$10 STM32 kit

'cloud' everything is a business model, it suits the accountants more then the engineers and may be selected for valid reasons.

If you have not realized, mbed is rapid prototyping tool, the clue is in the webpage title... code from here is supposed to be on the shelf after proving.

Talk about young dumb and living off mom. You are a moaning fanboy and should be banned from your mommy and daddy=92s modem.

tes

You're assuming the orgainization supports this flexibility of choice.

In some shops, The Hardware Guy has the lion's share of the processor choice; in others, The Software Guy does (especially for firms that don't design their own hardware); still others split this somewhat rationally.

Some shops place bias on "systems" that they have already used. Or, on those for which they have toolkits or existing library support investments.

Ah, but this analogy is incorrect! "Testing" shampoo/soap/perfume has very little cost *and* very little risk (if the soap "smells bad", you take another shower, etc.). OTOH, testing an MCU has a considerable cost (learning curve, etc.) -- which is why vendors try so hard to keep the real costs (development board + tools $$) and "intangible" costs (how quickly can you make that LED blink?) low. If you have to invest more than "petty cash", there is a political impediment in most companies -- namely, getting approval for the purchase. If you have to invest serious *time*, that impediment is even greater (hence the appeal to peddle things that an "enthusiastic engineer" can bring home and play with on his own time -- and own *dime*!).

The bigger concern is the *risk* attendant with such a "let's try this MCU" approach. Few firms want to "play" with new devices SOLELY for the sake of "getting experience" with yet another MCU. Why risk discovering "once you're pregnant" that the device has critical flaws? Or, they can't get their fab ramped up quickly enough? Or, that some other HUGE player has bought all of the available production for the next N months?

I don't think people are "naturally loyal". Rather, I think they are some mix of "lazy" (minimize effort) and "timid" (minimize risk). "Inertia" is a better word than "loyalty".

A "professional" usually sees the "$10" or "$50" cost as trivial in the greater scheme of things. He/she knows that the *true* cost of this venture is many orders of magnitude greater. So, the difference between 10 and 50 sits *in* the noise floor...

I see professional as open minded, I know that one cant be vary good at many platforms/devices. But having ability to test myths of devices(with debug) from other manufacturers is simply great. Currently, situation is not so great for developers, manufacturers make you specialize their tools, devices, spend big corp. money(or personal) which management(you) doesn't like, and later switch is quiet painful. Me as young engineer/student like to have broad knowledge on technology and devices. Specialize on one manufacturer device family could be bad decision(platform is OK, I think). Because many business have internal contract with different manufactures and their devices. When time comes I will have enough knowledge and money to buy some eval. board and quicker pass first steps... Hmmm, I hope that realistic scenario =3D)

It's arguable. For me, a $10 board that I don't use is a $10 loss. If I actually try some development with a board, I'll put in enough time to dwarf the price difference from a $50 board. I have a threshold somewhere below $100 where I start to care.

Mel.

I got the impression Microchip started with this model. Possibly the first with cheap / free tools, support for small users and hobbyists etc. Seems to have worked for them. (Never used them myself since the original PIC architecture was so ugly, and now there is ARM...)