[Csnd-dev] WASM: NaN in output buffer

classic Classic list List threaded Threaded
16 messages Options
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

[Csnd-dev] WASM: NaN in output buffer

hlolli
I'm getting NaN's with csound-wasm, both by running locally wasm/minimal.html and in trough my clojurescript on Nodejs setup

Error message is on both platforms exacly the same
NaN! outputPointer = 6633640
Those NaN seem have their origin in CsoundObj_getOutputBuffer coming from Csound. I think this is platform independent so I'm curious if others are getting these errors too?

Best,
Hlöðver
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: [Csnd-dev] WASM: NaN in output buffer

hlolli
This happens on all examples with the newest wasm build, some regression happened here I guess.

On 4 August 2017 at 13:53, Hlöðver Sigurðsson <[hidden email]> wrote:
I'm getting NaN's with csound-wasm, both by running locally wasm/minimal.html and in trough my clojurescript on Nodejs setup

Error message is on both platforms exacly the same
NaN! outputPointer = 6633640
Those NaN seem have their origin in CsoundObj_getOutputBuffer coming from Csound. I think this is platform independent so I'm curious if others are getting these errors too?

Best,
Hlöðver

Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: [Csnd-dev] WASM: NaN in output buffer

Victor Lazzarini-2
it does not happen here at all. 
Maybe it's the compiler? Emscripten had strange problems with NaNs before.

Victor Lazzarini
Dean of Arts, Celtic Studies, and Philosophy
Maynooth University
Ireland

On 4 Aug 2017, at 16:14, Hlöðver Sigurðsson <[hidden email]> wrote:

This happens on all examples with the newest wasm build, some regression happened here I guess.

On 4 August 2017 at 13:53, Hlöðver Sigurðsson <[hidden email]> wrote:
I'm getting NaN's with csound-wasm, both by running locally wasm/minimal.html and in trough my clojurescript on Nodejs setup

Error message is on both platforms exacly the same
NaN! outputPointer = 6633640
Those NaN seem have their origin in CsoundObj_getOutputBuffer coming from Csound. I think this is platform independent so I'm curious if others are getting these errors too?

Best,
Hlöðver

Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: [Csnd-dev] WASM: NaN in output buffer

hlolli
I try now updating emscripten from 1.37.16 to 1.37.18 (will take a while I guess).

On 4 August 2017 at 17:21, Victor Lazzarini <[hidden email]> wrote:
it does not happen here at all. 
Maybe it's the compiler? Emscripten had strange problems with NaNs before.

Victor Lazzarini
Dean of Arts, Celtic Studies, and Philosophy
Maynooth University
Ireland

On 4 Aug 2017, at 16:14, Hlöðver Sigurðsson <[hidden email]> wrote:

This happens on all examples with the newest wasm build, some regression happened here I guess.

On 4 August 2017 at 13:53, Hlöðver Sigurðsson <[hidden email]> wrote:
I'm getting NaN's with csound-wasm, both by running locally wasm/minimal.html and in trough my clojurescript on Nodejs setup

Error message is on both platforms exacly the same
NaN! outputPointer = 6633640
Those NaN seem have their origin in CsoundObj_getOutputBuffer coming from Csound. I think this is platform independent so I'm curious if others are getting these errors too?

Best,
Hlöðver


Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: [Csnd-dev] WASM: NaN in output buffer

Victor Lazzarini-2
or it could be the browser. I am running chrome 59 here. Did you try firefox?
========================
Prof. Victor Lazzarini
Dean of Arts, Celtic Studies, and Philosophy,
Maynooth University,
Maynooth, Co Kildare, Ireland
Tel: 00 353 7086936
Fax: 00 353 1 7086952

