freenode/#clim - IRC Chatlog
Search
9:42:27
jackdaniel
so I will reset --hard my changes. from things which make things look better you may want to incorporate: in "make-pane 'text-editor" replace :height 200 with :nlines 4; in layotus section add replace canvas with (+fill+ canvas), rest of my changes was mostly whitespace cleanup
10:16:50
jackdaniel
so answering my previous question about text-size and negative left-bearing, it was bogus
10:18:04
jackdaniel
because text-size gives only cursor progress width/height etc, it does not return x0/y0 (they are assumed to be 0,0)
10:18:40
jackdaniel
so fact that text-size-test shows width / height as if they had started at 0,0 is just a demo assumption, text-bounding-rectangle is for determining where text starts
10:37:41
scymtym
jackdaniel: if you have time, could you have a look at https://github.com/scymtym/McCLIM/tree/wip-text-size ? it is almost done, but there is one problem. the call to INVOKE-WITH-DRAWING-OPTIONS makes stuff disappear on repaint (not redisplay) while the equivalent use of WITH-DRAWING-OPTIONS works as expected. am i doing this wrong?
15:29:01
slyrus
fortunately the "go to sleep and fix the bugs in the morning" trick worked, only it took applying the trick twice.
17:03:59
jackdaniel
scymtym: here are my comments (with explanation why with-drawing-options behaves differently in your code): https://github.com/McCLIM/McCLIM/pull/577#pullrequestreview-165722393feel free to close the PR, I've opened it so it was easier to comment things
19:14:56
scymtym
jackdaniel: so, should i 1) remove the xrender-fonts change from the pull request 2) work around the problem with INVOKE-WITH-DRAWING-OPTIONS 3) address the other points raised in the review and update the pull request accordingly?
19:17:59
jackdaniel
I think that regarding 2) you could fix invoke-with-drawing-options to invoke thunk on the first argument (the sheet) while keeping with-medium-options called on its underlying medium
19:18:51
jackdaniel
according to a second paragraph medium can be a medium or a sheet (or a stream), so if we apply it to the first paragraph all fits well
19:26:24
scymtym
just to be clear: you are asking to add another commit to this pull request with the above change, right?
19:29:24
scymtym
btw, i'm thinking about combining the text size test and font selector demo into one since they are like 80 % identical. what do you think?
19:31:17
jackdaniel
the second one presents how to browse available font famiilies; the first one shows renderer features (like kerning, multiline, bounding box etc)
19:31:28
scymtym
i guess the real question is whether the font selector should be a gadget instead of a demo
19:32:59
scymtym
since the update-font-preview-in-response-to-gadget-change is common and accounts for a lot of the code in both cases
19:34:22
jackdaniel
font selector demo already has that as a response to selecting things in list. do you plan to add new gadgets to it?
19:57:00
jackdaniel
I'm referring "since the update-font-preview-in-response-to-gadget-change is common and accounts for a lot of the code in both cases" statement
19:59:18
jackdaniel
as of making it a gadget, I'm not sure if it is a gadget people miss in McCLIM (I mean - is it justified to add it?) - this demo presents how to list font families associated with a port etc, so even if we wanted to make a gadget from it, demo should stay (imo)
20:03:16
scymtym
sure. it is designed like a reusable dialog already, though, with the ok and cancel buttons
20:18:53
scymtym
jackdaniel: see you. spec annotation: http://bauhh.dyndns.org:8000/clim-spec/10-2.html#_526