Hi all -

I don't know whether this perhaps has been reported yet. I had  
problems using Csound 5.12 with QCS (no problems with Csound 5.11),  
and I saw now that this is related to PortMIDI. When I select (as I  
usuallly do) PortMIDI with -a, I can run just ONE csd. Then, after  
some seconds, QCS hangs, and I must (OSX) quit by violence.

When I use "none" as MIDI Modul, no problem. Funny enough, also no  
problem with Virtual MIDI.

By the way, Andrés: Console Printout is incredibly fast in the last  
svn!!! Wow - looks like sorcery ...

	joachim

Send bugs reports to the Sourceforge bug tracker
            https://sourceforge.net/tracker/?group_id=81968&atid=564599
Discussions of bugs and features can be posted here
To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"

I think I have seen this before. I have the impression PortMIDI is not  
reentrant and gives trouble like this.
I am sure I have reported a similar problem with qutecsound and  
portmidi.

Victor
On 26 Feb 2010, at 19:59, joachim heintz wrote:

> Hi all -
>
> I don't know whether this perhaps has been reported yet. I had  
> problems using Csound 5.12 with QCS (no problems with Csound 5.11),  
> and I saw now that this is related to PortMIDI. When I select (as I  
> usuallly do) PortMIDI with -a, I can run just ONE csd. Then, after  
> some seconds, QCS hangs, and I must (OSX) quit by violence.
>
> When I use "none" as MIDI Modul, no problem. Funny enough, also no  
> problem with Virtual MIDI.
>
> By the way, Andrés: Console Printout is incredibly fast in the last  
> svn!!! Wow - looks like sorcery ...
>
> 	joachim
>
> Send bugs reports to the Sourceforge bug tracker
>           https://sourceforge.net/tracker/?group_id=81968&atid=564599
> Discussions of bugs and features can be posted here
> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
> "unsubscribe csound"
>



Send bugs reports to the Sourceforge bug tracker
            https://sourceforge.net/tracker/?group_id=81968&atid=564599
Discussions of bugs and features can be posted here
To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"

yes, I just checked my mailbox and I have reported this very same  
problem with portmidi and qutecsound, on 16/Nov/09

Victor


On 26 Feb 2010, at 20:40, Victor Lazzarini wrote:

> I think I have seen this before. I have the impression PortMIDI is  
> not reentrant and gives trouble like this.
> I am sure I have reported a similar problem with qutecsound and  
> portmidi.
>
> Victor
> On 26 Feb 2010, at 19:59, joachim heintz wrote:
>
>> Hi all -
>>
>> I don't know whether this perhaps has been reported yet. I had  
>> problems using Csound 5.12 with QCS (no problems with Csound 5.11),  
>> and I saw now that this is related to PortMIDI. When I select (as I  
>> usuallly do) PortMIDI with -a, I can run just ONE csd. Then, after  
>> some seconds, QCS hangs, and I must (OSX) quit by violence.
>>
>> When I use "none" as MIDI Modul, no problem. Funny enough, also no  
>> problem with Virtual MIDI.
>>
>> By the way, Andrés: Console Printout is incredibly fast in the last  
>> svn!!! Wow - looks like sorcery ...
>>
>> 	joachim
>>
>> Send bugs reports to the Sourceforge bug tracker
>>          https://sourceforge.net/tracker/?group_id=81968&atid=564599
>> Discussions of bugs and features can be posted here
>> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
>> "unsubscribe csound"
>>
>
>
>
> Send bugs reports to the Sourceforge bug tracker
>           https://sourceforge.net/tracker/?group_id=81968&atid=564599
> Discussions of bugs and features can be posted here
> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
> "unsubscribe csound"
>



Send bugs reports to the Sourceforge bug tracker
            https://sourceforge.net/tracker/?group_id=81968&atid=564599
Discussions of bugs and features can be posted here
To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"

mmh, this is in so far different as I didn't have any problems with  
csound 5.11.
but perhaps you used already 5.12 from cvs at this time?

	joachim


Am 26.02.2010 um 21:48 schrieb Victor Lazzarini:

> yes, I just checked my mailbox and I have reported this very same  
> problem with portmidi and qutecsound, on 16/Nov/09
>
> Victor
>
>
> On 26 Feb 2010, at 20:40, Victor Lazzarini wrote:
>
>> I think I have seen this before. I have the impression PortMIDI is  
>> not reentrant and gives trouble like this.
>> I am sure I have reported a similar problem with qutecsound and  
>> portmidi.
>>
>> Victor
>> On 26 Feb 2010, at 19:59, joachim heintz wrote:
>>
>>> Hi all -
>>>
>>> I don't know whether this perhaps has been reported yet. I had  
>>> problems using Csound 5.12 with QCS (no problems with Csound  
>>> 5.11), and I saw now that this is related to PortMIDI. When I  
>>> select (as I usuallly do) PortMIDI with -a, I can run just ONE  
>>> csd. Then, after some seconds, QCS hangs, and I must (OSX) quit by  
>>> violence.
>>>
>>> When I use "none" as MIDI Modul, no problem. Funny enough, also no  
>>> problem with Virtual MIDI.
>>>
>>> By the way, Andrés: Console Printout is incredibly fast in the  
>>> last svn!!! Wow - looks like sorcery ...
>>>
>>> 	joachim
>>>
>>> Send bugs reports to the Sourceforge bug tracker
>>>         https://sourceforge.net/tracker/?group_id=81968&atid=564599
>>> Discussions of bugs and features can be posted here
>>> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
>>> "unsubscribe csound"
>>>
>>
>>
>>
>> Send bugs reports to the Sourceforge bug tracker
>>          https://sourceforge.net/tracker/?group_id=81968&atid=564599
>> Discussions of bugs and features can be posted here
>> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
>> "unsubscribe csound"
>>
>
>
>
> Send bugs reports to the Sourceforge bug tracker
>           https://sourceforge.net/tracker/?group_id=81968&atid=564599
> Discussions of bugs and features can be posted here
> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
> "unsubscribe csound"
>
>



Send bugs reports to the Sourceforge bug tracker
            https://sourceforge.net/tracker/?group_id=81968&atid=564599
Discussions of bugs and features can be posted here
To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"

That maybe, but I can't say now. It seems others weren't seeing the  
same error then. However, the portmidi module code (pmidi.c) has not  
changed in 14 months. My feeling this is related to the portmidi  
library itself.


Victor
On 26 Feb 2010, at 21:03, joachim heintz wrote:

> mmh, this is in so far different as I didn't have any problems with  
> csound 5.11.
> but perhaps you used already 5.12 from cvs at this time?
>
> 	joachim
>
>
> Am 26.02.2010 um 21:48 schrieb Victor Lazzarini:
>
>> yes, I just checked my mailbox and I have reported this very same  
>> problem with portmidi and qutecsound, on 16/Nov/09
>>
>> Victor
>>
>>
>> On 26 Feb 2010, at 20:40, Victor Lazzarini wrote:
>>
>>> I think I have seen this before. I have the impression PortMIDI is  
>>> not reentrant and gives trouble like this.
>>> I am sure I have reported a similar problem with qutecsound and  
>>> portmidi.
>>>
>>> Victor
>>> On 26 Feb 2010, at 19:59, joachim heintz wrote:
>>>
>>>> Hi all -
>>>>
>>>> I don't know whether this perhaps has been reported yet. I had  
>>>> problems using Csound 5.12 with QCS (no problems with Csound  
>>>> 5.11), and I saw now that this is related to PortMIDI. When I  
>>>> select (as I usuallly do) PortMIDI with -a, I can run just ONE  
>>>> csd. Then, after some seconds, QCS hangs, and I must (OSX) quit  
>>>> by violence.
>>>>
>>>> When I use "none" as MIDI Modul, no problem. Funny enough, also  
>>>> no problem with Virtual MIDI.
>>>>
>>>> By the way, Andrés: Console Printout is incredibly fast in the  
>>>> last svn!!! Wow - looks like sorcery ...
>>>>
>>>> 	joachim
>>>>
>>>> Send bugs reports to the Sourceforge bug tracker
>>>>        https://sourceforge.net/tracker/?group_id=81968&atid=564599
>>>> Discussions of bugs and features can be posted here
>>>> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
>>>> "unsubscribe csound"
>>>>
>>>
>>>
>>>
>>> Send bugs reports to the Sourceforge bug tracker
>>>         https://sourceforge.net/tracker/?group_id=81968&atid=564599
>>> Discussions of bugs and features can be posted here
>>> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
>>> "unsubscribe csound"
>>>
>>
>>
>>
>> Send bugs reports to the Sourceforge bug tracker
>>          https://sourceforge.net/tracker/?group_id=81968&atid=564599
>> Discussions of bugs and features can be posted here
>> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
>> "unsubscribe csound"
>>
>>
>
>
>
> Send bugs reports to the Sourceforge bug tracker
>           https://sourceforge.net/tracker/?group_id=81968&atid=564599
> Discussions of bugs and features can be posted here
> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
> "unsubscribe csound"
>



Send bugs reports to the Sourceforge bug tracker
            https://sourceforge.net/tracker/?group_id=81968&atid=564599
Discussions of bugs and features can be posted here
To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"

Maybe if qutecsound re-uses the same CSOUND instance, this might be  
avoided? Perhaps Andres can check this.

