Issue for 4.1.4 and MacOS

Previous Topic Next Topic
 
classic Classic list List threaded Threaded
93 messages Options
12345
Reply | Threaded
Open this post in threaded view
|

Issue for 4.1.4 and MacOS

FR web forum
Please have a look on:
https://bz.apache.org/ooo/show_bug.cgi?id=127568

---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]

Reply | Threaded
Open this post in threaded view
|

Re: Issue for 4.1.4 and MacOS

Larry Gusaas-4
On 2017-10-24, 4:19 AM FR web forum wrote:

> Please have a look on:
> https://bz.apache.org/ooo/show_bug.cgi?id=127568

I tested and confirm the problem. Posted my result on the bug report.


--
_________________________________

Larry I. Gusaas
Moose Jaw, Saskatchewan Canada
Website: http://larry-gusaas.com
"An artist is never ahead of his time but most people are far behind theirs." - Edgard Varese



---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]

Reply | Threaded
Open this post in threaded view
|

Re: Issue for 4.1.4 and MacOS

Patricia Shanahan
Since the issue appeared in 4.1.4, and involves document links, it might
be related to my changes.

I do not have a Mac, and do not have any Mac debug experience.

How do people think we should proceed with this? Is there a Mac
developer who can work with me on it?

On 10/24/2017 4:40 AM, Larry Gusaas wrote:
> On 2017-10-24, 4:19 AM FR web forum wrote:
>
>> Please have a look on:
>> https://bz.apache.org/ooo/show_bug.cgi?id=127568
>
> I tested and confirm the problem. Posted my result on the bug report.
>
>

---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]

Reply | Threaded
Open this post in threaded view
|

Re: Issue for 4.1.4 and MacOS

Rory O'Farrell
On Tue, 24 Oct 2017 05:35:34 -0700
Patricia Shanahan <[hidden email]> wrote:

> Since the issue appeared in 4.1.4, and involves document links, it might
> be related to my changes.
>
> I do not have a Mac, and do not have any Mac debug experience.
>
> How do people think we should proceed with this? Is there a Mac
> developer who can work with me on it?

I don't have access to a Mac, so cannot help there.  

I make one suggestion - a guess - based on loss of images in Writer (and Calc/Impress? I'm not certain of this) on other operating systems.  

John_Ha has done some work on loss on images in Writer and relates it to sometimes occurring at the time of OO saving an Autorecovery image.  See his note at
https://forum.openoffice.org/en/forum/viewtopic.php?f=71&t=86682

As this new bug is occurring at Save time, might it be a manifestation of the same problem?

Rory


>
> On 10/24/2017 4:40 AM, Larry Gusaas wrote:
> > On 2017-10-24, 4:19 AM FR web forum wrote:
> >
> >> Please have a look on:
> >> https://bz.apache.org/ooo/show_bug.cgi?id=127568
> >
> > I tested and confirm the problem. Posted my result on the bug report.
> >
> >
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [hidden email]
> For additional commands, e-mail: [hidden email]
>
>


--
Rory O'Farrell <[hidden email]>

---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]

Reply | Threaded
Open this post in threaded view
|

Re: Issue for 4.1.4 and MacOS

Patricia Shanahan


On 10/24/2017 5:49 AM, Rory O'Farrell wrote:

> On Tue, 24 Oct 2017 05:35:34 -0700
> Patricia Shanahan <[hidden email]> wrote:
>
>> Since the issue appeared in 4.1.4, and involves document links, it might
>> be related to my changes.
>>
>> I do not have a Mac, and do not have any Mac debug experience.
>>
>> How do people think we should proceed with this? Is there a Mac
>> developer who can work with me on it?
>
> I don't have access to a Mac, so cannot help there.
>
> I make one suggestion - a guess - based on loss of images in Writer (and Calc/Impress? I'm not certain of this) on other operating systems.
>
> John_Ha has done some work on loss on images in Writer and relates it to sometimes occurring at the time of OO saving an Autorecovery image.  See his note at
> https://forum.openoffice.org/en/forum/viewtopic.php?f=71&t=86682
>
> As this new bug is occurring at Save time, might it be a manifestation of the same problem?
>

I don't think there are any changes that relate to that bug between
4.1.3 and 4.1.4. This seems to be solid and repeatable with 4.1.4, and
not happening at all with 4.1.3.

