Ethics/Day-in-the-life software package

7:00am
I've been waiting up all morning, bundeled up and ready to be puchased by and sent away to the next lucky user. I just received my latest update that was finalized yesterday and now ready for consumers. This update will hopefully allow me to run faster on older machines and prevent any security hiccups. In the previous version, I was optimized specifically for parallel processing on Clusters with multi-core processors, and that seemed to have a negative effect on my performance with single core cpu machines, as I was receiving a substantial amount of negative feedback from some of my core users. Apparently I had strands of code that were looping many times trying to send data to another processor that was simply not available. This was causing between a 30 and 35 percent drop in performance to a semi-large group of my customer base. Apparently the whole "multi-core is the future" theme didn't hit home with my consumer base. In any case, the problem was fixed and now I'm ready to perform at a high level for both single-core and multi-core cluster systems.

9:30am
Well, ive just been packed up and sent of to the first user this morning. It was quite a bumpy ride home...you'd think my new owner lived in the rockies or something. Does he not understand that I am stored on a DVD, meaning my data can become corrupted by even the slightest scratch! What we he do if my data got ruined? Probably throw me in the trash, or complain to the store that my optical storage media protector(DVD) was "defected". I can already tell this guy has so respect for the little people in this world

12:30pm
Well its about time...ive been waiting to show off for my new owner but he decided to take a nap when he got home...Looks like this guy has a cluster set up in his basement, and yes his cluster seems to be outfitted with 4 year old sets of Intel P4 2.2ghz processors. Yes, single-core CPU's! This will be the perfect test of my new version capabilities...or fixes...however you look at it.

12:45pm
Ok I was just put into the head node minutes ago and am installing on the node as we speek. At least this guy seems to take good care of the cluster in terms of keeping the disk drives "roomy" and without fragments lying all over the place. I like that because it makes for a quick trasfer of my data onto the disk drive, or what this guy calls "installing" on the drive.

2:00pm
Ok, well im impressed with the install process. I know its already 2:00pm, but when you have to get your data onto 25 child nodes it can become very time consuming and exhausting. Not to metion having to set aside time in case, say, something goes wrong with the data transfer on any certain node. We did run into a problem with child node 12, as it seemed to be missing a specific software driver need for my installation. This child node actually thought I was an entirely different software package than what I actually am! I threw up a red flag when it asked me if I wanted to install in the default location C:\TotalWar\. Seriously, do I look like I have ANYTHING to do with war?...Anyways, turns out it didnt need a software driver update at all. There was malware on the node that was causing certain software to be installed into the above described default path, but my owner fixed it and now i am ready for action. Bring on the protien sequences!

4:30pm
Wow, just let me...catch...my......ok, that was exhausting! As is comparing any protien sequence. But??? Ask me. Ask me how i did...Ok I will just tell you staight up: I rocked that problem! People think the CPU does all the work. huh. Gag me. What about the software package? I'm the one who makes the CPU what it is; good or bad, and I sure made it look good for that run. The smartest people will tell you: hardware ain't nothin without software! I can tell this guy is pleased with how well that run went. He set the test up to compare the 2.18 Mb genome of Neisseria meningitis serogroup A against the 2.27 Mb genome of Neisseria meningitis serogroup B, and it executed in less than 2 seconds elapsed and CPU time! Now granted, thats a far cry from the eventual 3.098 GB human genome reference sequence that I will have to compare against itself in the coming days, but I'll consider this a short term win, and I play on enjoying this win.

8:30pm
A bigger test means more time and that is what I just experienced, but to my dismay, there were a few problems that I have to put sqarely on my shoulders. In short, this test was comparing 678mb genome files, which of course is a little over 300 times the size per file! Now I obviously dont expect 100% linearity here, but when the execution takes 3.5 hours to complete you're not quite as satisfied. The CPU was complaining to me the entire time to give it the data not only quickly, but give it the correct data. I seemed to have gotten confused somewhere in the 300mb range of the file. I'm not one to make excuses, but there were some uber-complex sequences in there that I just didn't seem to be able to process correctly, and therefore I made the CPU either wait too long for the data or give it stuff that was incorrect and need to be rehashed. Ok OK OK, I learned my lesson. I have been humbled. Maybe software packages are not always perfect, but we damn near try to be, and shame on you if you fault us for that, because you better believe that we are trying our best to make your life easier!

10:00pm
Ok, its bedtime. All in all it was a decent day of work. There were some up's and down's today, but when do computers (in general) ever work up to expectations? You're right, not very often. But tomorrow is another day to take another step forward. I'm going to have good days and bad days, but that won't prevent me from doing my job as best I can. I'm going to work my tail off to keep that cluster running at Mach 6 billion...ok there is no such thing, but you get the point.