Simulink for Octave

classic Classic list List threaded Threaded
23 messages Options
12
Reply | Threaded
Open this post in threaded view
|

Simulink for Octave

xavion

Any plans for a Simulink-like capability?  I think this would be hugely beneficial to the simulation community and entice more users to the Octave user base.  It would save people like myself thousands of dollars in Matlab/Simulink licenses.
Reply | Threaded
Open this post in threaded view
|

Re: Simulink for Octave

Sergei Steshenko


--- On Mon, 8/15/11, xavion <[hidden email]> wrote:

> From: xavion <[hidden email]>
> Subject: Simulink for Octave
> To: [hidden email]
> Date: Monday, August 15, 2011, 11:54 PM
>
> Any plans for a Simulink-like capability?  I think
> this would be hugely
> beneficial to the simulation community and entice more
> users to the Octave
> user base.  It would save people like myself thousands
> of dollars in
> Matlab/Simulink licenses.
>
> --
> View this message in context: http://octave.1599824.n4.nabble.com/Simulink-for-Octave-tp3746480p3746480.html
> Sent from the Octave - General mailing list archive at
> Nabble.com.
> _______________________________________________
> Help-octave mailing list
> [hidden email]
> https://mailman.cae.wisc.edu/listinfo/help-octave
>

Doesn't 'scilab' have similar to 'simulink' capabilities ?

Regards,
  Sergei.
_______________________________________________
Help-octave mailing list
[hidden email]
https://mailman.cae.wisc.edu/listinfo/help-octave
Reply | Threaded
Open this post in threaded view
|

Re: Simulink for Octave

c.-2
In reply to this post by xavion

On 16 Aug 2011, at 08:54, xavion wrote:

> Any plans for a Simulink-like capability?  I think this would be hugely
> beneficial to the simulation community and entice more users to the Octave
> user base.  It would save people like myself thousands of dollars in
> Matlab/Simulink licenses.

Would you be willing to invest part of those K$ to fund development of a
simulink-like module for Octave?

c.
_______________________________________________
Help-octave mailing list
[hidden email]
https://mailman.cae.wisc.edu/listinfo/help-octave
Reply | Threaded
Open this post in threaded view
|

Re: Simulink for Octave

Sergei Steshenko


--- On Tue, 8/16/11, c. <[hidden email]> wrote:

> From: c. <[hidden email]>
> Subject: Re: Simulink for Octave
> To: "xavion" <[hidden email]>
> Cc: [hidden email]
> Date: Tuesday, August 16, 2011, 2:36 AM
>
> On 16 Aug 2011, at 08:54, xavion wrote:
>
> > Any plans for a Simulink-like capability?  I
> think this would be hugely
> > beneficial to the simulation community and entice more
> users to the Octave
> > user base.  It would save people like myself
> thousands of dollars in
> > Matlab/Simulink licenses.
>
> Would you be willing to invest part of those K$ to fund
> development of a
> simulink-like module for Octave?
>
> c.
> _______________________________________________
> Help-octave mailing list
> [hidden email]
> https://mailman.cae.wisc.edu/listinfo/help-octave
>

If I'm not mistaken, http://www.scilab.org/products/xcos is an approximate
equivalent of 'simulink'.

So, maybe it has to be _ported_ to Octave - rather than developed from
scratch.

Regards,
  Sergei.

P.S. IMO 'scilab' has more features (I mean, with tools other than core
'scilab'), but inferior interpreter. It's a pity there is no FOSS project
combining the best of the two worlds.

'scilab' license is now compatible with GPL.
_______________________________________________
Help-octave mailing list
[hidden email]
https://mailman.cae.wisc.edu/listinfo/help-octave
Reply | Threaded
Open this post in threaded view
|

Re: Simulink for Octave

c.-2

On 16 Aug 2011, at 13:04, Sergei Steshenko wrote:

> If I'm not mistaken, http://www.scilab.org/products/xcos is an approximate
> equivalent of 'simulink'.
>
> So, maybe it has to be _ported_ to Octave - rather than developed from
> scratch.
>
> Regards,
>  Sergei.
>
> P.S. IMO 'scilab' has more features (I mean, with tools other than core
> 'scilab'), but inferior interpreter. It's a pity there is no FOSS project
> combining the best of the two worlds.
>
> 'scilab' license is now compatible with GPL.