---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]

Reply | Threaded
Open this post in threaded view
|

Re: Issue for 4.1.4 and MacOS

Jim Jagielski
Will also follow up on the Bugz report, but one change between 4.1.3
and 4.1.4 regarding macOS *builds* is that 4.1.4 uses '--with-alloc=system'
but the 4.1.3 build does not.

I will create a macOS build of AOO414 but adjust the configure script
to not include that and see what effect that may have.


> On Oct 24, 2017, at 8:52 AM, Patricia Shanahan <[hidden email]> wrote:
>
>
>
> On 10/24/2017 5:49 AM, Rory O'Farrell wrote:
>> On Tue, 24 Oct 2017 05:35:34 -0700
>> Patricia Shanahan <[hidden email]> wrote:
>>> Since the issue appeared in 4.1.4, and involves document links, it might
>>> be related to my changes.
>>>
>>> I do not have a Mac, and do not have any Mac debug experience.
>>>
>>> How do people think we should proceed with this? Is there a Mac
>>> developer who can work with me on it?
>> I don't have access to a Mac, so cannot help there.
>> I make one suggestion - a guess - based on loss of images in Writer (and Calc/Impress? I'm not certain of this) on other operating systems.
>> John_Ha has done some work on loss on images in Writer and relates it to sometimes occurring at the time of OO saving an Autorecovery image.  See his note at
>> https://forum.openoffice.org/en/forum/viewtopic.php?f=71&t=86682
>> As this new bug is occurring at Save time, might it be a manifestation of the same problem?
>
> I don't think there are any changes that relate to that bug between 4.1.3 and 4.1.4. This seems to be solid and repeatable with 4.1.4, and not happening at all with 4.1.3.
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [hidden email] <mailto:[hidden email]>
> For additional commands, e-mail: [hidden email] <mailto:[hidden email]>
Reply | Threaded
Open this post in threaded view
|

Re: Issue for 4.1.4 and MacOS

Peter Kovacs-3
I do own a Mac. But its me. ;) I hope that maybe with the actual guide I can build. So far I failed to install the old code on it. If I can manage to build, I could dig into the problem. I have a long weekend upcoming. Hope I can allocate time.

Am 24. Oktober 2017 17:36:16 MESZ schrieb Jim Jagielski <[hidden email]>:

>Will also follow up on the Bugz report, but one change between 4.1.3
>and 4.1.4 regarding macOS *builds* is that 4.1.4 uses
>'--with-alloc=system'
>but the 4.1.3 build does not.
>
>I will create a macOS build of AOO414 but adjust the configure script
>to not include that and see what effect that may have.
>
>
>> On Oct 24, 2017, at 8:52 AM, Patricia Shanahan <[hidden email]> wrote:
>>
>>
>>
>> On 10/24/2017 5:49 AM, Rory O'Farrell wrote:
>>> On Tue, 24 Oct 2017 05:35:34 -0700
>>> Patricia Shanahan <[hidden email]> wrote:
>>>> Since the issue appeared in 4.1.4, and involves document links, it
>might
>>>> be related to my changes.
>>>>
>>>> I do not have a Mac, and do not have any Mac debug experience.
>>>>
>>>> How do people think we should proceed with this? Is there a Mac
>>>> developer who can work with me on it?
>>> I don't have access to a Mac, so cannot help there.
>>> I make one suggestion - a guess - based on loss of images in Writer
>(and Calc/Impress? I'm not certain of this) on other operating systems.
>>> John_Ha has done some work on loss on images in Writer and relates
>it to sometimes occurring at the time of OO saving an Autorecovery
>image.  See his note at
>>> https://forum.openoffice.org/en/forum/viewtopic.php?f=71&t=86682
>>> As this new bug is occurring at Save time, might it be a
>manifestation of the same problem?
>>
>> I don't think there are any changes that relate to that bug between
>4.1.3 and 4.1.4. This seems to be solid and repeatable with 4.1.4, and
>not happening at all with 4.1.3.
>>
>> ---------------------------------------------------------------------
>> To unsubscribe, e-mail: [hidden email]
><mailto:[hidden email]>
>> For additional commands, e-mail: [hidden email]
><mailto:[hidden email]>

---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]

Reply | Threaded
Open this post in threaded view
|

Re: Issue for 4.1.4 and MacOS

Patricia Shanahan
Even if you can't get a Mac build going in time to help with this bug,
we obviously need more people building and testing on Mac.

