Forty Days: 2. Upgrade blog software

Day two in the Forty Days of NFPchart series:

The NFPchart blog software has been upgraded to the latest and greatest version.

The blog is powered by WordPress. Updates to the software are released as often as every few weeks to fix bugs and patch security flaws. We had been running a version over a year-and-a-half old. That’s a lifetime in blog software.

There may have been a brief moment when the blog was in maintenance mode. It wouldn’t have lasted more than a few seconds.

Forty Days: 1. Renew SSL certificate

Day one in the Forty Days of NFPchart series:

Yesterday I renewed the SSL certificate for NFPchart.

An SSL certificate is a code used to encrypt the secure https connection between your web browser and the NFPchart server. Having an encrypted connection is important when sharing personal information over the internet. After all, the information shared via NFPchart is quite personal.

The old SSL certificate expired yesterday at around 10 am Arizona time. I installed the new certificate at 8 am. There was no significant interruption in https service, although there may have been a few brief hiccups when I restarted the web server.

Announcement: Forty Days of NFPchart

Since putting NFPchart online over two years ago, I’ve admittedly done very little to improve it. I’m always looking forward to finding a month of free time where I can focus on some serious enhancements. It’s a month that never arrives.

After talking it over with Kathryn earlier this week, we agreed it would be better if I focused not on a major overhaul, but rather on a number of incremental enhancements. I came up with the idea of Forty Days of NFPchart. I’m proud to announce it today.

Here’s how it works: Every day, now through August 15, I’ll try to make some minor improvement to NFPchart. When I do, I’ll describe the change on this blog.

I already know there are a few days when I won’t have time to make even a simple change. Still, if I manage to create some modest enhancement on even half of these upcoming days, NFPchart will have improved a lot.

Resolved: charts not appearing

Some of you may have noticed overnight that charts weren’t appearing no matter how long you waited. The process for building the charts was dying because I failed to include a critical file in my latest build. Fortunately, it was an easy fix. The problem has been resolved, and charts should now appear normally.

Resolved: bleeding intensity

My wife noticed today that bleeding intensity could not be edited from the New Observation or Edit Observation forms, although it could be edited in place in the Observations table. I’ve fixed the problem, and bleeding intensity should now be editable both ways.

New features: peak days and pre-rise baselines

chart-prb-annotatedI’d like to introduce two new features I’ve brought online over the past several days.

Peak days are automatically calculated as the last day of the most fertile sign — either lubrication or egg-white mucus. The calculated peak days appear in the table and also on the chart. Optionally, you can override the automatically calculated peak days in any cycle on the Edit Cycle page. This is handy if the most fertile sign during a cycle is not lubrication or egg-white mucus.

Pre-rise baselines are also automatically calculated for each cycle, using the three-higher-than-six method. Disturbed observations are disregarded in the calculation, so be sure to mark them carefully. Like peak days, you can override the automatically calculated baseline temperature for any cycle on the Edit Cycle page.

Humbled by recent usage

Even though the site isn’t finished, a number of people have recently signed in and created new accounts. In particular, there’s been a noticeable uptick since the beginning of the year. Many have even started charting.

Thanks for the vote of confidence!

Since the site is still not feature complete, I’ve bumped the end date of the free preview period back to March 31, 2010.

The minimal set of features I’m envisioning includes:

  • Plotting baseline temperatures
  • Printable charts (PDF)
  • Optional sharing online with instructors

Until then, please continue to enjoy the free preview!

Facelift underway

I’m not much of a web designer, so I don’t always recognize when a website truly looks great. That said, after almost fifteen years of programming and developing websites, I do recognize when one looks downright awful.

When I personally develop web applications, I have only one real design goal: Keep it simple. After recently looking over what I’ve done so far with the application, I realize I’ve fallen far short of that goal.

In both the application and the blog, you’ll notice I’ve begun removing many of the thick rounded borders around various elements of the pages. Most of them were unnecessary, so getting rid them is a good thing.

Although my primary focus right now is on adding new functionality to the application, I’ll also be keeping an eye out for ways to simplify the look and feel. Visitors to the site should expect frequent changes.

As always, comments, criticisms, and suggestions are welcome. Fire away!

Resolved issue signing in with Yahoo!

For several hours yesterday, visitors trying to sign into the application using the Sign in with Yahoo! link were receiving the following error message:

Sorry, the OpenID server couldn’t be found

After some trial and error on my part, it appeared Yahoo! stopped responding to OpenID server requests at one domain but was still responding on another. I updated the domain name in the application, and it seems to be working properly now.

One of the features I’ve been developing will allow accounts to be associated with more than one OpenID. The feature was intended for spouses who want to sign into their accounts with their own identities. However, when the feature is released, an individual account owner could link more than one OpenID from different providers, preventing a lockout if one server goes offline.

Showstopper stopped

There was a bug in the charting feature that prevented a chart from being rendered if there was only one observation in a cycle and the temperature of the observation was left blank.

It’s admittedly an edge case, and perhaps I’m exaggerating to call it a showstopper. However, it wasn’t making a good first impression for visitors who created a new account, created a cycle, and then added a blank observation to see what would happen.

The bug appears to be fixed.

Older Posts »