Quantcast

P1s holding up 2.1.3 Release

classic Classic list List threaded Threaded
22 messages Options
12
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

P1s holding up 2.1.3 Release

Peter Sampson-2
I am getting increasingly concerned that we seem to be making poor progress currently
towards 2.1.3 Release - and note that today is the one-year anniversary of the release of
2.1.2 - not an anniversary to be proud of I'm thinking.

We are currently sitting with three P1s that are blocking release:

1304:  http://bugzilla.audacityteam.org/show_bug.cgi?id=1304
This remains still listed at P1 although it tests ok on all three platforms for the
originally reported issues.  It either needs degrading to P3 for the newly commented
upon residual issues, as Gale himself suggested - or closing as resolved and a fresh
P3 bug opened for the residual issues.
So this one should not be a blocker.


1577:  http://bugzilla.audacityteam.org/show_bug.cgi?id=1577
This is the one introduced by James' capitalization of Audacity in a Mac menu recently.
Kames is working on a fix for this with high confidence it will work
If it does not, then I suggest that we just  revert this change and that we live with the lower
case "audacity" in that menu, carried forward as a P3 for the next release in the interests
of expediting getting 2.1.3 out the door.  So this one should not be a blocker either.


Now we turn our attention to the potentially contentious one:
1567http://bugzilla.audacityteam.org/show_bug.cgi?id=1567
This only affects some Macs, Gale's in particular (ann mainly when he runs a debug version).
It works fine now on my Macbook Pro and on Cliff's Mac (after some tweaks for the Gatekeeper
changes were made by James following his experimentation.)

So we're holding up 2.1.3 for something (bug #1567) that only happens on some machines
and only happens on the very latest OS 10.12 Sierra from Apple - and for which there exist
known workarounds (published by Gale on the Forum). 

And let us not forget that a year ago we let 2.1.2 out the door without formal support for Microsoft's
latest operating system, Windows 10, and Windows is by far and away our major platform - so what
is different this time?

There is no crash and there are workarounds.

Also remember that any Mac user with Sierra who makes a fresh install of 2.1.2 will experience
these "enhanced" Gatekeeper issues (as I did earlier this week, resolved by the published
workarounds).


So I would strongly urge us to consider releasing 2.1.3 with bug #1567 unresolved and with
partial support for macOS Sierra (just as we had, and indeed stll have, partial, conditional,
support for W10 with 2.1.2.

Thanks,
Peter

------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
_______________________________________________
Audacity-quality mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/audacity-quality
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: P1s holding up 2.1.3 Release

Gale
Administrator
1304 has been closed with a new P3 bug
http://bugzilla.audacityteam.org/show_bug.cgi?id=1578

"TXT and XML exports are not to Documents and may
  require elevation to save".

Full steps to reproduce on all three platforms are included (I
tested on new Mac OS'es as first time Audacity user).

I would like to hear from Paul as to what we might be able to do
about the remaining bug 1567 Close Project problem (hoping that
is the only remaining problem).

2.1.2 partial support for Windows 10 isn't entirely comparable. Once
you fixed the problem with driver updates, or went back to 2.0.3,
the Win 10 problem was over.

With bug 1567, there is no obvious version of Audacity to go back to -
they all have the same problem AFAIK. And as we've seen, on
susceptible machines, once you've used xattr (or rebooted, or crashed
Audacity on launch, or other "solutions") the bug can come back
again.

The only cast iron solution is to go back to 10.12.0, which is a big
thing to ask. In contrast, new upgraders to Win 10 had a button to
click go back to previous Windows, as long as they did that within
30 days.



Gale


On 20 January 2017 at 14:07, Peter Sampson
<[hidden email]> wrote:

> I am getting increasingly concerned that we seem to be making poor progress
> currently
> towards 2.1.3 Release - and note that today is the one-year anniversary of
> the release of
> 2.1.2 - not an anniversary to be proud of I'm thinking.
>
> We are currently sitting with three P1s that are blocking release:
>
> 1304:  http://bugzilla.audacityteam.org/show_bug.cgi?id=1304
> This remains still listed at P1 although it tests ok on all three platforms
> for the
> originally reported issues.  It either needs degrading to P3 for the newly
> commented
> upon residual issues, as Gale himself suggested - or closing as resolved and
> a fresh
> P3 bug opened for the residual issues.
> So this one should not be a blocker.
>
>
> 1577:  http://bugzilla.audacityteam.org/show_bug.cgi?id=1577
> This is the one introduced by James' capitalization of Audacity in a Mac
> menu recently.
> Kames is working on a fix for this with high confidence it will work
> If it does not, then I suggest that we just  revert this change and that we
> live with the lower
> case "audacity" in that menu, carried forward as a P3 for the next release
> in the interests
> of expediting getting 2.1.3 out the door.  So this one should not be a
> blocker either.
>
>
> Now we turn our attention to the potentially contentious one:
> 1567:  http://bugzilla.audacityteam.org/show_bug.cgi?id=1567
> This only affects some Macs, Gale's in particular (ann mainly when he runs a
> debug version).
> It works fine now on my Macbook Pro and on Cliff's Mac (after some tweaks
> for the Gatekeeper
> changes were made by James following his experimentation.)
>
> So we're holding up 2.1.3 for something (bug #1567) that only happens on
> some machines
> and only happens on the very latest OS 10.12 Sierra from Apple - and for
> which there exist
> known workarounds (published by Gale on the Forum).
>
> And let us not forget that a year ago we let 2.1.2 out the door without
> formal support for Microsoft's
> latest operating system, Windows 10, and Windows is by far and away our
> major platform - so what
> is different this time?
>
> There is no crash and there are workarounds.
>
> Also remember that any Mac user with Sierra who makes a fresh install of
> 2.1.2 will experience
> these "enhanced" Gatekeeper issues (as I did earlier this week, resolved by
> the published
> workarounds).
>
>
> So I would strongly urge us to consider releasing 2.1.3 with bug #1567
> unresolved and with
> partial support for macOS Sierra (just as we had, and indeed stll have,
> partial, conditional,
> support for W10 with 2.1.2.
>
> Thanks,
> Peter
>
> ------------------------------------------------------------------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, SlashDot.org! http://sdm.link/slashdot
> _______________________________________________
> Audacity-quality mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/audacity-quality
>

------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
_______________________________________________
Audacity-quality mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/audacity-quality
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: P1s holding up 2.1.3 Release

James Crook
In reply to this post by Peter Sampson-2
I think only 1567 is actually blocking us in practice.

1304 is DEVEL-FIXMADE, and with 550 and 1305 being P2 it's already acknowledged / accepted that the  'residual' is intertwined in them, not 1304.   I'm prepared to go to RC with it DEVEL-FIXMADE if that is necessary to progress, but would not expect that to be necessary.
1577 can be fixed without reverting to 'Quit audacity'. 




1567 - Is a worry.  The timing 'fixes' haven't yet made enough progress to actually be a fix, as far as I understand it.  If a real fix is not progressing, then is announcing 'partial support for Sierra' enough to allow us to demote from P1 to P2?

--James.



On 1/20/2017 2:07 PM, Peter Sampson wrote:
I am getting increasingly concerned that we seem to be making poor progress
currently
towards 2.1.3 Release - and note that today is the one-year anniversary of
the release of
2.1.2 - not an anniversary to be proud of I'm thinking.