On 10/24/2017 4:01 PM, Peter kovacs wrote:

> I do own a Mac. But its me. ;) I hope that maybe with the actual guide I can build. So far I failed to install the old code on it. If I can manage to build, I could dig into the problem. I have a long weekend upcoming. Hope I can allocate time.
>
> Am 24. Oktober 2017 17:36:16 MESZ schrieb Jim Jagielski <[hidden email]>:
>> Will also follow up on the Bugz report, but one change between 4.1.3
>> and 4.1.4 regarding macOS *builds* is that 4.1.4 uses
>> '--with-alloc=system'
>> but the 4.1.3 build does not.
>>
>> I will create a macOS build of AOO414 but adjust the configure script
>> to not include that and see what effect that may have.
>>
>>
>>> On Oct 24, 2017, at 8:52 AM, Patricia Shanahan <[hidden email]> wrote:
>>>
>>>
>>>
>>> On 10/24/2017 5:49 AM, Rory O'Farrell wrote:
>>>> On Tue, 24 Oct 2017 05:35:34 -0700
>>>> Patricia Shanahan <[hidden email]> wrote:
>>>>> Since the issue appeared in 4.1.4, and involves document links, it
>> might
>>>>> be related to my changes.
>>>>>
>>>>> I do not have a Mac, and do not have any Mac debug experience.
>>>>>
>>>>> How do people think we should proceed with this? Is there a Mac
>>>>> developer who can work with me on it?
>>>> I don't have access to a Mac, so cannot help there.
>>>> I make one suggestion - a guess - based on loss of images in Writer
>> (and Calc/Impress? I'm not certain of this) on other operating systems.
>>>> John_Ha has done some work on loss on images in Writer and relates
>> it to sometimes occurring at the time of OO saving an Autorecovery
>> image.  See his note at
>>>> https://forum.openoffice.org/en/forum/viewtopic.php?f=71&t=86682
>>>> As this new bug is occurring at Save time, might it be a
>> manifestation of the same problem?
>>>
>>> I don't think there are any changes that relate to that bug between
>> 4.1.3 and 4.1.4. This seems to be solid and repeatable with 4.1.4, and
>> not happening at all with 4.1.3.
>>>
>>> ---------------------------------------------------------------------
>>> To unsubscribe, e-mail: [hidden email]
>> <mailto:[hidden email]>
>>> For additional commands, e-mail: [hidden email]
>> <mailto:[hidden email]>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [hidden email]
> For additional commands, e-mail: [hidden email]
>

---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]

Reply | Threaded
Open this post in threaded view
|

Re: Issue for 4.1.4 and MacOS

Dave Fisher
Hi -

I’ve been trying to get Mac builds going today. I am trying right now with trunk. I’m still stuck on configure and have an XCodeLegacy question out to Jim on another thread.

checking which jpeg to use... internal
checking which expat to use... internal
checking which libxslt to use... external
checking for xsltproc... /usr/bin/xsltproc
checking which libxml to use... external
checking checking SDK compatibility with OSX 10.9... configure: error: No SDK with OSX 10.9 compatibility found
../build_aooreleases.sh: line 153: ./bootstrap: No such file or directory

I have time to try tomorrow.

Regards,
Dave


On Oct 24, 2017, at 4:06 PM, Patricia Shanahan <[hidden email]> wrote:

Even if you can't get a Mac build going in time to help with this bug, we obviously need more people building and testing on Mac.

On 10/24/2017 4:01 PM, Peter kovacs wrote:
I do own a Mac. But its me. ;) I hope that maybe with the actual guide I can build. So far I failed to install the old code on it. If I can manage to build, I could dig into the problem. I have a long weekend upcoming. Hope I can allocate time.
Am 24. Oktober 2017 17:36:16 MESZ schrieb Jim Jagielski <[hidden email]>:
Will also follow up on the Bugz report, but one change between 4.1.3
and 4.1.4 regarding macOS *builds* is that 4.1.4 uses
'--with-alloc=system'
but the 4.1.3 build does not.

