ok, this sounds good could you post it at the end of the existing patch (iirc the original poster is MIA anyway)
i'll have a look at it ASAP, but no promises when that will be... quick question, why didn't you use the following definition for skip : "skip whatever it takes to bring me to the game itself" which seems more logical since skip is mainly used when you already know the story parts... the "double skip to reach game" logic you currently have seems a bit complicated On Mon, Apr 5, 2010 at 4:39 AM, Ed <icelus...@gmail.com> wrote: > There was a slight problem with this I noticed (inevitably) right > after I sent this mail. > > The problem is if you press skip rather than next on the last page of > the last story segment, it doesn't do anything. > > This isn't actually a trivial fix, because of the question of what > skip should really mean. > > The definition I used is this: skip means go to the end of the story > (making the fact that it contains parts hidden from the user) unless > already at the end of the story, in which case go to the game. > > The changes accomplish this by virtue of making the controller > slightly cleverer; it now knows whether its the last segment, and on > that basis will either skip internally (to find the end), skip > externally (to find another segment which is the end) or indicate the > game should be entered (if its the last and at its own end). > > The underlying assumption behind this patch really is that the story > screens are a single "book" as they appear to the naive user, grouped > into parts for implementation reasons and convenience. Essentially > this patch enforces that model for back and skip. > > There _should_ be a line using > > next_itor = itor+1 > > rather than > > next_itor = ++itor; ---itor; > > However touching config.hpp takes 2+ hours to rebuild on my machine, > so I'm sending you what I have for now (operator+ is unimplemented; i > found out the time requirement by implementing operator--). > > The more succinct > > next_itor = ++itor--; > > blows the compiler's mind (compiles but segfaults in resource tracking code). > > > > On 5 April 2010 02:52, Ed <icelus...@gmail.com> wrote: >> Here's a first attempt at a patch for the back button for story screens: >> >> http://wiki.wesnoth.org/Good_Ideas >> http://www.wesnoth.org/forum/viewtopic.php?t=8857 >> >> There is already a patch in the tracker for this: >> >> https://gna.org/patch/?1536 >> >> However I didn't realise someone had started when I wrote this, and >> this is rather further along than that one is. >> >> Problems with that patch: >> >> * Back doesn't work properly (try it on the last page of the intro >> where it shows you the map) - can't go back. >> * Doesn't wire up back button in all the situations next button is clickable. >> * Probably violates screen size in USE_TINY_GUI mode >> * There's no key shortcut for back. >> >> Features this patch has: >> >> * Handles back properly so that you can go all the way back to the >> start from the end. >> * Handles skip properly so that all segments are skipped not just per >> segment (but can still go back so this is ok!) >> * Adds functional quit button. >> * Adds key controls: RIGHT is added for next, LEFT and BACKSPACE for back. >> >> Improves the implementation: >> * no longer throws quit exception to quit (was just caught and >> suppressed anyway). >> * play story segments from either end so that back can work properly. >> * adds reverse direction iteration for config::const_child_iterator (was >> needed) >> > > _______________________________________________ > Wesnoth-dev mailing list > Wesnoth-dev@gna.org > https://mail.gna.org/listinfo/wesnoth-dev > > _______________________________________________ Wesnoth-dev mailing list Wesnoth-dev@gna.org https://mail.gna.org/listinfo/wesnoth-dev
