I believe that you raise two issues: 1) the mechanism of installing firmware; 2) the consequences of outdated firmware. Since the subject revolves around leap seconds, it would have been fun to delay this response by one day so that it would be posted on leap day, but I decided that the topic is too urgent to justify any delay in responding. It is only about 123.4 days until the next leap second will be added. (Note that I have now passed to you the opportunity to post a leap second reply on leap day.)
1) It is my intent to live with whatever firmware is installed by USGlobalSat. On October 21, 2011 you provided the instructions and files necessary to upgrade firmware, and on January 6, 2012 DaveGee reported having done that successfully. Considering the cautionary statements that were included, and my limited software skills, I do not intend to try that, nor is it considered practical for my customers (scattered around the globe) to do it for themselves. I believe that USGlobalSat should do the upgrade for free on any of the EM-406a returned to them, but that is not likely to be an attractive alternative to many.
2) Having outdated firmware means that the Universal Time reported by our product can be an integer number of seconds in error for up to 15 minutes after power is applied. Most of our customers are expected to use their timing device on an irregular schedule, such as once or twice a week. Consequently, in most instances their unit will have “forgotten” the latest almanac that was downloaded because that is retained for only a day or two. (The one older EM-406a that I have measured extensively “remembers” for about 35 hours.) If the firmware accounts for the current number of leap seconds, there will be no error at any time. However, in June 2012 I will be alerting all of my present customers that their timing devices will be expected to report time incorrectly by one second for the first 10 or 15 minutes after they apply power, beginning July 1, 2012. Our software detects when the reported time is corrected, informs the user of the precise time when that happens, and gives instructions on how the previous times are to be corrected. We would prefer that the error never be there, but if our users pay attention, their records will never include erroneous times.
It would be better if all of our future customers can be supplied with EM-406a incorporating firmware that accounts for all leap seconds through July 2012. Then they can expect that through June 2012 their unit will start with a one second error (which will automatically correct itself within 15 minutes), but which will be error-free after June 2012 – at least for a year or more until the next leap second is added. The time during which their device is likely to start with erroneous times is then at most a few months.
Our preference is for the EM-406a firmware to be updated as soon as possible.
Thanks for your continued pursuit of this issue.