Monthly Archives: May 2012

Anti-Pivots or Company Focus

Focus is a difficult thing as a start-up.  By definition you’re exploring a space: trying to find what you can build quickly that enough people will pay enough for. So you’re jumping around all the time.  We’ll do A which will lead to B which will…  No too long or too expensive, instead we’ll do C which requires an agreement with D.

For me it was who is the best customer for the value created in StickyVote.  State legislators; no, federal legislators; no, newspapers; big advocacy organizations; small ones.  Maybe users through this new Facebook thing?

The problem is that these shifts, while important, can confuse and wear down staff and pilot customers. Since focus and enthusiasm is going to shift, what would I do different next time?

Set more criteria for judging whether something is successful.  How can we prove whether this is the next right step?  Or what would demonstrate that this is the wrong early adopter?  It’s easy for some time to go by, attention to shift and to say, “This isn’t working.”  Which is an easy excuse to switch to whatever is exciting and new that week.

Ok, money isn’t pouring in; there are a million theories for why not. Before getting too far into those, lets look at why we thought it would work.  Have we put the new idea/software/pitch in front of enough of the target customers?  Is it just one first impression that we’re operating off of or is it really enough opinions to mean something.

Good things come from setting criteria. First, if I don’t have some initial criteria for measuring a new direction, then I shouldn’t talk to my staff about potentially shifting course. Second, refining the criteria with staff means they can hold me accountable when I next want to shift.  Just as a good agile process adds transparency to development and priorities, a good “direction process” should do the same for big questions.  Where are we going? Why? How do we know if this is a step on the right path?

Gigabyte P55-USB3 motherboard with 2000 MHz RAM

Just plugged 8GB of G.Skill 2133 mhz ram into my desktop.  The memory clock speed stayed at 1333 mhz when I did so which led to some fiddling with bios settings.

I have a Gigabyte P55-USB3 motherboard with an i5 2.8 Ghz processor and plugged in 2 x 4GB modules of the F3-17000CL11D Ripjaws ram from G.Skill (in a nice red color to boot).  Motherboard manual is here, http://www.gigabyte.com/products/product-page.aspx?pid=3440#manual

Once in the bios config in the ‘MB Intelligent Tweaker’, the ‘Advanced Memory Settings’ seemed to have what I needed, ‘Extreme Memory Profiles’ for different speeds.   Profile 2 put the speed at 2130 mhz and all the advanced timing settings matched the memory specs (11-11-11-30 2N).  Unfortunately, it wouldn’t boot.  Switching between the different  profiles (disabled, profile 1, profile 2) and ‘Performance Enhance’ modes (standard, turbo, extreme) didn’t help much, the only way  I could boot was at 1600 Mhz in the disabled profile with extreme performance enhance.

Fortunately I hit on a nice overclocking guide, http://www.overclockers.com/3-step-guide-overclock-core-i3-i5-i7/

Leaving the memory settings at the default values, I  went into the ‘Advanced Frequency Settings’ section and changed the Base Clock BCLK to 200Mz with  the CPU Clock Ratio at 16x.  This gave me an overclocked CPU speed of 3.2 Ghz (seemed reasonable and a nice bonus) and put the Memory Frequency at 2000 MHZ (good enough).

After a few hours of serious programming work, the system has been completely stable and seems a bit faster.  Hope that helps.