We are currently sitting with three P1s that are blocking release:

*1304:*  http://bugzilla.audacityteam.org/show_bug.cgi?id=1304
This remains still listed at P1 although it tests ok on all three platforms
for the
originally reported issues.  It either needs degrading to P3 for the newly
commented
upon residual issues, as Gale himself suggested - or closing as resolved
and a fresh
P3 bug opened for the residual issues.
So this one should not be a blocker.


*1577: * http://bugzilla.audacityteam.org/show_bug.cgi?id=1577
This is the one introduced by James' capitalization of Audacity in a Mac
menu recently.
Kames is working on a fix for this with high confidence it will work
If it does not, then I suggest that we just  revert this change and that we
live with the lower
case "audacity" in that menu, carried forward as a P3 for the next release
in the interests
of expediting getting 2.1.3 out the door.  So this one should not be a
blocker either.


Now we turn our attention to the potentially contentious one:
*1567*:  http://bugzilla.audacityteam.org/show_bug.cgi?id=1567
This only affects some Macs, Gale's in particular (ann mainly when he runs
a debug version).
It works fine now on my Macbook Pro and on Cliff's Mac (after some tweaks
for the Gatekeeper
changes were made by James following his experimentation.)

So we're holding up 2.1.3 for something (bug #1567) that only happens on
some machines
and only happens on the very latest OS 10.12 Sierra from Apple - and for
which there exist
known workarounds (published by Gale on the Forum).

And let us not forget that a year ago we let 2.1.2 out the door without
formal support for Microsoft's
latest operating system, Windows 10, and Windows is by far and away our
major platform - so what
is different this time?

There is no crash and there are workarounds.

Also remember that any Mac user with Sierra who makes a fresh install of
2.1.2 will experience
these "enhanced" Gatekeeper issues (as I did earlier this week, resolved by
the published
workarounds).


So I would strongly urge us to consider releasing 2.1.3 with bug #1567
unresolved and with
partial support for macOS Sierra (just as we had, and indeed stll have,
partial, conditional,
support for W10 with 2.1.2.

Thanks,
Peter



------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, SlashDot.org! http://sdm.link/slashdot


_______________________________________________
Audacity-quality mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/audacity-quality



------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
_______________________________________________
Audacity-quality mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/audacity-quality
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: P1s holding up 2.1.3 Release

Stevethefiddle
Just to note that it's not only Audacity that has problems with
plug-is on Sierra, and there are still several major companies saying
"check compatibility with your hardware / software before updating to
Sierra".

Steve

On 20 January 2017 at 15:48, James Crook <[hidden email]> wrote:

> I think only 1567 is actually blocking us in practice.
>
> 1304 is DEVEL-FIXMADE, and with 550 and 1305 being P2 it's already
> acknowledged / accepted that the  'residual' is intertwined in them, not
> 1304.   I'm prepared to go to RC with it DEVEL-FIXMADE if that is necessary
> to progress, but would not expect that to be necessary.
> 1577 can be fixed without reverting to 'Quit audacity'.
>
>
>
>
> 1567 - Is a worry.  The timing 'fixes' haven't yet made enough progress to
> actually be a fix, as far as I understand it.  If a real fix is not
> progressing, then is announcing 'partial support for Sierra' enough to allow
> us to demote from P1 to P2?
>
> --James.
>
>
>
> On 1/20/2017 2:07 PM, Peter Sampson wrote:
>
> I am getting increasingly concerned that we seem to be making poor progress
> currently
> towards 2.1.3 Release - and note that today is the one-year anniversary of
> the release of
> 2.1.2 - not an anniversary to be proud of I'm thinking.
>
> We are currently sitting with three P1s that are blocking release:
>
> *1304:*  http://bugzilla.audacityteam.org/show_bug.cgi?id=1304
> This remains still listed at P1 although it tests ok on all three platforms
> for the
> originally reported issues.  It either needs degrading to P3 for the newly
> commented
> upon residual issues, as Gale himself suggested - or closing as resolved
> and a fresh
> P3 bug opened for the residual issues.
> So this one should not be a blocker.
>
>
> *1577: * http://bugzilla.audacityteam.org/show_bug.cgi?id=1577
> This is the one introduced by James' capitalization of Audacity in a Mac
> menu recently.
> Kames is working on a fix for this with high confidence it will work
> If it does not, then I suggest that we just  revert this change and that we
> live with the lower
> case "audacity" in that menu, carried forward as a P3 for the next release
> in the interests
> of expediting getting 2.1.3 out the door.  So this one should not be a
> blocker either.
>
>
> Now we turn our attention to the potentially contentious one:
> *1567*:  http://bugzilla.audacityteam.org/show_bug.cgi?id=1567
> This only affects some Macs, Gale's in particular (ann mainly when he runs
> a debug version).
> It works fine now on my Macbook Pro and on Cliff's Mac (after some tweaks
> for the Gatekeeper
> changes were made by James following his experimentation.)
>
> So we're holding up 2.1.3 for something (bug #1567) that only happens on
> some machines
> and only happens on the very latest OS 10.12 Sierra from Apple - and for
> which there exist
> known workarounds (published by Gale on the Forum).
>
> And let us not forget that a year ago we let 2.1.2 out the door without
> formal support for Microsoft's
> latest operating system, Windows 10, and Windows is by far and away our
> major platform - so what
> is different this time?
>
> There is no crash and there are workarounds.
>
> Also remember that any Mac user with Sierra who makes a fresh install of
> 2.1.2 will experience
> these "enhanced" Gatekeeper issues (as I did earlier this week, resolved by
> the published
> workarounds).
>
>
> So I would strongly urge us to consider releasing 2.1.3 with bug #1567
> unresolved and with
> partial support for macOS Sierra (just as we had, and indeed stll have,
> partial, conditional,
> support for W10 with 2.1.2.
>
> Thanks,
> Peter
>
>
>
> ------------------------------------------------------------------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, SlashDot.org! http://sdm.link/slashdot
>
>
>
> _______________________________________________
> Audacity-quality mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/audacity-quality
>
>
>
> ------------------------------------------------------------------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, SlashDot.org! http://sdm.link/slashdot
> _______________________________________________
> Audacity-quality mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/audacity-quality
>

------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
_______________________________________________
Audacity-quality mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/audacity-quality
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: P1s holding up 2.1.3 Release

Peter Sampson-2
In reply to this post by Gale
Re 1567

Gale wrote:
>2.1.2 partial support for Windows 10 isn't entirely comparable. Once
>you fixed the problem with driver updates, or went back to 2.0.3,
>the Win 10 problem was over.

Well I would have regarded reverting from 2.1.2 to 2.0.3 as pretty darn
serious - they are 6 versions and three years of development apart,
resulting in loss of functionality if one followed that path.


>With bug 1567, there is no obvious version of Audacity to go back to -
>they all have the same problem AFAIK. And as we've seen, on
>susceptible machines, once you've used xattr (or rebooted, or crashed
>Audacity on launch, or other "solutions") the bug can come back
>again.

Agreed, but there is workaround and you can reuse these if necessary


>The only cast iron solution is to go back to 10.12.0, which is a big
>thing to ask. In contrast, new upgraders to Win 10 had a button to
>click go back to previous Windows, as long as they did that within
>30 days

Yes, as you say a big ask - but we would do well to remember that this
is a problem of Apple's making byr "enhancing" the Gatekeeper and
not a problem of our making.

Had we been able to release 2.1.3 in early October or so then this problem
would not have been occurrring up to and until Apple released the "enhanced"
Gatekeeper in macOS 10.12 Sierra.  At which point we would have logged
it as a bug to be fixed

So I would still support us releasing 2.1.3 with "limited support" (well documented
for Sierra - as otherwise
a) we extend the wait that our majority Widows users have for a new release
b) we inhibit even longer the new developments on hand and on hold.

