"ethics" (?) of forced supply purchases

This offer valid only at sea level! (not on your way SPIRALLING DOWN

*TO* IT!) :>

Not to mention the possibility of having to demonstrate, in a court of law, that the equipment was being operated in accordance with manufacturer's instructions.

--
"For a successful technology, reality must take precedence 
over public relations, for nature cannot be fooled."
                                       (Richard Feynman)

This offer not valid on days ending in 'Y'. ;-)

--
You can't have a sense of humor, if you have no sense.

A prime example of Don's off-topic posts. Absolutely nothing to do with embedded.

The device I'm designing is portable. Handheld. Battery powered. Uses a removable 16GB solid state disk to implement a DBMS. Has a gestural input system, speech output. It addresses a particular market segment for a particular application domain. It has I/O's particular to that domain.

Some would say it's an embedded system. (you could opt to call it a "really tiny desktop computer running a very specialized application with very unique hardware if you really want to avoid using that term!)

It has two clearly identifiable supplies/consumables: a battery and a storage device. Both are key to the use and operation of the device. I.e., the device won't function in the absence of either!

Questions:

- Do *we* supply the user with these replacements?

- Do we adopt COTS supplies so the user can purchase replacement supplies from other vendors?

- Are there concerns for the user's purchasing on the open market that could compromise the product's reliability?

- Our reputation?

- How does this affect the hardware and software design of the product?

Extra credit:

- How does this affect our profitability?

- What staffing consequences does it have for us?

- What attitude do we take towards pricing these supplies?

- How are our customers likely to view these pricing/availability policies?

- How are our competitors likely to *exploit* these issues?

- Do we want to be *in* the "supplies" business -- for a potentially long period of time?

- If we opt to be a/the supplier, what further risks do we take on if an alternative supplier comes onto the scene?

Is there something that prevents embedded system designers from being concerned about these issues? "Sorry, it's not my job!" Or, are we not allowed to discuss it because the same sorts of issues also apply to toilet paper dispensers and disposable razors?

[If you need a MORE descriptive statement of my application, I'm sorry but you'd have to contact my client and probably sign an NDA. I suspect they wouldn't be interested in talking to you, though -- lack of imagination!]

=46inally got around to asking.

?-/

It's complicated... Not helpful, I know, but it is.

With at least some consumables, you can move profit to the consumable, and sell the base product for less. The old razors-and-razorblades model. If this helps your sales, it can be a huge win. Or if nobody likes your gadget (and thus doesn't buy any consumables), a big loss.

With simple stuff, customers are more likely to be annoyed. Basic batteries, for example, or simple SD cards (more borderline, but arguable). What's reasonable to specify what is required, and that any warranty service caused by non-conforming supplies will be 100% billed to the customer, and that you accept no incidental losses caused by failures due to the third party supplies (assuming you do in any event). Further, you might want to produce and document some testing procedures (perhaps incorporate it into the format procedure when a new SD card is installed - better yet, add your own checksums to the data stored).

And then offer supplies *you've* verified with a reasonable markup. If the customer is worried, they'll buy your stuff. If they don't, and they have a problem, you can always suggest a buying a certified part on the phone as part of troubleshooting. And customers who want to go it on there own are happy too.

But for something like (common) batteries and SD cards, I'd really not worry too much about performance. With SD cards you might spec minimum and maximum sizes.

On the third hand, does the device *need* a 16GB storage to function? Then you should definitely ship one with it (perhaps with an option to upgrade it if appropriate, or remove it if you really think a customer would want to supply their own. But 16GB devices are not likely to cost that much, so go for simple.

Agreed. *My* situation adds to it (see below)

Yes.

All points understood.

Now, the complication -- that makes the issue important.

(Drawing on other posts that I've made recently, without repeating them here)

Imag "We need to make a *custom* battery/memory card/whatever to safeguard against the case where the user 'asynchronously' removes the device".

Or:

"We need to make a *custom* memory module that can be updated continuously INSTEAD OF using some off-the-shelf memory card technology."

