Theorising vs. Doing

Over the past many years I've found myself slowly moving (in my personal projects) from the realm of actually implementing code about which I've been curious to spending my overwhelming majority of my free time to theorising about various concepts and/or interests, primarily in the field of human language parsing, simulation and simplification of generic real-world object modelling.  Several of these have been mentioned and explored in previous posts over the last 7 years.  

I believe a great reason for this change in focus is partially age, and primarily due to the extent that my days are focused on building the next great startup application.  We recently (as of the New Year) pushed the past eleven months worth of work into existence as our new primary (Python/Django) product replacing the previous multiple iterations of our legacy (Java) product.  The focus and scope of the project limited my ability to go off on tangents wherein I could code my personal projects freely.  This was by no legal hand binding, it was simply a matter of wanting to focus my writing to building our product to the level it needs to be, to the high standards we require (and rightfully so).  

This leave one (me, namely) with little actual time to thunk down in front of my primary workstation to commit theory to codebase, hence I've found myself working it out on handheld whiteboards, sketchpads,  napkins, chalk on the ground whilst playing with my daughter or even simply working these theories out in my head.  While I would like in many cases to put my thoughts into actual runnable logic, I've found that the exercise of stepping through theoretical code in my mind has kept me sharper regarding my thought processes.  True, I cannot share this as easily with friends afar, but given that many of my personal friends are also senior level engineers, discussing (when we do) casually my theories and ideas, they get the gist quickly.  

The point I'm trying to get across here is that while today is perceived as one wherein a bunch of 20 year olds either in or fresh-out-of college spend an obscene amount of hours in incubators or at startups killing themselves churning out ridiculous amounts of code, there is most definitely no slow down in the thinking processes of the more senior of us out there.  We just don't find a necessity it raw churning out of code as it is the concepts that drive the field and aid in future innovation.  Let those who wish to "do" continue to run their path, but lets not overlook the value of those who've moved their focus (whether by situation or by intention) to the more abstract realm of theory as the two are inextricably bound.  We need ensure that these two distinct groups of people are in contact with one another because it will ultimately lead to the newer advances, bettering our field for all involved.

The Usefulness & Addictive Nature of Multiple Displays

It seems to be a divisive issue when dealing with developers, engineers and programmers alike... The usefulness or annoyance of multiple displays.  Recently having upgraded to my fifth display, I felt it was about time that I weigh in on the issue/share my opinions.

Five displays : One 1920x1080, Two 1920x1200, One 1280x1024 and One 480x800 :: 8.4 Mega Pixels in All

Back in 2001 at an interview with Health Market Science in Conshohocken, PA (now in King of Prussia, PA I believe), I had witnessed a key benefit of new up and coming startups for their developers...  multiple displays for each workstation/desktop.  Flash forward a few more years and one Mac Book Pro later and I find that ever company at which I would work, I would attach a secondary display so as to provide more desktop real estate.  Whether this was to provide a shell or two in which to run emacs while accessing a db frontend or browser preview window, it proved beneficial in terms of productivity with less switching between screens, windows and what not.

Ever since my first exposure to such setups (which were astronomical in price to setup back in the 90's and the beginning of the newest millennium), I was rather enamoured by them.  Now in 2010 I find myself working not on laptops, but full blown 64 bit Unix workstations.  A major benefit of such setups is the ability to run a considerable number of graphic cards and their associated displays.  Some wonder what could I possibly have running that dictates so many displays.

I personally like to keep a web browser open at all times so using the picture above, I will explain as best as possible.  So, as I was stating in the previous line, the left most display is running a full height copy of Google Chrome, flanked by three separate terminal windows running a mix of local and remote shells on  various servers.  The middle large display is running my current favourite IDE for python development in fullscreen (in this case, NetBeans 6.8).  The right hand most screen is used for additional remote shells used more so for large rsync'ing and process monitoring.  The upper left screen (a great displaylink/USB powered 480x800 (or 800x480) from Mimo) is used for Skype & Adium (and both are pinned to all spaces/virtual desktops due to content), and the upper rightmost monitor is a free for all to display whatever is needed in addition to all of the aforementioned items (pdf viewing, techtalk viewing, the occasional movie, db schemas or gui, etc.).

The biggest issue I currently have with the setup is physical layout.  A table roughly 144 cm (60 in) wide can only handle three widescreen displays (between 23" & 24" diagonally) when in a concave configuration.  Having to position the additional displays above the lower displays does cause the occasional confusion regarding pointer location.  A better layout would be an array of identical displays, preferably with VESA mounts in a 5 wide portrait layout or a two rows of three in landscape layout.

I highly recommend that others who haven't had the opportunity to work in multiple display setups to try it out as soon as possible.  Those simple keyboard shortcuts used to swap from window to window and/or virtual desktop to other virtual desktop do take up time and can interrupt one's flow.  Is it really worth it when displays are so cheap these days?

Caveat emptor:  The use of multiple displays causes varying ranges of discomfort when penned/cornered/shackled into a single screen machine, not to mention there is desire to acquire more screens after getting acclimated to the first addition, second addition, etc.

Till next time...