Daynote ,Svenson

MM-lv     Thursday

2000-02-24

It remains warm, 3°C in the morning but with far less sun and even a few short spells of rain.

I did solve the last of the II/DR problems and updated the documentation. I notice that OLI still has some programs to finish on the list so while it has taken a lot longer than I expected I am not the last to finish.

At around noon I got a call from Robert, Belgium, to ask if I had done something about the problem he reported yesterday. No, I hadn't done anything. I had recommended him to change the logging so we should get some more details, and I was waiting for this logging. I don't know if he actually change the logging but the reason for his call was that the problem had solved itself.
I am not principally opposed to self solving problems but it does not make me feel comfortable. Hey this is on an AS/400, not some sort of Windows thunking stuff, so there is no reason for software to behave Microsoft-y. My guess is that the DHK used some wrong setting or library list but without full comprehensive logging it is not possible to find out.

Theo asked me to pass the layout and a data sample from the code File (CD) to Rutger. He should then pass a selection of OPRS (I haven't got a clue what that stands for) data back to us pre formatted in the right file layout for importing to the AS/400. So I make a selection using the Query tool and mail that to Rutger. I even add a field definition file (xxx.FDF). Rutger is heading the Microsoft Competence Center (MCC) (I know it is a contradiction but I didn't make up that name) and of course doesn't recognized the fdf. Later I get into a discussion with one of his team and the end result is that we get an Excel sheet with three columns filled with data for the first three fields in our file. The rest is too difficult for them. So I convert it from XLS to TXT and send it over to the AS/400 where I make up the content for the other fields. Just a short conversion program is all it takes. The data that they supply is highly suspect however. It should contain a code-type, a code and a description but for about half of the records the code is blank. The code is the most important part of the file.
If we don't get a reasonable file I think I am going to persuade Theo to pass me the full OPRS file so I can make my own selection and extraction from it.
I guess the problem is that the people over at the MCC don't know what data to expect, for them anything could be acceptable, they are working completely out of context in this case

I didn't go to bed early yesterday and yet I didn't install anything on the new box. The evening came and went in a single instant it seems. Well I cannot remember what I actually did so it could well have been a single instant.

Adios