[GNC] Transactions in Placeholder

[David T.]

Michael, 

The OP problem had to do with the importer adding transactions to a placeholder account. So that's why I am discussing that, rather than bringing up the creation of new accounts in the process. I'm not even sure how that applies here; a newly-created account during the import process (if such is even achievable in the import process) would presumably not be set as a placeholder account. 

GnuCash already prevents a user from entering transactions into a placeholder account-- and even prevents you from opening such accounts without an explicit (non standard) open mode. The only way I know of entering transactions into a placeholder account is to turn off the setting first. Of course, then, it's not a placeholder account...

Again, in my example, you're missing the point. I'm not talking about the special case where a user has elevated rights and logs in specially to use them. I'm talking about a system allowing any user to change any file-- or worse, a system changing them as a result of some other action I took. If I, as a regular user, were to overwrite YOUR files, you'd be rather upset.

I'll repeat: adding transactions to a placeholder account during an import should not be possible, since it violates the GnuCash definition of a placeholder account. 

⁣David T. ​

On Apr 29, 2023, 6:17 PM, at 6:17 PM, Michael or Penny Novack <stepbystepfarm at comcast.net> wrote:
>On 4/29/2023 10:08 AM, David T. wrote:
>> Michael,
>>
>> I disagree. The importer shouldn't put transacting into an account 
>> that is--by definition-- write protected.
>>
>> My counter example would be a write-protected file folder. An 
>> operating system that allowed a user to put data into a 
>> write-protected folder would come in for serious criticism.
>>
>> Temporary records should go somewhere, for sure. It's been my 
>> experience that GnuCash uses Imbalance-xxx for such transactions. Why
>
>> would you ever expect to put them in write-protected accounts?
>>
>You are talking about the  behavior of the IMPORTER.
>
>Your proposed solution would affect ANY entry of transactions.
>
>And an operating system SHOULD allow writing into a write protected 
>folder whenever the person (or program) doing so has sufficient
>"rights" 
><< normally when I am logged in it is WITHOUT admin rights (even on my 
>home system) -- bear in mind decades in the cypher mines where I had to
>
>be very conscious of whether I was using my ordinary login or my
>"prod*" 
>log in. Like in the middle of the night emergency fix to a hanging 
>production program -- doing this during the day I would walk over to
>the 
>desk of whoever normally handled "builds" and have them do the 
>replacement of the program in the production library>>
>
>Michael
>
>PS --- By all means ask for a change to the behavior of the importer if
>
>you think that is in order. But actually what you want is that the 
>importer verify that it isn't specifying an account that does not exist
>
>OR one that is a placeholder. We used to call something like the the 
>"input editor" portion of a transaction handling program.