I wouldn't say the scilab interpreter is "inferior", but rather that it is less compatible with Matlab syntax as compared to Octave.
c.
_______________________________________________
Help-octave mailing list
[hidden email]
https://mailman.cae.wisc.edu/listinfo/help-octave
Reply | Threaded
Open this post in threaded view
|

Re: Simulink for Octave

Richard Crozier
In reply to this post by xavion
Under the hood, simulink uses the Matlab ode solver functions such as ode45, ode23, ode15s etc. All simulink really does is provide a pretty interface and some handy built-in blocks (such as SimPowerSystems).

Octave has equivalents to these ode functions (in an octave-forge package at least) so you can get the same functionality without being locked into simulink if you are willing to invest the time in learning to use these. They are significantly slower implementations in Octave though.


Reply | Threaded
Open this post in threaded view
|

Re: Simulink for Octave

Sergei Steshenko
In reply to this post by c.-2


--- On Tue, 8/16/11, c. <[hidden email]> wrote:

> From: c. <[hidden email]>
> Subject: Re: Simulink for Octave
> To: "Sergei Steshenko" <[hidden email]>
> Cc: "xavion" <[hidden email]>, [hidden email]
> Date: Tuesday, August 16, 2011, 5:14 AM

[snip]

> I wouldn't say the scilab interpreter is "inferior", but
> rather that it is less compatible with Matlab syntax as
> compared to Octave.
> c.

'scilab' interpreter truncates identifiers to 24 characters - already
an inferiority from my point of view.

Regards,
  Sergei.

_______________________________________________
Help-octave mailing list
[hidden email]
https://mailman.cae.wisc.edu/listinfo/help-octave
Reply | Threaded
Open this post in threaded view
|

Re: Simulink for Octave

Olaf Till
In reply to this post by Richard Crozier
On Tue, Aug 16, 2011 at 05:24:33AM -0700, Richard Crozier wrote:
> Under the hood, simulink uses the Matlab ode solver functions such as ode45,
> ode23, ode15s etc. All simulink really does is provide a pretty interface
> and some handy built-in blocks (such as SimPowerSystems).
>
> Octave has equivalents to these ode functions (in an octave-forge package at
> least) so you can get the same functionality without being locked into
> simulink if you are willing to invest the time in learning to use these.

Agree. And you get (probably, I don't use simulink) more flexibility
for organizing your code for larger or evolving problems.

> They are significantly slower implementations in Octave though.

Don't know this what regards ode45 and so on, but in most cases
you should use Octaves 'lsode' anyway. It should be rather fast and
accurate.

>
>
>
>
> --
> View this message in context: http://octave.1599824.n4.nabble.com/Simulink-for-Octave-tp3746480p3747054.html
> Sent from the Octave - General mailing list archive at Nabble.com.
> _______________________________________________
> Help-octave mailing list
> [hidden email]
> https://mailman.cae.wisc.edu/listinfo/help-octave
_______________________________________________
Help-octave mailing list
[hidden email]
https://mailman.cae.wisc.edu/listinfo/help-octave
Reply | Threaded
Open this post in threaded view
|

Re: Simulink for Octave

xavion
In reply to this post by c.-2

Can't afford to invest in anything right now.  I will say that if the Octave organization distributed a Simulink-like capability with Octave for free, then offered other "toolboxes" for $50 to $100, I would purchase those toolboxes.  For example, an Excel-link would be beneficial - Excel is the ultimate GUI.  Or perhaps a C autocode toolbox - hugely beneficial.

I believe than an open source & free Octave/Simulink with affordable "toolboxes" would take a huge bite out of Mathwork's bottom line.  Universities and companies could reduce (but not eliminate) their Mathworks overhead costs and reallocate those savings to more important ventures like research & development.


Reply | Threaded
Open this post in threaded view
|

Re: Simulink for Octave

Sergei Steshenko




----- Original Message -----
> From: xavion <[hidden email]>
> To: [hidden email]
> Cc:
> Sent: Friday, August 19, 2011 10:04 PM
> Subject: Re: Simulink for Octave
>
>
[snip]
> Excel is the ultimate GUI.
[snip]

Huh ?

