Hi,
I've just started playing with McCLIM.
Below are some issues for McCLIM CVS on SBCL 1.0.12.17.
Please treat this as a bug report or tell me what I'm doing wrong (e.g. if CVS is known to be unstable).
Thanks, Daniel
(clim-demo::run-test 'clim-demo::demodemo)
CLIM-Fig:
Selecting "ellipse" and then clicking on the canvas results in "Attempt to invert the probably singular transformation #<CLIM-INTERNALS::STANDARD-HAIRY-TRANSFORMATION 0 0 0 0 266 185>. [Condition of type CLIM:SINGULAR-TRANSFORMATION]"
Bezier curves are very difficult to draw (lots of errors).
Calculator:
The buttons frequently don't respond. Also, most calculators start a new number when a number is entered following; this calculator appends to the current result.
For example, the keys "3*4=5" generally show "3", "4", "7", and "5". But I'm seeing "3", "4", "7", and "75".
Draggable Graph:
If I resize the window (usually a single maximize works) and then click a node, it disappears and SBCL says "There is no applicable method for the generic function #<STANDARD-GENERIC-FUNCTION CLIM:SHEET-NATIVE-TRANSFORMATION (6)> when called with arguments (NIL). [Condition of type SIMPLE-ERROR]"
The node disappears for good (can't select it) even though its connections refresh when connected nodes are dragged. Only the first dragged node after the resize is affected (unless I select several nodes before acknowledging the error; then all selected nodes disappear).
Misc tests: Empty Records 3
The circle is in the same location as for Empty Records 1 & 2. The bottom and right of the box are close to the circle as before. However, the top left corner is now 2 diameters from the circle.
Daniel Herring dherring@tentpost.com writes:
Please treat this as a bug report or tell me what I'm doing wrong (e.g. if CVS is known to be unstable).
It's not, but I think most of the bugs you report (certainly the Calculator one, and probably also the CLIM-FIG one) are bugs in the demo applications, not McCLIM itself. Thus, while they should probably be fixed, they don't have highest priority. But thanks for the bug reports anyway, I don't think many McCLIM hackers spend a lot of time running the demo programs, so bugs in them can often go unnoticed.
Troels Henriksen athas@sigkill.dk writes:
It's not, but I think most of the bugs you report (certainly the Calculator one, and probably also the CLIM-FIG one) are bugs in the demo applications, not McCLIM itself.
So yeah, of course I immediately find a redisplay bug that affected the calculator.
On Jan 21, 2008 10:24 PM, Daniel Herring dherring@tentpost.com wrote:
Misc tests: Empty Records 3
The circle is in the same location as for Empty Records 1 & 2. The bottom and right of the box are close to the circle as before. However, the top left corner is now 2 diameters from the circle.
This one is a known failure which I've meant to fix for some time.
Daniel Herring dherring@tentpost.com writes:
Bezier curves are very difficult to draw (lots of errors).
The main error I encountered (really, the only one, disregarding redisplay-flickering) was how ease it was to end up with "null size beziers", which McCLIM really doesn't like. I made CLIM-FIG not attempt to draw these, which should make bezier drawing easier, but I don't know whether this is the right thing.
Daniel Herring dherring@tentpost.com writes:
Selecting "ellipse" and then clicking on the canvas results in "Attempt to invert the probably singular transformation #<CLIM-INTERNALS::STANDARD-HAIRY-TRANSFORMATION 0 0 0 0 266 185>. [Condition of type CLIM:SINGULAR-TRANSFORMATION]"
Your problem was trying to draw zero-radius ellipses. I fixed this so it's now possible (though they obviously aren't visible).
-
Andy Hefner -
Daniel Herring -
Troels Henriksen