after all the failed product launches and underperforming hardware you'd think people would be over the whole 'pre-order' fantasy of things being as good as initially advertised/marketed. i understand problems happen but items should be heavily discounted if a manufacturer wants money before there is hard evidence of performance available to buyers.
Excellent point Bart. One thing we discussed recently was how products seem to get to reviewers late in the cycle -- and usually after official release. It would be better for buyers and for the sellers to seed products with reviewers pre-release. But it seems as if the incredible speed of innovation in these markets is causing companies to announce features that don't really exist, to prototype hardware and just assume firmware can fix problems later, etc.
Reputation becomes very important as a result. A great example in my case recently was the Atomos Shogun. Atomos has a great reputation for both making excellent products and for solid service. So, many (including me) pre-ordered the Shogun. Then the release was delayed, the initial firmware was incomplete, and months later we finally have a solid and usable product (it's actually pretty amazing at this point with more features to come). Even the best out there are just guessing about the development calendar. And, increasingly, the software/firmware is the last piece of the puzzle (and almost always more difficult than assumed). This is a classic example of something happening just because it can: Firmware can be shoved in at the very last minute of development, unlike hardware which depends on outside vendors, tooling, PCB layout and stuffing, etc.
A common result of all of this, by the way, is over-built hardware (more capable micro controllers, more memory, etc.). The worst outcome for the builder is to find out down the road that the promised firmware features need memory that isn't there, or processing cycles that the hardware can't provide. So, better to overbuild a bit. This is in large contrast to the days when firmware largely didn't exist. Then the mark of a good engineer was the ideal and most efficient design.
Everything is in perpetual beta, and the choice is to be an unpaid beta tester with new hardware or to use yesterday's products. The right decision depends on what we need, when we need it, and how much we trust the company selling us that shiny new thing. Another example: We are at a place now where flight controllers and firmware from a year or two ago can be the best choice.
The gimbal problem is actually more interesting than it first appears. Demands are incredibly high. A little horizon drift and the internet blows up with complaints. But anyone with any experience knows that you can get amazing footage even with a little drift here and there. Maybe you touch up the roll axis heading into a shot, you edit around it, work the camera angle to avoid revealing the problem, or live with it and call it art. A gimbal that can deal with vibrations and produce smooth movements is the key, and that seems to have a lot to do with the physical as much as the software.
Anyway, long post. I remain very interested in what
@SleepyC has to say about the Gremsy H6.