Or:

...

This can conflict with "decision X" -- even if decision X had been "let's be in the supplies business"! And, it can interfere in a big way!

E.g., you may have decided there would be lots of markup on those COTS supplies that you are essentially rebranding and reselling. Suddenly, it's *not* a COTS device but, rather, some assembly that you are manufacturing! And, as the cost (DM+DL) has gone up considerably over those COMMODITY PARTS that you were drooling over previously, your profit margin either takes a beating -- or, the "supplies" suddenly are exhorbitantly priced (to satisfy your Y% markup):

"Who the heck is going to pay $100 for 16GB of 'memory' when they see 16GB memory cards selling for $10? (granted, this

*isn't* a COTS memory card, but the user doesn't know/care...) Our customers will scream if we expose this sort of cost to them! And, our competitors will make hay of it!" [I was purchasing a door checker for my other half's auto the other day. A woman came in with a broken ignition key (the kind with the "chip" inside, wireless locks, etc.). When told the new key -- including reprogramming of her old "spare" *and* the vehicle -- would set her back $280 (!) she nearly fell over! She walked away from the purchase!! I'm sure that experience did not help the car dealership *or* car manufacturer in that lady's opinion!]

Or, worse, imagine decision X was to NOT be in the supply business. Perhaps you had decided that users could purchase COTS supplies from 3rd parties -- without requiring any contractual requirements *with* those 3rd parties (i.e., "Install 3 AA batteries in the battery compartment located on the back of the device...").

Now, that option is not available to you! Instead, you're faced with "manufacture the devices and offer them as supplies" (i.e., you are now in the supplies business that you did NOT want to be in) *or* locate and contract with a third party to supply these items on your behalf (i.e., something you presumably didn't want to be bothered with!).

I.e., it's important (to me) to understand how clients are going to look at this issue so that I am prepared to counter their decisions/reactions to specific *design* proposals.

E.g., if the supplies business is a non-starter, then my fallback position might have to be, "There are no user replaceable supplies" (i.e., the battery is rechargeable and only replaced at the factory; the memory module is integrated, not expandable and not replaceable; etc.)

This, in turn, affects how the system treats those "non-supplies" (no need to worry about battery being suddenly disconnected -- since it can't be removed; just worry about it GRADUALLY discharging -- which you can anticipate!)

=20

Maybe, and then only if the pressure regulator to the external world is permanently part of the cartridge.

Just the same restaurants seem to be able to have plenty of trained personnel to handle large cylinders of CO2 (2200 psi full). Not all resituate personnel are the sharpest knives in the drawer. I am sure = that a civil suit or product liability suit defense lawyer could bring this = up.

?-)

As a customer, I have no real concern over stuff I can't replace, or can replace only with difficulty, being vendor-supplied only. I do get annoyed if I look at a part that's *obviously* COTS, or a part that could be COTS, and I'm being charged an outrageous price for it - but for sealed stuff, I don't see it, so I largely don't care.

A couple of cases for discussion, though.

Batteries in mid-sized* UPSs are almost always (user) replaceable. The batteries (usually several identical ones in a given UPS) from are one of perhaps a dozen different models, but are assembled**, wired and connectored differently between UPS manufacturers and models therein (although there's often several UPS models using the same battery pack).

Replacing UPS batteries is not particularly difficult, but not something anyone does on a regular basis. And I'd say the lifetime of the typical UPS is such that you'd change the batteries on it once. Invariably it requires a screwdriver and some other tools to accomplish, and there's obviously an electrical*** safety hazard handling big lead acid batteries of those capacities (hence the typical pre-connectorization).

Now in most cases there *are* third party vendors of replacement battery packs, who use compatible batteries and attach similar bits of wire, and they usually charge about half what the UPS vendors do. The UPS vendors are clearly not encouraging the practice, but they tolerate it with the usual "it might void your warrantee" warnings. That being said, I think most users buy the from the UPS vendors. Here we often cycle older UPS's down to less critical functions, and those get the cheapo third party batteries.

