 |
|
|
|
|
|
| |
| |
|
|
|
|
| |
| |
|
|
I start an overnight render. I wake up the next morning, expecting a beautiful,
polished image.
Instead, the damn machine has gone into sleep mode!
Sleep mode is good. It saves power.
BUT NOT WHILE THE COMPUTER IS IN THE MIDDLE OF A TASK.
Someone recently mentioned the principle of least surprise in p.o-t. Micro$hit
seems adamantly opposed to the whole concept.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Cousin Ricky <rickysttATyahooDOTcom> wrote:
> Instead, the damn machine has gone into sleep mode!
And I assume there's no way of saying "don't go to sleep mode if this
program is running"? Or even "don't go to sleep mode if there's significant
CPU activity"?
OTOH, probably no OS has such a feature (at least not out-of-the-box,
easily selectable via GUI).
--
- Warp
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Le 2011/09/27 14:58, Cousin Ricky a écrit :
> I start an overnight render. I wake up the next morning, expecting a beautiful,
> polished image.
>
> Instead, the damn machine has gone into sleep mode!
>
> Sleep mode is good. It saves power.
>
> BUT NOT WHILE THE COMPUTER IS IN THE MIDDLE OF A TASK.
>
> Someone recently mentioned the principle of least surprise in p.o-t. Micro$hit
> seems adamantly opposed to the whole concept.
>
>
>
One of the first thing I do on a freshly installed/new computer it to
set it as a presentation machine. This effectively disable automatic
sleep mode.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
On 27/09/2011 7:58 PM, Cousin Ricky wrote:
> Sleep mode is good. It saves power.
>
> BUT NOT WHILE THE COMPUTER IS IN THE MIDDLE OF A TASK.
Create a custom Power Plan to use when rendering when the 'putor never
goes to sleep.
Control Panel\All Control Panel Items\Power Options\Edit Plan Settings
--
Regards
Stephen
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
On 27-9-2011 23:10, Stephen wrote:
> On 27/09/2011 7:58 PM, Cousin Ricky wrote:
>> Sleep mode is good. It saves power.
>>
>> BUT NOT WHILE THE COMPUTER IS IN THE MIDDLE OF A TASK.
>
> Create a custom Power Plan to use when rendering when the 'putor never
> goes to sleep.
> Control Panel\All Control Panel Items\Power Options\Edit Plan Settings
>
>
Hey! thanks Stephen! I had not (yet) found that option either! I have
selected High Performance now.
By the way: Personally, I like Windows 7 quite a lot...
Thomas
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
On 28/09/2011 8:59 AM, Thomas de Groot wrote:
> On 27-9-2011 23:10, Stephen wrote:
>> Control Panel\All Control Panel Items\Power Options\Edit Plan Settings
>>
>>
> Hey! thanks Stephen! I had not (yet) found that option either! I have
> selected High Performance now.
>
An essential setting for all Povers IMO.
> By the way: Personally, I like Windows 7 quite a lot...
>
So do I. It is a bit resource hungry but my current machine has the
resources. I actually saw the memory usage go higher than 5.5 Gigs today
while I was experimenting with Poser hair.
--
Regards
Stephen
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
On 28-9-2011 12:05, Stephen wrote:
> On 28/09/2011 8:59 AM, Thomas de Groot wrote:
>> On 27-9-2011 23:10, Stephen wrote:
>
>>> Control Panel\All Control Panel Items\Power Options\Edit Plan Settings
>>>
>>>
>> Hey! thanks Stephen! I had not (yet) found that option either! I have
>> selected High Performance now.
>>
>
> An essential setting for all Povers IMO.
Absolutely. I had not looked at the power utility yet because I believed
it was not essential when on a PC. How wrong I was :-)
I should do something to my Windows 7 Curiosity Setting: Very Low; Low;
Average (default); High, Very High.
Mine was set at Low apparently... ;-)
>
>> By the way: Personally, I like Windows 7 quite a lot...
>>
>
> So do I. It is a bit resource hungry but my current machine has the
> resources. I actually saw the memory usage go higher than 5.5 Gigs today
> while I was experimenting with Poser hair.
Yes, that can be expected. Best to have enough resources.
Thomas
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Le 2011/09/27 14:58, Cousin Ricky a écrit :
> I start an overnight render. I wake up the next morning, expecting a beautiful,
> polished image.
>
> Instead, the damn machine has gone into sleep mode!
>
> Sleep mode is good. It saves power.
