I am of course referring to my underwear (see my previous blog).   My laundry was supposed to be returned before 9 pm today but so far no underwear.  Luckily I have some extra.  I‘m just wandering if the cirkus sized man that took my laundry is somewhere in the desert now, using my shorts as a tent and a sock as a sleeping bag.  OK, maybe he wasn‘t that small – but almost.  Börkur got his laundry.  He just spoke with me, he got returned someone elses underwear.  At least male ones.  Look how interesting our days are.  I blog of my laundry.  This is what real men do.  Really.

Today was somehow a boring day.  We were at work at 8 and back at the hotel at  6 which makes this Sunday the shortest workday so far.  Which was good.  We then went to the mall where my Creditcard got refused – the ATM transaction timed out 3 times until I was afraid to try again.  Lisa wanted to buy some beer – non-alcoholic of course, just for the taste of it....

I just paid another 500 Ryials for 1 weeks Internet usage (about 8.500 ISK ).  These hotels are a ripoff.  The best part is that this is per computer.  So if you have 2, the price doubles.  Börkur has 2.

Work went well.  I just made a small breakthrough this morning allowing me to continue our work full speed ahead.  Lisa will be leaving on Thursday so we really have to get our act together.  She‘s responsible for the Discoverer stuff.  I‘m mainly responsible for populating the data warehouse and Börkur handles the portal stuff  and access control.  Jon Mead will be coming on Saturday and spend the last week with us.  Mark Rittman will only keep phone contact. 

We tried compressing our fact table today.  It shrank 1/3 in size which is considerable.  Then it is bitmap indexed to hell and answers very quickly (instantly).  However, we will probably list partition it by indicator allowing us to partition exchange loading it without affecting anything else.  Now, I do a lot of deletes and inserts into the table while loading and re-loading.  I also need to do updates in some cases immediately after loading (as a part of the load process) which is not good for a data warehouse.  So, if we PEL the data and fix it completely loading it, that should be nice.  OWB handles PEL quite nicely but I have to verify if it does that automatically for list partitoning.  I‘ve done it with range partitioning but I‘m not sure about lists.  If not, we will implement it manually.

There are 3 times as many man in Qatar as there are women.  There are so many foreigners working here , mostly man, and this is the result.  Of the few women that are here all the Qatari women that are married are veiled.  No wonder Lisa gets eyed a lot.