Price of Javelin
Justin Shumaker
Posts: 22
I was curious what sort or volume on sales would need to be made in order to bring the pricetag of the javelin down to the $50 - $60 price range so that it was competitive with the basic stamps. Do you forsee this happening over time, or are the amount of sales preventing the current price from diminishing right now? Or does the cost of producing each javeline just cost so much that it can't be brought down much lower? I have a sneaky suspicion that sales would land somewhere between double and triple of where they currently are if the cost of this product hit the $49.95 mark because at that point it would become an economical alternative to the popular basic stamp and additionally if folks break a couple it won't send them to the poor house, thus reducing the risk factor of getting involved with this "new" product. Thoughts?
- Justin
- Justin
Comments
Just one.
I don't see Parralax under selling their "popular basic stamp".
Jack
The development costs for a product such as the Javelin are quite large, especially compared to a mature product such as the Stamp. It always costs much more to produce the first of something compared to revisions/updates/improvements. The best bet for such a mark down would be for them to come out with a newer edition of the Javelin, but seeing as how the sales are lackluster, I would imagine it would be difficult for anyone within the company to press for such a product.
Post Edited (Paul Baker) : 3/21/2005 8:32:37 PM GMT
- Justin
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Jon Williams
Applications Engineer, Parallax
Dallas, TX· USA
1. Not enough IO capability. The Javelin gives you all that variable space, a decently flexible language, and enough horsepower to do fairly complicated things. And then gives you 16 pins (not enough for large projects), only 6 of which can effectively be "fancy" (Uarts, PWM's etc). I ended up buying 2 to get a room automation project·running - and then asking myself why I didn't just buy a Chickadee or something, which would have given me more I/O and much more speed.
2. 16 bit ints are cute. But for serious projects, 32 should be available. Natively.
3. Compiler quality. "short" seems to cause the compiler to choke obscurely. Float and long could have been supported, and probably should be. And the compiler doesn't optimize as well as it should: (a? 1 : 0) compiles down to way too many instructions, for one memorable example. This hardware *demands* the ultimate a compiler can provide in optimization.
4. Fragility. The pins are cheap! A device like this (one that costs this much, basically) is going to see a LOT of insertions and removals, and I've already broken off a pin one one (fortunately, not an IO pin). I'm not comfortable plunking down that much cash for a chip that can't survive my clumsy fingers. There are better pin technologies out there, and an $89 unit should use them.
5. Speed. Contrary to the documentation, having 6 VPs chugging along at once DOES seem to affect program speed. I had to do a lot of hand tuning.
6. Fragility, part 2. Like most microprocessors, a casual brush with a wire is instant death for these things.·Output pins don't survive shorts, and heaven help you if a wire carrying 12v gets away from your fingers for a second. It would be NICE if the board was sealed over, and nicer if the pins had guard resistors so shorts didn't fry them. Yes, I know, that makes rctime operations harder, but sheesh. Even a 220 ohm resistance on each pin would be some protection from casual mistakes.
Don't get me wrong - A Javelin can do far more than a BS/2 ever can, and I have 2 Javelins running a sophisticated sound room, handling communication with a sound system, 4 channels of lighting, remote control, plus odd jobs like monitoring a phone line, doorbell, flood alarm and security alarms. You can't do that with 10 basic stamps. It's just that it's been misery to develop, the stamps have been pushed to their limits, it's not as fast as I'd like, it cost a fortune, I'm not done yet, and I don't know if I'd do it the same way again if I was starting from scratch.
But I'd pay lots more than double for a different form factor, 32 I/O pins, a 2nd, onboard processor with power to handle 32 VPs, pins that could handle 50mA output each, over-voltage input·and open shorts without dying, and no attempt at onboard power regulation·(leave that for dedicated power supplies!) A decent Java implementation in command of all that would be a Home Automator's dream system.
·
·
Instead, I'd rather have the pin be an output that signaled when JVM stops due to an unhandled exception. That might help in situations where I put my Javelins, basically on a board in a box without a console line to a PC to tell me why it has stopped responding. The output could have a big resistor there to prevent damage should someone put it in a socket with Vin on it.
The IO count is fine as near as I can tell. If I need more simple inputs or ouputs, a shift register does the job. More VPs would be nice, but I'm not sure they are really necessary. If I did start needing more, slave processors or other dedicated hardware seem to be more appropiate.
lBill
Yes, more VPs are necessary, the same way that 4+Ghz 64 bit processors are necessary - there's always a project that needs more, faster and wider.
I agree that an "I'm Dead" output could be useful, but in practice you only get exceptions when the code is wrong, and the answer is always to fix the code. Once you've fixed the code, the pin becomes useless (well, unless the VM starts throwing spurious exceptions, which it does quite readily if the power sags - I went NUTS for a weekend trying to figure out why my code was randomly dying, and the answer was, I was running 2 Javelins on the wall adapter power supply that comes with the BOE. Almost enough power, not quite.) Yes, "fix the code" is the classic pollyanna answer, but I'm a software engineer.
When the Javelin crashes, do all pins revert to inputs? You could set a pin Low when you start, pull it weakly high with a resistor, then watch for a high value to detect crashes, in that case. Or use plan B: use try/catch to catch all exceptions from main, and then set an output pin low, which is tied to RST... it's a hack and it eats a pin, but auto-reboot-on-any-error-and-hope-no-one-noticed has worked for Microsoft for years.
This, IMHO, makes the BS2 a GREAT product for beginners!
I'm by no means comfortable (let alone competant) in programming....and having lots of sample codes for the BS2 makes them great for 'grabbing & going'!!
But I'll admit....I don't have much of a clue what I'm doing with the Javelin!
If there was a WAM type manual for the Javelin or some "Javelins In Class" notes, this product might take off a bit more!
Sadly, if it doesn't pick up, you could probably expect to see the price come down (as parallax lets go of their stock!) [noparse]:([/noparse]
It's taking me time...but I'm slowly learning!! If I weren't already bald...well...there you have it! lol
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
·
Steve
http://ca.geocities.com/steve.brady@rogers.com/index.html
"Inside each and every one of us is our one, true authentic swing. Something we was born with. Something that's ours and ours alone. Something that can't be learned... something that's got to be remembered."