>
> BUT NOT WHILE THE COMPUTER IS IN THE MIDDLE OF A TASK.
>
> Someone recently mentioned the principle of least surprise in p.o-t. Micro$hit
> seems adamantly opposed to the whole concept.
>
>
>
By the way, power management have been there since windows 95 if I
remember corectly... Things like turning display off, powering off the
hard drives when idle, throtling down the CPU if the BIOS support it,...
It's only that some features got added over time as the hardware started
to offer the possibility, and the terminology used seems to have changed
with 7... to some extent.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
On 9/27/2011 13:36, Warp wrote:
> Cousin Ricky<rickysttATyahooDOTcom> wrote:
>> Instead, the damn machine has gone into sleep mode!
>
> And I assume there's no way of saying "don't go to sleep mode if this
> program is running"?
Windows asks each program if it's OK to go to sleep before it does so.
POV-Ray would have to be coded to recognize that message and deny sleeping
while a render is in progress. Most disk burners, for example, will do this.
--
Darren New, San Diego CA, USA (PST)
How come I never get only one kudo?
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
On 29-9-2011 21:15, Darren New wrote:
> Windows asks each program if it's OK to go to sleep before it does so.
> POV-Ray would have to be coded to recognize that message and deny
> sleeping while a render is in progress. Most disk burners, for example,
> will do this.
Am I right in assuming that we should put this is the Wish List?
Thomas
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
On 10/2/2011 23:48, Thomas de Groot wrote:
> Am I right in assuming that we should put this is the Wish List?
Not for me personally, but it's certainly something I could see many people
wanting, especially if you make it an option in the windows interface that
can be checked on and off: "Prevent sleep while rendering?"
I'm actually not sure how you make the machine sleep when the render is
finished if you've told it not to sleep, like whether Windows keeps asking
you every few minutes or something, or whether you have to say "OK, try
again" when you finish.
--
Darren New, San Diego CA, USA (PST)
How come I never get only one kudo?
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
On 10/2/2011 23:48, Thomas de Groot wrote:
> Am I right in assuming that we should put this is the Wish List?
Actually, it seems pretty easy, relatively speaking.
http://msdn.microsoft.com/en-us/library/windows/desktop/aa373208%28v=vs.85%29.aspx
It looks like adding a call to
SetThreadExecutionState(ES_SYSTEM_REQUIRED)
each time a work unit is dispatched in the renderer should be enough, as
long as at least one of those work units is going to finish before the sleep
timer runs out. Adding in ES_CONTINUOUS might be good, but you'd want to
take care that you don't fail to clear that at the end of the render, too.
Probably safer to keep telling it you're awake than to tell it to stay awake
until further notice.
HTH!
--
Darren New, San Diego CA, USA (PST)
How come I never get only one kudo?
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
On 10/3/2011 11:08, Darren New wrote:
> Actually, it seems pretty easy, relatively speaking.
http://msdn.microsoft.com/en-us/library/aa373233%28VS.85%29.aspx
Actually, it looks like XP and 2000 work the way I remember.
"""
The system initiates a sleep request by first sending a PBT_APMQUERYSUSPEND
event. Applications should not deny this request; doing so will prevent the
system from going to sleep.
"""
--
Darren New, San Diego CA, USA (PST)
How come I never get only one kudo?
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
On 3-10-2011 20:14, Darren New wrote:
> Actually, it looks like XP and 2000 work the way I remember.
I think XP did not go to sleep with a render under way. It maybe slowed
down though... I am not sure.
Thomas
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
On 28/09/2011 04:58, Cousin Ricky wrote:
> I start an overnight render. I wake up the next morning, expecting a beautiful,
> polished image.
>
> Instead, the damn machine has gone into sleep mode!
I've added a new 'Prevent sleep while rendering' option to avoid this
issue. It defaults to on, which is probably what most folks would want.
-- Chris
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Am 29.01.2013 01:40, schrieb Chris Cason:
> On 28/09/2011 04:58, Cousin Ricky wrote:
>> I start an overnight render. I wake up the next morning, expecting a beautiful,
>> polished image.
>>
>> Instead, the damn machine has gone into sleep mode!
>
> I've added a new 'Prevent sleep while rendering' option to avoid this
> issue. It defaults to on, which is probably what most folks would want.
