[GNC-dev] Normalizing live data, a suggestion for discussion

Wm wm_o_o_o at yahoo.co.uk
Sat Feb 2 13:49:33 EST 2019


On 02/02/2019 16:11, Geert Janssens wrote:
> But I don't know how feasible it is to effectively obfuscate that data withoug
> resorting to a complex script

The script will be seen by others that do understand sql before anyone 
innocent gets to use it, promise.

If the script is well documented (I don't see the point of obfuscated 
sql when we are doing something like this as time is not the major 
issue, getting the problem fixed is) then people that can read will use it.

Further, most of the actual gnc code is so fucking obfuscated it is 
acknowledged only a handful of people can read it, so do you really want 
to raise the issue of obfuscation, Geert?

Seriously, people that don't know how code works are already trusting 
their financial data to code they have no clue about.  Why is my 
suggestion going to increase or decrease trust or increase or decrease 
complexity?

Grrrrr.

 >> that may introduce its own set of bugs

My script cannot introduce a bug, we are normalizing data <-- read that 
again, please.

> or
> inadvertently also obfuscate the actual issue. 

That is a possibility.  I consider this a positive not a negative from a 
triage POV. the user says: "oops, my problem doesn't exist after I ran 
the normalizing script" <-- is this good or bad?  if the script is well 
documented the user can edit it and run it again, possibly solving the 
problem themselves.

 > > The latter is quickly tested,
> the former is a time waster.

This is a very good point and I repeat, this is not suggested as 
compulsory, this is intended to make things easier not harder for people 
that do want to report things that may be specific to them without 
exposing irrelevant details they may consider private or personal.

-- 
Wm



More information about the gnucash-devel mailing list