|
From: Pau G. <pau...@gm...> - 2010-12-21 22:10:28
|
Fabien, thanks for the answer and the pointer to your plug-ins, they will be very useful. After catching the kOfxStatFailed when fetching the images the examples work fine. I'll send a patch with the changes I've made. I'm using the last version of Nuke on a 64bits linux, but I've compulsively tried a bunch of versions from the Foundry's ftp, to see if the error happened only some of them. As it turned out, it was not a version problem but an oversight in the example. cheers, pau On Tue, Dec 21, 2010 at 5:10 PM, <fab...@fr...> wrote: > Hi, > The default examples are not really great. > Effectively, nuke often returns an empty image (I don't know why) and you have to return an error. And nuke does the process again, with a good image... > We have created an open source project for openfx plugins and a library to batch a graph of openfx nodes. The project is still under early development, but you can found a bunch of plugin examples, simple or advanced ones. > https://github.com/tuttleofx/TuttleOFX/tree/master/plugins/image/process > There is no documentation, so if you have any question, don't hesitate to ask me. > To build see INSTALL.scons (the bjam version is no more used and so not up to date). > > Regards, > Fabien Castan > > PS: which version of nuke are you using ? There is many openfx bug fixes in nuke >= 6.0. > > ----- Mail Original ----- > De: "Pau Gargallo" <pau...@gm...> > À: ope...@li... > Envoyé: Mardi 21 Décembre 2010 13h01:37 GMT +01:00 Amsterdam / Berlin / Berne / Rome / Stockholm / Vienne > Objet: [Openfx-developer] Openfx examples in Nuke, problems and fixes > > Hi all, > > I'm new to OpenFX. I've downloaded the headers and examples, compiled > them and loaded them into Nuke without problems. The problems arrive > when I actually use the plug-in examples. They work properly as long > as you do not move the sliders too much. But as soon as you move a > slider (any of them) a bit to fast, Nuke crashes. I've tried multiple > combinations of Nuke and OpenFX versions both on linux and mac, and > had always the same crash. > > The problem seems to be that the render method in the examples does > not check the validity of the source image. The call > > gEffectHost->clipGetImage(myData->sourceClip, time, NULL, &sourceImg); > > in line 667 of the basic.cpp example, is sporadically returning > kOfxStatFailed, but the return value is ignored. > > What is the proper way to handle such failure? > Should the code in the examples be modified to handle this? > > Thanks in advance, > pau > > ------------------------------------------------------------------------------ > Lotusphere 2011 > Register now for Lotusphere 2011 and learn how > to connect the dots, take your collaborative environment > to the next level, and enter the era of Social Business. > http://p.sf.net/sfu/lotusphere-d2d > _______________________________________________ > Openfx-developer mailing list > Ope...@li... > https://lists.sourceforge.net/lists/listinfo/openfx-developer > |
|
From: Pierre J. <ja...@re...> - 2010-12-22 00:11:44
|
On 12/21/2010 2:42 PM, Pau Gargallo wrote:
> Ok, for what it's worth, here is a patch against OpenFX version 1.2.1
> of Examples/Basic/basic.cpp to handle empty images and make the
> plug-in work with Nuke.
>
> To apply:
> $ cd Examples/Basic
> $ patch< basic.cpp.patch
>
Nuke does not like to be returned Failed, you have to catch that and
return kOfxStatOK;
FWIW, Here's my cookbook for all tested OFX hosts here (varies, not just
Nuke here)
* check main input is connected
gPropHost->propGetInt(props, kOfxImageClipPropConnected, 0, &connected);
(if not throw)
* check current time you are asked to render is within valid frame range
of clip
gPropHost->propGetDoubleN(props, kOfxImageEffectPropFrameRange, 2,
frameRange);
(if not throw)
* get handle
gEffectHost->clipGetHandle(effect, clipToGet, &clipHandle , 0);
* get image
ofxStatus = gEffectHost->clipGetImage(...,&img[i]);)
if (ofxStatus != kOfxStatOK || img[i] == NULL)
{ throw and cleanup }
* get image pointer
gPropHost->propGetPointer(imageHandle, kOfxImagePropData, 0, &r);
check again if data is null, if it is { throw and cleanup }
* check abort...
if ->abort (throw and cleanup)
> Cheers,
> pau
>
>
> On Tue, Dec 21, 2010 at 11:10 PM, Pau Gargallo<pau...@gm...> wrote:
>
>> Fabien,
>>
>> thanks for the answer and the pointer to your plug-ins, they will be
>> very useful.
>>
>> After catching the kOfxStatFailed when fetching the images the
>> examples work fine. I'll send a patch with the changes I've made.
>>
>> I'm using the last version of Nuke on a 64bits linux, but I've
>> compulsively tried a bunch of versions from the Foundry's ftp, to see
>> if the error happened only some of them. As it turned out, it was not
>> a version problem but an oversight in the example.
>>
>> cheers,
>> pau
>>
>>
>>
>> On Tue, Dec 21, 2010 at 5:10 PM,<fab...@fr...> wrote:
>>
>>> Hi,
>>> The default examples are not really great.
>>> Effectively, nuke often returns an empty image (I don't know why) and you have to return an error. And nuke does the process again, with a good image...
>>> We have created an open source project for openfx plugins and a library to batch a graph of openfx nodes. The project is still under early development, but you can found a bunch of plugin examples, simple or advanced ones.
>>> https://github.com/tuttleofx/TuttleOFX/tree/master/plugins/image/process
>>> There is no documentation, so if you have any question, don't hesitate to ask me.
>>> To build see INSTALL.scons (the bjam version is no more used and so not up to date).
>>>
>>> Regards,
>>> Fabien Castan
>>>
>>> PS: which version of nuke are you using ? There is many openfx bug fixes in nuke>= 6.0.
>>>
>>> ----- Mail Original -----
>>> De: "Pau Gargallo"<pau...@gm...>
>>> À: ope...@li...