I will create a macOS build of AOO414 but adjust the configure script
to not include that and see what effect that may have.


On Oct 24, 2017, at 8:52 AM, Patricia Shanahan <[hidden email]> wrote:



On 10/24/2017 5:49 AM, Rory O'Farrell wrote:
On Tue, 24 Oct 2017 05:35:34 -0700
Patricia Shanahan <[hidden email]> wrote:
Since the issue appeared in 4.1.4, and involves document links, it
might
be related to my changes.

I do not have a Mac, and do not have any Mac debug experience.

How do people think we should proceed with this? Is there a Mac
developer who can work with me on it?
I don't have access to a Mac, so cannot help there.
I make one suggestion - a guess - based on loss of images in Writer
(and Calc/Impress? I'm not certain of this) on other operating systems.
John_Ha has done some work on loss on images in Writer and relates
it to sometimes occurring at the time of OO saving an Autorecovery
image.  See his note at
https://forum.openoffice.org/en/forum/viewtopic.php?f=71&t=86682
As this new bug is occurring at Save time, might it be a
manifestation of the same problem?

I don't think there are any changes that relate to that bug between
4.1.3 and 4.1.4. This seems to be solid and repeatable with 4.1.4, and
not happening at all with 4.1.3.

---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
<[hidden email]>
For additional commands, e-mail: [hidden email]
<[hidden email]>
---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]

---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]



signature.asc (817 bytes) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: Issue for 4.1.4 and MacOS

Jim Jagielski
In reply to this post by Jim Jagielski
I did a quick rebuild but with the '--with-alloc=system'
line removed. No change in behavior.

Will try to track down the question Patricia
raised in the Bugz report.

> On Oct 24, 2017, at 11:36 AM, Jim Jagielski <[hidden email]> wrote:
>
> Will also follow up on the Bugz report, but one change between 4.1.3
> and 4.1.4 regarding macOS *builds* is that 4.1.4 uses '--with-alloc=system'
> but the 4.1.3 build does not.
>
> I will create a macOS build of AOO414 but adjust the configure script
> to not include that and see what effect that may have.
>
>
>> On Oct 24, 2017, at 8:52 AM, Patricia Shanahan <[hidden email]> wrote:
>>
>>
>>
>> On 10/24/2017 5:49 AM, Rory O'Farrell wrote:
>>> On Tue, 24 Oct 2017 05:35:34 -0700
>>> Patricia Shanahan <[hidden email]> wrote:
>>>> Since the issue appeared in 4.1.4, and involves document links, it might
>>>> be related to my changes.
>>>>
>>>> I do not have a Mac, and do not have any Mac debug experience.
>>>>
>>>> How do people think we should proceed with this? Is there a Mac
>>>> developer who can work with me on it?
>>> I don't have access to a Mac, so cannot help there.
>>> I make one suggestion - a guess - based on loss of images in Writer (and Calc/Impress? I'm not certain of this) on other operating systems.
>>> John_Ha has done some work on loss on images in Writer and relates it to sometimes occurring at the time of OO saving an Autorecovery image.  See his note at
>>> https://forum.openoffice.org/en/forum/viewtopic.php?f=71&t=86682
>>> As this new bug is occurring at Save time, might it be a manifestation of the same problem?
>>
>> I don't think there are any changes that relate to that bug between 4.1.3 and 4.1.4. This seems to be solid and repeatable with 4.1.4, and not happening at all with 4.1.3.
>>
>> ---------------------------------------------------------------------
>> To unsubscribe, e-mail: [hidden email] <mailto:[hidden email]>
>> For additional commands, e-mail: [hidden email] <mailto:[hidden email]>


---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]

Reply | Threaded
Open this post in threaded view
|

Re: Issue for 4.1.4 and MacOS

Jim Jagielski
I set breakpoints on both:

   (lldb) breakpoint set --file main/sfx2/source/appl/linkmgr2.cxx --line 288
   (lldb) breakpoint set -n LinkManager::GetUserAllowsLinkUpdate
Breakpoint 4: where = libsfx.dylib`sfx2::LinkManager::GetUserAllowsLinkUpdate(Window*), address = 0x000000010070a260

and when opening up the corrupted file, neither breakpoint
was reached. :/
---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]

Reply | Threaded
Open this post in threaded view
|

Re: Issue for 4.1.4 and MacOS

Patricia Shanahan
OK, forget that hypothesis. You have even done the next test I would
have asked for if the line 288 breakpoint was not reached.

