SDK for Max 5

Apr 29, 2008 at 4:08am

SDK for Max 5

Hi pals,

Are we going to have a new SDK for Max 5 shortly ? or should we keep on developing with the current SDK for new projects ?

Good Day.

yours,
William

#37362
Apr 29, 2008 at 11:36am

On 2008 Apr 28, at 11:08 PM, williamshome wrote:

> Hi pals,
>
> Are we going to have a new SDK for Max 5 shortly ? or should we keep
> on developing with the current SDK for new projects ?
>
> Good Day.

For the time being it is best to develop against the Max 4.6 SDK.
Peter Castine has made a nice list of caveats to keep in mind while
doing this. Now that Max 5 is released, we’re beginning to work on
the SDK. We hope to make it available in at least a preliminary form
by the summer.

best,
Tim

#129059
Apr 29, 2008 at 12:41pm

Can someone repost this list of caveats again?

Anthony

—– Original Message —–
From: “Timothy Place”
To:
Sent: Tuesday, April 29, 2008 6:36 AM
Subject: Re: [dev] SDK for Max 5

> On 2008 Apr 28, at 11:08 PM, williamshome wrote:
>
>> Hi pals,
>>
>> Are we going to have a new SDK for Max 5 shortly ? or should we keep
>> on developing with the current SDK for new projects ?
>>
>> Good Day.
>
> For the time being it is best to develop against the Max 4.6 SDK.
> Peter Castine has made a nice list of caveats to keep in mind while
> doing this. Now that Max 5 is released, we’re beginning to work on
> the SDK. We hope to make it available in at least a preliminary form
> by the summer.
>
> best,
> Tim
>
>
>
>

#129060
Apr 29, 2008 at 1:22pm

On 2008 Apr 29, at 7:42 PM, Anthony Palomba wrote:
> Can someone repost this list of caveats again?

Here’s a link:

http://www.cycling74.com/forums/index.php?t=msg&rid=0&S=8eecc48dde031a3654e68bcbd271bcb1&th=33065&goto=138609#msg_138629

best,
Tim

#129061
May 21, 2008 at 2:06pm

I’m having some luck hacking away at what I can in the new (Max 5) MaxAPI.framework, in particular the new jbox and jpatcher stuff.
Is anyone else doing this?

(e.g. I’ve managed to redo my patcher/box traversal routines).

BTW I’ll post what I have once I’m fairly sure it’s working though perhaps I shouldn’t be so impatient … and wait for the SDK… :)

#129062
May 21, 2008 at 3:59pm

We’re working on the SDK, and some preliminary form of it will be
ready in June.

If you are accessing the t_jbox and t_jpatcher struct members
directly, then you are asking for big trouble.

best,
Tim

On 2008 May 21, at 9:06 AM, Martin Robinson wrote:

>
> I’m having some luck hacking away at what I can in the new (Max 5)
> MaxAPI.framework, in particular the new jbox and jpatcher stuff.
> (e.g. I’ve managed to redo my patcher/box traversal routines).
> Is anyone else doing this?
>
> BTW I’ll post what I have once I’m fairly sure it’s working though
> perhaps I shouldn’t be so impatient … and wait for the SDK… :)
> –
> martin robinson
> max/msp java c
> installations/performance and systems

#129063
May 21, 2008 at 5:52pm

Quote: Timothy Place wrote on Wed, 21 May 2008 17:59
—————————————————-
> If you are accessing the t_jbox and t_jpatcher struct members
> directly, then you are asking for big trouble.
>

Hehe, ok ok.. got the message ;)

Mattijs

#129064
May 21, 2008 at 5:57pm

On 2008 May 21, at 12:52 PM, Mattijs Kneppers wrote:

> Quote: Timothy Place wrote on Wed, 21 May 2008 17:59
> —————————————————-
>> If you are accessing the t_jbox and t_jpatcher struct members
>> directly, then you are asking for big trouble.
>
> Hehe, ok ok.. got the message ;)

Just to drive it home… Cycling 74′s own externals in Max 5 do not
even access these struct members directly. We’re not just being
ornery about this subject.

best,
Tim

#129065
May 21, 2008 at 5:58pm

Quote: Timothy Place wrote on Wed, 21 May 2008 16:59
—————————————————-
> We’re working on the SDK, and some preliminary form of it will be
> ready in June.

Yes, I’m in no way trying to hurry things up …!

>
> If you are accessing the t_jbox and t_jpatcher struct members
> directly, then you are asking for big trouble.

