Software Upgrades

[Lance]

I just watched Dave's review of the new oscilloscope. Generally I personally can't decide whether or not I hate the idea of having to pay money to unlock features a device I own is fully capable of to begin with.

Pro: You don't need to go out and get a new piece of hardware.

Con: It's just such a silly concept. You have all these features on a device, but you can't use any of them unless you pay out the rear end for them. Why not offer a device with decent capabilities, and then sell it at a lower price? Need more features? Get a better device.

Going back to the example of the scope I think Agilent is really banking on people paying for these extra features down the road. It's good in theory, but then there's always people who are going to end up paying for the hardware but never end up unlocking it. I think it's wasted resources.

I can't decide whether or not it's a good thing. What I'm really afraid of is magpie managers ordering this to be put into devices that it has no business being in. Can you imagine if say, a programmer/debugger tool for micro-controllers had this going on? Similar to what the folks at Microchip talked about in their response to Dave's review of the PICkit3. What if companies really did start making people pay for silly little upgrades?

That's just my thoughts. What do you fellows think?

[Psi]

I don't like the idea at all, but if the end result is they're able to sell the product cheaper then i can't really complain.

It's hard to imagine it being cheaper to produce units that have hardware that will never be used in the entire life of some units but i guess that's just the way it is.

I suspect much of the savings that makes this possible, and counteracts the cost of the wasted parts that never get used, comes from all the internal business costs that are no longer needed. Costs associated with having a wide range of different products all requiring different service support, helpdesk, bug fixes, marketing etc..
When you're a company as big as HP/Agilent with lots of employees and branches around the world i imagine these costs are huge.

There is also the savings from only having to design and manufacture a small range of different units and all the cost savings associated with bulk discounts on the parts now used in higher quantities.

I think i would be happier about this sort of thing if they eventually released the full speed/functionality firmware for free X years after release, maybe when the next model arrived or something, but companies never seem to do that.

You can however look at the whole thing another way.  If you had a piece of test equipment with no disabled features or reduced speed then the chances someone will hack it and enable them is zero

It does, after all, give hardware hackers something fun to do.

I think this modified quote from Stargate SG1 matches up to this whole practice

Company: "We will put disabled features into these products you use, BUT. Be warned. Anyone attempting to enable them will be shown no legal mercy."

Hacker: "Well.. That certainly makes life more. Interesting..."

[ziq8tsi]

What if companies really did start making people pay for silly little upgrades?

Or, worse still, you have to rent the premium features.

Next step, your new device is sold with a "trial size" three month subscription, after which the hardware you bought automatically bricks itself to extort more money.

[alm]

I don't like it, it will make it necessary for companies to lock down their designs to prevent people from using options without paying. If it requires expensive parts (eg. different ADC's, or different PCB), you don't have to protect the details since people are less likely to implement them. It also promotes the concept that you should not be allowed to modify your own equipment. Another issue seen on equipment from the nineties is that they sometimes lose their configuration (dead backup battery or EEPROM). Since those options cost money and are often locked to the serial number, you can't just reset the memory. Only the manufacturer can re-enable the options (assuming they still have the software, i.e. it's no more than a few years old). But if we're lucky the equipment won't be around that long anyway .

Next step, your new device is sold with a "trial size" three month subscription, after which the hardware you bought automatically bricks itself to extort more money.

They already do. I know Tektronix ships scopes with some option enabled for a limited time. You have to pay if you want to keep using them.

[Uncle Vernon]

I understand companies need to recover R&D but even to this end crippling features or functions soon becomes immoral.

It's bad enough in test gear but when the concept is alive and well within medical electronics you really have to question it.

We have a successful company here that has pioneered world breaking implant technology. The products bought at reduced price by charities and third world medical services are set to offer poorer performance, to ensure none ever find their way back to first world countries via grey markets.

You really have to consider how managers can decide market protection of a more important than the hearing ability of a child in the third world.

Hardware and software disabling is a magnet for dubious hackers it just creates an environment where the greed of the illegal modifier is pitched against the greed of the corporations.

