Luboš Luňák wrote:
> I'm confused. I wrote that I find the API bad and that it would probably
> require writing it again from scratch. You call that tweaking?
>
Ain't that much code, so yeah, tweaks..
> How is this bike-shedding? Discussing the code will get much harder
> without handling the
On Monday 23 of March 2020, Thorsten Behrens wrote:
> Luboš Luňák wrote:
> > Replacing one error-prone API with another error-prone API doesn't solve
> > anything.
>
> True. But the basegfx classes are reasonably easy to tweak (and
> B2IRect has only few clients).
I'm confused. I wrote that I fi
Luboš Luňák wrote:
> Replacing one error-prone API with another error-prone API doesn't solve
> anything.
>
True. But the basegfx classes are reasonably easy to tweak (and
B2IRect has only few clients).
Perhaps best to let code speak? Happy to continue the bike-shedding on
gerrit (and perhaps l
Luboš Luňák wrote:
> [...] simpler to just write the class from scratch rather than try
> to be too smart and share the code with tools::Rectangle, but if it
> would be in the same header and had pretty much the same API, then
> for most practical purposes it would be still part of that
> already-e
On Mon, 23 Mar 2020 at 17:12, Luboš Luňák wrote:
> > I see the problem not in tools::Rectangle itself, but in the fact, that
> > it uses integer and not double. Using double makes width = right - left
> > in all cases
> > and would solve accuracy problems in manipulating shapes.
> > It would be u
On Friday 20 of March 2020, Regina Henschel wrote:
> Luboš Luňák schrieb am 19-Mar-20 um 16:13:
> > Hello,
> >
> > yes, this is about the tools::Rectangle nightmare of an API (in case
> > you don't know, it's this [1] ). I'm hunting an off-by-one error
> > somewhere in VCL, and it's hard to fin
On Friday 20 of March 2020, Tomaž Vajngerl wrote:
> I fear a bit introducing a new implementation as that will just mean one
> more implementation that we will have to deal long run, but I don't think
> it would make things much worse as they are now - at elast we will know
> what was audited to ha
On Friday 20 of March 2020, Thorsten Behrens wrote:
> Luboš Luňák wrote:
> > > One question: any reason not to tweak the basegfx classes to fit this
> > > model?
> >
> > Looking at those classes, I think their API serves some other
> > purpose than "simple point/size/rectangle", so the closest to
Hi,
On Fri, Mar 20, 2020 at 11:54 PM Regina Henschel
wrote:
> Hi,
>
> The problem for huge values is not Writer but Calc. For example, make a
> sheet with row height 10cm on all rows. Insert a shape in A2000 and look
> at the position information in the status bar. (Shapes are anchored to
> page
Hi,
Tomaž Vajngerl schrieb am 20-Mar-20 um 18:13:
Hi,
Do you know a specific case where that actually matters? My personal
position
on integers is "use int, unless you explicitly need something else".
We are dealing with 100th-mm or twip units, which can get quite big (in
writer
Hi Regina, *,
Regina Henschel wrote:
> These off-by-one problems occur earlier than in VCL. For example
> changes to maSnapRect when a shape is transformed by shear and
> rotation.
>
Yep. Also c.f. SwRect mentioned earlier.
> A new kind of rectangle does not solve the problem, that you have to
>
Hi Luboš,
Luboš Luňák schrieb am 19-Mar-20 um 16:13:
Hello,
yes, this is about the tools::Rectangle nightmare of an API (in case you
don't know, it's this [1] ). I'm hunting an off-by-one error somewhere in
VCL, and it's hard to find it when I can't even tell which parts of the code
are ri
FWIW there is also canvas/source/tools/surfacerect.hxx, which might be
the lowest hanging fruit to get rid of, and which I found when fixing
tdf#40534 :-)
If you replace this as a starter, we would at least not have the "just
and additional standard" situation and keep our Rectangle type count
equ
Hi,
Do you know a specific case where that actually matters? My personal
> position
> on integers is "use int, unless you explicitly need something else".
>
>
We are dealing with 100th-mm or twip units, which can get quite big (in
writer we are always referencing relative from page 1 AFAIK) and co
Hi,
On Thu, Mar 19, 2020 at 4:14 PM Luboš Luňák wrote:
>
> Hello,
>
> yes, this is about the tools::Rectangle nightmare of an API (in case you
> don't know, it's this [1] ). I'm hunting an off-by-one error somewhere in
> VCL, and it's hard to find it when I can't even tell which parts of the
>
Luboš Luňák wrote:
> > One question: any reason not to tweak the basegfx classes to fit this
> > model?
>
> Looking at those classes, I think their API serves some other
> purpose than "simple point/size/rectangle", so the closest to
> tweaking them I see is wrapping them in a completely new API.
On 2020/03/20 2:12 pm, Luboš Luňák wrote:
With one extra wrinkle: the current tools::Rectangle uses long, and lots of
places that work with tools::Rectangle also use long, which is particularly
pernicious because long is a different bit-size on Windows and Linux.
Do you know a specific ca
On Thursday 19 of March 2020, Thorsten Behrens wrote:
> Hey Luboš,
>
> Luboš Luňák wrote:
> > So, yeah, I'm proposing a new standard Rectangle class (and I know
> > xkcd, and I'm still serious). My idea is roughly that there will be
> > some tools::NewRectangle (or whatever usable name), it will b
On Thursday 19 of March 2020, Noel Grandin wrote:
> On Thu, 19 Mar 2020 at 17:14, Luboš Luňák wrote:
> > - internal representation will be whatever sane thing will work, e.g.
> > x,y,width,height , and it won't matter for the API
> > - empty rectangle is simply width == 0 || height == 0
>
> +1 on
Hey Luboš,
Luboš Luňák wrote:
> So, yeah, I'm proposing a new standard Rectangle class (and I know
> xkcd, and I'm still serious). My idea is roughly that there will be
> some tools::NewRectangle (or whatever usable name), it will be more
> or less like tools::Rectangle, but it'll make things cle
On Thu, 19 Mar 2020 at 17:14, Luboš Luňák wrote:
> - internal representation will be whatever sane thing will work, e.g.
> x,y,width,height , and it won't matter for the API
> - empty rectangle is simply width == 0 || height == 0
>
>
+1 on this part.
FWIW, the Java Rectangle API is pretty much t
Hello,
yes, this is about the tools::Rectangle nightmare of an API (in case you
don't know, it's this [1] ). I'm hunting an off-by-one error somewhere in
VCL, and it's hard to find it when I can't even tell which parts of the code
are right or wrong :(.
This has been already discussed a co
22 matches
Mail list logo