Victor
On 26 Feb 2010, at 21:13, Victor Lazzarini wrote:

> That maybe, but I can't say now. It seems others weren't seeing the  
> same error then. However, the portmidi module code (pmidi.c) has not  
> changed in 14 months. My feeling this is related to the portmidi  
> library itself.
>
>
> Victor
> On 26 Feb 2010, at 21:03, joachim heintz wrote:
>
>> mmh, this is in so far different as I didn't have any problems with  
>> csound 5.11.
>> but perhaps you used already 5.12 from cvs at this time?
>>
>> 	joachim
>>
>>
>> Am 26.02.2010 um 21:48 schrieb Victor Lazzarini:
>>
>>> yes, I just checked my mailbox and I have reported this very same  
>>> problem with portmidi and qutecsound, on 16/Nov/09
>>>
>>> Victor
>>>
>>>
>>> On 26 Feb 2010, at 20:40, Victor Lazzarini wrote:
>>>
>>>> I think I have seen this before. I have the impression PortMIDI  
>>>> is not reentrant and gives trouble like this.
>>>> I am sure I have reported a similar problem with qutecsound and  
>>>> portmidi.
>>>>
>>>> Victor
>>>> On 26 Feb 2010, at 19:59, joachim heintz wrote:
>>>>
>>>>> Hi all -
>>>>>
>>>>> I don't know whether this perhaps has been reported yet. I had  
>>>>> problems using Csound 5.12 with QCS (no problems with Csound  
>>>>> 5.11), and I saw now that this is related to PortMIDI. When I  
>>>>> select (as I usuallly do) PortMIDI with -a, I can run just ONE  
>>>>> csd. Then, after some seconds, QCS hangs, and I must (OSX) quit  
>>>>> by violence.
>>>>>
>>>>> When I use "none" as MIDI Modul, no problem. Funny enough, also  
>>>>> no problem with Virtual MIDI.
>>>>>
>>>>> By the way, Andrés: Console Printout is incredibly fast in the  
>>>>> last svn!!! Wow - looks like sorcery ...
>>>>>
>>>>> 	joachim
>>>>>
>>>>> Send bugs reports to the Sourceforge bug tracker
>>>>>       https://sourceforge.net/tracker/?group_id=81968&atid=564599
>>>>> Discussions of bugs and features can be posted here
>>>>> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
>>>>> "unsubscribe csound"
>>>>>
>>>>
>>>>
>>>>
>>>> Send bugs reports to the Sourceforge bug tracker
>>>>        https://sourceforge.net/tracker/?group_id=81968&atid=564599
>>>> Discussions of bugs and features can be posted here
>>>> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
>>>> "unsubscribe csound"
>>>>
>>>
>>>
>>>
>>> Send bugs reports to the Sourceforge bug tracker
>>>         https://sourceforge.net/tracker/?group_id=81968&atid=564599
>>> Discussions of bugs and features can be posted here
>>> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
>>> "unsubscribe csound"
>>>
>>>
>>
>>
>>
>> Send bugs reports to the Sourceforge bug tracker
>>          https://sourceforge.net/tracker/?group_id=81968&atid=564599
>> Discussions of bugs and features can be posted here
>> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
>> "unsubscribe csound"
>>
>
>
>
> Send bugs reports to the Sourceforge bug tracker
>           https://sourceforge.net/tracker/?group_id=81968&atid=564599
> Discussions of bugs and features can be posted here
> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
> "unsubscribe csound"
>



Send bugs reports to the Sourceforge bug tracker
            https://sourceforge.net/tracker/?group_id=81968&atid=564599
Discussions of bugs and features can be posted here
To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"

I have done some tests that seem to point at problems of reentrancy  
with portmidi; it seems to have some issues.
The second time Csound opens the MIDI output through portmidi, this  
message appears at the terminal

2010-02-26 21:40:51.498 cstest[13900:10b] *** CFMessagePort:  
bootstrap_register(): failed 1103 (0x44f) 'Service name already  
exists', port = 0x5f1b, name = 'Apple MIDI Client'

then the next time the csound instance is destroyed, it crashes.

Victor


On 26 Feb 2010, at 21:16, Victor Lazzarini wrote:

> Maybe if qutecsound re-uses the same CSOUND instance, this might be  
> avoided? Perhaps Andres can check this.
>
> Victor
> On 26 Feb 2010, at 21:13, Victor Lazzarini wrote:
>
>> That maybe, but I can't say now. It seems others weren't seeing the  
>> same error then. However, the portmidi module code (pmidi.c) has  
>> not changed in 14 months. My feeling this is related to the  
>> portmidi library itself.
>>
>>
>> Victor
>> On 26 Feb 2010, at 21:03, joachim heintz wrote:
>>
>>> mmh, this is in so far different as I didn't have any problems  
>>> with csound 5.11.
>>> but perhaps you used already 5.12 from cvs at this time?
>>>
>>> 	joachim
>>>
>>>
>>> Am 26.02.2010 um 21:48 schrieb Victor Lazzarini:
>>>
>>>> yes, I just checked my mailbox and I have reported this very same  
>>>> problem with portmidi and qutecsound, on 16/Nov/09
>>>>
>>>> Victor
>>>>
>>>>
>>>> On 26 Feb 2010, at 20:40, Victor Lazzarini wrote:
>>>>
>>>>> I think I have seen this before. I have the impression PortMIDI  
>>>>> is not reentrant and gives trouble like this.
>>>>> I am sure I have reported a similar problem with qutecsound and  
>>>>> portmidi.
>>>>>
>>>>> Victor
>>>>> On 26 Feb 2010, at 19:59, joachim heintz wrote:
>>>>>
>>>>>> Hi all -
>>>>>>
>>>>>> I don't know whether this perhaps has been reported yet. I had  
>>>>>> problems using Csound 5.12 with QCS (no problems with Csound  
>>>>>> 5.11), and I saw now that this is related to PortMIDI. When I  
>>>>>> select (as I usuallly do) PortMIDI with -a, I can run just ONE  
>>>>>> csd. Then, after some seconds, QCS hangs, and I must (OSX) quit  
>>>>>> by violence.
>>>>>>
>>>>>> When I use "none" as MIDI Modul, no problem. Funny enough, also  
>>>>>> no problem with Virtual MIDI.
>>>>>>
>>>>>> By the way, Andrés: Console Printout is incredibly fast in the  
>>>>>> last svn!!! Wow - looks like sorcery ...
>>>>>>
>>>>>> 	joachim
>>>>>>
>>>>>> Send bugs reports to the Sourceforge bug tracker
>>>>>>      https://sourceforge.net/tracker/?group_id=81968&atid=564599
>>>>>> Discussions of bugs and features can be posted here
>>>>>> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
>>>>>> "unsubscribe csound"
>>>>>>
>>>>>
>>>>>
>>>>>
>>>>> Send bugs reports to the Sourceforge bug tracker
>>>>>       https://sourceforge.net/tracker/?group_id=81968&atid=564599
>>>>> Discussions of bugs and features can be posted here
>>>>> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
>>>>> "unsubscribe csound"
>>>>>
>>>>
>>>>
>>>>
>>>> Send bugs reports to the Sourceforge bug tracker
>>>>        https://sourceforge.net/tracker/?group_id=81968&atid=564599
>>>> Discussions of bugs and features can be posted here
>>>> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
>>>> "unsubscribe csound"
>>>>
>>>>
>>>
>>>
>>>
>>> Send bugs reports to the Sourceforge bug tracker
>>>         https://sourceforge.net/tracker/?group_id=81968&atid=564599
>>> Discussions of bugs and features can be posted here
>>> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
>>> "unsubscribe csound"
>>>
>>
>>
>>
>> Send bugs reports to the Sourceforge bug tracker
>>          https://sourceforge.net/tracker/?group_id=81968&atid=564599
>> Discussions of bugs and features can be posted here
>> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
>> "unsubscribe csound"
>>
>
>
>
> Send bugs reports to the Sourceforge bug tracker
>           https://sourceforge.net/tracker/?group_id=81968&atid=564599
> Discussions of bugs and features can be posted here
> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
> "unsubscribe csound"
>



Send bugs reports to the Sourceforge bug tracker
            https://sourceforge.net/tracker/?group_id=81968&atid=564599
Discussions of bugs and features can be posted here
To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"

Hi,

The issue is new, because previously, you couldn't run several
instances of Csound simultaneously in QuteCsound, but now that you can
this becomes an issue.