>>> Envoyé: Mardi 21 Décembre 2010 13h01:37 GMT +01:00 Amsterdam / Berlin / Berne / Rome / Stockholm / Vienne
>>> Objet: [Openfx-developer] Openfx examples in Nuke, problems and fixes
>>>
>>> Hi all,
>>>
>>> I'm new to OpenFX. I've downloaded the headers and examples, compiled
>>> them and loaded them into Nuke without problems. The problems arrive
>>> when I actually use the plug-in examples. They work properly as long
>>> as you do not move the sliders too much. But as soon as you move a
>>> slider (any of them) a bit to fast, Nuke crashes. I've tried multiple
>>> combinations of Nuke and OpenFX versions both on linux and mac, and
>>> had always the same crash.
>>>
>>> The problem seems to be that the render method in the examples does
>>> not check the validity of the source image. The call
>>>
>>> gEffectHost->clipGetImage(myData->sourceClip, time, NULL,&sourceImg);
>>>
>>> in line 667 of the basic.cpp example, is sporadically returning
>>> kOfxStatFailed, but the return value is ignored.
>>>
>>> What is the proper way to handle such failure?
>>> Should the code in the examples be modified to handle this?
>>>
>>> Thanks in advance,
>>> pau
>>>
>>> ------------------------------------------------------------------------------
>>> Lotusphere 2011
>>> Register now for Lotusphere 2011 and learn how
>>> to connect the dots, take your collaborative environment
>>> to the next level, and enter the era of Social Business.
>>> http://p.sf.net/sfu/lotusphere-d2d
>>> _______________________________________________
>>> Openfx-developer mailing list
>>> Ope...@li...
>>> https://lists.sourceforge.net/lists/listinfo/openfx-developer
>>>
>>>
>>
>>
>>
>> ------------------------------------------------------------------------------
>> Forrester recently released a report on the Return on Investment (ROI) of
>> Google Apps. They found a 300% ROI, 38%-56% cost savings, and break-even
>> within 7 months. Over 3 million businesses have gone Google with Google Apps:
>> an online email calendar, and document program that's accessible from your
>> browser. Read the Forrester report: http://p.sf.net/sfu/googleapps-sfnew
>>
>>
>> _______________________________________________
>> Openfx-developer mailing list
>> Ope...@li...
>> https://lists.sourceforge.net/lists/listinfo/openfx-developer
>>
|
|
From: Pau G. <pau...@gm...> - 2010-12-22 08:55:02
|
Thanks a lot for the cookbook Pierre!
It will be very helpful.
pau
On Wed, Dec 22, 2010 at 12:57 AM, Pierre Jasmin <ja...@re...> wrote:
> On 12/21/2010 2:42 PM, Pau Gargallo wrote:
>
> Ok, for what it's worth, here is a patch against OpenFX version 1.2.1
> of Examples/Basic/basic.cpp to handle empty images and make the
> plug-in work with Nuke.
>
> To apply:
> $ cd Examples/Basic
> $ patch < basic.cpp.patch
>
>
> Nuke does not like to be returned Failed, you have to catch that and return
> kOfxStatOK;
>
> FWIW, Here's my cookbook for all tested OFX hosts here (varies, not just
> Nuke here)
>
> * check main input is connected
> gPropHost->propGetInt(props, kOfxImageClipPropConnected, 0, &connected);
> (if not throw)
>
> * check current time you are asked to render is within valid frame range of
> clip
> gPropHost->propGetDoubleN(props, kOfxImageEffectPropFrameRange, 2,
> frameRange);
> (if not throw)
>
> * get handle
> gEffectHost->clipGetHandle(effect, clipToGet, &clipHandle , 0);
>
> * get image
> ofxStatus = gEffectHost->clipGetImage(...,&img[i]);)
> if (ofxStatus != kOfxStatOK || img[i] == NULL)
> { throw and cleanup }
>
> * get image pointer
> gPropHost->propGetPointer(imageHandle, kOfxImagePropData, 0, &r);
> check again if data is null, if it is { throw and cleanup }
>
> * check abort...
> if ->abort (throw and cleanup)
>
>
> Cheers,
> pau
>
>
> On Tue, Dec 21, 2010 at 11:10 PM, Pau Gargallo <pau...@gm...>
> wrote:
>
>
> Fabien,
>
> thanks for the answer and the pointer to your plug-ins, they will be
> very useful.
>
> After catching the kOfxStatFailed when fetching the images the
> examples work fine. I'll send a patch with the changes I've made.
>
> I'm using the last version of Nuke on a 64bits linux, but I've
> compulsively tried a bunch of versions from the Foundry's ftp, to see
> if the error happened only some of them. As it turned out, it was not
> a version problem but an oversight in the example.
>
> cheers,
> pau
>
>
>
> On Tue, Dec 21, 2010 at 5:10 PM, <fab...@fr...> wrote:
>
>
> Hi,
> The default examples are not really great.
> Effectively, nuke often returns an empty image (I don't know why) and you
> have to return an error. And nuke does the process again, with a good
> image...
> We have created an open source project for openfx plugins and a library to
> batch a graph of openfx nodes. The project is still under early development,
> but you can found a bunch of plugin examples, simple or advanced ones.
> https://github.com/tuttleofx/TuttleOFX/tree/master/plugins/image/process
> There is no documentation, so if you have any question, don't hesitate to
> ask me.
> To build see INSTALL.scons (the bjam version is no more used and so not up
> to date).
>
> Regards,
> Fabien Castan
>
> PS: which version of nuke are you using ? There is many openfx bug fixes in
> nuke >= 6.0.
>
> ----- Mail Original -----
> De: "Pau Gargallo" <pau...@gm...>
> À: ope...@li...
> Envoyé: Mardi 21 Décembre 2010 13h01:37 GMT +01:00 Amsterdam / Berlin /
> Berne / Rome / Stockholm / Vienne
> Objet: [Openfx-developer] Openfx examples in Nuke, problems and fixes
>
> Hi all,
>
> I'm new to OpenFX. I've downloaded the headers and examples, compiled
> them and loaded them into Nuke without problems. The problems arrive
> when I actually use the plug-in examples. They work properly as long
> as you do not move the sliders too much. But as soon as you move a
> slider (any of them) a bit to fast, Nuke crashes. I've tried multiple
> combinations of Nuke and OpenFX versions both on linux and mac, and
> had always the same crash.
>
> The problem seems to be that the render method in the examples does
> not check the validity of the source image. The call
>
> gEffectHost->clipGetImage(myData->sourceClip, time, NULL, &sourceImg);
>
> in line 667 of the basic.cpp example, is sporadically returning
> kOfxStatFailed, but the return value is ignored.
