Ah, the day (and night) in the life of a developer. You know, there are more professional golfers in the world, than people who make a living making astronomy software and products. If only the pay were as good - LOL. Seriously though, I realize just how lucky I am to be working in this field, much less to be working at Software Bisque, at the top of the innovation food chain. I might be biased... but remember kiddies, being biased doesn't mean your wrong, it just makes being right more fun ;-)
So what is a typical day like for me? I couldn't tell you... but I can tell you about yesterday because it was just a really great day. I spent most of the morning doing emails, and working on some future projects that are still way down the road (and I'm not going to tell you about them just yet <g>). After lunch, I setup outside on my patio and worked on some focuser issues. I have several focusers and have used several of them for testing with @Focus3. A few users however have had problems with focuser movements and I could not figure out why. Then when setting up a new system for shooting the moon, I switched to a focuser I had not used in a quite a while, indeed never used it with @Focus3 at all. Guess what? Focuser movement commands when into never-never land. Sometimes sporadic code failures are really highlighted by grumpy hardware, or just a slightly different sequence of events (we call these code-paths). There is nothing wrong with this focuser brand mind you, nor the plug-in. Still, I spent most of the afternoon getting to the bottom of this, and it was one of those moments as a programmer when you wonder why this code ever worked at all then, if THAT little thing was really the problem.
Well, this was exciting and now I could try @Focus3 on the moon tonight with my super long focal length telescope. It got dark, and my plan was to "play" for about an hour in my back yard.
Ah, the plans of mice and men. If you've followed the @Focus3 threads, some beta testers have reported that @Focus3 crashes, and/or there are some WEIRD ADU readings of over 100,000, or even in the millions. Well, nothing like yet another new configuration to shake things out... and almost every time I tried to focus, it would crash. This is a wonderful opportunity as a developer to get to the bottom of things, but alas I would not be joining my wife to help clean a few episodes off the DVR tonight.
The iron was hot, and I could repeatably reproduce a problem some customers had reported, and until now I could not. So, in my development environment, I traced through the code carefully, watched what was going on... and it looked like memory corruption. I could not find anywhere that this was happening, and I instrumented the code six ways from Sunday... three hours later, BOOM. I found it.
Another... HOW DOES THIS EVER WORK, MUCH LESS ONLY SOMETIMES FAILS?!?! type bug.
This boys and girls is why @Focus3 is still a "Technical Preview" - LOL. But, I think this may actually be the last serious bug as what I found (it was in my new hot pixel removal routine btw) explains two of the three known occasional issues with @Focus3, and the focuser movement bug early in the day should fix the other. A couple of the automation clients are also testing the @Focus3 scripting, and I think the focuser movement issue was a problem for them).
Anyway, for a long days work, it's nearly midnight and I've fixed three different serious bugs today. LTI, and @Focus3 in TheSkyX Pro are both more stable to boot. Oh, and I've modified the wizard that computes the focuser search size. It's a different formula now that takes into account pixel size, and this was also a final test for that (it seems to work from f/3 to f/15, and with pixel sizes from 4 to 9 microns, so I think we are good).
Well, what to do at midnight with a clear sky, a beautiful moon, and a well-oiled imaging system? Of course, I focused and took some images of the Moon. It was then, I realized, I could actually slew from crater to crater with my long focal length and it was working! Gotta love TPoint! So, I had one more crazy idea. I exited LTI, and started TheSkyX Pro and used the mosaic tool to create a mosaic that covered the whole Moon (and followed the Moon). I took a single photo at each mosaic point, and assembled my masterpiece with Microsoft ICE. Yes, it was my first time (doing Mosaics), and it was wonderful. This is why we do what we do at Software Bisque. ;-)
12-05-2017 7:27 AM