I have a compile time macro that allows Csound to be reused or be
destroyed after each run. I have messed with this recently, so I'm not
sure what Joachim is testing with. It's currently set on svn to
destroy Csound on every run, so the problem should only appear when
trying to run a csd while another tab is running.

I can confirm that Csound hangs when trying to run a second instance
of Csound if using portmidi, but things are working fine on Linux if
Csound is stopped before running a new csd.

The hang (according to my debugger) occurs in pmidi.c line 252 when
attempting to open the midi device.

Cheers,
Andrés

On Fri, Feb 26, 2010 at 9:03 PM, joachim heintz  wrote:
> mmh, this is in so far different as I didn't have any problems with csound
> 5.11.
> but perhaps you used already 5.12 from cvs at this time?
>
>        joachim
>
>
> Am 26.02.2010 um 21:48 schrieb Victor Lazzarini:
>
>> yes, I just checked my mailbox and I have reported this very same problem
>> with portmidi and qutecsound, on 16/Nov/09
>>
>> Victor
>>
>>
>> On 26 Feb 2010, at 20:40, Victor Lazzarini wrote:
>>
>>> I think I have seen this before. I have the impression PortMIDI is not
>>> reentrant and gives trouble like this.
>>> I am sure I have reported a similar problem with qutecsound and portmidi.
>>>
>>> Victor
>>> On 26 Feb 2010, at 19:59, joachim heintz wrote:
>>>
>>>> Hi all -
>>>>
>>>> I don't know whether this perhaps has been reported yet. I had problems
>>>> using Csound 5.12 with QCS (no problems with Csound 5.11), and I saw now
>>>> that this is related to PortMIDI. When I select (as I usuallly do) PortMIDI
>>>> with -a, I can run just ONE csd. Then, after some seconds, QCS hangs, and I
>>>> must (OSX) quit by violence.
>>>>
>>>> When I use "none" as MIDI Modul, no problem. Funny enough, also no
>>>> problem with Virtual MIDI.
>>>>
>>>> By the way, Andrés: Console Printout is incredibly fast in the last
>>>> svn!!! Wow - looks like sorcery ...
>>>>
>>>>        joachim
>>>>
>>>> Send bugs reports to the Sourceforge bug tracker
>>>>        https://sourceforge.net/tracker/?group_id=81968&atid=564599
>>>> Discussions of bugs and features can be posted here
>>>> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe
>>>> csound"
>>>>
>>>
>>>
>>>
>>> Send bugs reports to the Sourceforge bug tracker
>>>         https://sourceforge.net/tracker/?group_id=81968&atid=564599
>>> Discussions of bugs and features can be posted here
>>> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe
>>> csound"
>>>
>>
>>
>>
>> Send bugs reports to the Sourceforge bug tracker
>>          https://sourceforge.net/tracker/?group_id=81968&atid=564599
>> Discussions of bugs and features can be posted here
>> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe
>> csound"
>>
>>
>
>
>
> Send bugs reports to the Sourceforge bug tracker
>           https://sourceforge.net/tracker/?group_id=81968&atid=564599
> Discussions of bugs and features can be posted here
> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe
> csound"
>
>



-- 


Andrés


Send bugs reports to the Sourceforge bug tracker
            https://sourceforge.net/tracker/?group_id=81968&atid=564599
Discussions of bugs and features can be posted here
To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"

Just for clarifying: The problems I reported occur with the "old"  
QuteCsound 0.4.6.1. This has nothing to do with your new developments.
The difference is between using Csound 5.11 or 5.12:
5.11 runs without problems with using portmidi; 5.12 behaves as I  
reported.
So I am asking myself: What has changed in Csound 5.12 that lets  
QuteCsound behave different?!?
If it's a general problem with portmidi, why hasn't it occurred in  
5.11 or earlier?

	joachim

Am 27.02.2010 um 10:27 schrieb Andres Cabrera:

> Hi,
>
> The issue is new, because previously, you couldn't run several
> instances of Csound simultaneously in QuteCsound, but now that you can
> this becomes an issue.
>
> I have a compile time macro that allows Csound to be reused or be
> destroyed after each run. I have messed with this recently, so I'm not
> sure what Joachim is testing with. It's currently set on svn to
> destroy Csound on every run, so the problem should only appear when
> trying to run a csd while another tab is running.
>
> I can confirm that Csound hangs when trying to run a second instance
> of Csound if using portmidi, but things are working fine on Linux if
> Csound is stopped before running a new csd.
>
> The hang (according to my debugger) occurs in pmidi.c line 252 when
> attempting to open the midi device.
>
> Cheers,
> Andrés
>
> On Fri, Feb 26, 2010 at 9:03 PM, joachim heintz  
>  wrote:
>> mmh, this is in so far different as I didn't have any problems with  
>> csound
>> 5.11.
>> but perhaps you used already 5.12 from cvs at this time?
>>
>>        joachim
>>
>>
>> Am 26.02.2010 um 21:48 schrieb Victor Lazzarini:
>>
>>> yes, I just checked my mailbox and I have reported this very same  
>>> problem
>>> with portmidi and qutecsound, on 16/Nov/09
>>>
>>> Victor
>>>
>>>
>>> On 26 Feb 2010, at 20:40, Victor Lazzarini wrote:
>>>
>>>> I think I have seen this before. I have the impression PortMIDI  
>>>> is not
>>>> reentrant and gives trouble like this.
>>>> I am sure I have reported a similar problem with qutecsound and  
>>>> portmidi.
>>>>
>>>> Victor
>>>> On 26 Feb 2010, at 19:59, joachim heintz wrote:
>>>>
>>>>> Hi all -
>>>>>
>>>>> I don't know whether this perhaps has been reported yet. I had  
>>>>> problems
>>>>> using Csound 5.12 with QCS (no problems with Csound 5.11), and I  
>>>>> saw now
>>>>> that this is related to PortMIDI. When I select (as I usuallly  
>>>>> do) PortMIDI
>>>>> with -a, I can run just ONE csd. Then, after some seconds, QCS  
>>>>> hangs, and I
>>>>> must (OSX) quit by violence.
>>>>>
>>>>> When I use "none" as MIDI Modul, no problem. Funny enough, also no
>>>>> problem with Virtual MIDI.
>>>>>
>>>>> By the way, Andrés: Console Printout is incredibly fast in the  
>>>>> last
>>>>> svn!!! Wow - looks like sorcery ...
>>>>>
>>>>>        joachim
>>>>>
>>>>> Send bugs reports to the Sourceforge bug tracker
>>>>>        https://sourceforge.net/tracker/?group_id=81968&atid=564599
>>>>> Discussions of bugs and features can be posted here
>>>>> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
>>>>> "unsubscribe
>>>>> csound"
>>>>>
>>>>
>>>>
>>>>
>>>> Send bugs reports to the Sourceforge bug tracker
>>>>         https://sourceforge.net/tracker/?group_id=81968&atid=564599
>>>> Discussions of bugs and features can be posted here
>>>> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
>>>> "unsubscribe
>>>> csound"
>>>>
>>>
>>>
>>>
>>> Send bugs reports to the Sourceforge bug tracker
>>>          https://sourceforge.net/tracker/?group_id=81968&atid=564599
>>> Discussions of bugs and features can be posted here
>>> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
>>> "unsubscribe
>>> csound"
>>>
>>>
>>
>>
>>
>> Send bugs reports to the Sourceforge bug tracker
>>           https://sourceforge.net/tracker/?group_id=81968&atid=564599
>> Discussions of bugs and features can be posted here
>> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
>> "unsubscribe
>> csound"
>>
>>
>
>
>
> -- 
>
>
> Andrés
>
>
> Send bugs reports to the Sourceforge bug tracker
>            https://sourceforge.net/tracker/?group_id=81968&atid=564599
> Discussions of bugs and features can be posted here
> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
> "unsubscribe csound"
>
>



Send bugs reports to the Sourceforge bug tracker
            https://sourceforge.net/tracker/?group_id=81968&atid=564599
Discussions of bugs and features can be posted here
To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"

Hi Joachim,

OH, OK, thanks for the clarification. Then something has probably
changed in portaudio, since pmidi.c in Csound was last modfied 14
months ago. Or there is some binary clash somewhere...