I'll need to think up a new hypothesis. I'm planning to go groom a
couple of horses this morning, but I'll work on this in the afternoon.

On 10/26/2017 5:01 AM, Jim Jagielski wrote:

> I set breakpoints on both:
>
>     (lldb) breakpoint set --file main/sfx2/source/appl/linkmgr2.cxx --line 288
>     (lldb) breakpoint set -n LinkManager::GetUserAllowsLinkUpdate
> Breakpoint 4: where = libsfx.dylib`sfx2::LinkManager::GetUserAllowsLinkUpdate(Window*), address = 0x000000010070a260
>
> and when opening up the corrupted file, neither breakpoint
> was reached. :/
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [hidden email]
> For additional commands, e-mail: [hidden email]
>

---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]

Reply | Threaded
Open this post in threaded view
|

Re: Issue for 4.1.4 and MacOS

Jim Jagielski
Yeah, I kinda figured out what your hypothesis was... I'll need
to think about this some more.

> On Oct 26, 2017, at 8:34 AM, Patricia Shanahan <[hidden email]> wrote:
>
> OK, forget that hypothesis. You have even done the next test I would have asked for if the line 288 breakpoint was not reached.
>
> I'll need to think up a new hypothesis. I'm planning to go groom a couple of horses this morning, but I'll work on this in the afternoon.
>
> On 10/26/2017 5:01 AM, Jim Jagielski wrote:
>> I set breakpoints on both:
>>    (lldb) breakpoint set --file main/sfx2/source/appl/linkmgr2.cxx --line 288
>>    (lldb) breakpoint set -n LinkManager::GetUserAllowsLinkUpdate
>> Breakpoint 4: where = libsfx.dylib`sfx2::LinkManager::GetUserAllowsLinkUpdate(Window*), address = 0x000000010070a260
>> and when opening up the corrupted file, neither breakpoint
>> was reached. :/
>> ---------------------------------------------------------------------
>> To unsubscribe, e-mail: [hidden email]
>> For additional commands, e-mail: [hidden email]
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [hidden email]
> For additional commands, e-mail: [hidden email]
>


---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]

Reply | Threaded
Open this post in threaded view
|

Re: Issue for 4.1.4 and MacOS

Patricia Shanahan
I at first thought this might be a load problem. It is now clear it is a
case of file corruption on save, only on Mac.

Unfortunately, that means I am out of ideas, other than checking that
your original test has been done on the file save. The fixes I know
about are intended to affect file loading, not saving.

On 10/26/2017 5:45 AM, Jim Jagielski wrote:

> Yeah, I kinda figured out what your hypothesis was... I'll need
> to think about this some more.
>> On Oct 26, 2017, at 8:34 AM, Patricia Shanahan <[hidden email]> wrote:
>>
>> OK, forget that hypothesis. You have even done the next test I would have asked for if the line 288 breakpoint was not reached.
>>
>> I'll need to think up a new hypothesis. I'm planning to go groom a couple of horses this morning, but I'll work on this in the afternoon.
>>
>> On 10/26/2017 5:01 AM, Jim Jagielski wrote:
>>> I set breakpoints on both:
>>>     (lldb) breakpoint set --file main/sfx2/source/appl/linkmgr2.cxx --line 288
>>>     (lldb) breakpoint set -n LinkManager::GetUserAllowsLinkUpdate
>>> Breakpoint 4: where = libsfx.dylib`sfx2::LinkManager::GetUserAllowsLinkUpdate(Window*), address = 0x000000010070a260
>>> and when opening up the corrupted file, neither breakpoint
>>> was reached. :/
>>> ---------------------------------------------------------------------
>>> To unsubscribe, e-mail: [hidden email]
>>> For additional commands, e-mail: [hidden email]
>>
>> ---------------------------------------------------------------------
>> To unsubscribe, e-mail: [hidden email]
>> For additional commands, e-mail: [hidden email]
>>
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [hidden email]
> For additional commands, e-mail: [hidden email]
>