>
> What is the proper way to handle such failure?
> Should the code in the examples be modified to handle this?
>
> Thanks in advance,
> pau
>
> ------------------------------------------------------------------------------
> Lotusphere 2011
> Register now for Lotusphere 2011 and learn how
> to connect the dots, take your collaborative environment
> to the next level, and enter the era of Social Business.
> http://p.sf.net/sfu/lotusphere-d2d
> _______________________________________________
> Openfx-developer mailing list
> Ope...@li...
> https://lists.sourceforge.net/lists/listinfo/openfx-developer
>
>
>
>
>
> ------------------------------------------------------------------------------
> Forrester recently released a report on the Return on Investment (ROI) of
> Google Apps. They found a 300% ROI, 38%-56% cost savings, and break-even
> within 7 months. Over 3 million businesses have gone Google with Google
> Apps:
> an online email calendar, and document program that's accessible from your
> browser. Read the Forrester report: http://p.sf.net/sfu/googleapps-sfnew
>
> _______________________________________________
> Openfx-developer mailing list
> Ope...@li...
> https://lists.sourceforge.net/lists/listinfo/openfx-developer
>
>
> ------------------------------------------------------------------------------
> Forrester recently released a report on the Return on Investment (ROI) of
> Google Apps. They found a 300% ROI, 38%-56% cost savings, and break-even
> within 7 months. Over 3 million businesses have gone Google with Google
> Apps:
> an online email calendar, and document program that's accessible from your
> browser. Read the Forrester report: http://p.sf.net/sfu/googleapps-sfnew
> _______________________________________________
> Openfx-developer mailing list
> Ope...@li...
> https://lists.sourceforge.net/lists/listinfo/openfx-developer
>
>
|
|
From: Bruno N. <br...@th...> - 2010-12-22 11:19:25
|
What is happening is that Nuke is interrupting all process when you move a slider about. Which means you plugin gets interrupted. So if a plugin fetches an image and part way through that Nuke interrupts, then the fetch image will fail, because all upstream processing has halted. This is not a bug in Nuke, it is the way Nuke works. If you check the interruption status after a fetch, you will see that you will have been interrupted. Interruption is not an error state, you plugin has not gone wrong. In which case you should return OK. b On 21 Dec 2010, at 22:10, Pau Gargallo wrote: > Fabien, > > thanks for the answer and the pointer to your plug-ins, they will be > very useful. > > After catching the kOfxStatFailed when fetching the images the > examples work fine. I'll send a patch with the changes I've made. > > I'm using the last version of Nuke on a 64bits linux, but I've > compulsively tried a bunch of versions from the Foundry's ftp, to see > if the error happened only some of them. As it turned out, it was not > a version problem but an oversight in the example. > > cheers, > pau > > > > On Tue, Dec 21, 2010 at 5:10 PM, <fab...@fr...> wrote: >> Hi, >> The default examples are not really great. >> Effectively, nuke often returns an empty image (I don't know why) and you have to return an error. And nuke does the process again, with a good image... >> We have created an open source project for openfx plugins and a library to batch a graph of openfx nodes. The project is still under early development, but you can found a bunch of plugin examples, simple or advanced ones. >> https://github.com/tuttleofx/TuttleOFX/tree/master/plugins/image/process >> There is no documentation, so if you have any question, don't hesitate to ask me. >> To build see INSTALL.scons (the bjam version is no more used and so not up to date). >> >> Regards, >> Fabien Castan >> >> PS: which version of nuke are you using ? There is many openfx bug fixes in nuke >= 6.0. >> >> ----- Mail Original ----- >> De: "Pau Gargallo" <pau...@gm...> >> À: ope...@li... >> Envoyé: Mardi 21 Décembre 2010 13h01:37 GMT +01:00 Amsterdam / Berlin / Berne / Rome / Stockholm / Vienne >> Objet: [Openfx-developer] Openfx examples in Nuke, problems and fixes >> >> Hi all, >> >> I'm new to OpenFX. I've downloaded the headers and examples, compiled >> them and loaded them into Nuke without problems. The problems arrive >> when I actually use the plug-in examples. They work properly as long >> as you do not move the sliders too much. But as soon as you move a >> slider (any of them) a bit to fast, Nuke crashes. I've tried multiple >> combinations of Nuke and OpenFX versions both on linux and mac, and >> had always the same crash. >> >> The problem seems to be that the render method in the examples does >> not check the validity of the source image. The call >> >> gEffectHost->clipGetImage(myData->sourceClip, time, NULL, &sourceImg); >> >> in line 667 of the basic.cpp example, is sporadically returning >> kOfxStatFailed, but the return value is ignored. >> >> What is the proper way to handle such failure? >> Should the code in the examples be modified to handle this? >> >> Thanks in advance, >> pau >> >> ------------------------------------------------------------------------------ >> Lotusphere 2011 >> Register now for Lotusphere 2011 and learn how >> to connect the dots, take your collaborative environment >> to the next level, and enter the era of Social Business. >> http://p.sf.net/sfu/lotusphere-d2d >> _______________________________________________ >> Openfx-developer mailing list >> Ope...@li... >> https://lists.sourceforge.net/lists/listinfo/openfx-developer >> > > ------------------------------------------------------------------------------ > Forrester recently released a report on the Return on Investment (ROI) of > Google Apps. They found a 300% ROI, 38%-56% cost savings, and break-even > within 7 months. Over 3 million businesses have gone Google with Google Apps: > an online email calendar, and document program that's accessible from your > browser. Read the Forrester report: http://p.sf.net/sfu/googleapps-sfnew > _______________________________________________ > Openfx-developer mailing list > Ope...@li... > https://lists.sourceforge.net/lists/listinfo/openfx-developer -- Bruno Nicoletti, Chief Technology Officer The Foundry 6th Floor, The Communications Building, 48, Leicester Square, London, WC2H 7LT, UK Skype: brunonicoletti Tel: +44 (0)20 7968 6828 - Fax: +44 (0)20 7930 8906 Web: www.thefoundry.co.uk Email: br...@th... The Foundry Visionmongers Ltd. Registered in England and Wales No: 4642027 |
