Stretches was using the set_x/yscale() APIs when it shouldn't have been.
We can't rely on scaling in place of stretches because some browsers
don't support CSS transforms. Besides being a shortcut, stretches has a
different meaning from scale: stretch my resource to fit my width/beight
vs. scale me and all my children.
Because of this, it's technically possible to use both stretches and
scaling, so I don't think we need a warning. And, since scaling isn't
available everywhere, we can't really simplify the kernel API either :(
In other news, I discovered why we used to maintain view._xscale and
_yscale properties: LzViewLinkage used them and is now causing
getAttributeRelative() to fail. I'll fix this also.
Regards,
Max Carlson
OpenLaszlo.org
On 8/17/10 3:53 PM, P T Withington wrote:
Can you explain what you mean here?
Are you saying that it should be an error at the LFC API to use scaling and
stretches? Or are you saying the error is at the kernel API?
In either case, it seems like there should be more to this change to fix it.
Either you should give an error at the LFC API if both a used, or you should
simplify the kernel API so (for instance) the scaling API is the only API and
the LFC API is just two different ways of specifying the same thing (as you did
with transform and tint).
I understand that the stretches API is a nice shortcut for what would be a
complex constraint on scale, so I'm fine with that additional API at the LFC
level. But we shouldn't have to be providing 'convenience' API's at the kernel
level. There we should strive for simplicity, to make it easier to port and
more likely to be correct.
On 2010-08-17, at 18:21, Max Carlson wrote:
Change maxcarlson-20100817-dBG by maxcarl...@friendly on 2010-08-17 15:12:43 PDT
in /Users/maxcarlson/openlaszlo/trunk2
for http://svn.openlaszlo.org/openlaszlo/trunk
Summary: Fix stretches behavior
Bugs Fixed: LPP-9292 - Add way to scale views
Technical Reviewer: hminsky
QA Reviewer: ptw
Details: I'm checking this in ahead of time, as it's a pretty serious
regression...
Scaling shouldn't be used in addition to stretches. Removed all scaling update
code, since it's now reserved for user calls, e.g. setAttribute('xscale', 2);
Tests: my-apps/copy-of-hello.lzx?lzr=dhtml&lzt=html&debug=true shows debugger
window correctly, testcase at LPP-9292 runs like before.
Files:
M WEB-INF/lps/lfc/views/LaszloView.lzs
Changeset:
http://svn.openlaszlo.org/openlaszlo/patches/maxcarlson-20100817-dBG.tar