[Psi]

Definitly agree with that.

Having a medical unit that doesn't perform as well as its capable of because the company wants to sell more advanced units is unethical. The lack of enabled features could easy result in a death or injury that would not have occurred otherwise.

[mikeselectricstuff]

Unfortunately the problem is that manufacturers are there to make money and answer to shareholders, not to make the coolest stuff as cheap as possible.
I think however there is a distinct difference between additional software features and deliberately crippled hardware - the latter leaves a very bad taste & feels like you're being ripped off.  
I have some sympathy for software authors whose software is ripped off, but none whatsoever for hardware manufacturers whose crippleware gets hacked.  

Next step, your new device is sold with a "trial size" three month subscription, after which the hardware you bought automatically bricks itself to extort more money.
They already do. I know Tektronix ships scopes with some option enabled for a limited time. You have to pay if you want to keep using them.

I wonder how many support calls they get complaining their scope has stopped working properly...

Agilent do some time-limited trial licenses, but only 2 weeks  - from the user's point of view this is probably better than shipping with a time limit, as you could use them if you have a 1-off job requirement, and you don't get that sinking feeling when your new scope suddenly gets less useful 3 months after you bought it.
And it shouldn't be _too_ hard to "influence" the RTC's idea of how long a week is...

[adam1213]

Definitly agree with that.

Having a medical unit that doesn't perform as well as its capable of because the company wants to sell more advanced units is unethical. The lack of enabled features could easy result in a death or injury that would not have occurred otherwise.

A company might produce a basic version of the software initially. If they can't make any additional money from it then what motivation do they have to add the extra features - so are they unethical for not wasting money? However if they have a basic machine that is very accurate and they deliberately make it less accurate (eg GPS) then this is another issue.

Alternatively the company could release the source code so they don't have to spend money on having it developed further.

[Psi]

Definitly agree with that.

Having a medical unit that doesn't perform as well as its capable of because the company wants to sell more advanced units is unethical. The lack of enabled features could easy result in a death or injury that would not have occurred otherwise.

A company might produce a basic version of the software initially. If they can't make any additional money from it then what motivation do they have to add the extra features - so are they unethical for not wasting money? However if they have a basic machine that is very accurate and they deliberately make it less accurate (eg GPS) then this is another issue.

yeah, i was meaning features that are disabled and where all the software or hardware already exists. Basically any feature that requires no effort to 'switch on' other than flipping a software switch in the firmware.

[Sionyn]

agreed its like this micro payment stuff

but a stupid move if as easy to hack as the rigol they have lost again

it will be interesting when dave opens the 3000 series

[Lance]

I have some sympathy for software authors whose software is ripped off, but none whatsoever for hardware manufacturers whose crippleware gets hacked.

Agreed. I don't pirate software on principle. Crippleware on the other hand, I wouldn't have an issue with. I'm paying for the hardware, why shouldn't I be allowed to use it?

[Uncle Vernon]

Unfortunately the problem is that manufacturers are there to make money and answer to shareholders, not to make the coolest stuff as cheap as possible.
I think however there is a distinct difference between additional software features and deliberately crippled hardware - the latter leaves a very bad taste & feels like you're being ripped off.  
I have some sympathy for software authors whose software is ripped off, but none whatsoever for hardware manufacturers whose crippleware gets hacked.  

I don't see the distinction between hardware and software, both are intellectual property both are due a fair reward. Most of what you pay for hardware is a return for expertise not the component cost.

Payment for upgrades and ongoing development new versions etc is entirely reasonable, but the idea of crippling functionality of hardware or software is not the creation of a designer or engineer. A fair price is a fair price it has nothing to do with what sectors of the market can reasonably accommodate.

M$crosoft and music publishers for example scream blind about pirate theft while at the same time crowing about obscene profits. Part of good design is matching a product to it's intended marketplace.  If M$crosoft priced it's retail packages at similar levels to what the offer it to Dell etc, I'd suggest their piracy problems would be a fraction of current levels and revenues would not suffer.