Peter.

------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
_______________________________________________
Audacity-quality mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/audacity-quality
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: P1s holding up 2.1.3 Release

Peter Sampson-2
In reply to this post by Stevethefiddle
Steve wrote:
>Just to note that it's not only Audacity that has problems with
>plug-is on Sierra, and there are still several major companies saying
>"check compatibility with your hardware / software before updating to
>Sierra".

Which is just what Windows 10 users had to do unless and until they got
the correct W10 drivers for their hardware.

And that did not stop us releasing 2.1.2.

Peter.

------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
_______________________________________________
Audacity-quality mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/audacity-quality
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: P1s holding up 2.1.3 Release

Gale
Administrator
In reply to this post by James Crook
As already stated, 1304 is closed (RESOLVED). It takes time to write
out and test full details for the new residual bug to replace it, which is
required before 1304 is RESOLVED.


Gale


On 20 January 2017 at 15:48, James Crook <[hidden email]> wrote:

> I think only 1567 is actually blocking us in practice.
>
> 1304 is DEVEL-FIXMADE, and with 550 and 1305 being P2 it's already
> acknowledged / accepted that the  'residual' is intertwined in them, not
> 1304.   I'm prepared to go to RC with it DEVEL-FIXMADE if that is necessary
> to progress, but would not expect that to be necessary.
> 1577 can be fixed without reverting to 'Quit audacity'.
>
>
>
>
> 1567 - Is a worry.  The timing 'fixes' haven't yet made enough progress to
> actually be a fix, as far as I understand it.  If a real fix is not
> progressing, then is announcing 'partial support for Sierra' enough to allow
> us to demote from P1 to P2?
>
> --James.
>
>
>
> On 1/20/2017 2:07 PM, Peter Sampson wrote:
>
> I am getting increasingly concerned that we seem to be making poor progress
> currently
> towards 2.1.3 Release - and note that today is the one-year anniversary of
> the release of
> 2.1.2 - not an anniversary to be proud of I'm thinking.
>
> We are currently sitting with three P1s that are blocking release:
>
> *1304:*  http://bugzilla.audacityteam.org/show_bug.cgi?id=1304
> This remains still listed at P1 although it tests ok on all three platforms
> for the
> originally reported issues.  It either needs degrading to P3 for the newly
> commented
> upon residual issues, as Gale himself suggested - or closing as resolved
> and a fresh
> P3 bug opened for the residual issues.
> So this one should not be a blocker.
>
>
> *1577: * http://bugzilla.audacityteam.org/show_bug.cgi?id=1577
> This is the one introduced by James' capitalization of Audacity in a Mac
> menu recently.
> Kames is working on a fix for this with high confidence it will work
> If it does not, then I suggest that we just  revert this change and that we
> live with the lower
> case "audacity" in that menu, carried forward as a P3 for the next release
> in the interests
> of expediting getting 2.1.3 out the door.  So this one should not be a
> blocker either.
>
>
> Now we turn our attention to the potentially contentious one:
> *1567*:  http://bugzilla.audacityteam.org/show_bug.cgi?id=1567
> This only affects some Macs, Gale's in particular (ann mainly when he runs
> a debug version).
> It works fine now on my Macbook Pro and on Cliff's Mac (after some tweaks
> for the Gatekeeper
> changes were made by James following his experimentation.)
>
> So we're holding up 2.1.3 for something (bug #1567) that only happens on
> some machines
> and only happens on the very latest OS 10.12 Sierra from Apple - and for
> which there exist
> known workarounds (published by Gale on the Forum).
>
> And let us not forget that a year ago we let 2.1.2 out the door without
> formal support for Microsoft's
> latest operating system, Windows 10, and Windows is by far and away our
> major platform - so what
> is different this time?
>
> There is no crash and there are workarounds.
>
> Also remember that any Mac user with Sierra who makes a fresh install of
> 2.1.2 will experience
> these "enhanced" Gatekeeper issues (as I did earlier this week, resolved by
> the published
> workarounds).
>
>
> So I would strongly urge us to consider releasing 2.1.3 with bug #1567
> unresolved and with
> partial support for macOS Sierra (just as we had, and indeed stll have,
> partial, conditional,
> support for W10 with 2.1.2.
>
> Thanks,
> Peter
>
>
>
> ------------------------------------------------------------------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, SlashDot.org! http://sdm.link/slashdot
>
>
>
> _______________________________________________
> Audacity-quality mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/audacity-quality
>
>
>
> ------------------------------------------------------------------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, SlashDot.org! http://sdm.link/slashdot
> _______________________________________________
> Audacity-quality mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/audacity-quality
>

------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
_______________________________________________
Audacity-quality mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/audacity-quality
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: P1s holding up 2.1.3 Release

Peter Sampson-2
In reply to this post by Peter Sampson-2
Gale wrote:
>As already stated, 1304 is closed (RESOLVED). It takes time to write
>out and test full details for the new residual bug to replace it, which is
>required before 1304 is RESOLVED.

I'm not sure what your point is here Gale :-//

We get that - it's the 1567 blocker that only occurs on some machines,
yours in particular and not on other like mine or Cliff's that we are talking
about now


------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
_______________________________________________
Audacity-quality mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/audacity-quality
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: P1s holding up 2.1.3 Release

Gale
Administrator
The point was just that James was still saying 1304 was
DEVEL-FIXED after I said I closed it (but not very long after).

I didn't want someone to think I might have changed my mind
or found another problem with it.


Gale


On 20 January 2017 at 16:36, Peter Sampson
<[hidden email]> wrote:

> Gale wrote:
>>As already stated, 1304 is closed (RESOLVED). It takes time to write
>>out and test full details for the new residual bug to replace it, which is
>>required before 1304 is RESOLVED.
>
> I'm not sure what your point is here Gale :-//
>
> We get that - it's the 1567 blocker that only occurs on some machines,
> yours in particular and not on other like mine or Cliff's that we are
> talking
> about now
>
>
> ------------------------------------------------------------------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, SlashDot.org! http://sdm.link/slashdot
> _______________________________________________
> Audacity-quality mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/audacity-quality
>

------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
_______________________________________________
Audacity-quality mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/audacity-quality
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: P1s holding up 2.1.3 Release

Gale
Administrator
In reply to this post by Peter Sampson-2
We've shown with "timing issues" that we can reduce incidence
of the bug by adding sleeps (usually, the longer the sleep the
better).

On my machine the bug seems to trigger about half the time that
I Close Project without saving changes. That seems too often to
me to be P2. Extra sleeps at an appropriate point seem to me to
be worth investigating. If we can make it only happen "occasionally"
it could be only P2.

Quit Audacity without saving changes doesn't seem to be a
problem, because the new launch of Audacity with the crash
hack always makes Audacity trusted again.

I've not myself found any developer-ID-signed apps other than
Audacity that has these kind of problems (or which can only be
be used once if Admin installs it but Standard/Guest user uses it).

Even the latter problem is P1/P2 borderline IMO, but I would go
for P2 from where we are now.