> On 4 Aug 2017, at 16:27, Hlöðver Sigurðsson <[hidden email]> wrote:
>
> I try now updating emscripten from 1.37.16 to 1.37.18 (will take a while I guess).
>
> On 4 August 2017 at 17:21, Victor Lazzarini <[hidden email]> wrote:
> it does not happen here at all.
> Maybe it's the compiler? Emscripten had strange problems with NaNs before.
>
> Victor Lazzarini
> Dean of Arts, Celtic Studies, and Philosophy
> Maynooth University
> Ireland
>
> On 4 Aug 2017, at 16:14, Hlöðver Sigurðsson <[hidden email]> wrote:
>
>> This happens on all examples with the newest wasm build, some regression happened here I guess.
>>
>> On 4 August 2017 at 13:53, Hlöðver Sigurðsson <[hidden email]> wrote:
>> I'm getting NaN's with csound-wasm, both by running locally wasm/minimal.html and in trough my clojurescript on Nodejs setup
>>
>> Error message is on both platforms exacly the same
>> NaN! outputPointer = 6633640
>> Those NaN seem have their origin in CsoundObj_getOutputBuffer coming from Csound. I think this is platform independent so I'm curious if others are getting these errors too?
>>
>> Best,
>> Hlöðver
>>
>

Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: [Csnd-dev] WASM: NaN in output buffer

hlolli
Tried Node/Firefox too, I tried the wasm examples hosted online on github.csound.io, and I served it locally, and it works in the csound online examples but not locally with with builds (on same browser).

On 4 August 2017 at 17:29, Victor Lazzarini <[hidden email]> wrote:
or it could be the browser. I am running chrome 59 here. Did you try firefox?
========================
Prof. Victor Lazzarini
Dean of Arts, Celtic Studies, and Philosophy,
Maynooth University,
Maynooth, Co Kildare, Ireland
Tel: 00 353 7086936
Fax: 00 353 1 7086952

> On 4 Aug 2017, at 16:27, Hlöðver Sigurðsson <[hidden email]> wrote:
>
> I try now updating emscripten from 1.37.16 to 1.37.18 (will take a while I guess).
>
> On 4 August 2017 at 17:21, Victor Lazzarini <[hidden email]> wrote:
> it does not happen here at all.
> Maybe it's the compiler? Emscripten had strange problems with NaNs before.
>
> Victor Lazzarini
> Dean of Arts, Celtic Studies, and Philosophy
> Maynooth University
> Ireland
>
> On 4 Aug 2017, at 16:14, Hlöðver Sigurðsson <[hidden email]> wrote:
>
>> This happens on all examples with the newest wasm build, some regression happened here I guess.
>>
>> On 4 August 2017 at 13:53, Hlöðver Sigurðsson <[hidden email]> wrote:
>> I'm getting NaN's with csound-wasm, both by running locally wasm/minimal.html and in trough my clojurescript on Nodejs setup
>>
>> Error message is on both platforms exacly the same
>> NaN! outputPointer = 6633640
>> Those NaN seem have their origin in CsoundObj_getOutputBuffer coming from Csound. I think this is platform independent so I'm curious if others are getting these errors too?
>>
>> Best,
>> Hlöðver
>>
>


Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: [Csnd-dev] WASM: NaN in output buffer

Victor Lazzarini-2
It could be the compiler then.
========================
Prof. Victor Lazzarini
Dean of Arts, Celtic Studies, and Philosophy,
Maynooth University,
Maynooth, Co Kildare, Ireland
Tel: 00 353 7086936
Fax: 00 353 1 7086952

