Skip to content

All posts by moore - 54. page

Using Munin To Track Business Values

Munin is a great piece of software that we use at my company to track overall trends in disk usage, CPU and other system purposes.  Now, we don’t have a ton of servers, so I’m not sure how munin scales for many machines, but it has been invaluable in troubleshooting problems and giving us historic context.

One thing we’ve started to do is to incorporate business specific metrics into munin.  This is good because it ties the technical operations more tightly to the business, making us aware when there are issues.

Anything you can run a sql query or do a wget for, you can graph in munin.  (Here’s something I wrote about writing munin plugins a year ago.)

I don’t think that munin is acceptable as a general purpose dashboard.  I’d probably look at Google Analytics if I was web drivingdriven (updated Feb 25 2012), and at statsmix if I needed to integrate a bunch of disparate services.  But for bringing additional business awareness to a technical team, writing a few custom munin plugins that will graph key business metrics can be very useful.

Business Tips Useful Tools | moore |

Hiring Tips

I’ve only hired a few times, but I just wanted to jot some notes about what worked for me in this process.

  • Use a bug tracker or issue tracker to keep track of resumes, emails and interactions
  • Respond to every person – there were some outsourcing firms that I didn’t respond to, but every other person got a response from me
  • Use Craigslist.
  • Use an email alias on your Craigslist post
  • Use other mailing lists (rmiug-jobs, cu cs jobs, even local neighborhood lists)
  • Ask your networks for candidates, but don’t expect too much of them
  • Pre-screen with a set of email questions if possible.  Don’t ask candidates to do too much, but asking them to do some work will allow some to self select out
  • When doing an interview, set the candidate up to succeed, by telling them what you are planning to ask them to do
  • Set deadlines for yourself, and share them with the candidates
  • Follow up with every candidate when you make a decision – I don’t think that it is fair to do otherwise
  • If you can point job seekers at another position, do so.  I recently did this with a QA position – in my search, I discovered another firm that was looking to hire, so I pointed all the candidates that didn’t work for us to that firms job posting
  • The web is full of sample job interview questions – use them!

 

Business Jobs | moore |

Moving away from Google Maps

A while back I wrote about Google Maps charging.  This caused an uproar around many map dependent sites.   (Brian Timoney, ever reliable, delivers snark on that front.)

I, for one, don’t begrudge Google the chance to charge for their excellent product.  What that does, however, is make certain business models that were previously viable now unviable.  That is, if they remain with Google Maps.  In my previous post, I went over some alternatives.

The company I work for, after evaluating all the alternatives, including paying Google for a license, has decided to go with Mapquest.  While the API is not as gorgeously documented (both on the company site and on the web), it has the same licensure as Google Maps used to (free for any publicly accessible website), with much of the same base functionality.

I’ll be very interested to see how many people are doing such high value work with Google maps that they can afford to pay for licenses or per use.

Google Maps | moore |

On Being Disrupted

This is a tough post to write.  I’m at the tail end of an evaluation process for my company that ended up with us deciding to go with a third party vendor for software which powers key area of our business.  It is augmentation rather than replacement, but still.

It was hard for me, because this particular key functionality was previously provided by custom software that I helped build over years.

I like to build things!  Like most other software developers, I get excited about building stuff.  One of the unmentioned frustrations of many software developers is building something and then seeing it shelved.

However, it was clear from the survey of solutions that there were three choices:

  1. buy something off the shelf
  2. get better as software developers and really, really accelerate our development
  3. have the business be negatively affected by this piece of software

Now, #3 is obviously not a good solution.  #2 is a great solution, but is hard to put into practice, especially in a short time frame with a large code base (though we are trying to use some of the agile methodologies to make our software development more productive).

#1 it is.

Was this a wise choice?  Talk to me at the end of the implementation, but I am hopeful.  We did take several steps to protect ourself (after all outsourcing core business functionality can be deadly), including:

  • a long, laborious evaluation
  • engagement with multiple vendors, and
  • building a set of criteria to help us determine if this outsourcing is meeting our needs

One thing that helped me take this decision a little less personally is to redefine in my mind the value of software to the company.  It’s not the particular implementation of the software that provides the value.

Unlike a software company, my company doesn’t exist to write software (see Five Worlds for more on different types of sofware development).  Instead, software exists to serve the company.  Having something off the shelf provides the similar functionality for much cheaper.  It also allows me and other members of the software team to write software that is unique.

Having been a contractor and having worked for a startup and consultancy, I’m used to being the disruptor.  In this scenario, I was the disrupted (ht, David Skinner).  It’s a humbling place to be, even if I wasn’t disrupted out of a job.

Business Programming Web Applications | moore |

Word to the Wise

I recently read this post about startup team success from Paul Graham.  Always fun to read Paul–I have a few sites I remember and type in periodically (haven’t used an RSS reader since Bloglines), and paulgraham.com is one of them.

The older I get, the more I see that being resourceful and having follow through are very very valuable traits.  This post just confirmed that.

The corollary to that is, if you don’t think you are going to follow through on a suggestion or a favor someone asks of you, it’s far better to say no up front than to fail to follow through.  That is going to be a tough lesson to integrate into my life, but it’s the flip side of that coin.