The Agilant pricing strategy is a invitation for their products to be hacked. How much of future development effort then goes into a endless battle
of hacker verses upgrade security?  Crippling your product from delivering it's best performance is just flawed short term marketing.
 

[Zero999]

I don't see the distinction between hardware and software, both are intellectual property both are due a fair reward

I see your point about deserving a reward for R&D but disagree with you about intellectual property being the same as hardware i.e. real property. If you run a business making oscilloscopes and someone steals 1000 units you loose the value of 1000 units. On the other hand, if you make software and 1000 people pirate it you don't necessarily loose the value of 1000 licences because that would be assuming they would've all bought it and that they haven't shown it to their friends or employers who might buy it.

Most of what you pay for hardware is a return for expertise not the component cost.

Is that really still true for oscilloscopes?

Think of all the code that goes into the FPGAs, MCUs and flash memory ICs. Come to thing of it, even with the ADCs, it wouldn't surprise me if you're paying for the mask and schematic more than for the silicon.

Then there's calibration: is a hacked 'sope as accurate as the genuine 'scope at the higher frequencies? It might be possible the premium units are subject to more calibration than cheaper models.

Though I do agree if all you're paying for is features which you've already bought the hardware for. I suppose it wouldn't be so bad if they demanded more money for something like FFT, integration or true RMS measurements.

[Uncle Vernon]

Most of what you pay for hardware is a return for expertise not the component cost.

Is that really still true for oscilloscopes?

Think of all the code that goes into the FPGAs, MCUs and flash memory ICs. Come to thing of it, even with the ADCs, it wouldn't surprise me if you're paying for the mask and schematic more than for the silicon.

Yes very true, The IP proportion isn't just limited to the finished product. Components are finished products in their own right. Equally their value is more in the IP effort taken to create them than in the value of the modified sand they comprise.

If crippleware migrates further into chip level where does it all end?

[mkissin]

If crippleware migrates further into chip level where does it all end?

Welcome to the future.

All those 3-core CPUs that AMD sells? Generally 4-core units with one otherwise functional core disabled.

All those "different" video cards that nVidia and AMD sell? Generally the same silicon with different clock speeds and some of the silicon disabled.

This is market segmentation, and it's how the world works. Sure, without it, the most expensive items would likely be cheaper, but the cheapest items would also be more expensive. It would have to meet in the middle to recoup all of the R&D costs, which will generally completely dwarf the costs of the physical item you're buying.

[Lance]

Most of what you pay for hardware is a return for expertise not the component cost.

Is that really still true for oscilloscopes?

Think of all the code that goes into the FPGAs, MCUs and flash memory ICs. Come to thing of it, even with the ADCs, it wouldn't surprise me if you're paying for the mask and schematic more than for the silicon.

Yes very true, The IP proportion isn't just limited to the finished product. Components are finished products in their own right. Equally their value is more in the IP effort taken to create them than in the value of the modified sand they comprise.

If crippleware migrates further into chip level where does it all end?

We can't really predict. We think like engineers, ie smart people. As I said earlier I see magpie managers attracted to this idea and sticking it on everything.

[EEVblog]

The Agilant pricing strategy is a invitation for their products to be hacked. How much of future development effort then goes into a endless battle
of hacker verses upgrade security?  Crippling your product from delivering it's best performance is just flawed short term marketing.

No, it's clever, because they know the vast majority of their customers either won't know about any hack, or wouldn't be bothered to implement it.
The majority of sales on these scopes will go to educational institution and businesses, not to people who follow forums and like to hack things.

Dave.

[EEVblog]

Then there's calibration: is a hacked 'sope as accurate as the genuine 'scope at the higher frequencies? It might be possible the premium units are subject to more calibration than cheaper models.

That's a myth that was perpetuated with the Rigol hack.
One of the key drivers behind the decision to offer these options is the fact they only have to build one product, and that naturally extends into the test and calibration realm as well.
In fact, every Agilent scope MUST be tested and calibrated for full bandwidth before it leaves the factory, because it is a user field upgradable option they are selling.