---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]

Reply | Threaded
Open this post in threaded view
|

Re: Issue for 4.1.4 and MacOS

Dave Fisher
Hi -

I’ve attached a diff of the differences between 4.1.3 and 4.1.4. I hope that this helps. If I had a working build I would be looking into undoing a change at a time to see which had the impact on the save of links on Mac.

I am also suspicious of tooling and what caused the link xml to be created, but empty. Is there a system library involved and an impact due to fixes and updates in Xcode? If aoo413 is built on the same tooling as aoo414 is it broken too?

Regards,
Dave

> On Oct 28, 2017, at 5:47 AM, Patricia Shanahan <[hidden email]> wrote:
>
> I at first thought this might be a load problem. It is now clear it is a case of file corruption on save, only on Mac.
>
> Unfortunately, that means I am out of ideas, other than checking that your original test has been done on the file save. The fixes I know about are intended to affect file loading, not saving.
>
> On 10/26/2017 5:45 AM, Jim Jagielski wrote:
>> Yeah, I kinda figured out what your hypothesis was... I'll need
>> to think about this some more.
>>> On Oct 26, 2017, at 8:34 AM, Patricia Shanahan <[hidden email]> wrote:
>>>
>>> OK, forget that hypothesis. You have even done the next test I would have asked for if the line 288 breakpoint was not reached.
>>>
>>> I'll need to think up a new hypothesis. I'm planning to go groom a couple of horses this morning, but I'll work on this in the afternoon.
>>>
>>> On 10/26/2017 5:01 AM, Jim Jagielski wrote:
>>>> I set breakpoints on both:
>>>>    (lldb) breakpoint set --file main/sfx2/source/appl/linkmgr2.cxx --line 288
>>>>    (lldb) breakpoint set -n LinkManager::GetUserAllowsLinkUpdate
>>>> Breakpoint 4: where = libsfx.dylib`sfx2::LinkManager::GetUserAllowsLinkUpdate(Window*), address = 0x000000010070a260
>>>> and when opening up the corrupted file, neither breakpoint
>>>> was reached. :/
>>>> ---------------------------------------------------------------------
>>>> To unsubscribe, e-mail: [hidden email]
>>>> For additional commands, e-mail: [hidden email]
>>>
>>> ---------------------------------------------------------------------
>>> To unsubscribe, e-mail: [hidden email]
>>> For additional commands, e-mail: [hidden email]
>>>
>> ---------------------------------------------------------------------
>> To unsubscribe, e-mail: [hidden email]
>> For additional commands, e-mail: [hidden email]
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [hidden email]
> For additional commands, e-mail: [hidden email]
>


signature.asc (817 bytes) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: Issue for 4.1.4 and MacOS

Patricia Shanahan
The original bug report said "This wasn't an issue of AOO 4.1.3, but is
an issue of AOO 4.1.4" so we are looking for something that changed.

I would have assumed that the same libraries would be used in writing
the main content.xml and the one in the 'Object 1' subdirectory. The
main content.xml is fine. The one in the 'Object 1' subdirectory is
empty, so I am confused and puzzled.

Can someone with Mac building and testing capability do the search for
the change that made it fail? In another project, I used binary search
on the SVN revision number to locate exactly when a regression appeared.

(For anyone who is not already familiar with the structure of the files,
it may be useful to look at an unzipped .ods file.)

