 |
|
|
|
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Does anyone know whether it is possible for a scene file to detect
whether or not it is being rendered by UberPOV, without bombing out if
it is rendered by official POV-Ray?
Why would I need to do this? Because I'm writing a macro that needs to
know whether or not true inverse square light attenuation is available.
For official POV-Ray, this is version >= 3.8, but for UberPOV, it is
version >= 3.71.
My first (and so far only) idea was to start with a test for the
existence of the variable 'unofficial', but it turns out POV-Ray 3.8
does the same test--but treats it as if it were a keyword. This screws
up the test for any SDL code running under 3.8.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Cousin Ricky <ric### [at] yahoo com> wrote:
> Does anyone know whether it is possible for a scene file to detect
> whether or not it is being rendered by UberPOV, without bombing out if
> it is rendered by official POV-Ray?
>
> Why would I need to do this? Because I'm writing a macro that needs to
> know whether or not true inverse square light attenuation is available.
> For official POV-Ray, this is version >= 3.8, but for UberPOV, it is
> version >= 3.71.
>
> My first (and so far only) idea was to start with a test for the
> existence of the variable 'unofficial', but it turns out POV-Ray 3.8
> does the same test--but treats it as if it were a keyword. This screws
> up the test for any SDL code running under 3.8.
Yes.
I believe you're searching for the combination of this :
#ifdef(unofficial)
with a more narrowed patch version number that belongs only to uber pov such as:
#if(unofficial = \"patch\") #if(patch(\"upov-reflection-roughness\") > 0)
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
"Mr" <nomail@nomail> wrote:
> > My first (and so far only) idea was to start with a test for the
> > existence of the variable 'unofficial', but it turns out POV-Ray 3.8
> > does the same test--but treats it as if it were a keyword. This screws
> > up the test for any SDL code running under 3.8.
>
> Yes.
> I believe you're searching for the combination of this :
>
> #ifdef(unofficial)
>
> with a more narrowed patch version number that belongs only to uber pov such as:
>
> #if(unofficial = \"patch\") #if(patch(\"upov-reflection-roughness\") > 0)
in qtpovray3.8:
#ifdef (unofficial)
Parse Warning: Trying to test whether a reserved keyword is defined. Test result
may not be what you expect.
#declare U = unofficial;
Parse Error: Expected 'RValue to declare', unofficial found instead
#declare P = patch;
Parse Error: Cannot pass uninitialized identifier to non-optional LValue.
#debug unofficial
Parse Error: Expected 'string expression', unofficial found instead
#debug patch
Parse Error: Expected 'string expression', undeclared identifier 'patch' found
instead
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
On 2021-03-06 6:53 PM (-4), Bald Eagle wrote:
> "Mr" <nomail@nomail> wrote:
>
>>> My first (and so far only) idea was to start with a test for the
>>> existence of the variable 'unofficial', but it turns out POV-Ray 3.8
>>> does the same test--but treats it as if it were a keyword. This screws
>>> up the test for any SDL code running under 3.8.
>>
>> Yes.
>> I believe you're searching for the combination of this :
>>
>> #ifdef(unofficial)
>>
>> with a more narrowed patch version number that belongs only to uber pov such as:
>>
>> #if(unofficial = \"patch\") #if(patch(\"upov-reflection-roughness\") > 0)
>
>
> in qtpovray3.8:
>
> #ifdef (unofficial)
> Parse Warning: Trying to test whether a reserved keyword is defined. Test result
> may not be what you expect.
> #declare U = unofficial;
> Parse Error: Expected 'RValue to declare', unofficial found instead
> #declare P = patch;
> Parse Error: Cannot pass uninitialized identifier to non-optional LValue.
> #debug unofficial
> Parse Error: Expected 'string expression', unofficial found instead
> #debug patch
> Parse Error: Expected 'string expression', undeclared identifier 'patch' found
> instead
Exactly. The results are the same in official POV-Ray 3.8. The problem
is that the test on 'unofficial' is reliable _only_ with a patch that
uses it--which is exactly what we don't know in advance; and any further
refined tests are pointless.