Regards,
  Sergei.

_______________________________________________
Help-octave mailing list
[hidden email]
https://mailman.cae.wisc.edu/listinfo/help-octave
Reply | Threaded
Open this post in threaded view
|

Re: Simulink for Octave

c.-2
In reply to this post by xavion

On 19 Aug 2011, at 21:04, xavion wrote:

> Can't afford to invest in anything right now.  I will say that if the Octave
> organization
There is no such thing as an "Octave organization",
it has been proposed to create one but it seems to be technically difficult.

Octave is completely developed and maintained by individual volunteers.

Some of them accept to do custom developement if someone is willing to hire
them for a specific task and pay for their time.

> distributed a Simulink-like capability with Octave for free,
> then offered other "toolboxes" for $50 to $100, I would purchase those
> toolboxes.  
Unfortunately, as far as I understand, it would be near impossible to collect that money at the moment.

> For example, an Excel-link would be beneficial
I think something (almost) like that already exists in the IO package of Octave-Forge:
http://octave.sourceforge.net/io/index.html

I know of at least one successful project where Excel was used as a GUI and Octave as a
backend computational engine.

Scilab seems to have the feature you would like:
http://atoms.scilab.org/toolboxes/Scilab_XLL

>  - Excel is the ultimate GUI.  
Well, I'm not sure I agree, but I understand your point of view is
shared by many others

> Or perhaps a C autocode toolbox - hugely beneficial.
That has been attempted in the past with little success:
http://www.program-transformation.org/Octave/WebHome

Scilab seems to have that as an external module:
http://atoms.scilab.org/toolboxes/scilab2c/


> I believe than an open source & free Octave/Simulink with affordable
> "toolboxes" would take a huge bite out of Mathwork's bottom line.
> Universities and companies could reduce (but not eliminate) their Mathworks
> overhead costs and reallocate those savings to more important ventures like
> research & development.
As far as I (and many others Octave users I know) am concerned,
price is not the reason why I prefer to use Octave, but rather its being "free":
http://www.gnu.org/philosophy/free-sw.html

You also must understand that the number of people using Octave makes little
difference to Octave developers, in particular none of the money that you state
universities and companies would save would go to Octave developers, so your argument
does not work to motivate them. Why would they want to spend THEIR time to save YOUR money?

What usually happens is that a feature ends up being added to Octave if someone who has
sufficient skills and time to implement it needs it for himself or is paid to do it.

Finally I see no reason why Octave developers would want to "take a bite out of Mathworks bottom line".
In my opinion Mathworks offers good products and, if you are OK with using proprieatry software,
it's also good value-for-money.

c.

_______________________________________________
Help-octave mailing list
[hidden email]
https://mailman.cae.wisc.edu/listinfo/help-octave
Reply | Threaded
Open this post in threaded view
|

Re: Simulink for Octave

xavion
I appreciate the helpful responses.  Some background on me... I'm a hardcore Matlab/Simulink user for 15 years and discovered Octave a few weeks ago after my Matlab student edition could not be installed on my new Windows 7 64bit PC.  I contacted Mathworks for a quote - $2500 for Matlab and $1000 for Simulink.  After that, I decided to search for alternatives.  

I'd like to address some of c.-2 statements:

c.-2 wrote
You also must understand that the number of people using Octave makes little
difference to Octave developers, in particular none of the money that you state
universities and companies would save would go to Octave developers, so your argument
does not work to motivate them. Why would they want to spend THEIR time to save YOUR money?
I'd say the Octave developers are spending THEIR time making ZERO money.  A healthy REWARD is deserved for developing an astonishing and powerful tool.  My humble recommendation... develop a Simulink-like capability compatible with Octave as an open source & free tool.  Then sell some additional toolboxes at an affordable price.  For me, I would not mind paying $50 to $100 dollars for a toolbox.  If you sell 100,000 tool boxes, that's $10M in revenue.  Does that motivate you?


c.-2 wrote
Finally I see no reason why Octave developers would want to "take a bite out of Mathworks bottom line".  In my opinion Mathworks offers good products and, if you are OK with using proprietary software, it's also good value-for-money.
You have proven that Mathworks does NOT offer good value-for-money because Octave does most of what basic Matlab can do at zero cost.  I think you are already biting into Mathworks bottom line through shear passion.  Its catching on at universities and small businesses.  Its time to collect a reward.  My only concern is that Matlab will soon realize this and offer to buy out Ocatve.  Please don't sell your soul.