|
From: Pau G. <pau...@gm...> - 2010-12-22 12:04:46
|
That clarifies my doubts, thanks Bruno! Does it make sense to update the openfx examples to handle such case and avoid the crash? I understand that they are not meant to be complete implementations, but it would be instructive for newbies like me ;-) pau On Wed, Dec 22, 2010 at 11:46 AM, Bruno Nicoletti <br...@th...> wrote: > What is happening is that Nuke is interrupting all process when you move a slider about. Which means you plugin gets interrupted. > > So if a plugin fetches an image and part way through that Nuke interrupts, then the fetch image will fail, because all upstream processing has halted. This is not a bug in Nuke, it is the way Nuke works. If you check the interruption status after a fetch, you will see that you will have been interrupted. > > Interruption is not an error state, you plugin has not gone wrong. In which case you should return OK. > > b > > > On 21 Dec 2010, at 22:10, Pau Gargallo wrote: > >> Fabien, >> >> thanks for the answer and the pointer to your plug-ins, they will be >> very useful. >> >> After catching the kOfxStatFailed when fetching the images the >> examples work fine. I'll send a patch with the changes I've made. >> >> I'm using the last version of Nuke on a 64bits linux, but I've >> compulsively tried a bunch of versions from the Foundry's ftp, to see >> if the error happened only some of them. As it turned out, it was not >> a version problem but an oversight in the example. >> >> cheers, >> pau >> >> >> >> On Tue, Dec 21, 2010 at 5:10 PM, <fab...@fr...> wrote: >>> Hi, >>> The default examples are not really great. >>> Effectively, nuke often returns an empty image (I don't know why) and you have to return an error. And nuke does the process again, with a good image... >>> We have created an open source project for openfx plugins and a library to batch a graph of openfx nodes. The project is still under early development, but you can found a bunch of plugin examples, simple or advanced ones. >>> https://github.com/tuttleofx/TuttleOFX/tree/master/plugins/image/process >>> There is no documentation, so if you have any question, don't hesitate to ask me. >>> To build see INSTALL.scons (the bjam version is no more used and so not up to date). >>> >>> Regards, >>> Fabien Castan >>> >>> PS: which version of nuke are you using ? There is many openfx bug fixes in nuke >= 6.0. >>> >>> ----- Mail Original ----- >>> De: "Pau Gargallo" <pau...@gm...> >>> À: ope...@li... >>> Envoyé: Mardi 21 Décembre 2010 13h01:37 GMT +01:00 Amsterdam / Berlin / Berne / Rome / Stockholm / Vienne >>> Objet: [Openfx-developer] Openfx examples in Nuke, problems and fixes >>> >>> Hi all, >>> >>> I'm new to OpenFX. I've downloaded the headers and examples, compiled >>> them and loaded them into Nuke without problems. The problems arrive >>> when I actually use the plug-in examples. They work properly as long >>> as you do not move the sliders too much. But as soon as you move a >>> slider (any of them) a bit to fast, Nuke crashes. I've tried multiple >>> combinations of Nuke and OpenFX versions both on linux and mac, and >>> had always the same crash. >>> >>> The problem seems to be that the render method in the examples does >>> not check the validity of the source image. The call >>> >>> gEffectHost->clipGetImage(myData->sourceClip, time, NULL, &sourceImg); >>> >>> in line 667 of the basic.cpp example, is sporadically returning >>> kOfxStatFailed, but the return value is ignored. >>> >>> What is the proper way to handle such failure? >>> Should the code in the examples be modified to handle this? >>> >>> Thanks in advance, >>> pau >>> >>> ------------------------------------------------------------------------------ >>> Lotusphere 2011 >>> Register now for Lotusphere 2011 and learn how >>> to connect the dots, take your collaborative environment >>> to the next level, and enter the era of Social Business. >>> http://p.sf.net/sfu/lotusphere-d2d >>> _______________________________________________ >>> Openfx-developer mailing list >>> Ope...@li... >>> https://lists.sourceforge.net/lists/listinfo/openfx-developer >>> >> >> ------------------------------------------------------------------------------ >> Forrester recently released a report on the Return on Investment (ROI) of >> Google Apps. They found a 300% ROI, 38%-56% cost savings, and break-even >> within 7 months. Over 3 million businesses have gone Google with Google Apps: >> an online email calendar, and document program that's accessible from your >> browser. Read the Forrester report: http://p.sf.net/sfu/googleapps-sfnew >> _______________________________________________ >> Openfx-developer mailing list >> Ope...@li... >> https://lists.sourceforge.net/lists/listinfo/openfx-developer > > -- > Bruno Nicoletti, > Chief Technology Officer > The Foundry > 6th Floor, The Communications Building, > 48, Leicester Square, > London, WC2H 7LT, UK > Skype: brunonicoletti > Tel: +44 (0)20 7968 6828 - Fax: +44 (0)20 7930 8906 > Web: www.thefoundry.co.uk > Email: br...@th... > > The Foundry Visionmongers Ltd. > Registered in England and Wales No: 4642027 > > > ------------------------------------------------------------------------------ > Forrester recently released a report on the Return on Investment (ROI) of > Google Apps. They found a 300% ROI, 38%-56% cost savings, and break-even > within 7 months. Over 3 million businesses have gone Google with Google Apps: > an online email calendar, and document program that's accessible from your > browser. Read the Forrester report: http://p.sf.net/sfu/googleapps-sfnew > _______________________________________________ > Openfx-developer mailing list > Ope...@li... > https://lists.sourceforge.net/lists/listinfo/openfx-developer > |
|
From: Bruno N. <br...@th...> - 2010-12-22 12:20:05
|