On 10/28/2017 10:01 AM, Dave Fisher wrote:

> Hi -
>
> I’ve attached a diff of the differences between 4.1.3 and 4.1.4. I hope that this helps. If I had a working build I would be looking into undoing a change at a time to see which had the impact on the save of links on Mac.
>
> I am also suspicious of tooling and what caused the link xml to be created, but empty. Is there a system library involved and an impact due to fixes and updates in Xcode? If aoo413 is built on the same tooling as aoo414 is it broken too?
>
> Regards,
> Dave
>
>> On Oct 28, 2017, at 5:47 AM, Patricia Shanahan <[hidden email]> wrote:
>>
>> I at first thought this might be a load problem. It is now clear it is a case of file corruption on save, only on Mac.
>>
>> Unfortunately, that means I am out of ideas, other than checking that your original test has been done on the file save. The fixes I know about are intended to affect file loading, not saving.
>>
>> On 10/26/2017 5:45 AM, Jim Jagielski wrote:
>>> Yeah, I kinda figured out what your hypothesis was... I'll need
>>> to think about this some more.
>>>> On Oct 26, 2017, at 8:34 AM, Patricia Shanahan <[hidden email]> wrote:
>>>>
>>>> OK, forget that hypothesis. You have even done the next test I would have asked for if the line 288 breakpoint was not reached.
>>>>
>>>> I'll need to think up a new hypothesis. I'm planning to go groom a couple of horses this morning, but I'll work on this in the afternoon.
>>>>
>>>> On 10/26/2017 5:01 AM, Jim Jagielski wrote:
>>>>> I set breakpoints on both:
>>>>>     (lldb) breakpoint set --file main/sfx2/source/appl/linkmgr2.cxx --line 288
>>>>>     (lldb) breakpoint set -n LinkManager::GetUserAllowsLinkUpdate
>>>>> Breakpoint 4: where = libsfx.dylib`sfx2::LinkManager::GetUserAllowsLinkUpdate(Window*), address = 0x000000010070a260
>>>>> and when opening up the corrupted file, neither breakpoint
>>>>> was reached. :/
>>>>> ---------------------------------------------------------------------
>>>>> To unsubscribe, e-mail: [hidden email]
>>>>> For additional commands, e-mail: [hidden email]
>>>>
>>>> ---------------------------------------------------------------------
>>>> To unsubscribe, e-mail: [hidden email]
>>>> For additional commands, e-mail: [hidden email]
>>>>
>>> ---------------------------------------------------------------------
>>> To unsubscribe, e-mail: [hidden email]
>>> For additional commands, e-mail: [hidden email]
>>
>> ---------------------------------------------------------------------
>> To unsubscribe, e-mail: [hidden email]
>> For additional commands, e-mail: [hidden email]
>>
>

---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]

Reply | Threaded
Open this post in threaded view
|

Re: Issue for 4.1.4 and MacOS

Dave Fisher
Hi -

> On Oct 28, 2017, at 10:30 AM, Patricia Shanahan <[hidden email]> wrote:
>
> The original bug report said "This wasn't an issue of AOO 4.1.3, but is an issue of AOO 4.1.4" so we are looking for something that changed.
>
> I would have assumed that the same libraries would be used in writing the main content.xml and the one in the 'Object 1' subdirectory. The main content.xml is fine. The one in the 'Object 1' subdirectory is empty, so I am confused and puzzled.
>
> Can someone with Mac building and testing capability do the search for the change that made it fail? In another project, I used binary search on the SVN revision number to locate exactly when a regression appeared.

I think that only Jim can build 4.1.4 on Mac. I am blocked waiting for a friend to give me a Mac JDK7. Or, I’ll apply Damjan’s javadoc JDK8 patches and see.

Keep in mind that 4.1.3 was built over a year ago. AFIAK it was not built by Jim. Apple has updated Xcode and switched MacOSX to macOS since then. We ought to factor out the build tooling by comparing a build based on 4.1.3 with the bug to rule out a tooling regression.

>
> (For anyone who is not already familiar with the structure of the files, it may be useful to look at an unzipped .ods file.)