> On 4 Aug 2017, at 16:40, Hlöðver Sigurðsson <[hidden email]> wrote:
>
> Tried Node/Firefox too, I tried the wasm examples hosted online on github.csound.io, and I served it locally, and it works in the csound online examples but not locally with with builds (on same browser).
>
> On 4 August 2017 at 17:29, Victor Lazzarini <[hidden email]> wrote:
> or it could be the browser. I am running chrome 59 here. Did you try firefox?
> ========================
> Prof. Victor Lazzarini
> Dean of Arts, Celtic Studies, and Philosophy,
> Maynooth University,
> Maynooth, Co Kildare, Ireland
> Tel: 00 353 7086936
> Fax: 00 353 1 7086952
>
> > On 4 Aug 2017, at 16:27, Hlöðver Sigurðsson <[hidden email]> wrote:
> >
> > I try now updating emscripten from 1.37.16 to 1.37.18 (will take a while I guess).
> >
> > On 4 August 2017 at 17:21, Victor Lazzarini <[hidden email]> wrote:
> > it does not happen here at all.
> > Maybe it's the compiler? Emscripten had strange problems with NaNs before.
> >
> > Victor Lazzarini
> > Dean of Arts, Celtic Studies, and Philosophy
> > Maynooth University
> > Ireland
> >
> > On 4 Aug 2017, at 16:14, Hlöðver Sigurðsson <[hidden email]> wrote:
> >
> >> This happens on all examples with the newest wasm build, some regression happened here I guess.
> >>
> >> On 4 August 2017 at 13:53, Hlöðver Sigurðsson <[hidden email]> wrote:
> >> I'm getting NaN's with csound-wasm, both by running locally wasm/minimal.html and in trough my clojurescript on Nodejs setup
> >>
> >> Error message is on both platforms exacly the same
> >> NaN! outputPointer = 6633640
> >> Those NaN seem have their origin in CsoundObj_getOutputBuffer coming from Csound. I think this is platform independent so I'm curious if others are getting these errors too?
> >>
> >> Best,
> >> Hlöðver
> >>
> >
>
>

Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: [Csnd-dev] WASM: NaN in output buffer

Victor Lazzarini-2
I have this one here (older?)

victor$ emcc --version
emcc (Emscripten gcc/clang-like replacement) 1.37.9 ()


========================
Prof. Victor Lazzarini
Dean of Arts, Celtic Studies, and Philosophy,
Maynooth University,
Maynooth, Co Kildare, Ireland
Tel: 00 353 7086936
Fax: 00 353 1 7086952

> On 4 Aug 2017, at 16:45, Victor Lazzarini <[hidden email]> wrote:
>
> [This sender failed our fraud detection checks and may not be who they appear to be. Learn about spoofing at http://aka.ms/LearnAboutSpoofing]
>
> It could be the compiler then.
> ========================
> Prof. Victor Lazzarini
> Dean of Arts, Celtic Studies, and Philosophy,
> Maynooth University,
> Maynooth, Co Kildare, Ireland
> Tel: 00 353 7086936
> Fax: 00 353 1 7086952
>
>> On 4 Aug 2017, at 16:40, Hlöðver Sigurðsson <[hidden email]> wrote:
>>
>> Tried Node/Firefox too, I tried the wasm examples hosted online on github.csound.io, and I served it locally, and it works in the csound online examples but not locally with with builds (on same browser).
>>
>> On 4 August 2017 at 17:29, Victor Lazzarini <[hidden email]> wrote:
>> or it could be the browser. I am running chrome 59 here. Did you try firefox?
>> ========================
>> Prof. Victor Lazzarini
>> Dean of Arts, Celtic Studies, and Philosophy,
>> Maynooth University,
>> Maynooth, Co Kildare, Ireland
>> Tel: 00 353 7086936
>> Fax: 00 353 1 7086952
>>
>>> On 4 Aug 2017, at 16:27, Hlöðver Sigurðsson <[hidden email]> wrote:
>>>
>>> I try now updating emscripten from 1.37.16 to 1.37.18 (will take a while I guess).
>>>
>>> On 4 August 2017 at 17:21, Victor Lazzarini <[hidden email]> wrote:
>>> it does not happen here at all.
>>> Maybe it's the compiler? Emscripten had strange problems with NaNs before.
>>>
>>> Victor Lazzarini
>>> Dean of Arts, Celtic Studies, and Philosophy
>>> Maynooth University
>>> Ireland
>>>
>>> On 4 Aug 2017, at 16:14, Hlöðver Sigurðsson <[hidden email]> wrote:
>>>
>>>> This happens on all examples with the newest wasm build, some regression happened here I guess.
>>>>
>>>> On 4 August 2017 at 13:53, Hlöðver Sigurðsson <[hidden email]> wrote:
>>>> I'm getting NaN's with csound-wasm, both by running locally wasm/minimal.html and in trough my clojurescript on Nodejs setup
>>>>
>>>> Error message is on both platforms exacly the same
>>>> NaN! outputPointer = 6633640
>>>> Those NaN seem have their origin in CsoundObj_getOutputBuffer coming from Csound. I think this is platform independent so I'm curious if others are getting these errors too?
>>>>
>>>> Best,
>>>> Hlöðver
>>>>
>>>
>>
>>
>

Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: [Csnd-dev] WASM: NaN in output buffer

hlolli
ok, https://github.com/kripken/emscripten/releases you're on an older version. Well, guess it would make sense you tried out too to update :) ?

