Clear and warm (11°C), turning .... yhea.

The problem with the program that wouldn't compile even though there were no errors could be the result of converting it from RPG/400 to RPG IV. I manually deleted the object and then compilation went as expected. When converting a program we usually mark the old version as Obsolete since <version-nbr>. Doing so deletes the existing object. I was not sure it the applied change would work so I wanted to test before making the current version obsolete. The worlds doesn't work that way, it's a bit like a Microsoft patch. Once you begin you are committed. No way back. Just like real life.
Testing tomorrow.

I sprayed the potatoes Saturday, against leaf fungus. Had to repeat that but against potato plague today. We never had a problem with either but since we planted a red variety some years ago we cannot get rid of the plague.
Maybe we should leave potatoes out of the garden for a few years. Sigh

Spain-Norway : 0-1
Yugoslavia - Slovene : 3-3
That's the football surprise.

Clear and warm (14°C), turning .... oeps, copied once too often here. Warm indeed but not clear but rather a hazy misty grey morning with the promise of a hot sunny day. In stead it stays grey hot and sweltering.

After rigging up a test situation for the validation program. I get interrupted by an impossible situation in Belgium.

Orders made in TeleSales get status 'VAL' when they are passed on to the validation program on the AS/400. If anything fails a validation in either the online or the background processing (two stage validation) the order status is turned to 'HLD'. If the validations pass the status is set to 'REL' and the order is released to Logistics. 'VAL' is a temporal situation that normally only lasts a few seconds to a few minutes. Now they report the existence of orders with status 'VAL' in the archived orders file (QH). There exist invoices for some of these orders. Now it is impossible to invoice an order is it is anything different from shipped (shp). Only Logistics turns the status in shipped (shp) and logistics will only pick up orders with status released (rel). After invoicing the order has status closed (cls) and gets archived. 'VAL' in archived orders is impossible to achieve using our programs. Another fact pointing to unacceptable manual FUBAR-ing is that on the invoices the order number has a sub-number indicating partial shipments. The 'VAL' records in the archive are all un-split.
Of course I cannot proof a thing.
And now they want me to find out how to reconstruct the fouled up records. And I know for sure I cannot do that.

And Koen helped out the Austrians with their 'impossible' problem of the missing delivery address. You cannot release an order if the delivery address isn't filled and so you cannot get an invoice. They have an invoice (the client has one) with the delivery address printed on it, and now they wanted to produce a 'reprint'. This is impossible because suddenly the delivery address is missing. both from the invoice-print file as well as from the order.

And ... well you get the type of day I had.

I can not log in into my account at Zanova anymore. Typing the wrong password produces an 'error, try again' message screen. Typing the correct password however simply dumps me back on the login screen, no messages or any other indication, just the cleared login screen. Over and over again.
I intended to pull off from Zanova early next year but I don't think I will wait that long. I can transfer to Masset, for the same price I get more mailboxes and 50% more bandwidth. I lose the editing and management tools but I never used them so ... And if things don't work as I expect I can walk in to his shop (do it about every week anyway) and demand an explanation. Talk about customer interaction :-)

later Ha, now. They changed support policy. the support@zanova etc returns the mail pointing me back to the support page on their site (which, luckily, doesn't require a login)
They reply I should check my user+password. Do that and ... presto, don't even get the login screen. Not that it matter much, I will move over to Masset as I planned anyway.

Same type of grey morning as yesterday (same 14°C as well) but with a few drizzle spells. Later in the morning the greyness and drizzle transformed into rain, even heavy at times while the afternoon became sunny. The forecast called it unsettled. Yea, that is always right.

I got one test run passed and the validations passed flawlessly. I do need a few moor runs testing various limits. Boring work requiring lots of concentration. And a good debugger.

Loek has tested (?) the transport cost implementation for V09 and has approved the change. So I applied the same change to V08. Of course this has to be tested as well.
By Loek.
He tries and doesn't succeed.
When I check in the L environment (for which I normally have no authorization) I notice that he simply didn't set up the condition correctly. I admit that setting the condition is not a trivial matter but while explaining how I notice he skipped some steps because he doesn't know how to set up the whole thing.
How the f*** did he test and approve the V09 adaptation?

Did I mention that my validations testing required concentration? Now I lost that here.

In the after noon I assist Jan with a few strange problems. There are a lot of records in the trigger file and in the temporal invoice files. Both should only contain a small set of records, lest say the amount of orders for one or two days. And the problem of the status-VAL records in the archived file I got yesterday. These three problems were reported from Belgium but it seems that in Germany and in America the same problems occur.
Which mean they are real problems.
I don't think they are related but all three are not explainable using the normal programs and they don't occur on our own system.
We don't come down to a solution so the fun will continue tomorrow.