No, for info I’m using the patcher_eachdo() function and in its callback method doing:

b = jpatcher_get_firstobject((t_object*)patcher);

…to get the first object, replacing the pre-Max 5 code:

b = patcher->p_box;

and using:

b = jbox_get_nextobject((t_object*)b);

…to traverse the objects until b is null, pre-Max 5 this was:

b = b->b_next;

Also I’m doing:

symbol = jbox_get_varname((t_object*)b);

To get the object name rather than the old:

patcher_boxname(p, b, &symbol);

…code. Anyway I’m stuck now and I shall probably have to leave this here until the SDK (again, no rush!).

I was using:

mess1((t_object *)b,gensym(“psave”),bb);

…to get an object (b) to save itself into a binbuf (bb). (I know jamie was using this too.)

I get that the object doesn’t understand ‘psave’, anyway I’m probably sending the message to the wrong object (I guess I’m traversing objects rather than boxes) or using the wrong message (tried ‘save’, ‘jsave’, ‘save2′).

I’m not expecting answers here so no need to respond…

#129066
May 21, 2008 at 6:19pm

On 2008 May 21, at 12:58 PM, Martin Robinson wrote:

> Quote: Timothy Place wrote on Wed, 21 May 2008 16:59
> —————————————————-
>> We’re working on the SDK, and some preliminary form of it will be
>> ready in June.
>
> Yes, I’m in no way trying to hurry things up …!
>
>> If you are accessing the t_jbox and t_jpatcher struct members
>> directly, then you are asking for big trouble.
>
> No, for info I’m using the patcher_eachdo() function and in its
> callback method doing:

Okay, good!

> b = jpatcher_get_firstobject((t_object*)patcher);
>
> …to get the first object, replacing the pre-Max 5 code:
>
> b = patcher->p_box;
>
> and using:
>
> b = jbox_get_nextobject((t_object*)b);
>
> …to traverse the objects until b is null, pre-Max 5 this was:
>
> b = b->b_next;
>
> Also I’m doing:
>
> symbol = jbox_get_varname((t_object*)b);
>
> To get the object name rather than the old:
>
> patcher_boxname(p, b, &symbol);
>
> …code. Anyway I’m stuck now and I shall probably have to leave
> this here until the SDK (again, no rush!).
>
> I was using:
>
> mess1((t_object *)b,gensym(“psave”),bb);
>
> …to get an object (b) to save itself into a binbuf (bb). (I know
> jamie was using this too.)
>
> I get that the object doesn’t understand ‘psave’, anyway I’m
> probably sending the message to the wrong object (I guess I’m
> traversing objects rather than boxes) or using the wrong message
> (tried ‘save’, ‘jsave’, ‘save2′).
>
> I’m not expecting answers here so no need to respond…

We aren’t ready to answer questions about an API where the rough draft
is still being assembled. However, there are some things that may
help you to help yourself in the Max 4 API. So I’ll just remind you a
couple of them.

First is the trick to find out what class a t_object* is. So if you
have a t_object* named “b”, then you can do this:
t_symbol *classname_of_b = b->o_messlist[-1].m_sym;
Then you will at least know what kind of object you are sending
messages to.

Second, while the mess1() macro should still work, the recommended
form introduced in Max 4.5 (see the Obex documentation) is to use
object_method(), object_method_typed(), etc. If a method has an
A_GIMME signature then you need to use object_method_typed().

Third, if you have the t_object*, then you can investigate the
o_messlist member to find out what messages the object does actually
understand.

We don’t really use binbufs to save things with the patcher anymore in
Max 5. We won’t be elucidating this matter until the API is documented.

best,
Tim

#129067
May 21, 2008 at 7:13pm

Quote: Timothy Place wrote on Wed, 21 May 2008 19:19
—————————————————-

> We don’t really use binbufs to save things with the patcher anymore in
> Max 5. We won’t be elucidating this matter until the API is documented.

Yes I realise the binbuf issue is the main stumbling block but temporarily forgot about that as I wrote my last post. (I was aware it was going to be an issue before Max 5 was released.)

Thanks.

#129068
Jul 24, 2008 at 2:23pm

Is the preliminary Max 5 SDK available yet?

#129069
Jul 24, 2008 at 9:50pm

On 24 juil. 08, at 16:23, Chris Graham wrote:

> Is the preliminary Max 5 SDK available yet?

Nope, it’s gonna be available at some point later in a not to distant
future.

ej

#129070
Sep 9, 2008 at 4:00pm

Any sign of it yet?

#129071
Sep 9, 2008 at 4:14pm

Yes indeed, I would like to know as well…

—- Jonny wrote:
>
> Any sign of it yet?

#129072
Sep 10, 2008 at 5:30pm

Or at least maybe a new estimate of release.

Rick

#129073
Sep 11, 2008 at 4:00am

count me in as to when this will be made available…

…it would be greatly appreciated.

Brandon.