On 4 August 2017 at 17:46, Victor Lazzarini <[hidden email]> wrote:
I have this one here (older?)

victor$ emcc --version
emcc (Emscripten gcc/clang-like replacement) 1.37.9 ()


========================
Prof. Victor Lazzarini
Dean of Arts, Celtic Studies, and Philosophy,
Maynooth University,
Maynooth, Co Kildare, Ireland
Tel: 00 353 7086936
Fax: 00 353 1 7086952

> On 4 Aug 2017, at 16:45, Victor Lazzarini <[hidden email]> wrote:
>
> [This sender failed our fraud detection checks and may not be who they appear to be. Learn about spoofing at http://aka.ms/LearnAboutSpoofing]
>
> It could be the compiler then.
> ========================
> Prof. Victor Lazzarini
> Dean of Arts, Celtic Studies, and Philosophy,
> Maynooth University,
> Maynooth, Co Kildare, Ireland
> Tel: 00 353 7086936
> Fax: 00 353 1 7086952
>
>> On 4 Aug 2017, at 16:40, Hlöðver Sigurðsson <[hidden email]> wrote:
>>
>> Tried Node/Firefox too, I tried the wasm examples hosted online on github.csound.io, and I served it locally, and it works in the csound online examples but not locally with with builds (on same browser).
>>
>> On 4 August 2017 at 17:29, Victor Lazzarini <[hidden email]> wrote:
>> or it could be the browser. I am running chrome 59 here. Did you try firefox?
>> ========================
>> Prof. Victor Lazzarini
>> Dean of Arts, Celtic Studies, and Philosophy,
>> Maynooth University,
>> Maynooth, Co Kildare, Ireland
>> Tel: 00 353 7086936
>> Fax: 00 353 1 7086952
>>
>>> On 4 Aug 2017, at 16:27, Hlöðver Sigurðsson <[hidden email]> wrote:
>>>
>>> I try now updating emscripten from 1.37.16 to 1.37.18 (will take a while I guess).
>>>
>>> On 4 August 2017 at 17:21, Victor Lazzarini <[hidden email]> wrote:
>>> it does not happen here at all.
>>> Maybe it's the compiler? Emscripten had strange problems with NaNs before.
>>>
>>> Victor Lazzarini
>>> Dean of Arts, Celtic Studies, and Philosophy
>>> Maynooth University
>>> Ireland
>>>
>>> On 4 Aug 2017, at 16:14, Hlöðver Sigurðsson <[hidden email]> wrote:
>>>
>>>> This happens on all examples with the newest wasm build, some regression happened here I guess.
>>>>
>>>> On 4 August 2017 at 13:53, Hlöðver Sigurðsson <[hidden email]> wrote:
>>>> I'm getting NaN's with csound-wasm, both by running locally wasm/minimal.html and in trough my clojurescript on Nodejs setup
>>>>
>>>> Error message is on both platforms exacly the same
>>>> NaN! outputPointer = 6633640
>>>> Those NaN seem have their origin in CsoundObj_getOutputBuffer coming from Csound. I think this is platform independent so I'm curious if others are getting these errors too?
>>>>
>>>> Best,
>>>> Hlöðver
>>>>
>>>
>>
>>
>


Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: [Csnd-dev] WASM: NaN in output buffer

Victor Lazzarini-2
Not if it’s going to create bad builds… it’s working now, so I am not too keen on breaking it.
========================
Prof. Victor Lazzarini
Dean of Arts, Celtic Studies, and Philosophy,
Maynooth University,
Maynooth, Co Kildare, Ireland
Tel: 00 353 7086936
Fax: 00 353 1 7086952