Gale


On 20 January 2017 at 16:20, Peter Sampson
<[hidden email]> wrote:

> Re 1567
>
> Gale wrote:
>>2.1.2 partial support for Windows 10 isn't entirely comparable. Once
>>you fixed the problem with driver updates, or went back to 2.0.3,
>>the Win 10 problem was over.
>
> Well I would have regarded reverting from 2.1.2 to 2.0.3 as pretty darn
> serious - they are 6 versions and three years of development apart,
> resulting in loss of functionality if one followed that path.
>
>
>>With bug 1567, there is no obvious version of Audacity to go back to -
>>they all have the same problem AFAIK. And as we've seen, on
>>susceptible machines, once you've used xattr (or rebooted, or crashed
>>Audacity on launch, or other "solutions") the bug can come back
>>again.
>
> Agreed, but there is workaround and you can reuse these if necessary
>
>
>>The only cast iron solution is to go back to 10.12.0, which is a big
>>thing to ask. In contrast, new upgraders to Win 10 had a button to
>>click go back to previous Windows, as long as they did that within
>>30 days
>
> Yes, as you say a big ask - but we would do well to remember that this
> is a problem of Apple's making byr "enhancing" the Gatekeeper and
> not a problem of our making.
>
> Had we been able to release 2.1.3 in early October or so then this problem
> would not have been occurrring up to and until Apple released the "enhanced"
> Gatekeeper in macOS 10.12 Sierra.  At which point we would have logged
> it as a bug to be fixed
>
> So I would still support us releasing 2.1.3 with "limited support" (well
> documented
> for Sierra - as otherwise
> a) we extend the wait that our majority Widows users have for a new release
> b) we inhibit even longer the new developments on hand and on hold.
>
> Peter.
>
> ------------------------------------------------------------------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, SlashDot.org! http://sdm.link/slashdot
> _______________________________________________
> Audacity-quality mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/audacity-quality
>

------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
_______________________________________________
Audacity-quality mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/audacity-quality
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: P1s holding up 2.1.3 Release

James Crook
In reply to this post by Gale
Thanks.  I'd sent my message before seeing yours about having closed 1304.
My comments on 1304 were to say to Peter that 1304 is not a worry (now).
--James.


On 1/20/2017 4:25 PM, Gale Andrews wrote:

> As already stated, 1304 is closed (RESOLVED). It takes time to write
> out and test full details for the new residual bug to replace it, which is
> required before 1304 is RESOLVED.
>
>
> Gale
>
>
> On 20 January 2017 at 15:48, James Crook <[hidden email]> wrote:
>> I think only 1567 is actually blocking us in practice.
>>
>> 1304 is DEVEL-FIXMADE, and with 550 and 1305 being P2 it's already
>> acknowledged / accepted that the  'residual' is intertwined in them, not
>> 1304.   I'm prepared to go to RC with it DEVEL-FIXMADE if that is necessary
>> to progress, but would not expect that to be necessary.
>> 1577 can be fixed without reverting to 'Quit audacity'.
>>
>>
>>
>>
>> 1567 - Is a worry.  The timing 'fixes' haven't yet made enough progress to
>> actually be a fix, as far as I understand it.  If a real fix is not
>> progressing, then is announcing 'partial support for Sierra' enough to allow
>> us to demote from P1 to P2?
>>
>> --James.
>>
>>
>>
>> On 1/20/2017 2:07 PM, Peter Sampson wrote:
>>
>> I am getting increasingly concerned that we seem to be making poor progress
>> currently
>> towards 2.1.3 Release - and note that today is the one-year anniversary of
>> the release of
>> 2.1.2 - not an anniversary to be proud of I'm thinking.
>>
>> We are currently sitting with three P1s that are blocking release:
>>
>> *1304:*  http://bugzilla.audacityteam.org/show_bug.cgi?id=1304
>> This remains still listed at P1 although it tests ok on all three platforms
>> for the
>> originally reported issues.  It either needs degrading to P3 for the newly
>> commented
>> upon residual issues, as Gale himself suggested - or closing as resolved
>> and a fresh
>> P3 bug opened for the residual issues.
>> So this one should not be a blocker.
>>
>>
>> *1577: * http://bugzilla.audacityteam.org/show_bug.cgi?id=1577
>> This is the one introduced by James' capitalization of Audacity in a Mac
>> menu recently.
>> Kames is working on a fix for this with high confidence it will work
>> If it does not, then I suggest that we just  revert this change and that we
>> live with the lower
>> case "audacity" in that menu, carried forward as a P3 for the next release
>> in the interests
>> of expediting getting 2.1.3 out the door.  So this one should not be a
>> blocker either.
>>
>>
>> Now we turn our attention to the potentially contentious one:
>> *1567*:  http://bugzilla.audacityteam.org/show_bug.cgi?id=1567
>> This only affects some Macs, Gale's in particular (ann mainly when he runs
>> a debug version).
>> It works fine now on my Macbook Pro and on Cliff's Mac (after some tweaks
>> for the Gatekeeper
>> changes were made by James following his experimentation.)
>>
>> So we're holding up 2.1.3 for something (bug #1567) that only happens on
>> some machines
>> and only happens on the very latest OS 10.12 Sierra from Apple - and for
>> which there exist
>> known workarounds (published by Gale on the Forum).
>>
>> And let us not forget that a year ago we let 2.1.2 out the door without
>> formal support for Microsoft's
>> latest operating system, Windows 10, and Windows is by far and away our
>> major platform - so what
>> is different this time?
>>
>> There is no crash and there are workarounds.
>>
>> Also remember that any Mac user with Sierra who makes a fresh install of
>> 2.1.2 will experience
>> these "enhanced" Gatekeeper issues (as I did earlier this week, resolved by
>> the published
>> workarounds).
>>
>>
>> So I would strongly urge us to consider releasing 2.1.3 with bug #1567
>> unresolved and with
>> partial support for macOS Sierra (just as we had, and indeed stll have,
>> partial, conditional,
>> support for W10 with 2.1.2.
>>
>> Thanks,
>> Peter
>>
>>
>>


------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
_______________________________________________
Audacity-quality mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/audacity-quality
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: P1s holding up 2.1.3 Release

Paul Licameli
In reply to this post by Gale
About the incidence of problems with 2.1.2 or 2.1.1 on Sierra (and I know some Mac users avoid 2.1.2 because some of the bugs we fixed in 2.1.3 were too bothersome), I see rather frequent complaints about it in the discussion group I manage, and have referred people to Gale's post on the Forum.

More disturbingly, one user, but only one so far, reported that recording started to suffer frequent drop-outs that didn't happen with GarageBand.  This was so bad for work he was on the point of switching to Reaper.  But I have since been told that the xattr workaround seems to have solved that problem.

Unscientifically, I have so far counted three such complainants, including the one with recording problems, out of a population of 1,256 in the group.  I don't know what portions of those are inactive, just lurkers, Windows users, not yet Sierra users, etc.

As for fixes for bug 1567, I understand the most promising thing so far was not the sleeps.  It was the weird fix involving a call to fork(), then crashing the parent process, and letting the child continue.  It did not work to crash the child and continue the parent.

The latest commit I pointed Gale to, was only meant to fix the incidental problem that the splash dialog was hidden behind other windows.  If the partial fix so far is to be merged, please let me squash it into one commit first and tidy a few things.

I looked into the sequence of events in closing with and without saving unsaved work, to understand what difference of disk operations there is, but I don't yet have a firm idea how to write a fix that works around that.

PRL


On Fri, Jan 20, 2017 at 12:04 PM, Gale Andrews <[hidden email]> wrote:
We've shown with "timing issues" that we can reduce incidence
of the bug by adding sleeps (usually, the longer the sleep the
better).

On my machine the bug seems to trigger about half the time that
I Close Project without saving changes. That seems too often to
me to be P2. Extra sleeps at an appropriate point seem to me to
be worth investigating. If we can make it only happen "occasionally"
it could be only P2.

