I like online forms. Apparently some people are hesitant to use them. They want to call me up and let me hit the magic button that just does what they want to happen. If I have to do more, that is okay too as long as what they want gets done by the time they go back to try again. Any longer makes a hindrance…. a bottleneck in the beaucratic processes that are stiffling everything that is good at work.

My online forms record the request to a data file so that should a review of the requests need to be done, there is an independent record against which to validate. The forms also kick off an email which my email program will file to to-do folder. Unread messages are bold with a number so I know just how much work awaits me. I try to stay on top of those requests and keep the folders as empty as possible.

None of these emails are something I would want to explain over the phone. They have been refined over the past five years to be concise and informative. I don’t want anyone having to email me for more information so everything must be there. I don’t want people emailing back because they could not find something, so it must be brief. Finding that balance took 3 years. Once my boss let me do the job, I revised his web page instructions so that they show everything they should.

For the most common and replicable requests, there is also a templated response. For example, a request for access to a web site has a response which tells the person how to access that web site. Before I started using the templates, I used to pop off emails to people that missed some detail or erroneously provided wrong information. The templates keep me consistent. Blame the early symptoms of senility.

I also made programs to expedite as much as possible all of those formerly monotonous tasks so that the computer mostly does it for me. Other tasks I have done so many times and so often, I know by heart hideously archaic commands (these are all from memory and may contain erros)…

• find . | wc -l; find . -user <username> | wc -l – to compare the number of files in a directory with the number of files owned by the owner. If there is a difference, then I…
• find . -user <username> -exec chown <newusername> {} \; – to selectively only modify the right files. I don’t have to figure out what those others files are and change them later back.
• wrote this one this morning to run the command, dump the screen data to a file and email me the output – NOW=`date +%Y-%m-%d-%H:%M`; <command to run> > output_$NOW.txt ; cat output_$NOW.txt | mailx -s “Command Output $NOW” myemail@domain.com
  In this one system we get about 5 lines of log data any time the system pops up a warning to users who are using an unvalidated web browser. I wrote this to identify who they are so that I might send them a message to check their web browser is properly set correctly – grep “Browser check found errors for user” logfile.log | awk -F: ‘{print $2}’ | sort | uniq – the odd thing was that it did not pull uniques when sort was after uniq. How odd?
• grep <class> *up.txt | awk -F, ‘{print $1 }’ | sort – The file has lines of users and the classes in which the user is enrolled. The grep finds only those users who are in the class. The awk just displays the username. The sort is the sexy part that alphabetizes the list.

Tite, huh?