> On 4 Aug 2017, at 16:50, Hlöðver Sigurðsson <[hidden email]> wrote:
>
> ok, https://github.com/kripken/emscripten/releases you're on an older version. Well, guess it would make sense you tried out too to update :) ?
>
> On 4 August 2017 at 17:46, Victor Lazzarini <[hidden email]> wrote:
> I have this one here (older?)
>
> victor$ emcc --version
> emcc (Emscripten gcc/clang-like replacement) 1.37.9 ()
>
>
> ========================
> Prof. Victor Lazzarini
> Dean of Arts, Celtic Studies, and Philosophy,
> Maynooth University,
> Maynooth, Co Kildare, Ireland
> Tel: 00 353 7086936
> Fax: 00 353 1 7086952
>
> > On 4 Aug 2017, at 16:45, Victor Lazzarini <[hidden email]> wrote:
> >
> > [This sender failed our fraud detection checks and may not be who they appear to be. Learn about spoofing at http://aka.ms/LearnAboutSpoofing]
> >
> > It could be the compiler then.
> > ========================
> > Prof. Victor Lazzarini
> > Dean of Arts, Celtic Studies, and Philosophy,
> > Maynooth University,
> > Maynooth, Co Kildare, Ireland
> > Tel: 00 353 7086936
> > Fax: 00 353 1 7086952
> >
> >> On 4 Aug 2017, at 16:40, Hlöðver Sigurðsson <[hidden email]> wrote:
> >>
> >> Tried Node/Firefox too, I tried the wasm examples hosted online on github.csound.io, and I served it locally, and it works in the csound online examples but not locally with with builds (on same browser).
> >>
> >> On 4 August 2017 at 17:29, Victor Lazzarini <[hidden email]> wrote:
> >> or it could be the browser. I am running chrome 59 here. Did you try firefox?
> >> ========================
> >> Prof. Victor Lazzarini
> >> Dean of Arts, Celtic Studies, and Philosophy,
> >> Maynooth University,
> >> Maynooth, Co Kildare, Ireland
> >> Tel: 00 353 7086936
> >> Fax: 00 353 1 7086952
> >>
> >>> On 4 Aug 2017, at 16:27, Hlöðver Sigurðsson <[hidden email]> wrote:
> >>>
> >>> I try now updating emscripten from 1.37.16 to 1.37.18 (will take a while I guess).
> >>>
> >>> On 4 August 2017 at 17:21, Victor Lazzarini <[hidden email]> wrote:
> >>> it does not happen here at all.
> >>> Maybe it's the compiler? Emscripten had strange problems with NaNs before.
> >>>
> >>> Victor Lazzarini
> >>> Dean of Arts, Celtic Studies, and Philosophy
> >>> Maynooth University
> >>> Ireland
> >>>
> >>> On 4 Aug 2017, at 16:14, Hlöðver Sigurðsson <[hidden email]> wrote:
> >>>
> >>>> This happens on all examples with the newest wasm build, some regression happened here I guess.
> >>>>
> >>>> On 4 August 2017 at 13:53, Hlöðver Sigurðsson <[hidden email]> wrote:
> >>>> I'm getting NaN's with csound-wasm, both by running locally wasm/minimal.html and in trough my clojurescript on Nodejs setup
> >>>>
> >>>> Error message is on both platforms exacly the same
> >>>> NaN! outputPointer = 6633640
> >>>> Those NaN seem have their origin in CsoundObj_getOutputBuffer coming from Csound. I think this is platform independent so I'm curious if others are getting these errors too?
> >>>>
> >>>> Best,
> >>>> Hlöðver
> >>>>
> >>>
> >>
> >>
> >
>
>

Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: [Csnd-dev] WASM: NaN in output buffer

Victor Lazzarini-2
Actually the version I have is the latest released binary. You might want to use it instead of tagged git releases.
It’s probably safer.
========================
Prof. Victor Lazzarini
Dean of Arts, Celtic Studies, and Philosophy,
Maynooth University,
Maynooth, Co Kildare, Ireland
Tel: 00 353 7086936
Fax: 00 353 1 7086952