Reply | Threaded
Open this post in threaded view
|

Re: Simulink for Octave

Jordi Gutiérrez Hermoso-2
On 19 August 2011 18:28, xavion <[hidden email]> wrote:

> I'd say the Octave developers are spending THEIR time making ZERO
> money.

There's more to life than making money, but money is nice, and I
personally am getting paid for working on Octave, which is also nice.
Companies do this sometime, they want to use free software, but the
software doesn't quite fit what they want it to do, so they pay
someone to improve it.

> A healthy REWARD is deserved for developing an astonishing
> and powerful tool. My humble recommendation... develop a
> Simulink-like capability compatible with Octave as an open source &
> free tool. Then sell some additional toolboxes at an affordable
> price. For me, I would not mind paying $50 to $100 dollars for a
> toolbox. If you sell 100,000 tool boxes, that's $10M in revenue.
> Does that motivate you?

Fantasy accounting? Not really. I'm motivated by other reasons.

You have repeated this idea many times, but let me tell you how that
would happen if it would. We wouldn't be selling toolboxes without
accompanying source and permission to redistribute them freely. It's
just how we do things here. We believe people's freedom to modify,
study, and redistribute code is much more important.

That being said, software can and has been sold freely, but it's
tricky to do so. I've seen several projects for example give the
source away for free but require you to pay for Windows and Mac OS X
binaries. Xchat and Ardour are two projects that do this:

     http://ardour.org/download
     http://xchat.org/windows/

I don't know how successful these two are at making money, but I doubt
they are making millions of dollars as you suggest.

> My only concern is that Matlab will soon realize this and offer to
> buy out Ocatve. Please don't sell your soul.

I think most of us working on Octave are too stubborn to sell out. I
know I am. I am quite disgusted by the idea of doing scientific
computation with license managers and hidden source code, and I
wouldn't want to be reponsible for making anyone suffer through that.
I personally want to help you do better scientific computation. I'm
not so interested in making sure you do it without pay.

- Jordi G. H.
_______________________________________________
Help-octave mailing list
[hidden email]
https://mailman.cae.wisc.edu/listinfo/help-octave
Reply | Threaded
Open this post in threaded view
|

Re: Simulink for Octave

Sven-Erik Tiberg
In reply to this post by xavion
As i remember you can use openmodula as a substitute to simulink even in RT applications.
wheter you can create a s-function from octave I'm not shure but probably with some custom API.

/Sven-Erik


Skickat från min iPhone

20 aug 2011 kl. 01:34 skrev "xavion" <[hidden email]>:

> I appreciate the helpful responses.  Some background on me... I'm a hardcore
> Matlab/Simulink user for 15 years and discovered Octave a few weeks ago
> after my Matlab student edition could not be installed on my new Windows 7
> 64bit PC.  I contacted Mathworks for a quote - $2500 for Matlab and $1000
> for Simulink.  After that, I decided to search for alternatives.  
>
> I'd like to address some of c.-2 statements:
>
>
> c.-2 wrote:
>>
>> You also must understand that the number of people using Octave makes
>> little
>> difference to Octave developers, in particular none of the money that you
>> state
>> universities and companies would save would go to Octave developers, so
>> your argument
>> does not work to motivate them. Why would they want to spend THEIR time to
>> save YOUR money?
>>
>
> I'd say the Octave developers are spending THEIR time making ZERO money.  A
> healthy REWARD is deserved for developing an astonishing and powerful tool.
> My humble recommendation... develop a Simulink-like capability compatible
> with Octave as an open source & free tool.  Then sell some additional
> toolboxes at an affordable price.  For me, I would not mind paying $50 to
> $100 dollars for a toolbox.  If you sell 100,000 tool boxes, that's $10M in
> revenue.  Does that motivate you?
>
>
>
> c.-2 wrote:
>>
>> Finally I see no reason why Octave developers would want to "take a bite
>> out of Mathworks bottom line".  In my opinion Mathworks offers good
>> products and, if you are OK with using proprietary software, it's also
>> good value-for-money.
>>
>
> You have proven that Mathworks does NOT offer good value-for-money because
> Octave does most of what basic Matlab can do at zero cost.  I think you are
> already biting into Mathworks bottom line through shear passion.  Its
> catching on at universities and small businesses.  Its time to collect a
> reward.  My only concern is that Matlab will soon realize this and offer to
> buy out Ocatve.  Please don't sell your soul.
>
>
>
> --
> View this message in context: http://octave.1599824.n4.nabble.com/Simulink-for-Octave-tp3746480p3756323.html
> Sent from the Octave - General mailing list archive at Nabble.com.
> _______________________________________________
> Help-octave mailing list
> [hidden email]
> https://mailman.cae.wisc.edu/listinfo/help-octave
_______________________________________________
Help-octave mailing list
[hidden email]
https://mailman.cae.wisc.edu/listinfo/help-octave
Reply | Threaded
Open this post in threaded view
|