Regards,
Dave


>
> On 10/28/2017 10:01 AM, Dave Fisher wrote:
>> Hi -
>> I’ve attached a diff of the differences between 4.1.3 and 4.1.4. I hope that this helps. If I had a working build I would be looking into undoing a change at a time to see which had the impact on the save of links on Mac.
>> I am also suspicious of tooling and what caused the link xml to be created, but empty. Is there a system library involved and an impact due to fixes and updates in Xcode? If aoo413 is built on the same tooling as aoo414 is it broken too?
>> Regards,
>> Dave
>>> On Oct 28, 2017, at 5:47 AM, Patricia Shanahan <[hidden email]> wrote:
>>>
>>> I at first thought this might be a load problem. It is now clear it is a case of file corruption on save, only on Mac.
>>>
>>> Unfortunately, that means I am out of ideas, other than checking that your original test has been done on the file save. The fixes I know about are intended to affect file loading, not saving.
>>>
>>> On 10/26/2017 5:45 AM, Jim Jagielski wrote:
>>>> Yeah, I kinda figured out what your hypothesis was... I'll need
>>>> to think about this some more.
>>>>> On Oct 26, 2017, at 8:34 AM, Patricia Shanahan <[hidden email]> wrote:
>>>>>
>>>>> OK, forget that hypothesis. You have even done the next test I would have asked for if the line 288 breakpoint was not reached.
>>>>>
>>>>> I'll need to think up a new hypothesis. I'm planning to go groom a couple of horses this morning, but I'll work on this in the afternoon.
>>>>>
>>>>> On 10/26/2017 5:01 AM, Jim Jagielski wrote:
>>>>>> I set breakpoints on both:
>>>>>>    (lldb) breakpoint set --file main/sfx2/source/appl/linkmgr2.cxx --line 288
>>>>>>    (lldb) breakpoint set -n LinkManager::GetUserAllowsLinkUpdate
>>>>>> Breakpoint 4: where = libsfx.dylib`sfx2::LinkManager::GetUserAllowsLinkUpdate(Window*), address = 0x000000010070a260
>>>>>> and when opening up the corrupted file, neither breakpoint
>>>>>> was reached. :/
>>>>>> ---------------------------------------------------------------------
>>>>>> To unsubscribe, e-mail: [hidden email]
>>>>>> For additional commands, e-mail: [hidden email]
>>>>>
>>>>> ---------------------------------------------------------------------
>>>>> To unsubscribe, e-mail: [hidden email]
>>>>> For additional commands, e-mail: [hidden email]
>>>>>
>>>> ---------------------------------------------------------------------
>>>> To unsubscribe, e-mail: [hidden email]
>>>> For additional commands, e-mail: [hidden email]
>>>
>>> ---------------------------------------------------------------------
>>> To unsubscribe, e-mail: [hidden email]
>>> For additional commands, e-mail: [hidden email]
>>>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [hidden email]
> For additional commands, e-mail: [hidden email]
>


signature.asc (817 bytes) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: Issue for 4.1.4 and MacOS

Andrea Pescetti-2
Dave Fisher wrote:
> We ought to factor out the build tooling by comparing a build based on 4.1.3 with the bug to rule out a tooling regression.

Very good suggestion. Building 4.1.3 on exactly the same system and with
exactly the same config used for 4.1.4 will tell us whether this is
related to the code or to the build options or the build machine.

Regards,
   Andrea.

---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]

Reply | Threaded
Open this post in threaded view
|

Re: Issue for 4.1.4 and MacOS

Dave Fisher

> On Oct 28, 2017, at 12:11 PM, Andrea Pescetti <[hidden email]> wrote:
>
> Dave Fisher wrote:
>> We ought to factor out the build tooling by comparing a build based on 4.1.3 with the bug to rule out a tooling regression.
>
> Very good suggestion. Building 4.1.3 on exactly the same system and with exactly the same config used for 4.1.4 will tell us whether this is related to the code or to the build options or the build machine.

A reminder from the 4.1.x is frozen and no updates thread. Macs builds are not done on VMs.

Regards,
Dave

>
> Regards,
>  Andrea.
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [hidden email]
> For additional commands, e-mail: [hidden email]
>


signature.asc (817 bytes) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: Issue for 4.1.4 and MacOS

Patricia Shanahan
In reply to this post by Andrea Pescetti-2
On 10/28/2017 12:11 PM, Andrea Pescetti wrote:
> Dave Fisher wrote:
>> We ought to factor out the build tooling by comparing a build based on
>> 4.1.3 with the bug to rule out a tooling regression.
>
> Very good suggestion. Building 4.1.3 on exactly the same system and with
> exactly the same config used for 4.1.4 will tell us whether this is
> related to the code or to the build options or the build machine.

+1

---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]

12345