Thanks a lot wolke, I'll let you deal with the undo stack fo rnow ;) I do think too that it's better to move tmp patterns from disk to memory.
About the virtual pattern, to be honest I discovered the functionality while working on the code, so I only tested it once ... seems pretty usefull! About the code, it seems to me that there is some code duplication, useless parts, and so on. I'll check it deeper. On Thursday 03 June 2010 06:12, [email protected] wrote : > Am Wed, 2 Jun 2010 22:12:20 +0200 > schrieb Jérémy Zurcher <[email protected]>: > > > Hi there, > > > > /***************/ > > right click on a pattern and load a new one, > > a pattern is saved through : int LocalFileMng::savePattern( ... ); > > mode = 4 => sPatternXmlFilename = patternname; > > > > so it's saved within current dir not tmp dir used for undo/redo. > > > > As I'm not working on this actually I'll let who wants to fix it ;) > we can remove mode 4. its the same than mode 2. and both modes use the > filename which > comes as third argument. > > the function who use mode 4 contains as third argument the pattern-filename. > for example in songeditor.cpp line 1463: > QString patternFilename = Preferences::get_instance()->getTmpDirectory() > +QString("%1").arg(thetime)+ QString( "PAT.xml" ); > > you see the name include the h2 tmp dir. also this dir will send as argument > to the undo stack. > line 1468: > SE_copyPatternAction *action = new SE_copyPatternAction( patternFilename , > pPatternList->get_size() ); > later its possible to undo this, because the clear filename is stored into > undostack. > > i think i have forgotten to revert the mode 4 into LocalFileMng if i have > finished subsequent undo functions. > sometimes the functions i have create for undo in a earlier stage of > development need a > rewrite because later on a other place in code arise new dependency's to hold > the undo > stack consistent. > > what ever. i also think it is better to move this tmp pattern from disk to > memory. if i create the function i think its important to save memory in undo > stack. > but patterns are very small and now i think its much better to copy this > tmp-pattern into the undo stack. > > > > > > /***************/ > > Questions about Pattern : > > > > std::set<Pattern*> virtual_pattern_set; > > keeps a list a virtual patterns. > > > > but what is exactly > > std::set<Pattern*> virtual_pattern_transitive_closure_set; > > for ?? > > > > void addEdges(H2Core::Pattern::virtual_patterns_t &patternSet); > > seems weird to me, is the recurtion usefull ??? > > > > I definitly miss something, but don't have time to bang my head on this. > > > > Someone could enlighten me please ?? > > hmm, > i don't understand your question. do you don't understand the use of the > virtual pattern function himself. if not i hope this help. > the virtual pattern thing is really magic. i like it very much. if you are > one, who > works with stacked pattern into a song creation, you will love this function > which > simplify a song-composition in songmode and save a lot of clicking. > you can define a pattern as virtual pattern/meta pattern and allocate him > some pattern you have already created and which contains parts of your song > composition. > > e.g.: > verse = pattern 1 + 2 = virtual-pattern 10 > verse variation = pattern 1 + 2 + 4 = virtual pattern 11 > chorus = pattern 1 + 3 = virtual-pattern 12 > chorus variation = 1 + 3 + 5 = virtual-pattern 13 > and so on... > now you can click only on your meta pattern to activate a stack for a > song-position. > > and if i am wrong and you already know this, i can not say so much about the > code. it was a patch from a user. > i only have rewrite some parts around to make stacked pattern undo able. > > greetings wolke > > > > > thanks in advance ... > > > > greetings > > > > ------------------------------------------------------------------------------ > ThinkGeek and WIRED's GeekDad team up for the Ultimate > GeekDad Father's Day Giveaway. ONE MASSIVE PRIZE to the > lucky parental unit. See the prize list and enter to win: > http://p.sf.net/sfu/thinkgeek-promo > _______________________________________________ > Hydrogen-devel mailing list > [email protected] > https://lists.sourceforge.net/lists/listinfo/hydrogen-devel > > !DSPAM:4c07112f243001314616602! > -- Jérémy Zurcher av General Guisan 1400 Yveron-les-bains +41 (0) 79 599 84 27 ------------------------------------------------------------------------------ ThinkGeek and WIRED's GeekDad team up for the Ultimate GeekDad Father's Day Giveaway. ONE MASSIVE PRIZE to the lucky parental unit. See the prize list and enter to win: http://p.sf.net/sfu/thinkgeek-promo _______________________________________________ Hydrogen-devel mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/hydrogen-devel