Sounds plausible indeed.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
clipka <ano### [at] anonymous org> wrote:
> Am 29.01.2013 01:40, schrieb Chris Cason:
> > On 28/09/2011 04:58, Cousin Ricky wrote:
> >> I start an overnight render. I wake up the next morning, expecting a beautiful,
> >> polished image.
> >>
> >> Instead, the damn machine has gone into sleep mode!
> >
> > I've added a new 'Prevent sleep while rendering' option to avoid this
> > issue. It defaults to on, which is probably what most folks would want.
>
> Sounds plausible indeed.
That is reasonable. Another thing I do immediatelly after having purchased a new
windows machine is to let it ask before downloading updates. Updates can cause a
reboot which is annoying just as well.
Best regards,
Michael
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Chris Cason <del### [at] deletethistoopovrayorg> wrote:
> I've added a new 'Prevent sleep while rendering' option to avoid this
> issue. It defaults to on, which is probably what most folks would want.
Which got me thinking...
How hard would it be for winpov to detect that it was terminated while
a rendering was in progress, save that info somewhere, and then the next
time it's started for it to ask the user if they want to continue the
render?
--
- Warp
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Warp <war### [at] tagpovrayorg> wrote:
> Chris Cason <del### [at] deletethistoopovrayorg> wrote:
> > I've added a new 'Prevent sleep while rendering' option to avoid this
> > issue. It defaults to on, which is probably what most folks would want.
>
> Which got me thinking...
>
> How hard would it be for winpov to detect that it was terminated while
> a rendering was in progress, save that info somewhere, and then the next
> time it's started for it to ask the user if they want to continue the
> render?
>
> --
> - Warp
I'm certainly no expert at this. But POV leaves a state file when interrupted
otherwise an image. So it should not be impossible. But with a multicore machine
the continue option works only if all threads have left no "whole" in the
picture. That means if a block is not finished but an other which is later in
the sequence the blocks are worked through from another thread is (may be the
first contains some glass the second not, so the later thread is ready before
the first) than "continue" continues behind the later thread overlooking the
"whole" from the first unfinished thread. So if you have to interrupt a
rendering since you need your machine for other purposes, you have to be lucky
to can use the +c. IMO this issue is more important then the occasional updates
or the sleeps, which can be avoided easily with the update options or the energy
options Stephen pointed out.
Best regards,
Michael
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
MichaelJF <mi-### [at] t-onlinede> wrote:
> I'm certainly no expert at this. But POV leaves a state file when interrupted
> otherwise an image. So it should not be impossible. But with a multicore machine
> the continue option works only if all threads have left no "whole" in the
> picture. That means if a block is not finished but an other which is later in
> the sequence the blocks are worked through from another thread is (may be the
> first contains some glass the second not, so the later thread is ready before
> the first) than "continue" continues behind the later thread overlooking the
> "whole" from the first unfinished thread. So if you have to interrupt a
> rendering since you need your machine for other purposes, you have to be lucky
> to can use the +c. IMO this issue is more important then the occasional updates
> or the sleeps, which can be avoided easily with the update options or the energy
> options Stephen pointed out.
Are you saying that +C doesn't currently work, or are you just guessing?
I wasn't talking about the +C option per se. I was just talking about a
possible convenience feature in winpov that would suggest the user to
continue the render if povray was closed the last time while it was
rendering.
--
- Warp
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Warp <war### [at] tagpovrayorg> wrote:
> MichaelJF <mi-### [at] t-onlinede> wrote:
> > I'm certainly no expert at this. But POV leaves a state file when interrupted
> > otherwise an image. So it should not be impossible. But with a multicore machine
> > the continue option works only if all threads have left no "whole" in the
> > picture. That means if a block is not finished but an other which is later in
> > the sequence the blocks are worked through from another thread is (may be the
> > first contains some glass the second not, so the later thread is ready before
> > the first) than "continue" continues behind the later thread overlooking the
> > "whole" from the first unfinished thread. So if you have to interrupt a
> > rendering since you need your machine for other purposes, you have to be lucky
> > to can use the +c. IMO this issue is more important then the occasional updates
> > or the sleeps, which can be avoided easily with the update options or the energy
> > options Stephen pointed out.
>
> Are you saying that +C doesn't currently work, or are you just guessing?
>
> I wasn't talking about the +C option per se. I was just talking about a
> possible convenience feature in winpov that would suggest the user to
> continue the render if povray was closed the last time while it was
> rendering.