Business | moore |

Building A Site With Google Sites

I recently built out a site for a local tax firm, Cahill, O’Kelly and Associates using Google Sites.  I’ve already talked a bit about using Google Sites for your web presence, but I wanted to share a bit more about my experience.

The reasons I chose Google Sites over something more flexible was pretty much based on what I determined was best for the client.

  • the client is not super technical and was wary of a website
  • I wanted the client to be able to update the site so they wouldn’t have to pay me to
  • they were not super picky about design of the site, so were willing to pick from a template for look and feel
  • site was small and functionality desired was limited (brochureware)

All of these added up to doing something that was quick, cheap, low risk, and easy.  If any of these decision points had been different, then Google Sites would have become a less attractive choice.

Good things about Google sites

  • easy to get going
  • easy to integrate with all of googles other services (analytics, webmaster tools, google forms, etc)
  • lots of templates to choose from
  • hosting is free

Things that suck about Google sites

  • the footer links are out of your control
  • some of the templates require photoshop knowledge to customize (background images)
  • no easy way to roll back changes to your site
  • some of the admin UI is clunky
  • scrolling around the look and feel customization page was difficult

Tips for developing on Google Sites

  • make sure your customer fits the profile mentioned above
  • educate your client on the limitations
  • make backup copies of your sites before you do big look and feel changes.  That way you can manually view the differences and roll back if needed
  • get to know google forms, as that is the only way to do interactive forms that I found
  • be prepared to spend some time tweaking the layout
  • be prepared to do some training on how to edit the site

All in all, Google Sites is a great solution for a certain type of client.  Consider adding it to your toolbox if you do site development.

Web Applications | moore |

Using Google Apps to generate a simple site from a form

I was blue skying with a friend the other day and came up with checklist to build a simple canned website from a form submission, using Google Apps and Google Apps Scripting.

  1. User goes to a form.  (More on creating a form)
  2. User submits images, video links, text, email, background color, etc.
  3. That all goes into a google spreadsheet, which triggers the building of a site.  Here’s trigger info.
  4. You can automatically build a google site.  Here’s the api: site creation, page creation.  If you want your own domain name (example.com) I think you can do that, but you’d need to integrate with a DNS service to set that up–this part bears a little more investigating.
  5. If desired, you can generate the QR code for the site.
  6. If you need to tell the form submitter that the site is created, you can email them a link to the QR code so they can print it out.
  7. You can even remove content from the site via cron (see triggers link up above), though you can’t automatically delete the site.
  8. It’s conceivable you could even use google checkout to charge users beforehand (though this is a bit less supported).

What’s great is that you don’t need to have any tool other than a browser to build this entire stack.  Google is truly creating a toolkit/macro language in the cloud.

Cloud Computing Google Apps Programming | moore |

Google Maps is now charging

…if you reach a certain number of mapviews.  You have up to 25K/day mapviews free, which works out to just over 9 million map views a year.  My understanding of those limits is that they are per-business, not per domain or site.  Here’s the announcement on their blog.

Alternatives for javascript map generation:

Google Maps has made mapping ubiquituous (I’m sure Brian Timoney would agree), and in doing so has done the web a great service.  It will be interesting to see what happens as they try to charge for this service.

GIS Google Maps | moore |

8z Real Estate Hackfest

Recently, at the company I work for, 8z Real Estate, we had a hackfest.  A hackfest, for those not in the know, is a chance for employees to spend time working on whatever they want to do but don’t have time to do during the normal business day.  It’s also known as ‘FedEx day’ because you build something to ship in one day.

The idea is to give everyone a chance to do something work related that they want to do, or try, or explore, but don’t have time to because of the hustle and bustle of work life.

From the post at Atlassian (as far as I know, the originators of FedEx day).

[The] task must be something “out of the ordinary”. This is hard to define – but basically the spirit is that you can’t do something you would normally do. It’s a chance to attack all those “I wonder if XYZ would work… “, “It would be nice if we could… ” small … tasks that always get pushed off in the heat of battle.”

We had about 11 employees and contractors gather at the office.  Our schedule:

  • 8:00 Employees arrives, normal work begins
  • 10:00 Employees cease normal work
  • 10:00 – 10:30 brainstorm session
  • 10:30 – 10:45 pick projects
  • 12:00-1:00 eat lunch (Snarfs!)
  • 10:45-4:15 continue work on project
  • 4:30-5:30 present projects (8 min presentation) and drink a beer

We had an 2 hour block of work at the beginning of the day because we needed to, but after that, almost no one did everyday work.  Phones were off, email was closed.

The type of projects selected varied.  Most folks weren’t developers, so we didn’t have a ton of shipped software.  But we had some really interesting ideas, ranging from investigation of interesting technologies (what’s coming down the pike with our e-newsletter sender, infographics) to outlines of business ideas, to refactoring of business processes.

The excitement of all working together, in one room, on different projects, for a fixed amount of time, with no interruptions was one highlight.  I also really enjoyed people’s varied takes on aspects of the business.  It was also impressive to see the skills that I didn’t know some people had (powerpoint, for one).  It was awesome how many good ideas we had, even though some of them would have taken a hackfest week to implement.

All in all, it was a worthy experiment and something every business should consider doing.

8z Business | moore |