Yes it would, I'll add in your patch when I get the chance later on today. b On 22 Dec 2010, at 12:04, Pau Gargallo wrote: > That clarifies my doubts, thanks Bruno! > > Does it make sense to update the openfx examples to handle such case > and avoid the crash? I understand that they are not meant to be > complete implementations, but it would be instructive for newbies like > me ;-) > > pau > > > On Wed, Dec 22, 2010 at 11:46 AM, Bruno Nicoletti > <br...@th...> wrote: >> What is happening is that Nuke is interrupting all process when you move a slider about. Which means you plugin gets interrupted. >> >> So if a plugin fetches an image and part way through that Nuke interrupts, then the fetch image will fail, because all upstream processing has halted. This is not a bug in Nuke, it is the way Nuke works. If you check the interruption status after a fetch, you will see that you will have been interrupted. >> >> Interruption is not an error state, you plugin has not gone wrong. In which case you should return OK. >> >> b >> >> >> On 21 Dec 2010, at 22:10, Pau Gargallo wrote: >> >>> Fabien, >>> >>> thanks for the answer and the pointer to your plug-ins, they will be >>> very useful. >>> >>> After catching the kOfxStatFailed when fetching the images the >>> examples work fine. I'll send a patch with the changes I've made. >>> >>> I'm using the last version of Nuke on a 64bits linux, but I've >>> compulsively tried a bunch of versions from the Foundry's ftp, to see >>> if the error happened only some of them. As it turned out, it was not >>> a version problem but an oversight in the example. >>> >>> cheers, >>> pau >>> >>> >>> >>> On Tue, Dec 21, 2010 at 5:10 PM, <fab...@fr...> wrote: >>>> Hi, >>>> The default examples are not really great. >>>> Effectively, nuke often returns an empty image (I don't know why) and you have to return an error. And nuke does the process again, with a good image... >>>> We have created an open source project for openfx plugins and a library to batch a graph of openfx nodes. The project is still under early development, but you can found a bunch of plugin examples, simple or advanced ones. >>>> https://github.com/tuttleofx/TuttleOFX/tree/master/plugins/image/process >>>> There is no documentation, so if you have any question, don't hesitate to ask me. >>>> To build see INSTALL.scons (the bjam version is no more used and so not up to date). >>>> >>>> Regards, >>>> Fabien Castan >>>> >>>> PS: which version of nuke are you using ? There is many openfx bug fixes in nuke >= 6.0. >>>> >>>> ----- Mail Original ----- >>>> De: "Pau Gargallo" <pau...@gm...> >>>> À: ope...@li... >>>> Envoyé: Mardi 21 Décembre 2010 13h01:37 GMT +01:00 Amsterdam / Berlin / Berne / Rome / Stockholm / Vienne >>>> Objet: [Openfx-developer] Openfx examples in Nuke, problems and fixes >>>> >>>> Hi all, >>>> >>>> I'm new to OpenFX. I've downloaded the headers and examples, compiled >>>> them and loaded them into Nuke without problems. The problems arrive >>>> when I actually use the plug-in examples. They work properly as long >>>> as you do not move the sliders too much. But as soon as you move a >>>> slider (any of them) a bit to fast, Nuke crashes. I've tried multiple >>>> combinations of Nuke and OpenFX versions both on linux and mac, and >>>> had always the same crash. >>>> >>>> The problem seems to be that the render method in the examples does >>>> not check the validity of the source image. The call >>>> >>>> gEffectHost->clipGetImage(myData->sourceClip, time, NULL, &sourceImg); >>>> >>>> in line 667 of the basic.cpp example, is sporadically returning >>>> kOfxStatFailed, but the return value is ignored. >>>> >>>> What is the proper way to handle such failure? >>>> Should the code in the examples be modified to handle this? >>>> >>>> Thanks in advance, >>>> pau >>>> >>>> ------------------------------------------------------------------------------ >>>> Lotusphere 2011 >>>> Register now for Lotusphere 2011 and learn how >>>> to connect the dots, take your collaborative environment >>>> to the next level, and enter the era of Social Business. >>>> http://p.sf.net/sfu/lotusphere-d2d >>>> _______________________________________________ >>>> Openfx-developer mailing list >>>> Ope...@li... >>>> https://lists.sourceforge.net/lists/listinfo/openfx-developer >>>> >>> >>> ------------------------------------------------------------------------------ >>> Forrester recently released a report on the Return on Investment (ROI) of >>> Google Apps. They found a 300% ROI, 38%-56% cost savings, and break-even >>> within 7 months. Over 3 million businesses have gone Google with Google Apps: >>> an online email calendar, and document program that's accessible from your >>> browser. Read the Forrester report: http://p.sf.net/sfu/googleapps-sfnew >>> _______________________________________________ >>> Openfx-developer mailing list >>> Ope...@li... >>> https://lists.sourceforge.net/lists/listinfo/openfx-developer >> >> -- >> Bruno Nicoletti, >> Chief Technology Officer >> The Foundry >> 6th Floor, The Communications Building, >> 48, Leicester Square, >> London, WC2H 7LT, UK >> Skype: brunonicoletti >> Tel: +44 (0)20 7968 6828 - Fax: +44 (0)20 7930 8906 >> Web: www.thefoundry.co.uk >> Email: br...@th... >> >> The Foundry Visionmongers Ltd. >> Registered in England and Wales No: 4642027 >> >> >> ------------------------------------------------------------------------------ >> Forrester recently released a report on the Return on Investment (ROI) of >> Google Apps. They found a 300% ROI, 38%-56% cost savings, and break-even >> within 7 months. Over 3 million businesses have gone Google with Google Apps: >> an online email calendar, and document program that's accessible from your >> browser. Read the Forrester report: http://p.sf.net/sfu/googleapps-sfnew >> _______________________________________________ >> Openfx-developer mailing list >> Ope...@li... >> https://lists.sourceforge.net/lists/listinfo/openfx-developer >> > > ------------------------------------------------------------------------------ > Forrester recently released a report on the Return on Investment (ROI) of > Google Apps. They found a 300% ROI, 38%-56% cost savings, and break-even > within 7 months. Over 3 million businesses have gone Google with Google Apps: > an online email calendar, and document program that's accessible from your > browser. Read the Forrester report: http://p.sf.net/sfu/googleapps-sfnew > _______________________________________________ > Openfx-developer mailing list > Ope...@li... > https://lists.sourceforge.net/lists/listinfo/openfx-developer -- Bruno Nicoletti, Chief Technology Officer The Foundry 6th Floor, The Communications Building, 48, Leicester Square, London, WC2H 7LT, UK Skype: brunonicoletti Tel: +44 (0)20 7968 6828 - Fax: +44 (0)20 7930 8906 Web: www.thefoundry.co.uk Email: br...@th... The Foundry Visionmongers Ltd. Registered in England and Wales No: 4642027 |
|
From: Pau G. <pau...@gm...> - 2010-12-22 15:24:08
Attachments:
basic.cpp.patch
|