I'll just go ahead and assume that the user is using official POV-Ray,
and have them set a flag variable if they are using a 3.7.1-based patch.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Am 08.03.2021 um 19:30 schrieb Cousin Ricky:
> Exactly. The results are the same in official POV-Ray 3.8. The problem
> is that the test on 'unofficial' is reliable _only_ with a patch that
> uses it--which is exactly what we don't know in advance; and any further
> refined tests are pointless.
Consider this a bug in the POV-Ray v3.8 betas.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Am 05.03.2021 um 16:54 schrieb Cousin Ricky:
> Does anyone know whether it is possible for a scene file to detect
> whether or not it is being rendered by UberPOV, without bombing out if
> it is rendered by official POV-Ray?
>
> Why would I need to do this? Because I'm writing a macro that needs to
> know whether or not true inverse square light attenuation is available.
> For official POV-Ray, this is version >= 3.8, but for UberPOV, it is
> version >= 3.71.
>
> My first (and so far only) idea was to start with a test for the
> existence of the variable 'unofficial', but it turns out POV-Ray 3.8
> does the same test--but treats it as if it were a keyword. This screws
> up the test for any SDL code running under 3.8.
Some background information:
The problem arises with recent development versions of POV-Ray that (1)
recognize `unofficial` as a proper keyword (as in `#version unofficial
FOO`, which official POV-Ray has been supporting for ages in the sense
that it prompts a very specific error message), and (2) report all
recognized keywords as "defined" with respect to the `#ifdef` statement
and its kin.
In v3.7 and early development versions identifying as v3.7.1, the
statement had been handled in a comparatively hackish manner, which did
not recognize `unofficial` as a proper keyword. Therefore, those
versions are unproblematic.
In later v3.7.1 development versions, starting with
v3.7.1-alpha.8449793, `unofficial` was promoted to a keyword proper.
However, `#ifdef` and its kin still reported keywords as "not defined",
so those versions are also unproblematic.
The problem was only introduced with the experimental
v3.8.0-tokenizer.9841908, and later with the mainline
v3.8.0-alpha.10008988, which changed the behaviour of `#ifdef` to report
keywords as "defined", including the `unofficial` keyword.
Note however that the change of the `#ifdef` statement also means that
support for the `patch` keyword can be tested directly, in a manner that
is also compatible with older versions. Also, whether or not `#ifdef`
works this way can be tested with a known keyword.
The following code will cause parse warnings with some versions, but
other than that it should work fine for any version:
---------------------------------------------------------------
// Tell POV-Ray that we'll be doing our own feature probing.
#version version;
// Test whether UberPOV-style `#patch`/`patch()` mechanism
// seems to be available
#ifdef(ifdef)
// New `#ifdef` behaviour, reporting supported keywords as defined.
// We can probe for support of the `patch` keyword directly.
#declare HAVE_PATCH = defined(patch);
#else
// Old `#ifdef` behaviour, reporting keywords as undefined.
// We can probe for support of the `unofficial` variable.
#ifdef(unofficial)
#declare HAVE_PATCH = (unofficial = "patch");
#else
#declare HAVE_PATCH = false;
#end
#end
// Probe for required features
#if(HAVE_PATCH)
// Probe features individually via `#patch`/`patch()`
// (or probe for UberPOV version via that same mechanism
// using "upov" as the feature name, though this is somewhat
// discouraged)
...
#else
// Probe via `#version`/`version´
...
#end
---------------------------------------------------------------
Note that support for the `patch` keyword per se does _not_ necessarily
imply that the version used is actually UberPOV. Most notably, the
details of the mechanism were specifically designed such that other
derivatives of POV-Ray could use the same mechanism without conflict,
and official POV-Ray might also eventually support it, if only to
further promote it as a standard mechanism for detecting derivatives and
probing for their features.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
|
|