>
> --
> - Warp
No, I'm not guessing, this is an observation. We had a discussion about this
issue some months ago and Thomas de Groot observed the same. Look at
http://news.povray.org/povray.beta-test/thread/%3Cweb.5033e17517f31266b62e53d00@news.povray.org%3E/
Best regards,
Michael
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Seems the link vanished. Hope it can be seen now:
http://news.povray.org/povray.beta-test/thread/%3Cweb.5033e17517f31266b62e53d00@news.povray.org%3E/
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
"MichaelJF" <mi-### [at] t-onlinede> wrote:
> Seems the link vanished. Hope it can be seen now:
>
"http://news.povray.org/povray.beta-test/thread/%3Cweb.5033e17517f31266b62e53d00@news.povray.org%3E/"
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
"MichaelJF" <mi-### [at] t-onlinede> wrote:
> "MichaelJF" <mi-### [at] t-onlinede> wrote:
> > Seems the link vanished. Hope it can be seen now:
> >
>
>
"http://news.povray.org/povray.beta-test/thread/%3Cweb.5033e17517f31266b62e53d00@news.povray.org%3E/"
Very mysterious, first time a link to the POV-newsgroups doesn't work. It was at
povray.beta-test, "Resuming renders in 3.7 RC6 is confusing at best, broken at
worst" by Chaanakya at 21 Aug 2012 19:30:00.
Best regards,
Michael
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Am 29.01.2013 21:53, schrieb MichaelJF:
> "MichaelJF" <mi-### [at] t-onlinede> wrote:
>> "MichaelJF" <mi-### [at] t-onlinede> wrote:
>>> Seems the link vanished. Hope it can be seen now:
>>>
>>
>>
"http://news.povray.org/povray.beta-test/thread/%3Cweb.5033e17517f31266b62e53d00@news.povray.org%3E/"
>
> Very mysterious, first time a link to the POV-newsgroups doesn't work. It was at
> povray.beta-test, "Resuming renders in 3.7 RC6 is confusing at best, broken at
> worst" by Chaanakya at 21 Aug 2012 19:30:00.
The very first link worked perfectly fine for me (I haven't tested the
others).
Did anyone ever file a bug report on this issue?
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
On 30/01/2013 20:59, clipka wrote:
>> Very mysterious, first time a link to the POV-newsgroups doesn't work. It was at
>> povray.beta-test, "Resuming renders in 3.7 RC6 is confusing at best, broken at
>> worst" by Chaanakya at 21 Aug 2012 19:30:00.
>
> The very first link worked perfectly fine for me (I haven't tested the
> others).
>
> Did anyone ever file a bug report on this issue?
I don't recall seeing one. It's definitely something we need to look
at as I can't see us releasing 3.7 if it's faulty.
-- Chris
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Le 2013-01-29 07:58, MichaelJF a écrit :
> clipka <ano### [at] anonymousorg> wrote:
>> Am 29.01.2013 01:40, schrieb Chris Cason:
>>> On 28/09/2011 04:58, Cousin Ricky wrote:
>>>> I start an overnight render. I wake up the next morning, expecting a beautiful,
>>>> polished image.
>>>>
>>>> Instead, the damn machine has gone into sleep mode!
>>>
>>> I've added a new 'Prevent sleep while rendering' option to avoid this
>>> issue. It defaults to on, which is probably what most folks would want.
>>
>> Sounds plausible indeed.
>
> That is reasonable. Another thing I do immediatelly after having purchased a new
> windows machine is to let it ask before downloading updates. Updates can cause a
> reboot which is annoying just as well.
>
> Best regards,
> Michael
>
>
When a Windows update need to reboot, it ask you and will wait untill
you sellect "Restart now" or "Ask again in...".
There have been a few times when I had that dialog open for a few days.
Any way, I prefer the setting that download the updated, then ask to
install them.
Alain
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Le 2013-01-30 13:44, Chris Cason a écrit :
> On 30/01/2013 20:59, clipka wrote:
>>> Very mysterious, first time a link to the POV-newsgroups doesn't work. It was at
>>> povray.beta-test, "Resuming renders in 3.7 RC6 is confusing at best, broken at
>>> worst" by Chaanakya at 21 Aug 2012 19:30:00.
>>
>> The very first link worked perfectly fine for me (I haven't tested the
>> others).
>>
>> Did anyone ever file a bug report on this issue?
>
> I don't recall seeing one. It's definitely something we need to look
> at as I can't see us releasing 3.7 if it's faulty.