http://csound.cvs.sourceforge.net/viewvc/csound/csound5/InOut/

Cheers,
Andrés

On Sat, Feb 27, 2010 at 9:02 PM, joachim heintz  wrote:
> Just for clarifying: The problems I reported occur with the "old" QuteCsound
> 0.4.6.1. This has nothing to do with your new developments.
> The difference is between using Csound 5.11 or 5.12:
> 5.11 runs without problems with using portmidi; 5.12 behaves as I reported.
> So I am asking myself: What has changed in Csound 5.12 that lets QuteCsound
> behave different?!?
> If it's a general problem with portmidi, why hasn't it occurred in 5.11 or
> earlier?
>
>        joachim
>
> Am 27.02.2010 um 10:27 schrieb Andres Cabrera:
>
>> Hi,
>>
>> The issue is new, because previously, you couldn't run several
>> instances of Csound simultaneously in QuteCsound, but now that you can
>> this becomes an issue.
>>
>> I have a compile time macro that allows Csound to be reused or be
>> destroyed after each run. I have messed with this recently, so I'm not
>> sure what Joachim is testing with. It's currently set on svn to
>> destroy Csound on every run, so the problem should only appear when
>> trying to run a csd while another tab is running.
>>
>> I can confirm that Csound hangs when trying to run a second instance
>> of Csound if using portmidi, but things are working fine on Linux if
>> Csound is stopped before running a new csd.
>>
>> The hang (according to my debugger) occurs in pmidi.c line 252 when
>> attempting to open the midi device.
>>
>> Cheers,
>> Andrés
>>
>> On Fri, Feb 26, 2010 at 9:03 PM, joachim heintz 
>> wrote:
>>>
>>> mmh, this is in so far different as I didn't have any problems with
>>> csound
>>> 5.11.
>>> but perhaps you used already 5.12 from cvs at this time?
>>>
>>>       joachim
>>>
>>>
>>> Am 26.02.2010 um 21:48 schrieb Victor Lazzarini:
>>>
>>>> yes, I just checked my mailbox and I have reported this very same
>>>> problem
>>>> with portmidi and qutecsound, on 16/Nov/09
>>>>
>>>> Victor
>>>>
>>>>
>>>> On 26 Feb 2010, at 20:40, Victor Lazzarini wrote:
>>>>
>>>>> I think I have seen this before. I have the impression PortMIDI is not
>>>>> reentrant and gives trouble like this.
>>>>> I am sure I have reported a similar problem with qutecsound and
>>>>> portmidi.
>>>>>
>>>>> Victor
>>>>> On 26 Feb 2010, at 19:59, joachim heintz wrote:
>>>>>
>>>>>> Hi all -
>>>>>>
>>>>>> I don't know whether this perhaps has been reported yet. I had
>>>>>> problems
>>>>>> using Csound 5.12 with QCS (no problems with Csound 5.11), and I saw
>>>>>> now
>>>>>> that this is related to PortMIDI. When I select (as I usuallly do)
>>>>>> PortMIDI
>>>>>> with -a, I can run just ONE csd. Then, after some seconds, QCS hangs,
>>>>>> and I
>>>>>> must (OSX) quit by violence.
>>>>>>
>>>>>> When I use "none" as MIDI Modul, no problem. Funny enough, also no
>>>>>> problem with Virtual MIDI.
>>>>>>
>>>>>> By the way, Andrés: Console Printout is incredibly fast in the last
>>>>>> svn!!! Wow - looks like sorcery ...
>>>>>>
>>>>>>       joachim
>>>>>>
>>>>>> Send bugs reports to the Sourceforge bug tracker
>>>>>>       https://sourceforge.net/tracker/?group_id=81968&atid=564599
>>>>>> Discussions of bugs and features can be posted here
>>>>>> To unsubscribe, send email sympa@lists.bath.ac.uk with body
>>>>>> "unsubscribe
>>>>>> csound"
>>>>>>
>>>>>
>>>>>
>>>>>
>>>>> Send bugs reports to the Sourceforge bug tracker
>>>>>        https://sourceforge.net/tracker/?group_id=81968&atid=564599
>>>>> Discussions of bugs and features can be posted here
>>>>> To unsubscribe, send email sympa@lists.bath.ac.uk with body
>>>>> "unsubscribe
>>>>> csound"
>>>>>
>>>>
>>>>
>>>>
>>>> Send bugs reports to the Sourceforge bug tracker
>>>>         https://sourceforge.net/tracker/?group_id=81968&atid=564599
>>>> Discussions of bugs and features can be posted here
>>>> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe
>>>> csound"
>>>>
>>>>
>>>
>>>
>>>
>>> Send bugs reports to the Sourceforge bug tracker
>>>          https://sourceforge.net/tracker/?group_id=81968&atid=564599
>>> Discussions of bugs and features can be posted here
>>> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe
>>> csound"
>>>
>>>
>>
>>
>>
>> --
>>
>>
>> Andrés
>>
>>
>> Send bugs reports to the Sourceforge bug tracker
>>           https://sourceforge.net/tracker/?group_id=81968&atid=564599
>> Discussions of bugs and features can be posted here
>> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe
>> csound"
>>
>>
>
>
>
> Send bugs reports to the Sourceforge bug tracker
>           https://sourceforge.net/tracker/?group_id=81968&atid=564599
> Discussions of bugs and features can be posted here
> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe
> csound"
>
>



-- 


Andrés


Send bugs reports to the Sourceforge bug tracker
            https://sourceforge.net/tracker/?group_id=81968&atid=564599
Discussions of bugs and features can be posted here
To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"

Hi Andrés -

thanks for looking at the age of the current pmidi.c. It's strange  
that the problem occurs right now, with the change from 5.11 to 5.12.

This brings me back to a suggestion which came up many times here on  
the list. Why not to split the releases in a "stable" and a "beta"  
release, like many other free software do?

With Csound 5.12, we have run in several serious problems. Victor made  
2 bug fixes releases (thanks for this!). And as you said, no one could  
foresee that portmidi crashes (or hangs) now Csound. These are the  
weeks in which users test the new Csound in their practice and report  
problems and bugs. Why not declare this process as beta testing?

On one person who reads the list, we have hopefully many persons who  
use Csound wthout being a part of the community here. They just go on  
the sourceforge page, download the last version, follow the (now  
fantastic simple) installing instructions, and simply want to USE  
Csound. Nothing more, but that's what Csound is for. When they do so  
with the latest OSX installers (5.12.2), and they have chosen portmidi  
as MIDI modul with option -a (which is recommended), they see Csound  
hang after the first run. What will they do? They will probably try it  
some times, then they will say: Csound is buggy and instable, I have  
to look for something else for realizing my music.

That's my concern, and I don't fantasize but I speak about what I  
experienced with students and others. If there was a stable version,  
the usual user can choose whether he goes a rather safe way (because  
it's a version which has been tested for some time and if necessary  
major bugs have been fixed), or risks to use the beta version. If then  
something is happening like now with the portmidi issue, he knows: OK,  
i can go back to the stable version.

I ask the main developers if this cannot be a better way to estimate  
their work by everyone (in-list and off-list), and to give a clear  
advice and greatest possible safety to unexperienced users.

	joachim


Am 28.02.2010 um 10:06 schrieb Andres Cabrera:

> Hi Joachim,
>
> OH, OK, thanks for the clarification. Then something has probably
> changed in portaudio, since pmidi.c in Csound was last modfied 14
> months ago. Or there is some binary clash somewhere...
>
> http://csound.cvs.sourceforge.net/viewvc/csound/csound5/InOut/
>
> Cheers,
> Andrés
>
> On Sat, Feb 27, 2010 at 9:02 PM, joachim heintz  
>  wrote:
>> Just for clarifying: The problems I reported occur with the "old"  
>> QuteCsound
>> 0.4.6.1. This has nothing to do with your new developments.
>> The difference is between using Csound 5.11 or 5.12:
>> 5.11 runs without problems with using portmidi; 5.12 behaves as I  
>> reported.
>> So I am asking myself: What has changed in Csound 5.12 that lets  
>> QuteCsound
>> behave different?!?
>> If it's a general problem with portmidi, why hasn't it occurred in  
>> 5.11 or
>> earlier?
>>
>>        joachim
>>
>> Am 27.02.2010 um 10:27 schrieb Andres Cabrera:
>>
>>> Hi,
>>>
>>> The issue is new, because previously, you couldn't run several
>>> instances of Csound simultaneously in QuteCsound, but now that you  
>>> can
>>> this becomes an issue.
>>>
>>> I have a compile time macro that allows Csound to be reused or be
>>> destroyed after each run. I have messed with this recently, so I'm  
>>> not
>>> sure what Joachim is testing with. It's currently set on svn to
>>> destroy Csound on every run, so the problem should only appear when
>>> trying to run a csd while another tab is running.
>>>
>>> I can confirm that Csound hangs when trying to run a second instance
>>> of Csound if using portmidi, but things are working fine on Linux if
>>> Csound is stopped before running a new csd.
>>>
>>> The hang (according to my debugger) occurs in pmidi.c line 252 when
>>> attempting to open the midi device.
>>>
>>> Cheers,
>>> Andrés
>>>
>>> On Fri, Feb 26, 2010 at 9:03 PM, joachim heintz  
>>> 
>>> wrote:
>>>>
>>>> mmh, this is in so far different as I didn't have any problems with
>>>> csound
>>>> 5.11.
>>>> but perhaps you used already 5.12 from cvs at this time?
>>>>
>>>>       joachim
>>>>
>>>>
>>>> Am 26.02.2010 um 21:48 schrieb Victor Lazzarini:
>>>>
>>>>> yes, I just checked my mailbox and I have reported this very same
>>>>> problem
>>>>> with portmidi and qutecsound, on 16/Nov/09
>>>>>
>>>>> Victor
>>>>>
>>>>>
>>>>> On 26 Feb 2010, at 20:40, Victor Lazzarini wrote:
>>>>>
>>>>>> I think I have seen this before. I have the impression PortMIDI  
>>>>>> is not
>>>>>> reentrant and gives trouble like this.
>>>>>> I am sure I have reported a similar problem with qutecsound and
>>>>>> portmidi.
>>>>>>
>>>>>> Victor
>>>>>> On 26 Feb 2010, at 19:59, joachim heintz wrote:
>>>>>>
>>>>>>> Hi all -
>>>>>>>
>>>>>>> I don't know whether this perhaps has been reported yet. I had
>>>>>>> problems
>>>>>>> using Csound 5.12 with QCS (no problems with Csound 5.11), and  
>>>>>>> I saw
>>>>>>> now
>>>>>>> that this is related to PortMIDI. When I select (as I usuallly  
>>>>>>> do)
>>>>>>> PortMIDI
>>>>>>> with -a, I can run just ONE csd. Then, after some seconds, QCS  
>>>>>>> hangs,
>>>>>>> and I
>>>>>>> must (OSX) quit by violence.
>>>>>>>
>>>>>>> When I use "none" as MIDI Modul, no problem. Funny enough,  
>>>>>>> also no
>>>>>>> problem with Virtual MIDI.
>>>>>>>
>>>>>>> By the way, Andrés: Console Printout is incredibly fast in the  
>>>>>>> last
>>>>>>> svn!!! Wow - looks like sorcery ...
>>>>>>>
>>>>>>>       joachim
>>>>>>>
>>>>>>> Send bugs reports to the Sourceforge bug tracker
>>>>>>>       https://sourceforge.net/tracker/? 
>>>>>>> group_id=81968&atid=564599
>>>>>>> Discussions of bugs and features can be posted here
>>>>>>> To unsubscribe, send email sympa@lists.bath.ac.uk with body
>>>>>>> "unsubscribe
>>>>>>> csound"
>>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Send bugs reports to the Sourceforge bug tracker
>>>>>>        https://sourceforge.net/tracker/? 
>>>>>> group_id=81968&atid=564599
>>>>>> Discussions of bugs and features can be posted here
>>>>>> To unsubscribe, send email sympa@lists.bath.ac.uk with body
>>>>>> "unsubscribe
>>>>>> csound"
>>>>>>
>>>>>
>>>>>
>>>>>
>>>>> Send bugs reports to the Sourceforge bug tracker
>>>>>         https://sourceforge.net/tracker/? 
>>>>> group_id=81968&atid=564599
>>>>> Discussions of bugs and features can be posted here
>>>>> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
>>>>> "unsubscribe
>>>>> csound"
>>>>>
>>>>>
>>>>
>>>>
>>>>
>>>> Send bugs reports to the Sourceforge bug tracker
>>>>          https://sourceforge.net/tracker/? 
>>>> group_id=81968&atid=564599
>>>> Discussions of bugs and features can be posted here
>>>> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
>>>> "unsubscribe
>>>> csound"
>>>>
>>>>
>>>
>>>
>>>
>>> --
>>>
>>>
>>> Andrés
>>>
>>>
>>> Send bugs reports to the Sourceforge bug tracker
>>>           https://sourceforge.net/tracker/? 
>>> group_id=81968&atid=564599
>>> Discussions of bugs and features can be posted here
>>> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
>>> "unsubscribe
>>> csound"
>>>
>>>
>>
>>
>>
>> Send bugs reports to the Sourceforge bug tracker
>>           https://sourceforge.net/tracker/?group_id=81968&atid=564599
>> Discussions of bugs and features can be posted here
>> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
>> "unsubscribe
>> csound"
>>
>>
>
>
>
> -- 
>
>
> Andrés
>
>
> Send bugs reports to the Sourceforge bug tracker
>            https://sourceforge.net/tracker/?group_id=81968&atid=564599
> Discussions of bugs and features can be posted here
> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
> "unsubscribe csound"
>
>



Send bugs reports to the Sourceforge bug tracker
            https://sourceforge.net/tracker/?group_id=81968&atid=564599
Discussions of bugs and features can be posted here
To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"

But the problem with MIDI is not 5.12 related. In my machine, it  
happened with 5.11 and I have reported it here.
In fact the problem with MIDI is not with the Csound code, but  
portmidi.  This is not solvable that easily. The best thing is to  
place, as with many software projects, a list of known bugs.

Victor


On 28 Feb 2010, at 11:36, joachim heintz wrote:

> Hi Andrés -
>
> thanks for looking at the age of the current pmidi.c. It's strange  
> that the problem occurs right now, with the change from 5.11 to 5.12.
>
> This brings me back to a suggestion which came up many times here on  
> the list. Why not to split the releases in a "stable" and a "beta"  
> release, like many other free software do?
>
> With Csound 5.12, we have run in several serious problems. Victor  
> made 2 bug fixes releases (thanks for this!). And as you said, no  
> one could foresee that portmidi crashes (or hangs) now Csound. These  
> are the weeks in which users test the new Csound in their practice  
> and report problems and bugs. Why not declare this process as beta  
> testing?
>
> On one person who reads the list, we have hopefully many persons who  
> use Csound wthout being a part of the community here. They just go  
> on the sourceforge page, download the last version, follow the (now  
> fantastic simple) installing instructions, and simply want to USE  
> Csound. Nothing more, but that's what Csound is for. When they do so  
> with the latest OSX installers (5.12.2), and they have chosen  
> portmidi as MIDI modul with option -a (which is recommended), they  
> see Csound hang after the first run. What will they do? They will  
> probably try it some times, then they will say: Csound is buggy and  
> instable, I have to look for something else for realizing my music.
>
> That's my concern, and I don't fantasize but I speak about what I  
> experienced with students and others. If there was a stable version,  
> the usual user can choose whether he goes a rather safe way (because  
> it's a version which has been tested for some time and if necessary  
> major bugs have been fixed), or risks to use the beta version. If  
> then something is happening like now with the portmidi issue, he  
> knows: OK, i can go back to the stable version.
>
> I ask the main developers if this cannot be a better way to estimate  
> their work by everyone (in-list and off-list), and to give a clear  
> advice and greatest possible safety to unexperienced users.
>
> 	joachim
>
>
> Am 28.02.2010 um 10:06 schrieb Andres Cabrera:
>
>> Hi Joachim,
>>
>> OH, OK, thanks for the clarification. Then something has probably
>> changed in portaudio, since pmidi.c in Csound was last modfied 14
>> months ago. Or there is some binary clash somewhere...
>>
>> http://csound.cvs.sourceforge.net/viewvc/csound/csound5/InOut/
>>
>> Cheers,
>> Andrés
>>
>> On Sat, Feb 27, 2010 at 9:02 PM, joachim heintz  
>>  wrote:
>>> Just for clarifying: The problems I reported occur with the "old"  
>>> QuteCsound
>>> 0.4.6.1. This has nothing to do with your new developments.
>>> The difference is between using Csound 5.11 or 5.12:
>>> 5.11 runs without problems with using portmidi; 5.12 behaves as I  
>>> reported.
>>> So I am asking myself: What has changed in Csound 5.12 that lets  
>>> QuteCsound
>>> behave different?!?
>>> If it's a general problem with portmidi, why hasn't it occurred in  
>>> 5.11 or
>>> earlier?
>>>
>>>       joachim
>>>
>>> Am 27.02.2010 um 10:27 schrieb Andres Cabrera:
>>>
>>>> Hi,
>>>>
>>>> The issue is new, because previously, you couldn't run several
>>>> instances of Csound simultaneously in QuteCsound, but now that  
>>>> you can
>>>> this becomes an issue.
>>>>
>>>> I have a compile time macro that allows Csound to be reused or be
>>>> destroyed after each run. I have messed with this recently, so  
>>>> I'm not
>>>> sure what Joachim is testing with. It's currently set on svn to
>>>> destroy Csound on every run, so the problem should only appear when
>>>> trying to run a csd while another tab is running.
>>>>
>>>> I can confirm that Csound hangs when trying to run a second  
>>>> instance
>>>> of Csound if using portmidi, but things are working fine on Linux  
>>>> if
>>>> Csound is stopped before running a new csd.
>>>>
>>>> The hang (according to my debugger) occurs in pmidi.c line 252 when
>>>> attempting to open the midi device.
>>>>
>>>> Cheers,
>>>> Andrés
>>>>
>>>> On Fri, Feb 26, 2010 at 9:03 PM, joachim heintz >>> >
>>>> wrote:
>>>>>
>>>>> mmh, this is in so far different as I didn't have any problems  
>>>>> with
>>>>> csound
>>>>> 5.11.
>>>>> but perhaps you used already 5.12 from cvs at this time?
>>>>>
>>>>>      joachim
>>>>>
>>>>>
>>>>> Am 26.02.2010 um 21:48 schrieb Victor Lazzarini:
>>>>>
>>>>>> yes, I just checked my mailbox and I have reported this very same
>>>>>> problem
>>>>>> with portmidi and qutecsound, on 16/Nov/09
>>>>>>
>>>>>> Victor
>>>>>>
>>>>>>
>>>>>> On 26 Feb 2010, at 20:40, Victor Lazzarini wrote:
>>>>>>
>>>>>>> I think I have seen this before. I have the impression  
>>>>>>> PortMIDI is not
>>>>>>> reentrant and gives trouble like this.
>>>>>>> I am sure I have reported a similar problem with qutecsound and
>>>>>>> portmidi.
>>>>>>>
>>>>>>> Victor
>>>>>>> On 26 Feb 2010, at 19:59, joachim heintz wrote:
>>>>>>>
>>>>>>>> Hi all -
>>>>>>>>
>>>>>>>> I don't know whether this perhaps has been reported yet. I had
>>>>>>>> problems
>>>>>>>> using Csound 5.12 with QCS (no problems with Csound 5.11),  
>>>>>>>> and I saw
>>>>>>>> now
>>>>>>>> that this is related to PortMIDI. When I select (as I  
>>>>>>>> usuallly do)
>>>>>>>> PortMIDI
>>>>>>>> with -a, I can run just ONE csd. Then, after some seconds,  
>>>>>>>> QCS hangs,
>>>>>>>> and I
>>>>>>>> must (OSX) quit by violence.
>>>>>>>>
>>>>>>>> When I use "none" as MIDI Modul, no problem. Funny enough,  
>>>>>>>> also no
>>>>>>>> problem with Virtual MIDI.
>>>>>>>>
>>>>>>>> By the way, Andrés: Console Printout is incredibly fast in  
>>>>>>>> the last
>>>>>>>> svn!!! Wow - looks like sorcery ...
>>>>>>>>
>>>>>>>>      joachim
>>>>>>>>
>>>>>>>> Send bugs reports to the Sourceforge bug tracker
>>>>>>>>      https://sourceforge.net/tracker/? 
>>>>>>>> group_id=81968&atid=564599
>>>>>>>> Discussions of bugs and features can be posted here
>>>>>>>> To unsubscribe, send email sympa@lists.bath.ac.uk with body
>>>>>>>> "unsubscribe
>>>>>>>> csound"
>>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> Send bugs reports to the Sourceforge bug tracker
>>>>>>>       https://sourceforge.net/tracker/? 
>>>>>>> group_id=81968&atid=564599
>>>>>>> Discussions of bugs and features can be posted here
>>>>>>> To unsubscribe, send email sympa@lists.bath.ac.uk with body
>>>>>>> "unsubscribe
>>>>>>> csound"
>>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Send bugs reports to the Sourceforge bug tracker
>>>>>>        https://sourceforge.net/tracker/? 
>>>>>> group_id=81968&atid=564599
>>>>>> Discussions of bugs and features can be posted here
>>>>>> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
>>>>>> "unsubscribe
>>>>>> csound"
>>>>>>
>>>>>>
>>>>>
>>>>>
>>>>>
>>>>> Send bugs reports to the Sourceforge bug tracker
>>>>>         https://sourceforge.net/tracker/? 
>>>>> group_id=81968&atid=564599
>>>>> Discussions of bugs and features can be posted here
>>>>> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
>>>>> "unsubscribe
>>>>> csound"
>>>>>
>>>>>
>>>>
>>>>
>>>>
>>>> --
>>>>
>>>>
>>>> Andrés
>>>>
>>>>
>>>> Send bugs reports to the Sourceforge bug tracker
>>>>          https://sourceforge.net/tracker/? 
>>>> group_id=81968&atid=564599
>>>> Discussions of bugs and features can be posted here
>>>> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
>>>> "unsubscribe
>>>> csound"
>>>>
>>>>
>>>
>>>
>>>
>>> Send bugs reports to the Sourceforge bug tracker
>>>          https://sourceforge.net/tracker/?group_id=81968&atid=564599
>>> Discussions of bugs and features can be posted here
>>> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
>>> "unsubscribe
>>> csound"
>>>
>>>
>>
>>
>>
>> -- 
>>
>>
>> Andrés
>>
>>
>> Send bugs reports to the Sourceforge bug tracker
>>           https://sourceforge.net/tracker/?group_id=81968&atid=564599
>> Discussions of bugs and features can be posted here
>> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
>> "unsubscribe csound"
>>
>>
>
>
>
> Send bugs reports to the Sourceforge bug tracker
>           https://sourceforge.net/tracker/?group_id=81968&atid=564599
> Discussions of bugs and features can be posted here
> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
> "unsubscribe csound"
>



Send bugs reports to the Sourceforge bug tracker
            https://sourceforge.net/tracker/?group_id=81968&atid=564599
Discussions of bugs and features can be posted here
To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"

I do agree wholeheartedly to the suggestion of "stable" and
"development" release versions.
The developers need to release to enable users to test and comment,
but it would be better for a users to be able to select (or even
switch) between stable and dev versions. Bugs are deemed to be
introduced with new (dev) releases, and the current situation will
eventually turn users away from our great tool.
Oeyvind

2010/2/28 joachim heintz :
>
> This brings me back to a suggestion which came up many times here on the
> list. Why not to split the releases in a "stable" and a "beta" release, like
> many other free software do?


Send bugs reports to the Sourceforge bug tracker
            https://sourceforge.net/tracker/?group_id=81968&atid=564599
Discussions of bugs and features can be posted here
To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"

Just to note, I had advocated both a test suite as well as a beta
release phase.  The former still does not seem to be used regularly by
anyone but myself as evidenced by bugs being marked fixed without a
test being committed to the suite.  The latter was tried out for a
release or two but later dropped for reasons of it being cumbersome or
not catching bugs as we did not get many people testing out the beta.

I would suggest strongly:

1. For bugs, we only accept bug reports with CSD's that can evidence
the break (not just code fragments).  Developers could then add it to
the test suite, run and have it fail, then fix until the test can
pass, with a final run of the test suite to test that nothing else has
broken.  I would strongly recommend only full CSD's be accepted for
bug reports as a means for a user to fully describe the problem and so
the developers aren't having to spend any more time than necessary to
fix the bug.

2. We begin again with beta releases.  Unlike last time, we should
start to take advantage of CVS's features and when we start a beta, we
should branch the trunk at beta time.  We issue the beta from the
branch, and if any fixes are necessary, we do the fix in the branch
and release from there.  We can re-merge any work into trunk.

3. Developers can just work with trunk.  When the branch occurs to
start the release cycle (betas -> release), development can continue
on the trunk for new features.

4. Users will need more visibility to betas (perhaps add a notice to
csounds.com as well as on csound.sourceforge.net).

There would be more work necessary on the test harness to allow for
running with either old or new parser, as well as allowing for
different suites of tests.

Thoughts?
steven


On Mon, Mar 1, 2010 at 3:21 AM, Oeyvind Brandtsegg  wrote:
> I do agree wholeheartedly to the suggestion of "stable" and
> "development" release versions.
> The developers need to release to enable users to test and comment,
> but it would be better for a users to be able to select (or even
> switch) between stable and dev versions. Bugs are deemed to be
> introduced with new (dev) releases, and the current situation will
> eventually turn users away from our great tool.
> Oeyvind
>
> 2010/2/28 joachim heintz :
>>
>> This brings me back to a suggestion which came up many times here on the
>> list. Why not to split the releases in a "stable" and a "beta" release, like
>> many other free software do?
>
>
> Send bugs reports to the Sourceforge bug tracker
>            https://sourceforge.net/tracker/?group_id=81968&atid=564599
> Discussions of bugs and features can be posted here
> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"
>
>


Send bugs reports to the Sourceforge bug tracker
            https://sourceforge.net/tracker/?group_id=81968&atid=564599
Discussions of bugs and features can be posted here
To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"

Hello Victor -

as to the MIDI problem, and whether it's related to 5.12 or not, I can  
just repeat to discribe the following:
1. When I run QuteCsound 0.4.6.1 with Csound 5.12.2, and I use  
portaudio, it hangs after finishing. This means that I can't use any  
MIDI input.
2. When I reinstall Csound 5.11, everything is fine.
If this is related to my machine, it's just my problem. If others have  
the same problem, it's related to an issue between QuteCsound 0.4.6.1  
and Csound 5.12. This means that both applications provided in the Mac  
installer can't come together in this point.

You analyzed the problem (portmidi not re-entrant) and as many others  
I do appreciate your care and support very much. But my question was  
more general: How can we point the UNexperienced user to a stable,  
long tested version? For the Mac, as far as I see, QuteCsound 0.4.6.1  
and Csound 5.11 are such a stable package. But Csound 5.11 comes with  
QCS 0.4.5 (which has some problems), and Csound 5.12 comes with QCS  
0.4.6.1 (which has at least for me the reported problems).

