[PATCH] misc fixes

[Qian Yun]

This patch contains 2 dozen one line misc fixes,
I think they are pretty self explanatory.

If you have questions on any of them, I can explain.

- Qian

[Waldek Hebisch]

On Mon, May 25, 2026 at 06:04:04PM +0800, Qian Yun wrote:
> This patch contains 2 dozen one line misc fixes,
> I think they are pretty self explanatory.

Most yes. However, AFAICS we can simply remove declaration of 'mode'
in edible.h and make declaration of 'MODE' in edible.c static
and remove declaration in spadbuf.c.

Change to 'view2D' looks fishy: if 'transform' has no side effects
we can completely get rid of it, if it may have side effects, then
loop is needed.

--
Waldek Hebisch

[Qian Yun]

In your recent change to view2D.spad, you remove the "numberCheck",
is that intentional? Now the loop and 'transform' is useless?

for aList in graf.llPoints repeat
for p in aList repeat
aPoint := transform p
- numberCheck aPoint
graf

- Qian

[Waldek Hebisch]

Yes, removal is intentional. Since w compute using real type
w should never get complex values, so test done by 'numberCheck'
is useless (in principle, given enough declarations sbcl should
be able to replace 'numberCheck' by no operation).

AFAICS 'transform' at least in principle may throw errors, and
the puspose of this loop seem to be to ensure that computation
can be done without throwing errors. Looks silly to do the
same computation twice, but it is not clear to me if this
loop is completely useless (error when sending data to
HyperDoc program would cause loss of synchronization).

--
Waldek Hebisch

[Waldek Hebisch]

We probably should modify 'sendGraphImage' to transform all
points in bulk before sending the data. Then errors (if any)
will appear before interaction with viewman, so no danger of
loss of cynchronization.


--
Waldek Hebisch