The keys.ss framework test suite seems to confirm this fix, altho
I'm not sure how exactly.
Also tweak the stress test
original commit: 9888c7f496bbb48978f46cf91fe2638a93b2c68b
Unfortunately, not the bug I was hoping to find.
This bug appears to be very old, however; perhaps
as old as the colorer itself.
original commit: 5aeca646d463ba4b065d0301465fdcf93415f7b5
also make keys.rkt test suite be less likely to corrupt the
preferences file (when run in standalone mode)
original commit: 593ab2d51fa5d44e107fa626d93c4106ef1ff33b
The `future-parent` argument is actually ignored now and
can be any value, so document it as such.
original commit: 8c7a6709b01790f3a920cd8d509a8a85a2545583
Add `dimension-integer?', `position-integer?',
`spacing-integer?', and `positive-dimension-integer?' to make
the documentation, especially, more parameterized over the
limit.
Also, make the slider limits match the dimension/position limit,
since they're often related.
Things go wrong with, for example, a canvas that is
larger than 32k in a dimension. It looks like there's
a problem on each platform, although not exactly the
same problem. Meanwhile, even larger scrollable panels
can work ok on platforms other than Windows.
Closes PR 13849
original commit: 2b1b8696f02fa59dea4278c2786fd21444204fc5
I'm not sure if the precise details of how this function behaves
(specifically that it copies snips that aren't cacheable) is
important, but I suspect the implementation is more authoritative
that the docs in this case, so document what it really does.
closes PR 13956
original commit: fb91397d24551dbef0db7c40a9d12042e6c7326b
when switching to white-on-black mode (or back).
this causes the cursor in embedded editors to pick up the right colors
(since it is selected by the inverse of the background color of the style
of the enclosing editor snip)
closes PR 13943
original commit: 492ddab085a1d87a70843df772b044ceb84adb3d
In particular, using a shut down eventspace actually returns
the exn:fail exception instead of exn:misc, which does not
exist. Example:
#(struct:exn:fail start in timer%: current eventspace is shutdown:
#<eventspace> #<continuation-mark-set>)
original commit: 26b7c98356919d340d77b9ead76862eaea07da05
when there is no containing paren
it used to return the beginning of the region (usually 0)
(this adjusts backward-containing-sexp to agree with the docs)
original commit: e1f4547ea99200c6ad0e9c34a76d3eb19eb906b6
The change avoids a collision with PLaneT fields.
Also, add checking for the form of a `release-note-files' value.
original commit: 8f195a1ff65a05f7679af852f2cbfa58ea8f4bfc
A package like "gui" does not depend on a package like "gui-test",
which means that you have to specifically ask for a test package if
you want the tests. A new "main-distribution-test" package
aggregates the tests for packages that are in "main-distribution".
The "plt-services" package, meanwhile, depends on the
"main-distribution-test" package, which means that all tests are
compiled by default for an in-place build.
original commit: d50d0f8bcab260f404049029b77407e0222f588a
It looks like button-press information is not available on 10.5
and earlier, so so try to report it in that case.
Closes PR 13780
original commit: d695e8a15afff1da7c3306a717ef7d102483873f
In many cases, `path->collects-relative' is used instead of
`path->main-collects-relative' to generalize existing support
for collection-relative paths.
original commit: f87981f96057c26cb311e786a91229bf4a874e5b
Left one dependency broken: "drracket" currently depends on "htdp" for
a test. That needs to be fixed by removing the dependency (moving the
test to "htdp?), instead of changing the declared dependencies.
original commit: 51290fd2a95def6bb3b6d3d735cb62444e157553
