIL:paolini@dmf.unicatt.it svn path=/trunk/kdeedu/kig/; revision=300637
-
- 28 Mar, 2004 1 commit
-
-
Dominique Devriese authored
This adds: - point styles - some i18n fixes - more work on the drgeo import filter CCMAIL:Pino Toscano <toscano.pino@tiscali.it> svn path=/trunk/kdeedu/kig/; revision=299221
-
- 18 Jan, 2004 1 commit
-
-
Dominique Devriese authored
* bump the version number to 0.7.1 * Work on the Cabri import filter, so that it actually becomes usable for some easier files.. * Improve the errors given by the "New Script Wizard", by making it get a proper error description from the python interpreter. * Add a lot of documentation to the new classes. * Fix the moving system again, it now only redraws exactly those objects that need to be redrawn. E.g. when a constrained point was moved, before it was assumed that all of its parents, and their children would move, whereas in reality, a constrained point does not move the curve it is constrained to. This is now taken into account for. This much optimizes the case where we move the constrained point in examples/sine-curve.kig. * Added a DESIGN document, documenting the Kig object system design. I think that is rather finished now, and it's probably about time to try and make some other people than myself get it ;) * Remove support for the ancient pre-0.4 file format that we still supported opening. If you still have old files around using it, you should convert them to the new format, by opening them with a Kig version between 0.4 to 0.6, and re-saving them. Maintaining compatibility with these old files doesn't seem very useful, because I don't think there are many files in this format available, and therefore I didn't think it was worth the trouble of porting the code to the new object system. * Introduce a new file format that matches the new object system better. Files in the old format can still be opened seamlessly. * Add undo/redo support for changing visible aspects of an object ( size, color, shown state ). In the new system, this was as easy as replacing the ObjectDrawer of an ObjectHolder with another one. * Another ( hopefully the last ) major change to the object system. Decouple the link between how an object is calced and how it is drawn. We now have a hierarchy structure of ObjectCalcer's describing various objects and their interrelations. On top of that, there are the ObjectHolder's, which hold a link to an ObjectCalcer from the hierarchy, and keep an ObjectDrawer describing how to draw it. The document only keeps a list of ObjectHolder's, nothing else.. Maurizio, err, the changes that I think are relevant to you are: 1 improvements for the cabri file format filter 2 redesign of the classes, and a lot of work on documentation 3 improvement of the python scripting errors Have fun with it ;) cheers domi CCMAIL:paolini@dmf.unicatt.it svn path=/trunk/kdeedu/kig/; revision=280616
-
- 31 Aug, 2003 1 commit
-
-
Dominique Devriese authored
svn path=/trunk/kdeedu/kig/; revision=247401
-
- 26 Jul, 2003 1 commit
-
-
Dirk Mueller authored
svn path=/trunk/kdeedu/kig/; revision=239302
-
- 03 Jul, 2003 1 commit
-
-
Dominique Devriese authored
support a python compiled script ObjectImp in the future.. svn path=/trunk/kdeedu/kig/; revision=235247
-
- 04 Jun, 2003 1 commit
-
-
Dominique Devriese authored
* clean up the object parent-child relation mechanism. Rather large code cleanup, that simplifies a lot of code.. Needed a backwards-compatible file format extension. This commit now also adds proper treatment of internal objects, because it no longer relies on the inherently wrong isInternal() hack, but features The Correct Fix(tm). svn path=/trunk/kdeedu/kig/; revision=229606
-
- 08 May, 2003 1 commit
-
-
Dominique Devriese authored
svn path=/trunk/kdeedu/kig/; revision=224586
-
- 07 May, 2003 1 commit
-
-
Dominique Devriese authored
efficient now 2 some more undo support: redefining points and text labels is now undoable.. svn path=/trunk/kdeedu/kig/; revision=224476
-
- 06 May, 2003 1 commit
-
-
Dominique Devriese authored
construction dialog. this makes it support multi-line labels, and changing the parameters svn path=/trunk/kdeedu/kig/; revision=224249
-
- 22 Mar, 2003 1 commit
-
-
Dominique Devriese authored
svn path=/branches/make_kdeedu_cool/kdeedu/kig/; revision=215576
-
- 13 Feb, 2003 1 commit
-
-
Dominique Devriese authored
svn path=/branches/make_kdeedu_cool/kdeedu/kig/; revision=206737
-
- 08 Feb, 2003 1 commit
-
-
Dominique Devriese authored
svn path=/branches/make_kdeedu_cool/kdeedu/kig/; revision=205734
-
- 03 Feb, 2003 1 commit
-
-
Dominique Devriese authored