Something mitigating against sealing the devices is the shipping weight of these things, which makes shipping them back to the factory cost prohibitive, plus the safety requirements *during* shipping, which the mid-sized and larger UPS meet (in part), by initially shipping the batteries disconnected (although almost always in the battery compartment).

UPS vendors also used to use oddly wired RS-232 serial cables for connecting their monitoring ports. Again they'd sell you those for a silly amount of money, but third party cables were available, and you could always wire your own - and besides, the (odd) pinouts were invariably available. These days most of that is USB, and while non-standard cables are possible, it usually doesn't happen.

As a second interesting example, I'd cite Apple and and Android phone makers. Most 'droids have user replaceable batteries and SD slots. Apple has neither (although most Apple devices can have their batteries replaced by Apple). Both user bases seem fairly happy, although a certain portion of the Android market is clearly people for whom those features matter.

I guess what I'm saying is that if you use four AAs to power your device, and you're going to charge me $100 to replace them, you better make it hard for me to discover the truth.

*And by "mid-sized" I mean most of the ones in the 1000-5000VA+ range

**And "assembled* in some cases means no more than a wrap of duct tape fastening a pair of batteries together.

***Obviously they're also heavy and contain hazardous substances

Think about the two examples I posited: batteries and memory.

Damn near everyone can understand "battery needs replacing". And, people have some notion of what a battery "should" cost (whether it's a couple of bucks for a classic cell or ~$80 for a laptop).

Likewise, memory cards in cameras, phones, PDAs, etc. have become commodities in the minds of consumers. People *expect* that they can increase the memory in their just by shelling out a "few" extra bucks (for varying values of "few").

I.e., we're not talking about replacing the throttle position sensor in an automobile and arguing about whether to go with "factory original" or "aftermarket" parts. These aren't obscure items that we have little awareness of, "under the hood"

Returning to the car key example I cited... *I* could understand the justification behind the $280 charges for that poor woman's replacement key. First, there's the (overpriced) "blank". And, it's got to be "cut". Nothing funny there (though when you subconsciously compare that to buying a new HOUSE KEY involving a blank and "cutting" at the hardware store, we're already off by a couple orders of magnitude!). And, when you consider how the key *works*, it's obvious that it has to be paired with the vehicle that it will "talk to" (or, vice versa). And, that could potentially involve reprogramming the *other*, existing keys to co-operate with this new/foreign one. And, those grease monkeys *need* $100/hr (brudened rate) to wield those tools... :>

OTOH, to that lady looking for the replacement key, she looks at the piece of broken plastic + metal key shaft in her hands and does NOT think: "This is $280 worth of (broken) stuff in my hands!". Instead, she thinks, "It's just a ^%&@!* KEY!! Why the heck is it so expensive??" (hardware store)

OToOH, if you told her that the "bypass caps for the logic supply regulator in her ECU" had failed (20c parts?) and that it would cost her $1200 for the repair, she would grimace... and ask you how soon she could get the car back!

Batteries and memory "smell" more like keys than ECUs.

So, do you see the UPS vendors as making their money primarily on the UPS? Or, the supplies? :> (Note that the size range of UPS you addressed is targeted more at small businesses and small data centers -- places with budgets for UPS batteries)

I suspect there are three classes of users wrt UPS batteries.

- Individuals/SOHO users who probably toss the UPS when it "stops working"

- Large firms (with regular maintenance budgets) that err on the side of just buying from the OEM ("No one was ever fired for buying from IBM" mentality)

- Small firms with "clever" support personnel who have "discovered" that they can "save money" by replacing these supplies with "equivalent" 3rd party offerings -- even if they don't understand the issues that determine whether a particular offering is good value or not.

Note that the effectivity of search engines means anyone who is motivated can probably find out how to do these sorts of things on their own -- outside of the sanction of the manufacturer.