Would it be an idea to pack QCS 0.4.6.1 with Csound 5.11 and call it  
"Latest stable version for OSX (intel)"? And the same for Windows?  
Andrés is developing rapidly QCS, so we will surely experience many  
problems at surprising places, like this portmidi issue (though there  
was no change in portmidi since 14  month).

I think: If you - the developers - release from time to time (say once  
or twice a year) a long tested, stable version, you are more free,  
because when then any issues appear in the development version, these  
are issues of the development version, not of the stable version.

Regards -

	joachim


Am 28.02.2010 um 16:17 schrieb Victor Lazzarini:

> But the problem with MIDI is not 5.12 related. In my machine, it  
> happened with 5.11 and I have reported it here.
> In fact the problem with MIDI is not with the Csound code, but  
> portmidi.  This is not solvable that easily. The best thing is to  
> place, as with many software projects, a list of known bugs.
>
> Victor
>
>
> On 28 Feb 2010, at 11:36, joachim heintz wrote:
>
>> Hi Andrés -
>>
>> thanks for looking at the age of the current pmidi.c. It's strange  
>> that the problem occurs right now, with the change from 5.11 to 5.12.
>>
>> This brings me back to a suggestion which came up many times here  
>> on the list. Why not to split the releases in a "stable" and a  
>> "beta" release, like many other free software do?
>>
>> With Csound 5.12, we have run in several serious problems. Victor  
>> made 2 bug fixes releases (thanks for this!). And as you said, no  
>> one could foresee that portmidi crashes (or hangs) now Csound.  
>> These are the weeks in which users test the new Csound in their  
>> practice and report problems and bugs. Why not declare this process  
>> as beta testing?
>>
>> On one person who reads the list, we have hopefully many persons  
>> who use Csound wthout being a part of the community here. They just  
>> go on the sourceforge page, download the last version, follow the  
>> (now fantastic simple) installing instructions, and simply want to  
>> USE Csound. Nothing more, but that's what Csound is for. When they  
>> do so with the latest OSX installers (5.12.2), and they have chosen  
>> portmidi as MIDI modul with option -a (which is recommended), they  
>> see Csound hang after the first run. What will they do? They will  
>> probably try it some times, then they will say: Csound is buggy and  
>> instable, I have to look for something else for realizing my music.
>>
>> That's my concern, and I don't fantasize but I speak about what I  
>> experienced with students and others. If there was a stable  
>> version, the usual user can choose whether he goes a rather safe  
>> way (because it's a version which has been tested for some time and  
>> if necessary major bugs have been fixed), or risks to use the beta  
>> version. If then something is happening like now with the portmidi  
>> issue, he knows: OK, i can go back to the stable version.
>>
>> I ask the main developers if this cannot be a better way to  
>> estimate their work by everyone (in-list and off-list), and to give  
>> a clear advice and greatest possible safety to unexperienced users.
>>
>> 	joachim
>>
>>
>> Am 28.02.2010 um 10:06 schrieb Andres Cabrera:
>>
>>> Hi Joachim,
>>>
>>> OH, OK, thanks for the clarification. Then something has probably
>>> changed in portaudio, since pmidi.c in Csound was last modfied 14
>>> months ago. Or there is some binary clash somewhere...
>>>
>>> http://csound.cvs.sourceforge.net/viewvc/csound/csound5/InOut/
>>>
>>> Cheers,
>>> Andrés
>>>
>>> On Sat, Feb 27, 2010 at 9:02 PM, joachim heintz  
>>>  wrote:
>>>> Just for clarifying: The problems I reported occur with the "old"  
>>>> QuteCsound
>>>> 0.4.6.1. This has nothing to do with your new developments.
>>>> The difference is between using Csound 5.11 or 5.12:
>>>> 5.11 runs without problems with using portmidi; 5.12 behaves as I  
>>>> reported.
>>>> So I am asking myself: What has changed in Csound 5.12 that lets  
>>>> QuteCsound
>>>> behave different?!?
>>>> If it's a general problem with portmidi, why hasn't it occurred  
>>>> in 5.11 or
>>>> earlier?
>>>>
>>>>      joachim
>>>>
>>>> Am 27.02.2010 um 10:27 schrieb Andres Cabrera:
>>>>
>>>>> Hi,
>>>>>
>>>>> The issue is new, because previously, you couldn't run several
>>>>> instances of Csound simultaneously in QuteCsound, but now that  
>>>>> you can
>>>>> this becomes an issue.
>>>>>
>>>>> I have a compile time macro that allows Csound to be reused or be
>>>>> destroyed after each run. I have messed with this recently, so  
>>>>> I'm not
>>>>> sure what Joachim is testing with. It's currently set on svn to
>>>>> destroy Csound on every run, so the problem should only appear  
>>>>> when
>>>>> trying to run a csd while another tab is running.
>>>>>
>>>>> I can confirm that Csound hangs when trying to run a second  
>>>>> instance
>>>>> of Csound if using portmidi, but things are working fine on  
>>>>> Linux if
>>>>> Csound is stopped before running a new csd.
>>>>>
>>>>> The hang (according to my debugger) occurs in pmidi.c line 252  
>>>>> when
>>>>> attempting to open the midi device.
>>>>>
>>>>> Cheers,
>>>>> Andrés
>>>>>
>>>>> On Fri, Feb 26, 2010 at 9:03 PM, joachim heintz >>>> >
>>>>> wrote:
>>>>>>
>>>>>> mmh, this is in so far different as I didn't have any problems  
>>>>>> with
>>>>>> csound
>>>>>> 5.11.
>>>>>> but perhaps you used already 5.12 from cvs at this time?
>>>>>>
>>>>>>     joachim
>>>>>>
>>>>>>
>>>>>> Am 26.02.2010 um 21:48 schrieb Victor Lazzarini:
>>>>>>
>>>>>>> yes, I just checked my mailbox and I have reported this very  
>>>>>>> same
>>>>>>> problem
>>>>>>> with portmidi and qutecsound, on 16/Nov/09
>>>>>>>
>>>>>>> Victor
>>>>>>>
>>>>>>>
>>>>>>> On 26 Feb 2010, at 20:40, Victor Lazzarini wrote:
>>>>>>>
>>>>>>>> I think I have seen this before. I have the impression  
>>>>>>>> PortMIDI is not
>>>>>>>> reentrant and gives trouble like this.
>>>>>>>> I am sure I have reported a similar problem with qutecsound and
>>>>>>>> portmidi.
>>>>>>>>
>>>>>>>> Victor
>>>>>>>> On 26 Feb 2010, at 19:59, joachim heintz wrote:
>>>>>>>>
>>>>>>>>> Hi all -
>>>>>>>>>
>>>>>>>>> I don't know whether this perhaps has been reported yet. I had
>>>>>>>>> problems
>>>>>>>>> using Csound 5.12 with QCS (no problems with Csound 5.11),  
>>>>>>>>> and I saw
>>>>>>>>> now
>>>>>>>>> that this is related to PortMIDI. When I select (as I  
>>>>>>>>> usuallly do)
>>>>>>>>> PortMIDI
>>>>>>>>> with -a, I can run just ONE csd. Then, after some seconds,  
>>>>>>>>> QCS hangs,
>>>>>>>>> and I
>>>>>>>>> must (OSX) quit by violence.
>>>>>>>>>
>>>>>>>>> When I use "none" as MIDI Modul, no problem. Funny enough,  
>>>>>>>>> also no
>>>>>>>>> problem with Virtual MIDI.
>>>>>>>>>
>>>>>>>>> By the way, Andrés: Console Printout is incredibly fast in  
>>>>>>>>> the last
>>>>>>>>> svn!!! Wow - looks like sorcery ...
>>>>>>>>>
>>>>>>>>>     joachim
>>>>>>>>>
>>>>>>>>> Send bugs reports to the Sourceforge bug tracker
>>>>>>>>>     https://sourceforge.net/tracker/? 
>>>>>>>>> group_id=81968&atid=564599
>>>>>>>>> Discussions of bugs and features can be posted here
>>>>>>>>> To unsubscribe, send email sympa@lists.bath.ac.uk with body
>>>>>>>>> "unsubscribe
>>>>>>>>> csound"
>>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> Send bugs reports to the Sourceforge bug tracker
>>>>>>>>      https://sourceforge.net/tracker/? 
>>>>>>>> group_id=81968&atid=564599
>>>>>>>> Discussions of bugs and features can be posted here
>>>>>>>> To unsubscribe, send email sympa@lists.bath.ac.uk with body
>>>>>>>> "unsubscribe
>>>>>>>> csound"
>>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> Send bugs reports to the Sourceforge bug tracker
>>>>>>>       https://sourceforge.net/tracker/? 
>>>>>>> group_id=81968&atid=564599
>>>>>>> Discussions of bugs and features can be posted here
>>>>>>> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
>>>>>>> "unsubscribe
>>>>>>> csound"
>>>>>>>
>>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Send bugs reports to the Sourceforge bug tracker
>>>>>>        https://sourceforge.net/tracker/? 
>>>>>> group_id=81968&atid=564599
>>>>>> Discussions of bugs and features can be posted here
>>>>>> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
>>>>>> "unsubscribe
>>>>>> csound"
>>>>>>
>>>>>>
>>>>>
>>>>>
>>>>>
>>>>> --
>>>>>
>>>>>
>>>>> Andrés
>>>>>
>>>>>
>>>>> Send bugs reports to the Sourceforge bug tracker
>>>>>         https://sourceforge.net/tracker/? 
>>>>> group_id=81968&atid=564599
>>>>> Discussions of bugs and features can be posted here
>>>>> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
>>>>> "unsubscribe
>>>>> csound"
>>>>>
>>>>>
>>>>
>>>>
>>>>
>>>> Send bugs reports to the Sourceforge bug tracker
>>>>         https://sourceforge.net/tracker/?group_id=81968&atid=564599
>>>> Discussions of bugs and features can be posted here
>>>> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
>>>> "unsubscribe
>>>> csound"
>>>>
>>>>
>>>
>>>
>>>
>>> -- 
>>>
>>>
>>> Andrés
>>>
>>>
>>> Send bugs reports to the Sourceforge bug tracker
>>>          https://sourceforge.net/tracker/?group_id=81968&atid=564599
>>> Discussions of bugs and features can be posted here
>>> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
>>> "unsubscribe csound"
>>>
>>>
>>
>>
>>
>> Send bugs reports to the Sourceforge bug tracker
>>          https://sourceforge.net/tracker/?group_id=81968&atid=564599
>> Discussions of bugs and features can be posted here
>> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
>> "unsubscribe csound"
>>
>
>
>
> Send bugs reports to the Sourceforge bug tracker
>           https://sourceforge.net/tracker/?group_id=81968&atid=564599
> Discussions of bugs and features can be posted here
> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
> "unsubscribe csound"
>
>



