[GNC] Error While Saving in GnuCash 3.5 on Win7-64

Fri Apr 5 23:00:51 EDT 2019

Greg,

No worries, but yes, you’ve guessed correctly.

There is a hidden header with a subject ID that is carried over when you hit ‘reply’. This is what threading is based off of, (for most mail clients) not the text characters in the subject line. (you can see this header line if you view the raw message source) Gmail sometimes tries to outsmart that header, sometimes successfully, sometimes not.

To guarantee that you start a new thread, simply start a new e-mail and address it to gnucash-user at gnucash.org.

After doing this once, Gmail should offer the sending address up in full as you start typing gnuc... to make entering it faster.

If I had a nickel for every time someone e-mailed me by pulling up an ancient message and hitting reply instead of just starting to type my name and let their mail app dutifully auto-suggest my already stored address... Then end result is getting new messages threaded into subjects that are months, and in some cases, years, yes years, old. (this is really baffling in the age of smart phones where the normal (?) workflow is to pull up the contact first, then hit either the call, text, or e-mail icons)

In short, unless you are *actually* and *literally* (rare proper double usage there) replying to the content of a specific message, don’t use the reply button. Use the new message button. (this is for any message, not just here on the GnuCash list. You’ll save your correspondents lots of potential grief, or at least, avoid having your new message end up lost or passed over.)

Regards,
Adrien

> On Apr 5, 2019, at 8:31 PM, Greg Feneis <mfeneis at gmail.com> wrote:
> 
> Adrien, Geert,
> 
> From the gmail web interface, I hit reply to the latest email I received
> from this list and changed the subject line to be what I wanted, and
> deleted the entire body before typing up what I wanted to say.  I thought
> it was the subject text that directed emails into various conversations
> (threads).  It seems to in gmail.  Apparently there's some stuff in the
> email's headers that make the email appear to belong to another thread.
> Knowing this, from now on, I'll start threads using virgin emails.
> 
> Kind regards,
> 
> Greg
> 
> 
> 
> 
> On Fri, Apr 5, 2019 at 10:20 AM Adrien Monteleone <
> adrien.monteleone at lusfiber.net> wrote:
> 
>> It threaded that way for me as well. But then once the first reply came
>> back, it somehow magically decoupled. I chalked it up to a Mail.app hiccup.
>> Now I’m intrigued as to how it got separated. Curious to see what the
>> Mailman archive page shows.
>> 
>> Regards,
>> Adrien
>> 
>>> On Apr 5, 2019, at 11:06 AM, Geert Janssens <geert.gnucash at kobaltwit.be>
>> wrote:
>>> 
>>> Hi Greg,
>>> 
>>> It was "hijacked" in the sense that you seem to have started it by
>> replying to
>>> John Ralls' reply on the "help lost data file" thread. So my mail client
>> did
>>> hide your message in that thread. Apparently that also happened for
>> Frank.
>>> 
>>> Of course I can't *know* what you did exactly. It just *looks* like this
>> is
>>> what you did: take an existing message, hit reply (or reply to
>> list/all),
>>> change subject and body, send. That would cause such a message to end up
>>> hidden in the original thread.
>>> 
>>> In any case I also didn't immediately see your message.
>>> 
>>> Regards,
>>> 
>>> Geert
>>> 
>>> Op vrijdag 5 april 2019 15:56:24 CEST schreef Greg Feneis:
>>>> Hi Frank,
>>>> 
>>>> I started this thread. I've been driving this thread since I started it.
>>>> What made you think I hijacked?
>>>> 
>>>> I was aware of bugzilla, but it's not an email list.  I had some concern
>>>> that the process had changed from when I last monitored this list many
>>>> months ago.  Previously an error like this gets discussed, and someone,
>> a
>>>> developer or person who works more closely with the project, asks the
>>>> reporter to log it to bugzilla.
>>>> 
>>>> Kind regards,
>>>> 
>>>> Greg Feneis <http://www.linkedin.com/in/electromechanical>
>>>> 
>>>> 
>>>> 
>>>> 
>>>> On Thu, Apr 4, 2019 at 10:43 PM Frank H. Ellenberger <
>>>> 
>>>> frank.h.ellenberger at gmail.com> wrote:
>>>>> Am 04.04.19 um 17:28 schrieb Greg Feneis:
>>>>>> Is there some other gnucash list I should use to report program
>> errors?
>>>>> 
>>>>> Yes, Greg,
>>>>> read https://wiki.gnucash.org/wiki/Bugzilla for the details.
>>>>> 
>>>>> BTW: Next time instead of hijacking other threads start a new one. ;-)
>>>>> 
>>>>> Regards frank
>>>>> 
>>>>>> Kind regards,
>>>>>> 
>>>>>> Greg Feneis <http://www.linkedin.com/in/electromechanical>
>>>>>> 
>>>>>> On Tue, Apr 2, 2019 at 5:41 PM Greg Feneis <mfeneis at gmail.com> wrote:
>>>>>>> Hi,
>>>>>>> 
>>>>>>> I just upgraded to 3.5 from 2.6.21.
>>>>>>> 
>>>>>>> I've been using 2.6.x for a few years, and never had trouble saving
>> to
>>>>>>> this directory, but after upgrading to 3.5, I tried to save and it
>>>>> 
>>>>> gives me
>>>>> 
>>>>>>> an error message.
>>>>>>> 
>>>>>>> "You attempted to save in "can't be displayed" or a subdirectory
>>>>> 
>>>>> thereof.
>>>>> 
>>>>>>> This is not allowed as GnuCash reserves that director for internal
>> use.
>>>>>>> 
>>>>>>> Please try again in a different directory."  (close)
>>>>>>> 
>>>>>>> 
>>>>>>> Here's a link to a screen shot of the error message:
>>>>>>> 
>> https://www.dropbox.com/s/1opqys3wxuyw4rv/2019-04-02_17-05-34.png?dl=0
>>>>>>> 
>>>>>>> Note that where I wrote "can't be displayed" is actually a strange
>>>>> 
>>>>> graphic
>>>>> 
>>>>>>> like when an emoji can't be displayed.  See the screen shot.
>>>>>>> 
>>>>>>> It appears when the error message was generated, the file path
>> couldn't
>>>>> 
>>>>> be
>>>>> 
>>>>>>> parsed.  This makes me think GnuCash isn't capable of parsing the
>> save
>>>>> 
>>>>> to
>>>>> 
>>>>>>> path.  I suspect it's a path length issue.  I tried shorter paths and
>>>>>>> it
>>>>>>> worked.  FSR, the new GnuCash seem like it can't save to the same
>>>>>>> length
>>>>>>> path that the old series was capable of.
>>>>>>> 
>>>>>>> Sorry if this is already a known issue
>>>>>>> 
>>>>>>> Kind regards,
>>>>>>> 
>>>>>>> Greg Feneis <http://www.linkedin.com/in/electromechanical>