Quit Audacity without saving changes doesn't seem to be a
problem, because the new launch of Audacity with the crash
hack always makes Audacity trusted again.

I've not myself found any developer-ID-signed apps other than
Audacity that has these kind of problems (or which can only be
be used once if Admin installs it but Standard/Guest user uses it).

Even the latter problem is P1/P2 borderline IMO, but I would go
for P2 from where we are now.


Gale


On 20 January 2017 at 16:20, Peter Sampson
<[hidden email]> wrote:
> Re 1567
>
> Gale wrote:
>>2.1.2 partial support for Windows 10 isn't entirely comparable. Once
>>you fixed the problem with driver updates, or went back to 2.0.3,
>>the Win 10 problem was over.
>
> Well I would have regarded reverting from 2.1.2 to 2.0.3 as pretty darn
> serious - they are 6 versions and three years of development apart,
> resulting in loss of functionality if one followed that path.
>
>
>>With bug 1567, there is no obvious version of Audacity to go back to -
>>they all have the same problem AFAIK. And as we've seen, on
>>susceptible machines, once you've used xattr (or rebooted, or crashed
>>Audacity on launch, or other "solutions") the bug can come back
>>again.
>
> Agreed, but there is workaround and you can reuse these if necessary
>
>
>>The only cast iron solution is to go back to 10.12.0, which is a big
>>thing to ask. In contrast, new upgraders to Win 10 had a button to
>>click go back to previous Windows, as long as they did that within
>>30 days
>
> Yes, as you say a big ask - but we would do well to remember that this
> is a problem of Apple's making byr "enhancing" the Gatekeeper and
> not a problem of our making.
>
> Had we been able to release 2.1.3 in early October or so then this problem
> would not have been occurrring up to and until Apple released the "enhanced"
> Gatekeeper in macOS 10.12 Sierra.  At which point we would have logged
> it as a bug to be fixed
>
> So I would still support us releasing 2.1.3 with "limited support" (well
> documented
> for Sierra - as otherwise
> a) we extend the wait that our majority Widows users have for a new release
> b) we inhibit even longer the new developments on hand and on hold.
>
> Peter.
>
> ------------------------------------------------------------------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, SlashDot.org! http://sdm.link/slashdot
> _______________________________________________
> Audacity-quality mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/audacity-quality
>

------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
_______________________________________________
Audacity-quality mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/audacity-quality


------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
_______________________________________________
Audacity-quality mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/audacity-quality
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: P1s holding up 2.1.3 Release

Gale
Administrator
On 20 January 2017 at 19:56, Paul Licameli <[hidden email]> wrote:

> About the incidence of problems with 2.1.2 or 2.1.1 on Sierra (and I know
> some Mac users avoid 2.1.2 because some of the bugs we fixed in 2.1.3 were
> too bothersome), I see rather frequent complaints about it in the discussion
> group I manage, and have referred people to Gale's post on the Forum.
>
> More disturbingly, one user, but only one so far, reported that recording
> started to suffer frequent drop-outs that didn't happen with GarageBand.
> This was so bad for work he was on the point of switching to Reaper.  But I
> have since been told that the xattr workaround seems to have solved that
> problem.

That's another interesting angle. Does Gatekeeper throttle back apps it does
not trust?  I see no evidence myself right now of a recording or playback
problem with previous fix attempts that made the bug worse - assuming of
course Audacity's "Audio to buffer" setting is well below the default of 100.


> Unscientifically, I have so far counted three such complainants, including
> the one with recording problems, out of a population of 1,256 in the group.
> I don't know what portions of those are inactive, just lurkers, Windows
> users, not yet Sierra users, etc.
>
> As for fixes for bug 1567, I understand the most promising thing so far was
> not the sleeps.  It was the weird fix involving a call to fork(), then
> crashing the parent process, and letting the child continue.  It did not
> work to crash the child and continue the parent.

That's correct, but sleeps have sometimes had an ameliorating effect
if made long enough or in the "right place".


Gale

> The latest commit I pointed Gale to, was only meant to fix the incidental
> problem that the splash dialog was hidden behind other windows.  If the
> partial fix so far is to be merged, please let me squash it into one commit
> first and tidy a few things.
>
> I looked into the sequence of events in closing with and without saving
> unsaved work, to understand what difference of disk operations there is, but
> I don't yet have a firm idea how to write a fix that works around that.
>
> PRL
>
>
> On Fri, Jan 20, 2017 at 12:04 PM, Gale Andrews <[hidden email]>
> wrote:
>>
>> We've shown with "timing issues" that we can reduce incidence
>> of the bug by adding sleeps (usually, the longer the sleep the
>> better).
>>
>> On my machine the bug seems to trigger about half the time that
>> I Close Project without saving changes. That seems too often to
>> me to be P2. Extra sleeps at an appropriate point seem to me to
>> be worth investigating. If we can make it only happen "occasionally"
>> it could be only P2.
>>
>> Quit Audacity without saving changes doesn't seem to be a
>> problem, because the new launch of Audacity with the crash
>> hack always makes Audacity trusted again.
>>
>> I've not myself found any developer-ID-signed apps other than
>> Audacity that has these kind of problems (or which can only be
>> be used once if Admin installs it but Standard/Guest user uses it).
>>
>> Even the latter problem is P1/P2 borderline IMO, but I would go
>> for P2 from where we are now.
>>
>>
>> Gale
>>
>>
>> On 20 January 2017 at 16:20, Peter Sampson
>> <[hidden email]> wrote:
>> > Re 1567
>> >
>> > Gale wrote:
>> >>2.1.2 partial support for Windows 10 isn't entirely comparable. Once
>> >>you fixed the problem with driver updates, or went back to 2.0.3,
>> >>the Win 10 problem was over.
>> >
>> > Well I would have regarded reverting from 2.1.2 to 2.0.3 as pretty darn
>> > serious - they are 6 versions and three years of development apart,
>> > resulting in loss of functionality if one followed that path.
>> >
>> >
>> >>With bug 1567, there is no obvious version of Audacity to go back to -
>> >>they all have the same problem AFAIK. And as we've seen, on
>> >>susceptible machines, once you've used xattr (or rebooted, or crashed
>> >>Audacity on launch, or other "solutions") the bug can come back
>> >>again.
>> >
>> > Agreed, but there is workaround and you can reuse these if necessary
>> >
>> >
>> >>The only cast iron solution is to go back to 10.12.0, which is a big
>> >>thing to ask. In contrast, new upgraders to Win 10 had a button to
>> >>click go back to previous Windows, as long as they did that within
>> >>30 days
>> >
>> > Yes, as you say a big ask - but we would do well to remember that this
>> > is a problem of Apple's making byr "enhancing" the Gatekeeper and
>> > not a problem of our making.
>> >
>> > Had we been able to release 2.1.3 in early October or so then this
>> > problem
>> > would not have been occurrring up to and until Apple released the
>> > "enhanced"
>> > Gatekeeper in macOS 10.12 Sierra.  At which point we would have logged
>> > it as a bug to be fixed
>> >
>> > So I would still support us releasing 2.1.3 with "limited support" (well
>> > documented
>> > for Sierra - as otherwise
>> > a) we extend the wait that our majority Widows users have for a new
>> > release
>> > b) we inhibit even longer the new developments on hand and on hold.
>> >
>> > Peter.
>> >
>> >
>> > ------------------------------------------------------------------------------
>> > Check out the vibrant tech community on one of the world's most
>> > engaging tech sites, SlashDot.org! http://sdm.link/slashdot
>> > _______________________________________________
>> > Audacity-quality mailing list
>> > [hidden email]
>> > https://lists.sourceforge.net/lists/listinfo/audacity-quality
>> >
>>
>>
>> ------------------------------------------------------------------------------
>> Check out the vibrant tech community on one of the world's most
>> engaging tech sites, SlashDot.org! http://sdm.link/slashdot
>> _______________________________________________
>> Audacity-quality mailing list
>> [hidden email]
>> https://lists.sourceforge.net/lists/listinfo/audacity-quality
>
>
>
> ------------------------------------------------------------------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, SlashDot.org! http://sdm.link/slashdot
> _______________________________________________
> Audacity-quality mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/audacity-quality
>

------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
_______________________________________________
Audacity-quality mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/audacity-quality
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: P1s holding up 2.1.3 Release

Peter Sampson-2
Gale, Paul,

I'm sorry but you are hijacking this high-level decision making the discussion
into a technical one on haow to solve #1567 - and we already have an
ongoing thread for just that.

