Best of blog.melski.net — October 2011

Here are the posts that got the most views in October 2011:

  1. Halloween 2010 Haunted Graveyard: 28% of views
  2. What’s new in ElectricAccelerator 6.0: 17% of views
  3. Electric Cloud Customer Summit 2011 by the Numbers: 15% of views
  4. Shell commands in GNU make: 7% of views
  5. Exceptions to conflict detection in ElectricMake: 6% of views

The article about last year’s Halloween graveyard generated a lot of traffic, all thanks to search results — lots of people looking for ideas for their own Halloween decorations. I hope they found some inspiration from my meager attempt! If you liked that post, you’ll probably enjoy the post-mortem for this year’s graveyard!

Top posts for December 2010

The following articles generated the most traffic in November. Unfortunately I don’t have stats for my articles on the Electric Cloud Blog this month, but here’s the data for the top five posts on blog.melski.net:

  1. Makefile hacks: print the value of any variable: 89% of page views
  2. Shell commands in GNU make: 4% of page views
  3. HOWTO: install kernel debuginfo packages on SUSE Linux Enterprise Server 11: 2% of page views
  4. Cloud computing for traditional dev/test: 1% of page views
  5. Public versus private clouds for dev/test: 1% of page views

The top article, Makefile hacks: print the value of any variable was wildly popular, which is great, but also skewed the numbers pretty heavily. The price of (small) success, I suppose.

Top posts for November 2010

The following articles generated the most traffic in November. Still mostly articles that I wrote for the Electric Cloud Blog, but the number one entry is actually from blog.melski.net, and I think there would have been another, except that I only published the article on November 30:

  1. Shell commands in GNU make: 38% of page views
  2. Makefile performance: $(shell): 19% of page views
  3. A second look at SCons performance: 6% of page views
  4. What’s new in GNU make 3.82: 6% of page views
  5. The last word on SCons performance: 4% of page views

What’s most interesting to me this month is the search performance of Shell commands in GNU make. I wrote that article specifically because I see a lot of search traffic from queries like “makefile shell”, “shell command in makefile” and other variants of that theme. To my disappointment, all of those searches still bring people to Makefile performance: $(shell), rather than to the new article. Not sure what I did wrong there. Perhaps it was the inclusion of the word “GNU” in my new article? Anybody with more SEO experience care to comment?

Top posts for October 2010

The following articles generated the most traffic in October. These are mostly articles that I wrote for the Electric Cloud Blog, but I’m pleased to see one entry from blog.melski.net here:

  1. Makefile performance: $(shell): 24% of page views
  2. A second look at SCons performance: 14% of page views
  3. The last word on SCons performance: 11% of page views
  4. Cloud computing for traditional dev/test: 6% of page views
  5. How scalable is SCons?: 5% of page views

Top posts for September 2010

The following articles generated the most traffic in September. These are all articles that I wrote for the Electric Cloud Blog, because (for now!) those articles get more views than those on blog.melski.net:

  1. Makefile performance: $(shell): 28% of page views
  2. A second look at SCons performance: 18% of page views
  3. The last word on SCons performance: 13% of page views
  4. What’s new in GNU make 3.82: 7% of page views
  5. How scalable is SCons?: 6% of page views

The popularity of the $(shell) article is a bit surprising — it’s just one of several articles about makefile performance, after all, and none of the others show up on this list. Also, it’s a really old article (from March 2009). What’s going on there? Look at the search queries that bring people to that post:

  1. makefile shell
  2. makefile shell command
  3. make shell
  4. shell command in makefile
  5. makefile $(shell

Conspicuously absent from these queries: any mention of performance. I think people are just looking for help understanding how gmake and the shell fit together in general, rather than specifically how to get the best performance. Maybe I should write a tutorial — the definitive guide to gmake and the shell?