And, just because you *can* find some information about a particular subject/issue, doesn't mean the information is *correct*! It often represents someone's "best guess" or OVERCONFIDENT GUESS as to how something *appears* to work.

So, you can see some "yahoo" telling folks that you can increase the size of the memory card by replacing it with XXXXX -- just because it *seems* to work (in the limited testing that the yahoo conducted). With no concern for durability, access speed, etc. ("Huh? What are those things? Isn't memory just memory??")

Or, suggesting that you can replace one battery technology/capacity with another -- without realizing that the device might implicitly rely on characteristics of the original battery choice ("OK, once the cell voltage drops to XXX with YYY load, I know I have ZZZ seconds of useful device operation remaining so I had better start backing up the data in volatile memory...")

And, you can see people feeling good about themselves (smug!) because they have stumbled on some little secret that the manufacturer didn't expect to be disclosed -- without realizing that the information is dubious, at best!

I think those are quick turnaround products. You *expect* to replace either of them relatively frequently. You *want* to replace them frequently so you can indulge yourself with the latest and greatest model/feature set.

They're also fairly inexpensive. How would attitudes change if the devices cost $2-5K?

And, how would you (the user) react to learn you had to live without the use of your Apple/Android product for a week while it travels to a repair depot, is repaired and later returned? (*Can* you turn it that quickly?) This adds to the cost of supporting the supply(ies) as those folks have to get paid in addition to supplying the replacement parts!

If you don't want the user poking around inside the device (to *see* those COTS components *or* try to install a 3rd party alternative), then the enclosure wants to be "resistant" to easy opening (yet all of these devices have detailed pictorial teardowns posted on the web). Do you *preserve* the original case (complete with the assorted stickers with which the user may have adorned it, markings, etc.)? Or,

*replace* it (another cost item -- depending on how it is fabbed and how the guts fasten to it, etc.)? Does the user complain that his cherished stickers are now gone? Or, relish the fact that he has a "brand new" (looking) device?

But you *know* it's "just a battery (of some sort)". Regardless of what it looks/feels/tastes like, you "feel" like a battery shouldn't cost you $100. Look at how many laptops have bad batteries and their owners are content to live with them in that state (granted, the laptop has a few years on it by that time but a new battery would be cheaper than a new laptop -- yet they'll keep the old laptop with bad battery instead of buying

*either*!) [I am not claiming laptop batteries are priced incorrectly wrt their cost. Rather, priced incorrectly wrt their perceived value as "batteries"! -- I can buy a CAR BATTERY for less :-/ ]

Personally, I would prefer to see supplies handled as an aftermarket issue. Make your money on the initial sale. Come up with realistic pricing/etc support those decisions. Then, step aside and let "whoever" deal with the little issues of maintenance. Spend your effort on making the product better since you *know* it's easy for someone to undercut you in the commodity market (i.e., someone who might not care about actual quality but appeals to consumers strictly on *price*)

The (big) downside (to me as a system developer) is that I now have to be prepared to deal with an activity that I can't control -- the user "asynchronously" doing something that I don't want him to do! (Apple doesn't have to worry about users replacing batteries in their kit -- even if the user

*does* manage to do so!)

Answer from another correspondent posted directly to s.e.d.

Repeating it here:

?-)

Those *are* the OEM Canon ones! And they are overpriced too! Almost everything hitech is on an exchange rate of GBP £1 = $1 USD.

On a good day I can get authentic BCI-6 Bk for £6, CMY for £7 in the UK and the more expensive PM and PC for £10. I can sometimes do even better if there is a buy one get one half price deal on. Clones are much cheaper, but they don't work at all and I don't run sufficient print volume to justify a bottled ink conversion kit (which does work).

$15 each for BCI-6 does not look like a good price to me.

Amazon.com puts a bound on original Canon cartridges of $53/6 ~ $9

Some cheaper clone cartridges alson come up on a wider search but the ones I have tested in the UK have been worse than useless. Thankfully they didn't damage the print head but neither did they work well.

--
Regards,
Martin Brown