Send bugs reports to the Sourceforge bug tracker
            https://sourceforge.net/tracker/?group_id=81968&atid=564599
Discussions of bugs and features can be posted here
To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"

I agree with everything Steven says except branching. I would prefer
"stable" releases and "beta" releases, even and odd numbered.

Regards,
Mike

On Mon, Mar 1, 2010 at 4:06 AM, Steven Yi  wrote:
> Just to note, I had advocated both a test suite as well as a beta
> release phase.  The former still does not seem to be used regularly by
> anyone but myself as evidenced by bugs being marked fixed without a
> test being committed to the suite.  The latter was tried out for a
> release or two but later dropped for reasons of it being cumbersome or
> not catching bugs as we did not get many people testing out the beta.
>
> I would suggest strongly:
>
> 1. For bugs, we only accept bug reports with CSD's that can evidence
> the break (not just code fragments).  Developers could then add it to
> the test suite, run and have it fail, then fix until the test can
> pass, with a final run of the test suite to test that nothing else has
> broken.  I would strongly recommend only full CSD's be accepted for
> bug reports as a means for a user to fully describe the problem and so
> the developers aren't having to spend any more time than necessary to
> fix the bug.
>
> 2. We begin again with beta releases.  Unlike last time, we should
> start to take advantage of CVS's features and when we start a beta, we
> should branch the trunk at beta time.  We issue the beta from the
> branch, and if any fixes are necessary, we do the fix in the branch
> and release from there.  We can re-merge any work into trunk.
>
> 3. Developers can just work with trunk.  When the branch occurs to
> start the release cycle (betas -> release), development can continue
> on the trunk for new features.
>
> 4. Users will need more visibility to betas (perhaps add a notice to
> csounds.com as well as on csound.sourceforge.net).
>
> There would be more work necessary on the test harness to allow for
> running with either old or new parser, as well as allowing for
> different suites of tests.
>
> Thoughts?
> steven
>
>
> On Mon, Mar 1, 2010 at 3:21 AM, Oeyvind Brandtsegg  wrote:
>> I do agree wholeheartedly to the suggestion of "stable" and
>> "development" release versions.
>> The developers need to release to enable users to test and comment,
>> but it would be better for a users to be able to select (or even
>> switch) between stable and dev versions. Bugs are deemed to be
>> introduced with new (dev) releases, and the current situation will
>> eventually turn users away from our great tool.
>> Oeyvind
>>
>> 2010/2/28 joachim heintz :
>>>
>>> This brings me back to a suggestion which came up many times here on the
>>> list. Why not to split the releases in a "stable" and a "beta" release, like
>>> many other free software do?
>>
>>
>> Send bugs reports to the Sourceforge bug tracker
>>            https://sourceforge.net/tracker/?group_id=81968&atid=564599
>> Discussions of bugs and features can be posted here
>> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"
>>
>>
>
>
> Send bugs reports to the Sourceforge bug tracker
>            https://sourceforge.net/tracker/?group_id=81968&atid=564599
> Discussions of bugs and features can be posted here
> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"
>
>



-- 
Michael Gogins
Irreducible Productions
http://www.michael-gogins.com
Michael dot Gogins at gmail dot com


Send bugs reports to the Sourceforge bug tracker
            https://sourceforge.net/tracker/?group_id=81968&atid=564599
Discussions of bugs and features can be posted here
To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"

On Mon, Mar 1, 2010 at 06:06, Steven Yi  wrote:
> 2. We begin again with beta releases.  Unlike last time, we should
> start to take advantage of CVS's features and when we start a beta, we
> should branch the trunk at beta time.  We issue the beta from the
> branch, and if any fixes are necessary, we do the fix in the branch
> and release from there.  We can re-merge any work into trunk.

Ugh. CVS is notoriously cumbersome to work with branches. I would
propose a different workflow. When a beta release is tagged, a code
freeze is started. No new code can be added, only bugs can be fixed.
That way, although it slows development a bit, you do not have to deal
with branches, etc.

-- 

Saludos,
Felipe Sateler


Send bugs reports to the Sourceforge bug tracker
            https://sourceforge.net/tracker/?group_id=81968&atid=564599
Discussions of bugs and features can be posted here
To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"

Yes, I agree with this.

Regards,
MIke

On Mon, Mar 1, 2010 at 9:51 AM, Felipe Sateler  wrote:
> On Mon, Mar 1, 2010 at 06:06, Steven Yi  wrote:
>> 2. We begin again with beta releases.  Unlike last time, we should
>> start to take advantage of CVS's features and when we start a beta, we
>> should branch the trunk at beta time.  We issue the beta from the
>> branch, and if any fixes are necessary, we do the fix in the branch
>> and release from there.  We can re-merge any work into trunk.
>
> Ugh. CVS is notoriously cumbersome to work with branches. I would
> propose a different workflow. When a beta release is tagged, a code
> freeze is started. No new code can be added, only bugs can be fixed.
> That way, although it slows development a bit, you do not have to deal
> with branches, etc.
>
> --
>
> Saludos,
> Felipe Sateler
>
>
> Send bugs reports to the Sourceforge bug tracker
>            https://sourceforge.net/tracker/?group_id=81968&atid=564599
> Discussions of bugs and features can be posted here
> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"
>
>



-- 
Michael Gogins
Irreducible Productions
http://www.michael-gogins.com
Michael dot Gogins at gmail dot com


Send bugs reports to the Sourceforge bug tracker
            https://sourceforge.net/tracker/?group_id=81968&atid=564599
Discussions of bugs and features can be posted here
To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"

True, CVS branching is not the best. :)  The workflow is something I
have used with SVN and it's not so bad with CVS I think, but with the
Csound project I think it's fine to do code freezes.

On Mon, Mar 1, 2010 at 9:51 AM, Felipe Sateler  wrote:
> On Mon, Mar 1, 2010 at 06:06, Steven Yi  wrote:
>> 2. We begin again with beta releases.  Unlike last time, we should
>> start to take advantage of CVS's features and when we start a beta, we
>> should branch the trunk at beta time.  We issue the beta from the
>> branch, and if any fixes are necessary, we do the fix in the branch
>> and release from there.  We can re-merge any work into trunk.
>
> Ugh. CVS is notoriously cumbersome to work with branches. I would
> propose a different workflow. When a beta release is tagged, a code
> freeze is started. No new code can be added, only bugs can be fixed.
> That way, although it slows development a bit, you do not have to deal
> with branches, etc.
>
> --
>
> Saludos,
> Felipe Sateler
>
>
> Send bugs reports to the Sourceforge bug tracker
>            https://sourceforge.net/tracker/?group_id=81968&atid=564599
> Discussions of bugs and features can be posted here
> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"
>
>


Send bugs reports to the Sourceforge bug tracker
            https://sourceforge.net/tracker/?group_id=81968&atid=564599
Discussions of bugs and features can be posted here
To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"