Re: Simulink for Octave

Sergei Steshenko




----- Original Message -----

> From: Sven-Erik Tiberg <[hidden email]>
> To: xavion <[hidden email]>
> Cc: "[hidden email]" <[hidden email]>
> Sent: Saturday, August 20, 2011 1:29 PM
> Subject: Re: Simulink for Octave
>
> As i remember you can use openmodula as a substitute to simulink even in RT
> applications.
> wheter you can create a s-function from octave I'm not shure but probably
> with some custom API.
>
> /Sven-Erik

I can't find "openmodula" on the web.

Could you please post the exact URL ?

Thanks,
  Sergei.

_______________________________________________
Help-octave mailing list
[hidden email]
https://mailman.cae.wisc.edu/listinfo/help-octave
Reply | Threaded
Open this post in threaded view
|

Re: Simulink for Octave

c.-2

On 20 Aug 2011, at 14:13, Sergei Steshenko wrote:

> I can't find "openmodula" on the web.
>
> Could you please post the exact URL ?

Maybe he meant OpenModelica?
http://www.openmodelica.org/

> Thanks,
>   Sergei.
c.
_______________________________________________
Help-octave mailing list
[hidden email]
https://mailman.cae.wisc.edu/listinfo/help-octave
Reply | Threaded
Open this post in threaded view
|

Re: Simulink for Octave

Sergei Steshenko




----- Original Message -----

> From: c. <[hidden email]>
> To: Sergei Steshenko <[hidden email]>
> Cc: Sven-Erik Tiberg <[hidden email]>; xavion <[hidden email]>; "[hidden email]" <[hidden email]>
> Sent: Saturday, August 20, 2011 3:24 PM
> Subject: Re: Simulink for Octave
>
>
> On 20 Aug 2011, at 14:13, Sergei Steshenko wrote:
>
>>  I can't find "openmodula" on the web.
>>
>>  Could you please post the exact URL ?
>
> Maybe he meant OpenModelica?
> http://www.openmodelica.org/
>
>>  Thanks,
>>    Sergei.
> c.
>

Isn't it what is used by 'scilab' ?

Thanks,
  Sergei.

_______________________________________________
Help-octave mailing list
[hidden email]
https://mailman.cae.wisc.edu/listinfo/help-octave
Reply | Threaded
Open this post in threaded view
|

Re: Simulink for Octave

c.-2

On 20 Aug 2011, at 14:31, Sergei Steshenko wrote:

> Isn't it what is used by 'scilab' ?

I don't think so, Modelica

https://www.modelica.org/

is a  standardized language meant to program simulators of dynamic systems
and OpenModelica and Modelicac (which is included with Scilab) are two among
its many implementations:

https://www.modelica.org/tools

c.

_______________________________________________
Help-octave mailing list
[hidden email]
https://mailman.cae.wisc.edu/listinfo/help-octave
Reply | Threaded
Open this post in threaded view
|

Re: Simulink for Octave

c.-2
In reply to this post by xavion

On 20 Aug 2011, at 01:28, xavion wrote:

> I appreciate the helpful responses.  Some background on me... I'm a hardcore
> Matlab/Simulink user for 15 years and discovered Octave a few weeks ago
> after my Matlab student edition could not be installed on my new Windows 7
> 64bit PC.  I contacted Mathworks for a quote - $2500 for Matlab and $1000
> for Simulink.
That is one of the risks involved with using proprietary software.