This thread was originated to discuss whethe or not we could release 2.1.3
with some bugs that are currently listed as P1 and thus blocking this (very
late) release.

#1304 is now resolved and James believes he can clear #1577

This leaves the troublesome #1567 - so let me repeat my original post and
proposal to bring this thread back onto its release management track

1567http://bugzilla.audacityteam.org/show_bug.cgi?id=1567
This only affects some Macs, Gale's in particular (ann mainly when he runs a debug version).
It works fine now on my Macbook Pro and on Cliff's Mac (after some tweaks for the Gatekeeper
changes were made by James following his experimentation.)

So we're holding up 2.1.3 for something (bug #1567) that only happens on some machines
and only happens on the very latest OS 10.12 Sierra from Apple - and for which there exist
known workarounds (published by Gale on the Forum). 
...
There is no crash and there are workarounds.

Also remember that any Mac user with Sierra who makes a fresh install of 2.1.2 will experience
these "enhanced" Gatekeeper issues ...

So I would strongly urge us to consider releasing 2.1.3 with bug #1567 unresolved and with
partial support for macOS Sierra, with suitable workaronds clearly published.

Peter

------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
_______________________________________________
Audacity-quality mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/audacity-quality
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: P1s holding up 2.1.3 Release

Gale
Administrator
And you (Peter) are repeating yourself, I believe. -1 to your proposal from me.


Gale


On 21 January 2017 at 11:42, Peter Sampson
<[hidden email]> wrote:

> Gale, Paul,
>
> I'm sorry but you are hijacking this high-level decision making the
> discussion
> into a technical one on haow to solve #1567 - and we already have an
> ongoing thread for just that.
>
> This thread was originated to discuss whethe or not we could release 2.1.3
> with some bugs that are currently listed as P1 and thus blocking this (very
> late) release.
>
> #1304 is now resolved and James believes he can clear #1577
>
> This leaves the troublesome #1567 - so let me repeat my original post and
> proposal to bring this thread back onto its release management track
>
> 1567:  http://bugzilla.audacityteam.org/show_bug.cgi?id=1567
> This only affects some Macs, Gale's in particular (ann mainly when he runs a
> debug version).
> It works fine now on my Macbook Pro and on Cliff's Mac (after some tweaks
> for the Gatekeeper
> changes were made by James following his experimentation.)
>
> So we're holding up 2.1.3 for something (bug #1567) that only happens on
> some machines
> and only happens on the very latest OS 10.12 Sierra from Apple - and for
> which there exist
> known workarounds (published by Gale on the Forum).
> ...
> There is no crash and there are workarounds.
>
> Also remember that any Mac user with Sierra who makes a fresh install of
> 2.1.2 will experience
> these "enhanced" Gatekeeper issues ...
>
> So I would strongly urge us to consider releasing 2.1.3 with bug #1567
> unresolved and with
> partial support for macOS Sierra, with suitable workaronds clearly
> published.
>
> Peter
>
> ------------------------------------------------------------------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, SlashDot.org! http://sdm.link/slashdot
> _______________________________________________
> Audacity-quality mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/audacity-quality
>

------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
_______________________________________________
Audacity-quality mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/audacity-quality
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: P1s holding up 2.1.3 Release

Peter Sampson-2
Indeed I was repeating myself, quite deliberately so.

I repeated myself to bring this thread back on track.

Peter

------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
_______________________________________________
Audacity-quality mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/audacity-quality
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: P1s holding up 2.1.3 Release

Paul Licameli
In reply to this post by Peter Sampson-2


On Sat, Jan 21, 2017 at 6:42 AM, Peter Sampson <[hidden email]> wrote:
Gale, Paul,

I'm sorry but you are hijacking this high-level decision making the discussion
into a technical one on haow to solve #1567 - and we already have an
ongoing thread for just that.

Peter, I meant no "hijack."  I meant to point out what I thought was pertinent to the discussion:

1) At least one report of drop-outs during recording that I know.  I learned of this only lately.. I had not heard of this symptom from Gale.  But if this is a possible symptom, I say it is a very severe one, even if only for a small minority of people, that should influence the decision whether to release without a fix.  However, anecdotally, the xattr workaround appears to fix the problem.

2) James seemed to me to misunderstand the nature of the fixes on offer, and that needed correction.  The possible fixes are not (just) about sleeps, but also involve the forking and crashing.

PRL
 

This thread was originated to discuss whethe or not we could release 2.1.3
with some bugs that are currently listed as P1 and thus blocking this (very
late) release.

#1304 is now resolved and James believes he can clear #1577

This leaves the troublesome #1567 - so let me repeat my original post and
proposal to bring this thread back onto its release management track

1567http://bugzilla.audacityteam.org/show_bug.cgi?id=1567
This only affects some Macs, Gale's in particular (ann mainly when he runs a debug version).
It works fine now on my Macbook Pro and on Cliff's Mac (after some tweaks for the Gatekeeper
changes were made by James following his experimentation.)

So we're holding up 2.1.3 for something (bug #1567) that only happens on some machines
and only happens on the very latest OS 10.12 Sierra from Apple - and for which there exist
known workarounds (published by Gale on the Forum). 
...
There is no crash and there are workarounds.

Also remember that any Mac user with Sierra who makes a fresh install of 2.1.2 will experience
these "enhanced" Gatekeeper issues ...

So I would strongly urge us to consider releasing 2.1.3 with bug #1567 unresolved and with
partial support for macOS Sierra, with suitable workaronds clearly published.

Peter

------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
_______________________________________________
Audacity-quality mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/audacity-quality



------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
_______________________________________________
Audacity-quality mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/audacity-quality
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: P1s holding up 2.1.3 Release

Gale
Administrator
On 21 January 2017 at 13:37, Paul Licameli <[hidden email]> wrote:

>
>
> On Sat, Jan 21, 2017 at 6:42 AM, Peter Sampson
> <[hidden email]> wrote:
>>
>> Gale, Paul,
>>
>> I'm sorry but you are hijacking this high-level decision making the
>> discussion
>> into a technical one on haow to solve #1567 - and we already have an
>> ongoing thread for just that.
>
>
> Peter, I meant no "hijack."  I meant to point out what I thought was
> pertinent to the discussion:

+1. We have to discuss symptoms, timescales for fixes and alternatives
that are open to users in order to make a decision on releasing 2.1.2
with "provisional support for Sierra" or not. Of course it is ultimately
James's decision but needs to decide on full knowledge of the
evidence.

I don't see our workarounds are good if affected users have to reapply
them every other time they close project without saving changes.
Remember that it's "normal" on Mac to close apps rather than quit.