use this one instead then, it returns OK. thanks, pau On Wed, Dec 22, 2010 at 1:19 PM, Bruno Nicoletti <br...@th...> wrote: > Yes it would, I'll add in your patch when I get the chance later on today. > > b > |
|
From: Bruno N. <br...@th...> - 2010-12-23 18:39:12
|
I'm being a bit slow on this, in theory I am on holidays and I am sharing child care duties. I'm doing a slightly more general fix to all the example plugins, not just basic. Should be up soon ish. b On 22 Dec 2010, at 15:24, Pau Gargallo wrote: > use this one instead then, it returns OK. > > thanks, > pau > > On Wed, Dec 22, 2010 at 1:19 PM, Bruno Nicoletti <br...@th...> wrote: >> Yes it would, I'll add in your patch when I get the chance later on today. >> >> b >> > <basic.cpp.patch>------------------------------------------------------------------------------ > Forrester recently released a report on the Return on Investment (ROI) of > Google Apps. They found a 300% ROI, 38%-56% cost savings, and break-even > within 7 months. Over 3 million businesses have gone Google with Google Apps: > an online email calendar, and document program that's accessible from your > browser. Read the Forrester report: http://p.sf.net/sfu/googleapps-sfnew_______________________________________________ > Openfx-developer mailing list > Ope...@li... > https://lists.sourceforge.net/lists/listinfo/openfx-developer -- Bruno Nicoletti, Chief Technology Officer The Foundry 6th Floor, The Communications Building, 48, Leicester Square, London, WC2H 7LT, UK Skype: brunonicoletti Tel: +44 (0)20 7968 6828 - Fax: +44 (0)20 7930 8906 Web: www.thefoundry.co.uk Email: br...@th... The Foundry Visionmongers Ltd. Registered in England and Wales No: 4642027 |
|
From: Pau G. <pau...@gm...> - 2010-12-23 20:59:41
|
Thanks a lot Bruno, sorry for inciting more work :( Merry Christmas to everyone, pau On Thu, Dec 23, 2010 at 7:39 PM, Bruno Nicoletti <br...@th...> wrote: > I'm being a bit slow on this, in theory I am on holidays and I am sharing child care duties. > > I'm doing a slightly more general fix to all the example plugins, not just basic. Should be up soon ish. > > b > > On 22 Dec 2010, at 15:24, Pau Gargallo wrote: > >> use this one instead then, it returns OK. >> >> thanks, >> pau >> >> On Wed, Dec 22, 2010 at 1:19 PM, Bruno Nicoletti <br...@th...> wrote: >>> Yes it would, I'll add in your patch when I get the chance later on today. >>> >>> b >>> >> <basic.cpp.patch>------------------------------------------------------------------------------ >> Forrester recently released a report on the Return on Investment (ROI) of >> Google Apps. They found a 300% ROI, 38%-56% cost savings, and break-even >> within 7 months. Over 3 million businesses have gone Google with Google Apps: >> an online email calendar, and document program that's accessible from your >> browser. Read the Forrester report: http://p.sf.net/sfu/googleapps-sfnew_______________________________________________ >> Openfx-developer mailing list >> Ope...@li... >> https://lists.sourceforge.net/lists/listinfo/openfx-developer > > -- > Bruno Nicoletti, > Chief Technology Officer > The Foundry > 6th Floor, The Communications Building, > 48, Leicester Square, > London, WC2H 7LT, UK > Skype: brunonicoletti > Tel: +44 (0)20 7968 6828 - Fax: +44 (0)20 7930 8906 > Web: www.thefoundry.co.uk > Email: br...@th... > > The Foundry Visionmongers Ltd. > Registered in England and Wales No: 4642027 > > > ------------------------------------------------------------------------------ > Learn how Oracle Real Application Clusters (RAC) One Node allows customers > to consolidate database storage, standardize their database environment, and, > should the need arise, upgrade to a full multi-node Oracle RAC database > without downtime or disruption > http://p.sf.net/sfu/oracle-sfdevnl > _______________________________________________ > Openfx-developer mailing list > Ope...@li... > https://lists.sourceforge.net/lists/listinfo/openfx-developer > |
|
From: Bruno N. <br...@th...> - 2010-12-24 09:36:12
|
OK, fixes are up in CVS and as a new file release. b On 23 Dec 2010, at 20:59, Pau Gargallo wrote: > Thanks a lot Bruno, sorry for inciting more work :( > Merry Christmas to everyone, > pau > > On Thu, Dec 23, 2010 at 7:39 PM, Bruno Nicoletti <br...@th...> wrote: >> I'm being a bit slow on this, in theory I am on holidays and I am sharing child care duties. >> >> I'm doing a slightly more general fix to all the example plugins, not just basic. Should be up soon ish. >> >> b >> >> On 22 Dec 2010, at 15:24, Pau Gargallo wrote: >> >>> use this one instead then, it returns OK. >>> >>> thanks, >>> pau >>> >>> On Wed, Dec 22, 2010 at 1:19 PM, Bruno Nicoletti <br...@th...> wrote: >>>> Yes it would, I'll add in your patch when I get the chance later on today. >>>> >>>> b >>>> >>> <basic.cpp.patch>------------------------------------------------------------------------------ >>> Forrester recently released a report on the Return on Investment (ROI) of >>> Google Apps. They found a 300% ROI, 38%-56% cost savings, and break-even >>> within 7 months. Over 3 million businesses have gone Google with Google Apps: >>> an online email calendar, and document program that's accessible from your >>> browser. Read the Forrester report: http://p.sf.net/sfu/googleapps-sfnew_______________________________________________ >>> Openfx-developer mailing list >>> Ope...@li... >>> https://lists.sourceforge.net/lists/listinfo/openfx-developer >> >> -- >> Bruno Nicoletti, >> Chief Technology Officer >> The Foundry >> 6th Floor, The Communications Building, >> 48, Leicester Square, >> London, WC2H 7LT, UK >> Skype: brunonicoletti >> Tel: +44 (0)20 7968 6828 - Fax: +44 (0)20 7930 8906 >> Web: www.thefoundry.co.uk >> Email: br...@th... >> >> The Foundry Visionmongers Ltd. >> Registered in England and Wales No: 4642027 >> >> >> ------------------------------------------------------------------------------ >> Learn how Oracle Real Application Clusters (RAC) One Node allows customers >> to consolidate database storage, standardize their database environment, and, >> should the need arise, upgrade to a full multi-node Oracle RAC database >> without downtime or disruption >> http://p.sf.net/sfu/oracle-sfdevnl >> _______________________________________________ >> Openfx-developer mailing list >> Ope...@li... >> https://lists.sourceforge.net/lists/listinfo/openfx-developer >> > > ------------------------------------------------------------------------------ > Learn how Oracle Real Application Clusters (RAC) One Node allows customers > to consolidate database storage, standardize their database environment, and, > should the need arise, upgrade to a full multi-node Oracle RAC database > without downtime or disruption > http://p.sf.net/sfu/oracle-sfdevnl > _______________________________________________ > Openfx-developer mailing list > Ope...@li... > https://lists.sourceforge.net/lists/listinfo/openfx-developer -- Bruno Nicoletti, Chief Technology Officer The Foundry 6th Floor, The Communications Building, 48, Leicester Square, London, WC2H 7LT, UK Skype: brunonicoletti Tel: +44 (0)20 7968 6828 - Fax: +44 (0)20 7930 8906 Web: www.thefoundry.co.uk Email: br...@th... The Foundry Visionmongers Ltd. Registered in England and Wales No: 4642027 |
