On Sun, Jan 5, 2014 at 2:51 PM, Hagar Delest <hagar.del...@laposte.net> wrote:
> Top posting.
> OK, you win. I'm fed up with this kind of discussion, so last post in this
> discussion.
> I won't ask your questions because when we ask even a basic one, very often
> we get no reply (I guess that frustration is for something).
>

Clearly users are mainly interested in getting their document back.
But if your immediately response is "kiss your document good bye" then
you will not hear from them again.  If we want to help the project,
then we need to handle users in a certain way.  Think of a policeman
interviewing a robbery victim.  We need to ask the questions even
though the person is upset.  We need to tell them that this
information is important and that the only way we can improve is by
having this info.

> After almost 8 years supporting users in forums, I think that we have a
> reasonable 6th sense to spot the smoking guns in such problems. I'll
> continue to record the cases in the forum. Even if you're the most vocal one
> on the mailing list, I hope that some dev will try to have a look.
>

You forget that I did professional technical support for Lotus years
ago, supporting 1-2-3 and other products.  So I am entitled to claim
reasonable sense in this area as well.  And what I sense is that over
multiple years what we're doing now has not helped in narrowing down
the problem or getting it fixed.  What we're doing now has failed to
resolve the issue.  You're welcome to continue in the same way, but
I'm making suggestions for how we can improve, not just repeat the
past.  If there is a problem I'd like to see it narrowed down and
fixed.  If it is user error then I'd like to establish that firmly so
we can improve documentation to reduce its occurrence.  What I don't
want to do is spend another 5 years to turn a list of 150 sparsely
documented examples into a list of 300 sparsely documented examples
and be no better off on resolving this either way.  I hope you would
agree that "more of the same" is not going to work here.

> If you think that crashed filed are like UFOs then let users experience
> UFOs.
>

The comparison should be more like this:  Determining the existence of
nonexistence of UFO's requires more than just compiling a list of
incidents.  Large lists of UFO sightings has proven nothing and will
not prove anything.  The sa,e is true of software crashes.

Regards,

-Rob


> Hagar
>
>
> Le 05/01/2014 19:45, Rob Weir a écrit :
>
>
>> On Sun, Jan 5, 2014 at 12:35 PM, Hagar Delest <hagar.del...@laposte.net>
>> wrote:
>>>
>>> Le 05/01/2014 17:05, Rob Weir a écrit :
>>>
>>>> It would be great to ask for information like this whenever someone
>>>> reports this kind of problem.  150 reports without this detail are
>>>> useless.  But even 10 reports with this detail might indicate a
>>>> pattern.
>>>>
>>>> 1) What AOO version is in use?
>>>>
>>>> 2) What OS version?
>>>>
>>>> 3) What file type (extension) was being saved?
>>>>
>>>> 4) Where was the file being saved?  USB?  Network drive?
>>>>
>>>> 5) Is autosaved enabled?
>>>>
>>>> 6) When you returned to your computer was it in the same state?  For
>>>> example, had you lost power?  Did the OS force a reboot?  Did your
>>>> laptop hibernate?
>>>
>>>
>>>
>>> Have you at least read the topic and the post with all the cases I've
>>> recorded???
>>> You'll see that most of these information are available. And each linked
>>> topic usually have also these information (with more details of course).
>>> -> https://forum.openoffice.org/en/forum/viewtopic.php?f=6&t=17677#p81363
>>>
>>
>> Well, obviously reports with partial information from 5 years ago are
>> not really useful.  As I pointed out, I can show over 300 UFO
>> sightings from just December.  So what?
>>
>> I'm suggesting collecting this information systematically, for new
>> reports, in 2014 using AOO 4.0.1.  For example, you did not ask all of
>> these questions when you responded to the user on this list just now.
>>
>>>
>>>
>>>> Just anecdotally, and without deeper analysis, I see a number of
>>>> reports on OpenOffice and with Microsoft Office, where a USB memory
>>>> stick is being used.   Savvy users know how to properly remove a
>>>> memory stick.  But not all users do.  This can cause problems.
>>>
>>>
>>>
>>> None of the 150+ cases I've recorded involve USB drives. Of course we
>>> know
>>> that case. That's why I've discarded them systematically.
>>> NB: one more case this very day.
>>>
>>
>> This isn't really true.  I did a spot check of the reports and some of
>> them did involve USB drives and for many of them the question was not
>> even asked.
>>
>>>
>>>
>>>> Another case to watch out for is old Wordperfect files.   A user saves
>>>> a WPD file, upgrades OOo 3.3.0 to AOO 4.0.1 and now their file won't
>>>> open.  But this is due to the loss of WPD support, not due to damage
>>>> to the file, though the symptoms look the same at first.
>>>
>>>
>>>
>>> Again, just read the topic. Most of the files are ODF, some .doc. No
>>> exotic
>>> format like WPD.
>>>
>>>
>>>
>>>> Another thing to look for is a forced reboot, the kind that recent
>>>> versions of Windows do when installing a critical security patch.
>>>> Some antivirus software does this as well.  If you have a document
>>>> loaded in OpenOffice with unsaved changed, and have autosave enabled,
>>>> and leave your machine on for a week, with OpenOffice running, and a
>>>> system restart is forced, what will happen?  Is there a correlation to
>>>> problems in that scenario?
>>>
>>>
>>>
>>> And?
>>> Even if the boot is forced, as I said, it would be understandable to lose
>>> the version that is currently under edition. But it is unacceptable to
>>> lose
>>> the saved file.
>>>
>>
>> And it would be unacceptable to see a UFO.  The question is whether
>> any actually have or whether it is user error.
>>
>>> I know that this is a difficult problem but please stop trying to find a
>>> scapegoat.
>>> If nobody wants to investigate, I don't have any problem as long as I
>>> don't
>>> suffer from this issue.
>>> But then, just accept the rants from users who lost important data.
>>> NB: I've switched from MS Word to OOo after I lost data. If I ever face
>>> this
>>> problem, that would be the end of my use of AOO. Users want reliability
>>> first. All the nice features you can put in an application will never
>>> counterbalance any important data loss.
>>>
>>
>> No one is scapegoating.  I'm just saying with 100 million users and 5
>> years and not a single reproducible error?  Really?  Seriously?
>>
>> -Rob
>>
>>
>>> Hagar
>>>
>>>
>>> ---------------------------------------------------------------------
>>> To unsubscribe, e-mail: users-unsubscr...@openoffice.apache.org
>>> For additional commands, e-mail: users-h...@openoffice.apache.org
>>>
>>
>> ---------------------------------------------------------------------
>> To unsubscribe, e-mail: users-unsubscr...@openoffice.apache.org
>> For additional commands, e-mail: users-h...@openoffice.apache.org
>>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: users-unsubscr...@openoffice.apache.org
> For additional commands, e-mail: users-h...@openoffice.apache.org
>

---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscr...@openoffice.apache.org
For additional commands, e-mail: users-h...@openoffice.apache.org

Reply via email to