I thought until a few days ago that we were getting close enough to
a complete fix for 1567 (as exhibited on my machine) that the balance
of argument was to see the fix through.


> 1) At least one report of drop-outs during recording that I know.  I learned
> of this only lately.. I had not heard of this symptom from Gale.  But if
> this is a possible symptom, I say it is a very severe one, even if only for
> a small minority of people, that should influence the decision whether to
> release without a fix.  However, anecdotally, the xattr workaround appears
> to fix the problem.

As I said, I cannot reproduce that, recording from Soundflower after 1567
strikes. I'd need to know more about what device the user is recording
from and what Mac they have - you can tell me off list if you like.

In reality there are several reasons, not just our mistaken buffer setting, why
recordings on Mac may have dropouts or seem clicky:
http://alphamanual.audacityteam.org/man/FAQ:Recording_-_Troubleshooting#mac_crackle
.


> 2) James seemed to me to misunderstand the nature of the fixes on offer, and
> that needed correction.  The possible fixes are not (just) about sleeps, but
> also involve the forking and crashing.

And a reminder that the fork and crash on launch is totally successful on
my machine. Audacity never *launches* with the bug present. Now we just
have to stop the bug happening while Audacity is running, for which I have
only so far found the one trigger - close project without saving changes.

Some more information. I built debug of Paul's "weird 2" with the fork
and crash, and (so far) the bug doesn't seem to trigger on File > Close
without saving changes. Another indication we have to slow something
down?


Gale


>> This thread was originated to discuss whethe or not we could release 2.1.3
>> with some bugs that are currently listed as P1 and thus blocking this
>> (very
>> late) release.
>>
>> #1304 is now resolved and James believes he can clear #1577
>>
>> This leaves the troublesome #1567 - so let me repeat my original post and
>> proposal to bring this thread back onto its release management track
>>
>> 1567:  http://bugzilla.audacityteam.org/show_bug.cgi?id=1567
>> This only affects some Macs, Gale's in particular (ann mainly when he runs
>> a debug version).
>> It works fine now on my Macbook Pro and on Cliff's Mac (after some tweaks
>> for the Gatekeeper
>> changes were made by James following his experimentation.)
>>
>> So we're holding up 2.1.3 for something (bug #1567) that only happens on
>> some machines
>> and only happens on the very latest OS 10.12 Sierra from Apple - and for
>> which there exist
>> known workarounds (published by Gale on the Forum).
>> ...
>> There is no crash and there are workarounds.
>>
>> Also remember that any Mac user with Sierra who makes a fresh install of
>> 2.1.2 will experience
>> these "enhanced" Gatekeeper issues ...
>>
>> So I would strongly urge us to consider releasing 2.1.3 with bug #1567
>> unresolved and with
>> partial support for macOS Sierra, with suitable workaronds clearly
>> published.
>>
>> Peter
>>
>>
>> ------------------------------------------------------------------------------
>> Check out the vibrant tech community on one of the world's most
>> engaging tech sites, SlashDot.org! http://sdm.link/slashdot
>> _______________________________________________
>> Audacity-quality mailing list
>> [hidden email]
>> https://lists.sourceforge.net/lists/listinfo/audacity-quality
>>
>
>
> ------------------------------------------------------------------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, SlashDot.org! http://sdm.link/slashdot
> _______________________________________________
> Audacity-quality mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/audacity-quality
>

------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
_______________________________________________
Audacity-quality mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/audacity-quality
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: P1s holding up 2.1.3 Release

Paul Licameli


On Sat, Jan 21, 2017 at 9:54 AM, Gale Andrews <[hidden email]> wrote:
On 21 January 2017 at 13:37, Paul Licameli <[hidden email]> wrote:
>
>
> On Sat, Jan 21, 2017 at 6:42 AM, Peter Sampson
> <[hidden email]> wrote:
>>
>> Gale, Paul,
>>
>> I'm sorry but you are hijacking this high-level decision making the
>> discussion
>> into a technical one on haow to solve #1567 - and we already have an
>> ongoing thread for just that.
>
>
> Peter, I meant no "hijack."  I meant to point out what I thought was
> pertinent to the discussion:

+1. We have to discuss symptoms, timescales for fixes and alternatives
that are open to users in order to make a decision on releasing 2.1.2
with "provisional support for Sierra" or not. Of course it is ultimately
James's decision but needs to decide on full knowledge of the
evidence.

I don't see our workarounds are good if affected users have to reapply
them every other time they close project without saving changes.
Remember that it's "normal" on Mac to close apps rather than quit.

Just to be sure, by "close apps rather than quit" you mean closing the main window, but leaving Audacity present in the alt-TAB cycle?  I call that closing a window, not closing an app.
 

I thought until a few days ago that we were getting close enough to
a complete fix for 1567 (as exhibited on my machine) that the balance
of argument was to see the fix through.


> 1) At least one report of drop-outs during recording that I know.  I learned
> of this only lately.. I had not heard of this symptom from Gale.  But if
> this is a possible symptom, I say it is a very severe one, even if only for
> a small minority of people, that should influence the decision whether to
> release without a fix.  However, anecdotally, the xattr workaround appears
> to fix the problem.

As I said, I cannot reproduce that, recording from Soundflower after 1567
strikes. I'd need to know more about what device the user is recording
from and what Mac they have - you can tell me off list if you like.

In reality there are several reasons, not just our mistaken buffer setting, why
recordings on Mac may have dropouts or seem clicky:
http://alphamanual.audacityteam.org/man/FAQ:Recording_-_Troubleshooting#mac_crackle
.


> 2) James seemed to me to misunderstand the nature of the fixes on offer, and
> that needed correction.  The possible fixes are not (just) about sleeps, but
> also involve the forking and crashing.

And a reminder that the fork and crash on launch is totally successful on
my machine. Audacity never *launches* with the bug present. Now we just
have to stop the bug happening while Audacity is running, for which I have
only so far found the one trigger - close project without saving changes.

Some more information. I built debug of Paul's "weird 2" with the fork
and crash, and (so far) the bug doesn't seem to trigger on File > Close
without saving changes. Another indication we have to slow something
down?

You mean debug appears not to suffer the problem, while release build still does?

With this bug and its mysterious file system interacitons -- who knows, perhaps the cause there is that the debug build writes more to stdout and log, rather than timing differences.

PRL
 


Gale


>> This thread was originated to discuss whethe or not we could release 2.1.3
>> with some bugs that are currently listed as P1 and thus blocking this
>> (very
>> late) release.
>>
>> #1304 is now resolved and James believes he can clear #1577
>>
>> This leaves the troublesome #1567 - so let me repeat my original post and
>> proposal to bring this thread back onto its release management track
>>
>> 1567:  http://bugzilla.audacityteam.org/show_bug.cgi?id=1567
>> This only affects some Macs, Gale's in particular (ann mainly when he runs
>> a debug version).
>> It works fine now on my Macbook Pro and on Cliff's Mac (after some tweaks
>> for the Gatekeeper
>> changes were made by James following his experimentation.)
>>
>> So we're holding up 2.1.3 for something (bug #1567) that only happens on
>> some machines
>> and only happens on the very latest OS 10.12 Sierra from Apple - and for
>> which there exist
>> known workarounds (published by Gale on the Forum).
>> ...
>> There is no crash and there are workarounds.
>>
>> Also remember that any Mac user with Sierra who makes a fresh install of
>> 2.1.2 will experience
>> these "enhanced" Gatekeeper issues ...
>>
>> So I would strongly urge us to consider releasing 2.1.3 with bug #1567
>> unresolved and with
>> partial support for macOS Sierra, with suitable workaronds clearly
>> published.
>>
>> Peter
>>
>>
>> ------------------------------------------------------------------------------
>> Check out the vibrant tech community on one of the world's most
>> engaging tech sites, SlashDot.org! http://sdm.link/slashdot
>> _______________________________________________
>> Audacity-quality mailing list
>> [hidden email]
>> https://lists.sourceforge.net/lists/listinfo/audacity-quality
>>
>
>
> ------------------------------------------------------------------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, SlashDot.org! http://sdm.link/slashdot
> _______________________________________________
> Audacity-quality mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/audacity-quality
>

