ArcEm Display refactoring

2006-03-10 Thread Matthew Howkins
I have just committed to CVS a patch to refactor some of the display code in ArcEm: Summary: Factor out common parts of functions DisplayKbd_Init() and DisplayKbd_XPoll(), (and rename DisplayKbd_XPoll() to DisplayKbd_Poll()). Details: 1. Rename DisplayKbd_XPoll() to DisplayKbd_PollHost() in e

Re: ArcEm Display refactoring

2006-03-10 Thread Ralph Corderoy
Hi Matthew, > PS Anyone know why my CVS commit generated so many postings to the > arcem-cvs mailing list? Commits under CVS aren't `atomic'. That's long been stated as a CVS flaw since if something goes wrong half-way through then your changes can be half-committed. It does it a directory at

Re: ArcEm Display refactoring

2006-03-10 Thread Ian Jeffray
Matthew Howkins wrote: I have just committed to CVS a patch to refactor some of the display code in ArcEm: Good job. Could I request that Michael Dales and Ian Jeffray check that the MacOS, RISC OS, and GP2x ports are still working, and that I haven't broken them. Shall do so soon. PS Any

Re: ArcEm Display refactoring

2006-03-10 Thread Ian Jeffray
Matthew Howkins wrote: I have just committed to CVS a patch to refactor some of the display code in ArcEm: 5. Move static variables enodes and xpollenode, and #define POLLGAP from each */DispKbd.c to DispKbdShared.c Uh... 7. Common code moved from DisplayKbd_PollHost() to DisplayKbd_Poll()

Re: ArcEm Display refactoring

2006-03-10 Thread Ian Jeffray
Ian Jeffray wrote: The code you removed was *NOT* common as I had tweaked around with it heavily to try and get at least some semblence of performance out of the gp2x port. I've made the PollHost return a flag to say if the queue should be run, however... The RISC OS port also appeas to be b

Re: ArcEm Display refactoring

2006-03-10 Thread Ralph Corderoy
Hi Ian, > That list is screwed anyway. I, and others, cannot subscribe. After I heard nothing more from you I assumed you'd subscribed. I've now tried subscribing from three different emails dotted around the Internet. All worked as expected. I'm filling in the http://lists.sourceforge.net/l

Re: ArcEm Display refactoring

2006-03-10 Thread Ian Jeffray
Ian Jeffray wrote: Ian Jeffray wrote: Nor does it on the gp2x. I think you may have broken something fundamental here. :-( D'oh. Doesn't compile for Windows any more either! Some missing includes or something. Looking at that, too... Actually, it appears to be simply a missing include

Re: ArcEm Display refactoring

2006-03-10 Thread Ian Jeffray
Ian Jeffray wrote: Nor does it on the gp2x. I think you may have broken something fundamental here. :-( D'oh. Doesn't compile for Windows any more either! Some missing includes or something. Looking at that, too... I. --- This SF.Ne

Re: ArcEm Display refactoring

2006-03-10 Thread Peter Howkins
On Fri, Mar 10, 2006 at 06:37:51PM +, Ian Jeffray wrote: > Ian Jeffray wrote: > >Ian Jeffray wrote: > > > >>Nor does it on the gp2x. I think you may have broken something > >>fundamental here. :-( > > > >D'oh. Doesn't compile for Windows any more either! Some missing > >includes or somethin

Re: ArcEm Display refactoring

2006-03-10 Thread Peter Howkins
On Fri, Mar 10, 2006 at 05:32:31PM +, Ian Jeffray wrote: > > The RISC OS port also appeas to be broken... keyboard/mouse input > doesn't work now. Checking the recent diffs of the riscos and generic code, I spotted a line that was missing in generic, yet was in the risc os version. http://cv

Re: ArcEm Display refactoring

2006-03-10 Thread Matthew Howkins
> Inside DisplayKbd_Init() try adding this line to the shared version > DC.AutoRefresh=AUTOREFRESHPOLL; /* Surely this is set somewhere??? */ > > Might be a place to start? The same line is missing from the gp2x version > too, so that might be the same problem. I spent a long time studying that li