Dave.

[Lance]

The Agilant pricing strategy is a invitation for their products to be hacked. How much of future development effort then goes into a endless battle
of hacker verses upgrade security?  Crippling your product from delivering it's best performance is just flawed short term marketing.

No, it's clever, because they know the vast majority of their customers either won't know about any hack, or wouldn't be bothered to implement it.
The majority of sales on these scopes will go to educational institution and businesses, not to people who follow forums and like to hack things.

Dave.

I guess that makes sense. I personally don't like the idea of buying crippled hardware. For schools I guess it makes sense for schools to get pretty decent scopes that don't do a whole ton of stuff. They don't really need the fancy functionality.

[Zero999]

If I were a lecturer I would add 'scope hacking to the syllabus. lol

I don't think it would be a bad is the cheap end of the market became more expensive and the higher end cheaper. Customers would get a better deal that there'd be less stuff thrown away.

[metalphreak]

From a business standpoint, it makes a whole lot of sense.

Agilent can group their costs into two categories: "production" and "research and development/support"

I can only speculate, but I'm assuming that the base 70Mhz 2ch scope for $1200 more than covers their cost of production, while also paying for some of the R&D. The R&D costs are the same regardless of the number of units sold, so as long as they are making a profit on the hardware costs, any sales of the lower end scopes are a bonus. Especially if those customers never actually *need* a higher end scope which they would *not* have purchased if it was the only available option (they would buy other brands etc). So by having the lower models, they can gain a larger market share without eating into their sales of other scopes whatsoever. Those who require the extra features will pay for them.

From a buyers standpoint, it also makes a whole lot of sense.

You are basically being given a rather large discount on the hardware by forgoing some of the features. Later on, if you need them, you can simply purchase the licences to enable them. You're not missing out on features that you've already paid for by buying the hardware, because you haven't *paid* for them. Now, if there were features that the board was capable of, but Agilent refused to even offer them as an upgrade option, I would be more pissed off  (ie the hardware is there, and I have *no* legitimate way to enable it)

[Psi]

I guess it all depends how you look at it. If when you buy something you view what you're buying as a collection of components or as a unit with X number of features.

If you think of it as a collection of components and it includes components that can do feature X and it has the firmware to do X then, since you own the item, you should be able to use feature X because you paid for the components that perform the function.

On the other hand, if you view it as a unit with a set of features that you purchased then it all seems normal and you wouldn't expect to use features you didn't pay for.

Engineers i think, tend to look at it the first way.

[Lance]

I guess it all depends how you look at it. If when you buy something you view what you're buying as a collection of components or as a unit with X number of features.

If you think of it as a collection of components and it includes components that can do feature X and has the firmware to do X as well then, since you own it, you should be able to use feature X because you paid for the components that do it.

On the other hand, if you view it a unit with a set of features that you purchased then it all seems normal and you wouldn't expect to use features you didn't pay for.

Engineers i think, tend to look at it the first way.

I know I do. If I pay for the hardware I should be able to use it. I'm sure the base $1200 covers the hardware and some R&D. I imagine everyone would flock to their scopes if they offered such an amazing deal. Easily recoup the costs.

[Psi]

Just out of interest, how many people would have no issue with disabled features if they had rented the item instead of bought it outright?  (ignoring the fact that you can't exactly solder hacks onto something you dont own)

Personally i would feel fine with disabled features if i was renting something.

[mkissin]

I know I do. If I pay for the hardware I should be able to use it. I'm sure the base $1200 covers the hardware and some R&D. I imagine everyone would flock to their scopes if they offered such an amazing deal. Easily recoup the costs.

Ah, and therein lies the rub, I think. You've paid for the hardware, and indeed you can do whatever you like it (except that you might void your warranty). That's absolutely your right. What you haven't paid for, and thus cannot use, is the Agilent software than operates that hardware.