> On 4 Aug 2017, at 16:55, Victor Lazzarini <[hidden email]> wrote:
>
> [This sender failed our fraud detection checks and may not be who they appear to be. Learn about spoofing at http://aka.ms/LearnAboutSpoofing]
>
> Not if it’s going to create bad builds… it’s working now, so I am not too keen on breaking it.
> ========================
> Prof. Victor Lazzarini
> Dean of Arts, Celtic Studies, and Philosophy,
> Maynooth University,
> Maynooth, Co Kildare, Ireland
> Tel: 00 353 7086936
> Fax: 00 353 1 7086952
>
>> On 4 Aug 2017, at 16:50, Hlöðver Sigurðsson <[hidden email]> wrote:
>>
>> ok, https://github.com/kripken/emscripten/releases you're on an older version. Well, guess it would make sense you tried out too to update :) ?
>>
>> On 4 August 2017 at 17:46, Victor Lazzarini <[hidden email]> wrote:
>> I have this one here (older?)
>>
>> victor$ emcc --version
>> emcc (Emscripten gcc/clang-like replacement) 1.37.9 ()
>>
>>
>> ========================
>> Prof. Victor Lazzarini
>> Dean of Arts, Celtic Studies, and Philosophy,
>> Maynooth University,
>> Maynooth, Co Kildare, Ireland
>> Tel: 00 353 7086936
>> Fax: 00 353 1 7086952
>>
>>> On 4 Aug 2017, at 16:45, Victor Lazzarini <[hidden email]> wrote:
>>>
>>> [This sender failed our fraud detection checks and may not be who they appear to be. Learn about spoofing at http://aka.ms/LearnAboutSpoofing]
>>>
>>> It could be the compiler then.
>>> ========================
>>> Prof. Victor Lazzarini
>>> Dean of Arts, Celtic Studies, and Philosophy,
>>> Maynooth University,
>>> Maynooth, Co Kildare, Ireland
>>> Tel: 00 353 7086936
>>> Fax: 00 353 1 7086952
>>>
>>>> On 4 Aug 2017, at 16:40, Hlöðver Sigurðsson <[hidden email]> wrote:
>>>>
>>>> Tried Node/Firefox too, I tried the wasm examples hosted online on github.csound.io, and I served it locally, and it works in the csound online examples but not locally with with builds (on same browser).
>>>>
>>>> On 4 August 2017 at 17:29, Victor Lazzarini <[hidden email]> wrote:
>>>> or it could be the browser. I am running chrome 59 here. Did you try firefox?
>>>> ========================
>>>> Prof. Victor Lazzarini
>>>> Dean of Arts, Celtic Studies, and Philosophy,
>>>> Maynooth University,
>>>> Maynooth, Co Kildare, Ireland
>>>> Tel: 00 353 7086936
>>>> Fax: 00 353 1 7086952
>>>>
>>>>> On 4 Aug 2017, at 16:27, Hlöðver Sigurðsson <[hidden email]> wrote:
>>>>>
>>>>> I try now updating emscripten from 1.37.16 to 1.37.18 (will take a while I guess).
>>>>>
>>>>> On 4 August 2017 at 17:21, Victor Lazzarini <[hidden email]> wrote:
>>>>> it does not happen here at all.
>>>>> Maybe it's the compiler? Emscripten had strange problems with NaNs before.
>>>>>
>>>>> Victor Lazzarini
>>>>> Dean of Arts, Celtic Studies, and Philosophy
>>>>> Maynooth University
>>>>> Ireland
>>>>>
>>>>> On 4 Aug 2017, at 16:14, Hlöðver Sigurðsson <[hidden email]> wrote:
>>>>>
>>>>>> This happens on all examples with the newest wasm build, some regression happened here I guess.
>>>>>>
>>>>>> On 4 August 2017 at 13:53, Hlöðver Sigurðsson <[hidden email]> wrote:
>>>>>> I'm getting NaN's with csound-wasm, both by running locally wasm/minimal.html and in trough my clojurescript on Nodejs setup
>>>>>>
>>>>>> Error message is on both platforms exacly the same
>>>>>> NaN! outputPointer = 6633640
>>>>>> Those NaN seem have their origin in CsoundObj_getOutputBuffer coming from Csound. I think this is platform independent so I'm curious if others are getting these errors too?
>>>>>>
>>>>>> Best,
>>>>>> Hlöðver
>>>>>>
>>>>>
>>>>
>>>>
>>>
>>
>>
>

Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: [Csnd-dev] WASM: NaN in output buffer

hlolli
Ok good point, 30% left of this build, roll back to 1.37.9 if the latest is also causing problems.

