PolarLava::Blog

PolarBlog V1.10.1 has been released.  This is primarily a minor security update, bug fix and spam reduction release.

Enhancements

• Added support for auto blocking of Akismet flagged messages.
• Added additional protection against comment spam.
• New configuration directive RBL_AKISMET_AUTODEL.

• Prevent hacking of input variables to eliviate potential path disclosures.
• Prevent hacking of session cookie to eliviate potential path disclosures.  Read more about these two changes in the PolarBlog Potential Path Disclosure entry.
• User IP detection could return a comma delimited list of addresses instead of just one.
• RSS feed not updating when multiple entires entered with the TTL.
• Entries in deleted topic cause fatal errors.

• Added RBL_AKISMET_AUTODEL

• None

This update adds the additional data points that are now collected by my new Blackburn Delphi 6.0 cyclocomputer.  That being, average and maximum cadence and heart rate as well as the total number of feet climbed on a ride.  I've also added some units to the ride display pages (i.e. mph, rpm, ft, etc.) as these previously weren't there.

There was a so a bug in the CycleLog  Calendar where by multiple weekly summaries were being displayed under various circumstances.  Thus the weekly summaries now display a single entry for each week as it should.

I came across this a while ago, but it got lost in the shuffle.  I just want anyone who uses any of my software to know that I stand behind it with this certification.  :)

I've made some additional display changes to the CycleLog  Calendar.  This consisted of fixing the ride entries to display using the same fonts as the weekly summaries.  For months that have no rides on a particular day of the week, when displayed in low resolution or in a narrow view, the day would get squeezed to a narrow size.  A spacer in now placed in one of those empty days of the month that is not used to maintain the cell width.  From the previous PolarLava CycleLog V2.4.1 update, I fixed a stray table row that was causing the XHTML validation to fail.

Not specific to the CycleLog, but more of a site wide issue, I've fixed a Google Analytics javascript loading problem that would cause errors to appear in some versions of IE.  Some day I wonder if I end up breaking more things that I actually fix.  "Good grief, Charlie Brown."

I was reading something earlier today about software development and it linked to this article on The Next Big Thing in software development.  This is a couple of years old, but it's still very relevant today.  Having worked for both big and small software organizations, I can tell you that regardless of size, all of them are capable of churning out some pretty shitty work.  As this article points out, "What other industry ships the kind of crap to their customers that we so regularly do?".  Welcome to the truth ladies and gentlemen.

I know this firsthand from my projects.  There are times when I wonder if I'm nothing more than a one-man open source wrecking machine!  ;)  But I've seen the silly mistakes I've made happen to much larger groups than my group of one.  Groups that have dozens of developers , QA people and even user (customer) acceptance testing!  It's pretty astounding that regardless, sloppy things slip through.

The only real answer —as the article states, it simply that people need to write better software.  It's not about putting in hours, but about taking care in what you do.  It's about accepting criticism without being offended and using that feedback loop to make a better product.

It's what I try do with my projects in hopes that I won't make the same mistakes.  It's a mindset that I try to instill in the people I work with so we get better at what we do.  So take care and have pride in what you're doing.  Don't churn out stuff that you wouldn't personally want your name on.  But when you do produce a dud, take the heat and learn from it.  Be professional, be a software craftsman.