X-Authentication-Warning: delorie.com: mail set sender to geda-user-bounces using -f X-Recipient: geda-user AT delorie DOT com Date: Sun, 21 Feb 2016 16:01:20 +0100 (CET) X-X-Sender: igor2 AT igor2priv To: geda-user AT delorie DOT com X-Debug: to=geda-user AT delorie DOT com from="gedau AT igor2 DOT repo DOT hu" From: gedau AT igor2 DOT repo DOT hu Subject: [geda-user] libstroke revival Message-ID: User-Agent: Alpine 2.00 (DEB 1167 2008-08-23) MIME-Version: 1.0 Content-Type: TEXT/PLAIN; format=flowed; charset=US-ASCII Note-from-DJ: This may be spam Reply-To: geda-user AT delorie DOT com Hi all, during my pcb-rnd cleanup I decided to give libstroke a try. I found the code slightly bitrotten and defunct. Some of the problems most probably present in mainline as well: 1. libstroke's record function is not called at all when using the gtk hid; it's because ghid_note_event_location() doesn't call EventMoveCrosshair() but reproduces it's non-libstroke part locally. 2. the zoom related events that used to call SetZoom() are removed in mainline and marked with XXX comments; this is a minor problem as the rest of the libstroke features should work. 3. in the lesstif hid the middle button acts as pan and libstroke never gets the events (mid_stroke stays false) 4. libstroke type bug: libstroke uses type float for some internal calculations in the record call while the api accepts int; pcb sends it too large coordinates even on small boards; this makes libstroke fall in an infinite loop (!); can be fixed by shifting down the coordinates in the call to stroke_record(); I used >>16 My conclusion is that I am moving libstroke support into a plugin and it would partially work with the gtk hid. Regards, Igor2