|
From: Fabien C. <fc...@qu...> - 2011-01-04 11:50:08
|
Hi Pierre, > Juut a short note: > I did a quick look at your source directory. I am noticing that you are > using an LPGL license v3 and you as well amend it. Yes, we choose LGPL after a long discussion with all the initial partners, so it will be difficult to change it. > It's your project but, it might be more useful if you used an MIT > x11/Modified BSD/NCSA permissive license (like OFX itself)? > http://www.gnu.org/licenses/license-list.html#Expat OpenFX is an API, so it would be strange not to be fully permissive. TuttleOFX is a library, so it's a little different. > (at least for the core library, someone might have a reason for GPL- > e.g. for a useful example.cpp or a specialized codec...) Sorry, I don't understand the idea. If someone would write GPL code, he can use an LGPL library. > Reason is some companies don't accept to use LPGL licensed code as it > becomes too complicated with their legal dept. > It's also becoming more common these days to just offer dual-licensing > -- http://en.wikipedia.org/wiki/Multi-licensing > (choose your own) so it doesn't have to become a religion. I'm not sure to understand why companies wouldn't accept to use LGPL. LGPL has only one real constrain: to redistribute the modifications in the library itself if you redistribute a product using it. That has nothing to do with GPL contaminating rule. We doesn't choose this licence by religious belief, but it's a way to manage the collaboration between the different partners. All partners can use this library in commercial products (applications or new plugins) as we do, but have to publish if they fix a bug in the core library, supports a new kind of parameter, etc. So all partners are really free to use it as they want but in the same time it's a way to keep the core library as best as possible in the interest of all. Correct me if I am wrong in my reading of the license. It's important to know if some companies don't accept to use LGPL licensed code. Could you explain me why ? Regards, Fabien |
|
From: Paul M. <pa...@fx...> - 2011-01-05 13:39:09
|
On 1/4/2011 5:15 AM, Fabien Castan wrote: > Correct me if I am wrong in my reading of the license. > It's important to know if some companies don't accept to use LGPL > licensed code. Could you explain me why ? I haven't look at it yet but my question would be how is this library getting distributed, and where would it be installed? If each plugin vendor distributes the library, wouldn't there be symbol contamination if multiple versions are used? |
|
From: Fabien C. <fc...@qu...> - 2011-01-05 16:26:49
|
Hi Paul, > I haven't look at it yet but my question would be how is this library > getting distributed, and where would it be installed? If each plugin > vendor distributes the library, wouldn't there be symbol contamination > if multiple versions are used? I'm embarrassed, I wouldn't like to pollute the OpenFX mailing list with Tuttle questions... I just activated some ways to contact us: * The public mailing-list tut...@go... * To contact the team in private tut...@go... * The github issues module, so you can use the repository for bug report http://github.com/tuttleofx/TuttleOFX/issues To quickly answer, Tuttle is composed of multiple parts: 1) host library Image processing library to manipulate a graph of openfx nodes 2) plugin library A customization of the OpenFX C++ plugin wrapping. It's very small, just to get directly boost::gil views. 3) some plugins 4) some applications (nothing substantial yet) About symbol contamination: 1) Applications using tuttle library, can link in static or dynamic with a particular version, like with all libraries. In this case, there is no reason to use multiple versions of the library. 2) Plugins are compiled in static with the c++ plugin wrapping, so each plugin can be compiled with a custom version. About distribution and installation: There is nothing special for this. You can install all the targets in a custom directory. Next you only need to add this path to the OFX_PLUGIN_PATH environnent variable. Tell me if I'm not answering your question. Really sorry for the advertisement here... Regards, Fabien |
|
From: Pau G. <pau...@gm...> - 2010-12-21 22:42:28
Attachments:
basic.cpp.patch
|
Ok, for what it's worth, here is a patch against OpenFX version 1.2.1 of Examples/Basic/basic.cpp to handle empty images and make the plug-in work with Nuke. To apply: $ cd Examples/Basic $ patch < basic.cpp.patch Cheers, pau On Tue, Dec 21, 2010 at 11:10 PM, Pau Gargallo <pau...@gm...> wrote: > Fabien, > > thanks for the answer and the pointer to your plug-ins, they will be > very useful. > > After catching the kOfxStatFailed when fetching the images the > examples work fine. I'll send a patch with the changes I've made. > > I'm using the last version of Nuke on a 64bits linux, but I've > compulsively tried a bunch of versions from the Foundry's ftp, to see > if the error happened only some of them. As it turned out, it was not > a version problem but an oversight in the example. > > cheers, > pau > > > > On Tue, Dec 21, 2010 at 5:10 PM, <fab...@fr...> wrote: >> Hi, >> The default examples are not really great. >> Effectively, nuke often returns an empty image (I don't know why) and you have to return an error. And nuke does the process again, with a good image... >> We have created an open source project for openfx plugins and a library to batch a graph of openfx nodes. The project is still under early development, but you can found a bunch of plugin examples, simple or advanced ones. >> https://github.com/tuttleofx/TuttleOFX/tree/master/plugins/image/process >> There is no documentation, so if you have any question, don't hesitate to ask me. >> To build see INSTALL.scons (the bjam version is no more used and so not up to date). >> >> Regards, >> Fabien Castan >> >> PS: which version of nuke are you using ? There is many openfx bug fixes in nuke >= 6.0. >> >> ----- Mail Original ----- >> De: "Pau Gargallo" <pau...