------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
_______________________________________________
Audacity-quality mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/audacity-quality


------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
_______________________________________________
Audacity-quality mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/audacity-quality
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: P1s holding up 2.1.3 Release

Gale
Administrator
On 21 January 2017 at 15:33, Paul Licameli <[hidden email]> wrote:

>
>
> On Sat, Jan 21, 2017 at 9:54 AM, Gale Andrews <[hidden email]> wrote:
>>
>> On 21 January 2017 at 13:37, Paul Licameli <[hidden email]>
>> wrote:
>> >
>> >
>> > On Sat, Jan 21, 2017 at 6:42 AM, Peter Sampson
>> > <[hidden email]> wrote:
>> >>
>> >> Gale, Paul,
>> >>
>> >> I'm sorry but you are hijacking this high-level decision making the
>> >> discussion
>> >> into a technical one on haow to solve #1567 - and we already have an
>> >> ongoing thread for just that.
>> >
>> >
>> > Peter, I meant no "hijack."  I meant to point out what I thought was
>> > pertinent to the discussion:
>>
>> +1. We have to discuss symptoms, timescales for fixes and alternatives
>> that are open to users in order to make a decision on releasing 2.1.2
>> with "provisional support for Sierra" or not. Of course it is ultimately
>> James's decision but needs to decide on full knowledge of the
>> evidence.
>>
>> I don't see our workarounds are good if affected users have to reapply
>> them every other time they close project without saving changes.
>> Remember that it's "normal" on Mac to close apps rather than quit.
>
>
> Just to be sure, by "close apps rather than quit" you mean closing the main
> window, but leaving Audacity present in the alt-TAB cycle?

Yes.

> I call that closing a window, not closing an app.

I call it "closing the app" only if we close the last project window. But 1567
occurs on close whether closing the last window or not.


>> I thought until a few days ago that we were getting close enough to
>> a complete fix for 1567 (as exhibited on my machine) that the balance
>> of argument was to see the fix through.
>>
>>
>> > 1) At least one report of drop-outs during recording that I know.  I
>> > learned
>> > of this only lately.. I had not heard of this symptom from Gale.  But if
>> > this is a possible symptom, I say it is a very severe one, even if only
>> > for
>> > a small minority of people, that should influence the decision whether
>> > to
>> > release without a fix.  However, anecdotally, the xattr workaround
>> > appears
>> > to fix the problem.
>>
>> As I said, I cannot reproduce that, recording from Soundflower after 1567
>> strikes. I'd need to know more about what device the user is recording
>> from and what Mac they have - you can tell me off list if you like.
>>
>> In reality there are several reasons, not just our mistaken buffer
>> setting, why
>> recordings on Mac may have dropouts or seem clicky:
>>
>> http://alphamanual.audacityteam.org/man/FAQ:Recording_-_Troubleshooting#mac_crackle
>> .
>>
>>
>> > 2) James seemed to me to misunderstand the nature of the fixes on offer,
>> > and
>> > that needed correction.  The possible fixes are not (just) about sleeps,
>> > but
>> > also involve the forking and crashing.
>>
>> And a reminder that the fork and crash on launch is totally successful on
>> my machine. Audacity never *launches* with the bug present. Now we just
>> have to stop the bug happening while Audacity is running, for which I have
>> only so far found the one trigger - close project without saving changes.
>>
>> Some more information. I built debug of Paul's "weird 2" with the fork
>> and crash, and (so far) the bug doesn't seem to trigger on File > Close
>> without saving changes. Another indication we have to slow something
>> down?
>
>
> You mean debug appears not to suffer the problem, while release build still
> does?

It looks that way right now. I don't see 1567 in a debug build of your "weird2".


Gale


> With this bug and its mysterious file system interacitons -- who knows,
> perhaps the cause there is that the debug build writes more to stdout and
> log, rather than timing differences.
>
> PRL
>
>>
>>
>>
>> Gale
>>
>>
>> >> This thread was originated to discuss whethe or not we could release
>> >> 2.1.3
>> >> with some bugs that are currently listed as P1 and thus blocking this
>> >> (very
>> >> late) release.
>> >>
>> >> #1304 is now resolved and James believes he can clear #1577
>> >>
>> >> This leaves the troublesome #1567 - so let me repeat my original post
>> >> and
>> >> proposal to bring this thread back onto its release management track
>> >>
>> >> 1567:  http://bugzilla.audacityteam.org/show_bug.cgi?id=1567
>> >> This only affects some Macs, Gale's in particular (ann mainly when he
>> >> runs
>> >> a debug version).
>> >> It works fine now on my Macbook Pro and on Cliff's Mac (after some
>> >> tweaks
>> >> for the Gatekeeper
>> >> changes were made by James following his experimentation.)
>> >>
>> >> So we're holding up 2.1.3 for something (bug #1567) that only happens
>> >> on
>> >> some machines
>> >> and only happens on the very latest OS 10.12 Sierra from Apple - and
>> >> for
>> >> which there exist
>> >> known workarounds (published by Gale on the Forum).
>> >> ...
>> >> There is no crash and there are workarounds.
>> >>
>> >> Also remember that any Mac user with Sierra who makes a fresh install
>> >> of
>> >> 2.1.2 will experience
>> >> these "enhanced" Gatekeeper issues ...
>> >>
>> >> So I would strongly urge us to consider releasing 2.1.3 with bug #1567
>> >> unresolved and with
>> >> partial support for macOS Sierra, with suitable workaronds clearly
>> >> published.
>> >>
>> >> Peter
>> >>
>> >>
>> >>
>> >> ------------------------------------------------------------------------------
>> >> Check out the vibrant tech community on one of the world's most
>> >> engaging tech sites, SlashDot.org! http://sdm.link/slashdot
>> >> _______________________________________________
>> >> Audacity-quality mailing list
>> >> [hidden email]
>> >> https://lists.sourceforge.net/lists/listinfo/audacity-quality
>> >>
>> >
>> >
>> >
>> > ------------------------------------------------------------------------------
>> > Check out the vibrant tech community on one of the world's most
>> > engaging tech sites, SlashDot.org! http://sdm.link/slashdot
>> > _______________________________________________
>> > Audacity-quality mailing list
>> > [hidden email]
>> > https://lists.sourceforge.net/lists/listinfo/audacity-quality
>> >
>>
>>
>> ------------------------------------------------------------------------------
>> Check out the vibrant tech community on one of the world's most
>> engaging tech sites, SlashDot.org! http://sdm.link/slashdot
>> _______________________________________________
>> Audacity-quality mailing list
>> [hidden email]
>> https://lists.sourceforge.net/lists/listinfo/audacity-quality
>
>
>
> ------------------------------------------------------------------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, SlashDot.org! http://sdm.link/slashdot
> _______________________________________________
> Audacity-quality mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/audacity-quality
>

------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
_______________________________________________
Audacity-quality mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/audacity-quality
12
Loading...