On Mon, Apr 28, 2008 at 9:08 PM, williamshome wrote:

>
> Hi pals,
>
> Are we going to have a new SDK for Max 5 shortly ? or should we keep on
> developing with the current SDK for new projects ?
>
> Good Day.
>
> yours,
> William
>

#129074
Sep 11, 2008 at 11:01am

yeah, is it still summer? :)

#129075
Sep 11, 2008 at 11:01am

yeah, is it still summer? :)

#129076
Sep 11, 2008 at 1:06pm

It’s still summer here in Texas, so…
this might still count as summer.

Keith

On Thu, Sep 11, 2008 at 6:01 AM, Martin Robinson <
martinrobinson_mail@yahoo.co.uk> wrote:

>
> yeah, is it still summer? :)
> –
> martin robinson
> max/msp java c
> installations/performance and systems
>

#129077
Sep 12, 2008 at 10:51pm

summer has yet to start in the southern hemisphere.
sorry, i am in brazil this week and that is what people keep telling me when they talk about the weather.
–NW

#129078
Oct 20, 2008 at 1:57pm

Any update on the Max5 SDK?

#129079
Oct 24, 2008 at 7:10am

?????????
> On 2008 Apr 28, at 11:08 PM, xx wrote:
> Now that Max 5 is released, we’re beginning to work on the SDK. We
> hope to make it available in at least a preliminary form by the
> summer.

Any update on the Max5 SDK?

#129080
Oct 24, 2008 at 10:13am

must be done by the ircam max dev event next month…

#129081
Oct 24, 2008 at 10:59am

On 24 oct. 08, at 12:13, Martin Robinson wrote:

> must be done by the ircam max dev event next month…

btw, there’s still a few place left for this workshop.

http://www.cycling74.com/forums/index.php?t=msg&th=35728&start=0&rid=0&S=9cd566bf4ca4ea046641e97453cc7a97

ej

#129082
Nov 13, 2008 at 6:37pm

Another developer said the SDK is not going to be available until the Spring now. Is that true? I am about to purchase Max, but it appears I will also be doing some development in C, because Max does not have all the features I will need for my designs. But the SDK only works for v4.6.

Googling for Max 4.6 I find an ‘OEM version’ offered for $40.

http://www.bestoemsoftware.net/sale-Cycling74_Max_Msp_4.57.htm

Is this what I need to buy in order to access the SDK? Full-price retailers offer Max 5 only. The SDK is not available for Max 5, and I cant find any other way to buy Max 4. I am reluctant to purchase this ‘OEM version,’ I suspect it is not fully licensed software.

#129083
Nov 13, 2008 at 6:48pm

Dear Ernest,

own externals developed using the SDK for MaxMSP 4 work perfectly in
Max 5 (at least this my experience with them :-) ).
If you have a look on Cycling ’74 website, you’ll find that the SDK is
completely free for any software (MaxMSP and/or Jitter).
Also consider that you don’t need to buy Max 4 separately, as the full
licence of Max 5 allows you to run any previous version of MaxMSP, at
least this was told me at Cycling ’74 when I purchased Max 5 (and this
must certainly be a real information, as I have got both Max 4 and Max
5 on my machine with valid licences).

Hope this helps,
Adam

On 2008.11.13., at 19:37, Ernest wrote:

>
> Another developer said the SDK is not going to be available until
> the Spring now. Is that true? I am about to purchase Max, but it
> appears I will also be doing some development in C, because Max does
> not have all the features I will need for my designs. But the SDK
> only works for v4.6.
>
> Googling for Max 4.6 I find an ‘OEM version’ offered for $40.
>
> http://www.bestoemsoftware.net/sale-Cycling74_Max_Msp_4.57.htm
>
> Is this what I need to buy in order to access the SDK? Full-price
> retailers offer Max 5 only. The SDK is not available for Max 5, and
> I cant find any other way to buy Max 4. I am reluctant to purchase
> this ‘OEM version,’ I suspect it is not fully licensed software.
>
>
>

#129084
Nov 13, 2008 at 6:56pm

Thank you, it was unclear that a Max 5 license included license for previous versions also.

#129085
Nov 14, 2008 at 11:02am

Quote: sadam wrote on Thu, 13 November 2008 19:48
—————————————————-
> own externals developed using the SDK for MaxMSP 4 work perfectly in
> Max 5 (at least this my experience with them :-) ).
—————————————————-

For plain-vanilla external programming this is the case, but there are quite a few gotchas that have been discussed several times in the past, including this very thread.

There is a fairly complete list of incompatibilities that I posted when 5.0 came out. It’s linked earlier in this thread, scroll back to postings from April 29.

#129086

You must be logged in to reply to this topic.