refactoring: remove the fixedArgs stuff, and replace it by a DataObject. This turns our dependency tree in something which is more easy to work with, as there are no more special cases for fixedArgs things... svn path=/branches/make_kdeedu_cool/kdeedu/kig/; revision=204599
-
- 02 Feb, 2003 2 commits
-
-
Dominique Devriese authored
revert the refactoring i did some time ago, which gave every object its own ObjectType, as this is Wrong (tm) + add a new CustomTypes class that keeps track of user-created types.. svn path=/branches/make_kdeedu_cool/kdeedu/kig/; revision=204454
-
Dominique Devriese authored
svn path=/branches/make_kdeedu_cool/kdeedu/kig/; revision=204379
-
- 20 Jan, 2003 1 commit
-
-
Dominique Devriese authored
Important changes are that: Object is split up into three parts: * ObjectImp is a class that does represents the Object's data. It does the drawing, stuff like contains() etc. * ObjectType is a class that calcs ObjectImp's from a certain number of parents.. * Object itself now is a simple class that contains one ObjectImp and one ObjectType, some data like width, color etc., and that forwards requests to one of the above.. Another set of changes is the representation of types. The type class is split up into: 1 a GUIAction class that is Kig's equivalent of a KAction. Will prolly make this inherit KAction directly, instead of havind that done by a special KigAction 2 a Constructible class. This represents something that wants a certain number of args of certain types, and constructs some objects with them.. 3 well, ObjectType above does part of the work previously done by the Type class, namely drawPrelim and friends.. The rest of the changes adapt the rest of kig to the new design. These changes are mainly for two things: 1 cleanup of the code: there was quite some boilerplate code and undue templates in kig, that are now removed. E.g. the artificial difference between drawPrelim and calc+draw is gone.. 2 necessary for the clean implementation new features: e.g. transformations of a circle can dynamically change type to conic when the transformation at hand is not a homothety, i'll add a bool enabled to constructibles some time in the future, so they can be disabled by the user ( e.g. some people have asked me for kig to only show ruler-compass stuff ), i'm thinking of adding scripting to kig some time in the future, and i think the current design allows this to be done easily... The redesign is mostly done, and what remains to be done is to reenable the things i disabled in the process of doing the change.. 1 i still need to hook up the GUIActions to the kigDocument again, so you actually see them 2 file loading/saving is disabled 3 macro's are disabled 4 locuses are disabled 5 transformations aren't finished 6 i need to make sure that i haven't changed any i18n'able strings when that wasn't necessary.. Because those things aren't finished yet, I'm putting this all in a new make_it_cool branch, where i can finish these things before they break HEAD.. I'd also feel more comfortable with this stuff in CVS, where it's harder to accidentally delete or something like that ;) Well, I think that about sums it up.. CCMAIL:paolini@dmf.unicatt.it Maurizio, I'm just CC'ing you to let you know what i'm doing, if you want, you can take a look at the new code, but it still needs a lot of work, and kig doesn't actually do anything useful atm ;) cheers domi svn path=/branches/make_kdeedu_cool/kdeedu/kig/; revision=201478
-
- 13 Oct, 2002 1 commit
-
-
Dominique Devriese authored
svn path=/trunk/kdeedu/kig/; revision=183702
-
- 25 Sep, 2002 2 commits
-
-
Dominique Devriese authored
Fixing one of the calcPath()'s.. Really strange how i never noticed how broken this algorithm really was... svn path=/trunk/kdeedu/kig/; revision=180413
-
Dominique Devriese authored
svn path=/trunk/kdeedu/kig/; revision=180347
-
- 18 Sep, 2002 1 commit
-
-
Dominique Devriese authored
CVS_SILENT and now changing my email address with a scripty too ( damn, i forgot to CVS_SILENT the previous commit... svn path=/trunk/kdeedu/kig/; revision=178736
-
- 03 Sep, 2002 1 commit
-
-
Dominique Devriese authored
"Curve" (??) ) 2 replaced ObjectHierarchy::calc() ( which resulted in certain objects being calced five times ( the locus in locustest.kig, most notably :( ) by an idea from KSeg. What we do is now to just sort the objects we need to calc in a suitable way. This is a bit more difficult than it sounds, and in the meanwhile, i think i've hit some sort of bug in my STL implementation (??), but it works now :) 3 indentation fixes svn path=/trunk/kdeedu/kig/; revision=175592
-