>  After that, I decided to search for alternatives.  

a few viable alternatives have been proposed previously in this thread.
In particular I would say, to summarize

- if you really need the GUI, but can give up on strict Matlab compatibility
Scilab + Xcos is the way to go

- if you don't really nedd the GUI, Octave can do with dynamic systems most
 of the things taht Matlab can

I would also suggest giving a look at OpenModelica:
http://www.openmodelica.org/

which seems very promising, although the frequent references to "open source"
on the web site make me wonder whether that is really free software.


> I'd say the Octave developers are spending THEIR time making ZERO money.
Fortunately it is not so :) most Octave developers are first of all Octave USERS,
they are paid for the work they do WITH Octave not on Octave.
Some are (ore have been)  paid to work on Octave.

>  A healthy REWARD is deserved for developing an astonishing and powerful tool.
part of this reward is that working cooperatively to improve Octave gives users
a better tool and makes them more productive in their everyday work.

> My humble recommendation... develop a Simulink-like capability compatible
> with Octave as an open source & free tool.  Then sell some additional
> toolboxes at an affordable price.  For me, I would not mind paying $50 to
> $100 dollars for a toolbox.  If you sell 100,000 tool boxes, that's $10M in
> revenue.  Does that motivate you?
The recommendation is very welcome and appreciated but unfortunately the approach you suggest
does not fit with the way Octave (and many other free software projects) work and cannot be implemented.
You seem to think of Octave as a company similar to the Mathworks, but it isn't.
Octave is developed by many indepent individuals who mostly develop those extensions they need
for themselves (or that someone else pays them to implement) and then agree to share their work with others.
One of the issues involved with selling Octave (or add-ons) would be how to distribute the
revenue deriving from its sales among the many contributors.


> You have proven that Mathworks does NOT offer good value-for-money because
> Octave does most of what basic Matlab can do at zero cost.  
I was trying to explain that Octave does not come at zero cost either, it is
a common mistake to confuse the term "free" in "free-software" with "gratis".

> I think you are
> already biting into Mathworks bottom line through shear passion.  Its
> catching on at universities and small businesses.  Its time to collect a
> reward.  My only concern is that Matlab will soon realize this and offer to
> buy out Ocatve.  Please don't sell your soul.
Oh, well. I don't think there is a way to actually "buy" Octave, the GPL license
is especially designed to prevent anyone from taking away freedom from the software.
On the other hand, Octave developers could be hired by a proprietary software company, but, as Jordi
said, some would be reluctant to accept.

c.



_______________________________________________
Help-octave mailing list
[hidden email]
https://mailman.cae.wisc.edu/listinfo/help-octave
Reply | Threaded
Open this post in threaded view
|

Re: Simulink for Octave

Jordi Gutiérrez Hermoso-2
On 20 August 2011 08:25, c. <[hidden email]> wrote:
> I would also suggest giving a look at OpenModelica:
> http://www.openmodelica.org/
>
> which seems very promising, although the frequent references to "open source"
> on the web site make me wonder whether that is really free software.

I looked into it a few days ago. Looks like it is. They have several
licensing options, but one of them is GPLv3 + additional restriction
(*everything* that gets near to OpenModelica should also be GPLv3,
much wider copyleft than what the actual GPLv3 says, is what they are
basically trying to impose). You can see their license terms here:

     https://openmodelica.org/svn/OpenModelica/trunk/OSMC-License.txt

User is "anonymous", password is "none".

However, GPLv3 says in Section 7,

     If the Program as you received it, or any part of it, contains a
     notice stating that it is governed by this License along with a
     term that is a further restriction, you may remove that term.

so it looks like it's an ineffective attempt anyways, and the code is
GPLv3, like Octave.

- Jordi G. H.
_______________________________________________
Help-octave mailing list
[hidden email]
https://mailman.cae.wisc.edu/listinfo/help-octave
12