Hi Greg

There are some very good PS courses on MVA, right from the horse's (Snover's) 
mouth.

I have moments with PS where I'm stunned how much I get done and how fast. 
Other days, I want to poke my eyes out with a sharp stick after trying to do 
something that seemed trivial.

Regards,

Greg

Dr Greg Low
1300SQLSQL (1300 775 775) office | +61 419201410 mobile│ +61 3 8676 4913 fax
SQL Down Under | Web: www.sqldownunder.com<http://www.sqldownunder.com>
________________________________
From: ozdotnet-boun...@ozdotnet.com <ozdotnet-boun...@ozdotnet.com> on behalf 
of Greg Keogh <gfke...@gmail.com>
Sent: Saturday, July 8, 2017 3:28:01 PM
To: ozDotNet
Subject: Re: Powershell testing

I think I found the cause of the weird behaviour. You have to put function 
definitions at the top, before the outer level code. It's not like C/C++/C# or 
a real language, it's read from the top down (event BAT files are better than 
that). I don't know how it intermittently worked, but I supposed it's cached 
the functions, it's the only crazy explanation.

Calling COM functions from PS is really fragile work. The slightest whim of a 
mistake will result in a cascade of incomprehensible errors.

Last night I wondered why the script was failing for over half an hour without 
any useful message. It turns out I had the MSI file open in Orca and couldn't 
see the window, which was causing the script to crash with a generic COM error.

GK

On 8 July 2017 at 15:10, Greg Keogh 
<gfke...@gmail.com<mailto:gfke...@gmail.com>> wrote:
How to people write PowerShell scripts in a productive way?

I'm writing my first non-trivial one to query and update the tables in MSI 
files. I reckon I have wasted more than half the coding time (many hours)on 
utterly infuriating and bewildering problems. I'm writing the script in ISE and 
I hit F5 to run it and check the results.

After making a code change I hit F5 and it runs the code before the change, the 
second F5 runs with the change. Even saving the script makes do different and I 
have to hit F5 twice to see the results.

If I make a small error the script crashes. Sometimes when I fix the error it 
continues to crash with an unrelated error as if I never corrected it. It's 
like an error gets "stuck" and can't be undone. I made a tiny correctional 
change about 15 minutes ago and the script crashes non-stop with a "not 
recognised" error, which I know is bullshit. I've even tried rebooting, but the 
correct script is still crashing.

Is this a joke. What century am I in?

Greg K

Reply via email to