On 4 August 2017 at 18:01, Victor Lazzarini <[hidden email]> wrote:
Actually the version I have is the latest released binary. You might want to use it instead of tagged git releases.
It’s probably safer.
========================
Prof. Victor Lazzarini
Dean of Arts, Celtic Studies, and Philosophy,
Maynooth University,
Maynooth, Co Kildare, Ireland
Tel: 00 353 7086936
Fax: 00 353 1 7086952

> On 4 Aug 2017, at 16:55, Victor Lazzarini <[hidden email]> wrote:
>
> [This sender failed our fraud detection checks and may not be who they appear to be. Learn about spoofing at http://aka.ms/LearnAboutSpoofing]
>
> Not if it’s going to create bad builds… it’s working now, so I am not too keen on breaking it.
> ========================
> Prof. Victor Lazzarini
> Dean of Arts, Celtic Studies, and Philosophy,
> Maynooth University,
> Maynooth, Co Kildare, Ireland
> Tel: 00 353 7086936
> Fax: 00 353 1 7086952
>
>> On 4 Aug 2017, at 16:50, Hlöðver Sigurðsson <[hidden email]> wrote:
>>
>> ok, https://github.com/kripken/emscripten/releases you're on an older version. Well, guess it would make sense you tried out too to update :) ?
>>
>> On 4 August 2017 at 17:46, Victor Lazzarini <[hidden email]> wrote:
>> I have this one here (older?)
>>
>> victor$ emcc --version
>> emcc (Emscripten gcc/clang-like replacement) 1.37.9 ()
>>
>>
>> ========================
>> Prof. Victor Lazzarini
>> Dean of Arts, Celtic Studies, and Philosophy,
>> Maynooth University,
>> Maynooth, Co Kildare, Ireland
>> Tel: 00 353 7086936
>> Fax: 00 353 1 7086952
>>
>>> On 4 Aug 2017, at 16:45, Victor Lazzarini <[hidden email]> wrote:
>>>
>>> [This sender failed our fraud detection checks and may not be who they appear to be. Learn about spoofing at http://aka.ms/LearnAboutSpoofing]
>>>
>>> It could be the compiler then.
>>> ========================
>>> Prof. Victor Lazzarini
>>> Dean of Arts, Celtic Studies, and Philosophy,
>>> Maynooth University,
>>> Maynooth, Co Kildare, Ireland
>>> Tel: 00 353 7086936
>>> Fax: 00 353 1 7086952
>>>
>>>> On 4 Aug 2017, at 16:40, Hlöðver Sigurðsson <[hidden email]> wrote:
>>>>
>>>> Tried Node/Firefox too, I tried the wasm examples hosted online on github.csound.io, and I served it locally, and it works in the csound online examples but not locally with with builds (on same browser).
>>>>
>>>> On 4 August 2017 at 17:29, Victor Lazzarini <[hidden email]> wrote:
>>>> or it could be the browser. I am running chrome 59 here. Did you try firefox?
>>>> ========================
>>>> Prof. Victor Lazzarini
>>>> Dean of Arts, Celtic Studies, and Philosophy,
>>>> Maynooth University,
>>>> Maynooth, Co Kildare, Ireland
>>>> Tel: 00 353 7086936
>>>> Fax: 00 353 1 7086952
>>>>
>>>>> On 4 Aug 2017, at 16:27, Hlöðver Sigurðsson <[hidden email]> wrote:
>>>>>
>>>>> I try now updating emscripten from 1.37.16 to 1.37.18 (will take a while I guess).
>>>>>
>>>>> On 4 August 2017 at 17:21, Victor Lazzarini <[hidden email]> wrote:
>>>>> it does not happen here at all.
>>>>> Maybe it's the compiler? Emscripten had strange problems with NaNs before.
>>>>>
>>>>> Victor Lazzarini
>>>>> Dean of Arts, Celtic Studies, and Philosophy
>>>>> Maynooth University
>>>>> Ireland
>>>>>
>>>>> On 4 Aug 2017, at 16:14, Hlöðver Sigurðsson <[hidden email]> wrote:
>>>>>
>>>>>> This happens on all examples with the newest wasm build, some regression happened here I guess.
>>>>>>
>>>>>> On 4 August 2017 at 13:53, Hlöðver Sigurðsson <[hidden email]> wrote:
>>>>>> I'm getting NaN's with csound-wasm, both by running locally wasm/minimal.html and in trough my clojurescript on Nodejs setup
>>>>>>
>>>>>> Error message is on both platforms exacly the same
>>>>>> NaN! outputPointer = 6633640
>>>>>> Those NaN seem have their origin in CsoundObj_getOutputBuffer coming from Csound. I think this is platform independent so I'm curious if others are getting these errors too?
>>>>>>
>>>>>> Best,
>>>>>> Hlöðver
>>>>>>
>>>>>
>>>>
>>>>
>>>
>>
>>
>


Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: [Csnd-dev] WASM: NaN in output buffer

Francesco Porta
In reply to this post by hlolli
Hi,
just to say that in ububtu 16.04 it seems to work with emcc 1.37.13.
I tried examples locally.
But i'm not an expert :)

ciao,
francesco
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: [Csnd-dev] WASM: NaN in output buffer

Steven Yi
I also compiled and ran here with 1.37.13.

On Mon, Aug 7, 2017 at 3:54 AM, Francesco Porta
<[hidden email]> wrote:
> Hi,
> just to say that in ububtu 16.04 it seems to work with emcc 1.37.13.
> I tried examples locally.
> But i'm not an expert :)
>
> ciao,
> francesco
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: [Csnd-dev] WASM: NaN in output buffer

hlolli
So emcc 1.37.18 works! I had painful time getting the emscripten to run again after trying updateing, good to keep in ming when using emsdk that "incoming" and "master" should not be mixed as it causes strage bugs. I removed all in the end and used incoming and will I think stick to the incoming versions.

Another possible conclusion here is that I installed 1.37.16 incorrectly, not sure how I could have done it incorrectly, but it's safer to avoid that version I guess.

On 7 August 2017 at 17:23, Steven Yi <[hidden email]> wrote:
I also compiled and ran here with 1.37.13.

On Mon, Aug 7, 2017 at 3:54 AM, Francesco Porta
<[hidden email]> wrote:
> Hi,
> just to say that in ububtu 16.04 it seems to work with emcc 1.37.13.
> I tried examples locally.
> But i'm not an expert :)
>
> ciao,
> francesco

Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: [Csnd-dev] WASM: NaN in output buffer

Steven Yi
I had issues using master in the past. I'd recommend just using "latest" which I think Victor and I and others are using for general compilation. 

On Tue, Aug 8, 2017, 7:50 AM Hlöðver Sigurðsson <[hidden email]> wrote:
So emcc 1.37.18 works! I had painful time getting the emscripten to run again after trying updateing, good to keep in ming when using emsdk that "incoming" and "master" should not be mixed as it causes strage bugs. I removed all in the end and used incoming and will I think stick to the incoming versions.

Another possible conclusion here is that I installed 1.37.16 incorrectly, not sure how I could have done it incorrectly, but it's safer to avoid that version I guess.

On 7 August 2017 at 17:23, Steven Yi <[hidden email]> wrote:
I also compiled and ran here with 1.37.13.

On Mon, Aug 7, 2017 at 3:54 AM, Francesco Porta
<[hidden email]> wrote:
> Hi,
> just to say that in ububtu 16.04 it seems to work with emcc 1.37.13.
> I tried examples locally.
> But i'm not an expert :)
>
> ciao,
> francesco

Loading...