@gm...> >> À: ope...@li... >> Envoyé: Mardi 21 Décembre 2010 13h01:37 GMT +01:00 Amsterdam / Berlin / Berne / Rome / Stockholm / Vienne >> Objet: [Openfx-developer] Openfx examples in Nuke, problems and fixes >> >> Hi all, >> >> I'm new to OpenFX. I've downloaded the headers and examples, compiled >> them and loaded them into Nuke without problems. The problems arrive >> when I actually use the plug-in examples. They work properly as long >> as you do not move the sliders too much. But as soon as you move a >> slider (any of them) a bit to fast, Nuke crashes. I've tried multiple >> combinations of Nuke and OpenFX versions both on linux and mac, and >> had always the same crash. >> >> The problem seems to be that the render method in the examples does >> not check the validity of the source image. The call >> >> gEffectHost->clipGetImage(myData->sourceClip, time, NULL, &sourceImg); >> >> in line 667 of the basic.cpp example, is sporadically returning >> kOfxStatFailed, but the return value is ignored. >> >> What is the proper way to handle such failure? >> Should the code in the examples be modified to handle this? >> >> Thanks in advance, >> pau >> >> ------------------------------------------------------------------------------ >> Lotusphere 2011 >> Register now for Lotusphere 2011 and learn how >> to connect the dots, take your collaborative environment >> to the next level, and enter the era of Social Business. >> http://p.sf.net/sfu/lotusphere-d2d >> _______________________________________________ >> Openfx-developer mailing list >> Ope...@li... >> https://lists.sourceforge.net/lists/listinfo/openfx-developer >> > |
|
From: Pierre J. <ja...@re...> - 2011-01-04 04:38:07
|
Fabien, Juut a short note: I did a quick look at your source directory. I am noticing that you are using an LPGL license v3 and you as well amend it. It's your project but, it might be more useful if you used an MIT x11/Modified BSD/NCSA permissive license (like OFX itself)? http://www.gnu.org/licenses/license-list.html#Expat (at least for the core library, someone might have a reason for GPL- e.g. for a useful example.cpp or a specialized codec...) Reason is some companies don't accept to use LPGL licensed code as it becomes too complicated with their legal dept. It's also becoming more common these days to just offer dual-licensing -- http://en.wikipedia.org/wiki/Multi-licensing (choose your own) so it doesn't have to become a religion. Pierre On 12/21/2010 2:10 PM, Pau Gargallo wrote: > Fabien, > > thanks for the answer and the pointer to your plug-ins, they will be > very useful. > > After catching the kOfxStatFailed when fetching the images the > examples work fine. I'll send a patch with the changes I've made. > > I'm using the last version of Nuke on a 64bits linux, but I've > compulsively tried a bunch of versions from the Foundry's ftp, to see > if the error happened only some of them. As it turned out, it was not > a version problem but an oversight in the example. > > cheers, > pau > > > > On Tue, Dec 21, 2010 at 5:10 PM,<fab...@fr...> wrote: > >> Hi, >> The default examples are not really great. >> Effectively, nuke often returns an empty image (I don't know why) and you have to return an error. And nuke does the process again, with a good image... >> We have created an open source project for openfx plugins and a library to batch a graph of openfx nodes. The project is still under early development, but you can found a bunch of plugin examples, simple or advanced ones. >> https://github.com/tuttleofx/TuttleOFX/tree/master/plugins/image/process >> There is no documentation, so if you have any question, don't hesitate to ask me. >> To build see INSTALL.scons (the bjam version is no more used and so not up to date). >> >> Regards, >> Fabien Castan >> >> PS: which version of nuke are you using ? There is many openfx bug fixes in nuke>= 6.0. >> >> ----- Mail Original ----- >> De: "Pau Gargallo"<pau...@gm...> >> À:ope...@li... >> Envoyé: Mardi 21 Décembre 2010 13h01:37 GMT +01:00 Amsterdam / Berlin / Berne / Rome / Stockholm / Vienne >> Objet: [Openfx-developer] Openfx examples in Nuke, problems and fixes >> >> Hi all, >> >> I'm new to OpenFX. I've downloaded the headers and examples, compiled >> them and loaded them into Nuke without problems. The problems arrive >> when I actually use the plug-in examples. They work properly as long >> as you do not move the sliders too much. But as soon as you move a >> slider (any of them) a bit to fast, Nuke crashes. I've tried multiple >> combinations of Nuke and OpenFX versions both on linux and mac, and >> had always the same crash. >> >> The problem seems to be that the render method in the examples does >> not check the validity of the source image. The call >> >> gEffectHost->clipGetImage(myData->sourceClip, time, NULL,&sourceImg); >> >> in line 667 of the basic.cpp example, is sporadically returning >> kOfxStatFailed, but the return value is ignored. >> >> What is the proper way to handle such failure? >> Should the code in the examples be modified to handle this? >> >> Thanks in advance, >> pau >> >> ------------------------------------------------------------------------------ >> Lotusphere 2011 >> Register now for Lotusphere 2011 and learn how >> to connect the dots, take your collaborative environment >> to the next level, and enter the era of Social Business. >> http://p.sf.net/sfu/lotusphere-d2d >> _______________________________________________ >> Openfx-developer mailing list >> Ope...@li... >> https://lists.sourceforge.net/lists/listinfo/openfx-developer >> >> > ------------------------------------------------------------------------------ > Forrester recently released a report on the Return on Investment (ROI) of > Google Apps. They found a 300% ROI, 38%-56% cost savings, and break-even > within 7 months. Over 3 million businesses have gone Google with Google Apps: > an online email calendar, and document program that's accessible from your > browser. Read the Forrester report:http://p.sf.net/sfu/googleapps-sfnew > _______________________________________________ > Openfx-developer mailing list > Ope...@li... > https://lists.sourceforge.net/lists/listinfo/openfx-developer > > |