Saturday, December 18, 2004

I give up, really...

I have just had one of those days. The bash script I wrote and ran on the test server to upgrade to the latest QTAC distribution failed. It ran correctly on the test server, on the main server it failed because of the shite that is Ingres. No details here; ask me, via comments, if you are interested.

Unfortunately the script was one of those run once only type things as it destroys the original structures and creates new ones in their place. I have a backup, needless to say, but that is beside the point. It failed, I fixed, I fixed again, I looked at the log, I undid, I fixed and it ran. It failed on one old database(which I will address on Monday), but it is all in place.

Execute the new version of the application - all is fine, look into the obscure logging for that application and see "cannot cache forced_offer_reasons table does not exist or is not owned by you". And hey, it's right, the table doesn't exist. QTAC failed to mention that one, now we have trouble calculating fields(yes, I know, but if I explained first stage normalisation to QTAC their heads would explode). No errors at the front end, that all looks fine. So now, I have to write a script to create that table and run it against 11 databases.

Questions asked are:

  • If they failed to tell us this, what else do they leave out?

  • If the errors aren't reported fully, what are we missing?

  • Can I convince QTAC to normalise/redesign before I go completely insane?

  • What is the point of having the "cache_control" table, specifically designed to slow the system down and cuase us more headaches, when there is no entry on it that says to cache the forced_offer_reason table?

  • Do QTAC add tables without thinking (Yes) and then loose track of their purpose and forget about them?

  • Why do we have yet another table which is a simple valid values look up table there are dozens on each database and one would do?

  • Why the smeg are we still using Ingres?

And once more,

  • Why don't UniSA et al consider OASys a threat to our core business?


It is a testimony to the staff at SATAC that they get it right every year.

-medge

This post is quite slanderous, I know, but just give me a chance to show everyone in a public forum.

0 Comments:

Post a Comment

<< Home