>
> -- Chris
>
One possibility would be to check at some point if some area are still
blank, or have only radiosity pretrace data. It could be done on a
resume or after the trace is otherwise finished.
Another way, would be to keep a per-tread last rendered block and not
just a last rendered block.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Thomas de Groot <tenDOTlnDOTretniATtoorgedDOTt> wrote:
> On 3-10-2011 20:14, Darren New wrote:
> > Actually, it looks like XP and 2000 work the way I remember.
>
> I think XP did not go to sleep with a render under way. It maybe slowed
> down though... I am not sure.
>
I've definitely had it go to sleep on me, with a render in progress. Had to turn
the sleep mode off completely, to prevent this (which is probably not good for
the ol' hard drive.) I always thought it was just an odd Windows bug. (But if I
recall correctly, POV-Ray was the only application this happened with...although
it's also the only app I use that churns through work for hours on end.) So
adding a fix to POV-Ray is a great idea!
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Le 30/01/2013 20:51, Alain nous fit lire :
> When a Windows update need to reboot, it ask you and will wait untill
> you sellect "Restart now" or "Ask again in...".
> There have been a few times when I had that dialog open for a few days.
> Any way, I prefer the setting that download the updated, then ask to
> install them.
YMMV, but I happened more than once to get an XP rebooted without
consent because it was updated and unattended long enough (coffee
break... oh, login again!)
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
"MichaelJF" <mi-### [at] t-onlinede> wrote:
> > > Seems the link vanished. Hope it can be seen now:
Nope, none of the your three tries are showing up at my end (on latest Firefox.)
BUT, when I went to post this message, the link DID come up in your own message.
I see that it's in quotes. Should it have them? Here's the same link, posted
without the quotes; see if it shows up successfully...
http://news.povray.org/povray.beta-test/thread/%3Cweb.5033e17517f31266b62e53d00@news.povray.org%3E/
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
"Kenneth" <kdw### [at] gmailcom> wrote:
> Here's the same link, posted
> without the quotes; see if it shows up successfully...
Nope, it still didn't work. :-/
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
On 30/01/2013 9:25 PM, Kenneth wrote:
> "Kenneth" <kdw### [at] gmailcom> wrote:
>> Here's the same link, posted
>> without the quotes; see if it shows up successfully...
>
> Nope, it still didn't work. :-/
>
>
>
>
Not in FireFox nor IE9 but it does in Thunderbird.
Comeback uncle Ken
--
Regards
Stephen
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Am 30.01.2013 22:13, schrieb Le_Forgeron:
> Le 30/01/2013 20:51, Alain nous fit lire :
>> When a Windows update need to reboot, it ask you and will wait untill
>> you sellect "Restart now" or "Ask again in...".
>> There have been a few times when I had that dialog open for a few days.
>> Any way, I prefer the setting that download the updated, then ask to
>> install them.
>
> YMMV, but I happened more than once to get an XP rebooted without
> consent because it was updated and unattended long enough (coffee
> break... oh, login again!)
Same here.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
"Kenneth" <kdw### [at] gmailcom> wrote:
> "MichaelJF" <mi-### [at] t-onlinede> wrote:
>
>
> > > > Seems the link vanished. Hope it can be seen now:
>
> Nope, none of the your three tries are showing up at my end (on latest Firefox.)
>
> BUT, when I went to post this message, the link DID come up in your own message.
> I see that it's in quotes. Should it have them? Here's the same link, posted
> without the quotes; see if it shows up successfully...
>
>
http://news.povray.org/povray.beta-test/thread/%3Cweb.5033e17517f31266b62e53d00@news.povray.org%3E/
The quoting was an idea to show it after two attemps without quoting in vain.
Best regards,
Michael
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
clipka <ano### [at] anonymousorg> wrote:
> Am 30.01.2013 22:13, schrieb Le_Forgeron:
> > Le 30/01/2013 20:51, Alain nous fit lire :
> >> When a Windows update need to reboot, it ask you and will wait untill
> >> you sellect "Restart now" or "Ask again in...".
> >> There have been a few times when I had that dialog open for a few days.
> >> Any way, I prefer the setting that download the updated, then ask to
> >> install them.
> >
> > YMMV, but I happened more than once to get an XP rebooted without
> > consent because it was updated and unattended long enough (coffee
> > break... oh, login again!)
>
> Same here.
For all I experienced: With "important" updates you have some 15 minutes to
answer the dialog otherwise the machine is shut down regardless of the
applications running. Only if you enter the update option and change this
behaviour you have a chance to save your work. It's a Windows issue since XP I
think. And it has nothing do to with the running application. Recently I had a
shutdown of this kind with a machine which is usually not connected to the net
and I was in need to use a network printer... No POV running at this time.
The issue with the missing link seems to be a browser issue (I use IE 9 so far).
The problem with the +c is of course a problem with POV and a solution would be
nice, but IMO not a severe problem.
Best regards,
Michael
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
On 30-1-2013 19:44, Chris Cason wrote:
> On 30/01/2013 20:59, clipka wrote:
>> Did anyone ever file a bug report on this issue?
>
> I don't recall seeing one. It's definitely something we need to look
> at as I can't see us releasing 3.7 if it's faulty.
afaic, I thought it was normal behaviour :-)
Thomas
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Am 30.01.2013 19:44, schrieb Chris Cason:
> On 30/01/2013 20:59, clipka wrote:
>>> Very mysterious, first time a link to the POV-newsgroups doesn't work. It was at
>>> povray.beta-test, "Resuming renders in 3.7 RC6 is confusing at best, broken at
>>> worst" by Chaanakya at 21 Aug 2012 19:30:00.
>>
>> The very first link worked perfectly fine for me (I haven't tested the
>> others).
>>
>> Did anyone ever file a bug report on this issue?
>
> I don't recall seeing one. It's definitely something we need to look
> at as I can't see us releasing 3.7 if it's faulty.
I tested it, and although I didn't actually manage to reproduce the
visible symptoms at first, stepping through the code revealed some bad
internal behaviour that could well lead to the phenomenon mentioned.
It turns out that we did have a fencepost error in the code that checks
whether a rendered block is consecutive to the previous ones or not.
(I also noticed that the respective code is pretty braindead anyway, but
as it won't lead to errors - to the contrary, there's nothing much to go
wrong in there except for that one fencepost - I think we can live with
it for 3.7.0.)
BTW, the bug is now officially FS#270.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
clipka <ano### [at] anonymousorg> wrote:
> I think we can live with
> it for 3.7.0.)
>
>
> BTW, the bug is now officially FS#270.
I think so as well. But one should issue an comment to the docs. At the times I
had a one core machine I used the c+ option frequently to stop a render and use
the machine for other purposes. After finishing the other job I resumed POV with
+c. The most programs I use today are not really supporting multithreading and
so I have not to stop POV but use the +wt4 (or +wt6) option to have enough
resources for my other applications. That work around is the reason why I judged
this problem as not severe. But a hint within the docs to this issue would be
fine especially for newbies.
Best regards,
Michael
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Am 01.02.2013 17:24, schrieb MichaelJF:
> clipka <ano### [at] anonymousorg> wrote:
>> I think we can live with
>> it for 3.7.0.)
>>
>>
>> BTW, the bug is now officially FS#270.
>
> I think so as well. But one should issue an comment to the docs. At the times I
> had a one core machine I used the c+ option frequently to stop a render and use
> the machine for other purposes. After finishing the other job I resumed POV with
> +c. The most programs I use today are not really supporting multithreading and
> so I have not to stop POV but use the +wt4 (or +wt6) option to have enough
> resources for my other applications. That work around is the reason why I judged
> this problem as not severe. But a hint within the docs to this issue would be
> fine especially for newbies.
When I wrote "I think we can live with it [...]", what I meant was not
the bug, but other quirks of the code in which the bug was emplaced.
Those quirks prevent the render continue feature from being as efficient
as intended (especially with respect to memory consumption), but have no
other drawbacks.
Thus, there is no reason to add anything to the docs: The bug won't stay
in there (in other words, the next release will no longer skip any
blocks when continuing a previously aborted render), and the other
quirks will remain invisible to the user.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
On 30/01/2013 07:53, MichaelJF wrote:
> Very mysterious, first time a link to the POV-newsgroups doesn't work. It was at
The missing link in the web view of the post was a result of an issue
in the code we use to sanitize email addresses; the PHP function in
question was returning NULL upon failure rather than the original
unchanged string. I've worked around this now.
-- Chris
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
On 2-2-2013 2:26, clipka wrote:
> Thus, there is no reason to add anything to the docs: The bug won't stay
> in there (in other words, the next release will no longer skip any
> blocks when continuing a previously aborted render), and the other
> quirks will remain invisible to the user.
>
You guys are truly amazing! Thanks :-)
Thomas
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
