Tim,
> > On the subject of how much support for $500, I
would answer NONE.
> >
> > Companies that fail to realize this don't tend to last that long.
>
> This is completely not my experience as the owner of a small company that
> develops embedded products. Our software suppliers (notably Rowley, and
> our Altium CAD reseller) give us amazing software support. What type of
> business are you in which makes you think this?
There are three type of customer:
#1. The competent customer
Can get along fine most of the time, only needing a small amount of support
when shit hits fan or when being particularly clueless until the bulb
illuminates. We have lots of this type of customer. Supporting them is a
pleasure.
#2. The incompetent customer
Needs lots of assistance, and usually doesn't know arse from elbow.
They
are best dealt with by asking them to read the manual, or referring them to
an FAE for seriously dump questions.
#3. The unreasonable customer
Has paid $1,500 for a license and thinks they have purchased a chunk of you
for that outlay, with unlimited immediate-response support as they are under
"severe time pressure". They think they know better than you, they think
their problems are absolutely the highest priority, and they never say sorry
when they're wrong. They are best dealt with by refunding their payment
and
asking them to look elsewhere. We've done this a number of times in the
past. Most of the time they can't put a coherent question together
because
they are "under time pressure". Then they don't do what you ask them to
do
and do something else, helpfully report a new set of observations, and ask
why things still don't work.
Thankfully, most of our customers fall into the first category. You really,
really don't want a customer that starts out a conversation "I'm under
time
pressure to..." as the last category.
--
Paul Curtis, Rowley Associates Ltd http://www.rowley.co.uk
SolderCore running Defender... http://www.vimeo.com/25709426
ARM Development Tools
Started by ●March 29, 2012
Reply by ●April 2, 20122012-04-02
Reply by ●April 2, 20122012-04-02
On 02/04/2012 15:46, Tim Mitchell wrote:
> ----Original Message----
> From: l...
> [mailto:l... ] On
> Behalf Of Phil Young
> Sent: 02 April 2012 15:33 To: l...
>
> Subject: RE: [lpc2000] Re: ARM Development Tools
>
> > On the subject of how much support for $500, I would
> > answer NONE.
> >
> > Companies that fail to realize this don't tend to last
> > that long.
>
> This is completely not my experience as the owner of a small company
> that develops embedded products. Our software suppliers (notably Rowley,
> and our Altium CAD reseller) give us amazing software support. What type
> of business are you in which makes you think this?
I don't know about the Altium resellers, but Altium themselves have had
severe financial problems for many years:
http://www.google.com/finance?q=ASX%3AALU&hl=en
They claim to have sold about 18,000 licenses. With 285 staff to pay I
can't see how they manage to keep trading.
Leon
--
Leon Heller
G1HSM
> ----Original Message----
> From: l...
> [mailto:l... ] On
> Behalf Of Phil Young
> Sent: 02 April 2012 15:33 To: l...
>
> Subject: RE: [lpc2000] Re: ARM Development Tools
>
> > On the subject of how much support for $500, I would
> > answer NONE.
> >
> > Companies that fail to realize this don't tend to last
> > that long.
>
> This is completely not my experience as the owner of a small company
> that develops embedded products. Our software suppliers (notably Rowley,
> and our Altium CAD reseller) give us amazing software support. What type
> of business are you in which makes you think this?
I don't know about the Altium resellers, but Altium themselves have had
severe financial problems for many years:
http://www.google.com/finance?q=ASX%3AALU&hl=en
They claim to have sold about 18,000 licenses. With 285 staff to pay I
can't see how they manage to keep trading.
Leon
--
Leon Heller
G1HSM
Reply by ●April 2, 20122012-04-02
Hi Tim,
Working in the past as FAE and AE for various semiconductor companies I've
seen this happen to many small companies, and in several start-ups I've
worked in I've seen the same occurring.
There are exceptions to any rule, but a customer that needs a lot of support
needs to be a very serious customer.
In my experience in NemeriX I saw this many times, with huge delays in
product development caused by customers asking too many technical questions
but not having the production volume to justify the effort answering them,
this always involved the development team getting involved.
This was eventually fixed but not before it did a lot of damage.
A typical SOC development these days costs in excess of $10,000,000 so
getting the big customer to revenue generation has to be the biggest
priority.
The last SOC I was working on, until just a few weeks ago has a development
budget several times that and at these costs and complexities you really
cannot afford to support more than 1 or 2 customers properly.
That's why companies like nVidia for example simply produce 1 reference
design that everybody copies, including all the physical design criteria and
usually even gerber output and all SW, they simply cannot afford to support
people with hundreds of questions.
Regards
Phil.
From: l... [mailto:l...] On Behalf Of
Tim Mitchell
Sent: 02 April 2012 15:46
To: l...
Subject: RE: [lpc2000] Re: ARM Development Tools
----Original Message----
From: l...
[mailto:l... ] On
Behalf Of Phil Young
Sent: 02 April 2012 15:33 To: l...
Subject: RE: [lpc2000] Re: ARM Development Tools
> On the subject of how much support for $500, I would
> answer NONE.
>
> Companies that fail to realize this don't tend to last
> that long.
This is completely not my experience as the owner of a small company that
develops embedded products. Our software suppliers (notably Rowley, and our
Altium CAD reseller) give us amazing software support. What type of business
are you in which makes you think this?
--
Tim Mitchell
Working in the past as FAE and AE for various semiconductor companies I've
seen this happen to many small companies, and in several start-ups I've
worked in I've seen the same occurring.
There are exceptions to any rule, but a customer that needs a lot of support
needs to be a very serious customer.
In my experience in NemeriX I saw this many times, with huge delays in
product development caused by customers asking too many technical questions
but not having the production volume to justify the effort answering them,
this always involved the development team getting involved.
This was eventually fixed but not before it did a lot of damage.
A typical SOC development these days costs in excess of $10,000,000 so
getting the big customer to revenue generation has to be the biggest
priority.
The last SOC I was working on, until just a few weeks ago has a development
budget several times that and at these costs and complexities you really
cannot afford to support more than 1 or 2 customers properly.
That's why companies like nVidia for example simply produce 1 reference
design that everybody copies, including all the physical design criteria and
usually even gerber output and all SW, they simply cannot afford to support
people with hundreds of questions.
Regards
Phil.
From: l... [mailto:l...] On Behalf Of
Tim Mitchell
Sent: 02 April 2012 15:46
To: l...
Subject: RE: [lpc2000] Re: ARM Development Tools
----Original Message----
From: l...
[mailto:l... ] On
Behalf Of Phil Young
Sent: 02 April 2012 15:33 To: l...
Subject: RE: [lpc2000] Re: ARM Development Tools
> On the subject of how much support for $500, I would
> answer NONE.
>
> Companies that fail to realize this don't tend to last
> that long.
This is completely not my experience as the owner of a small company that
develops embedded products. Our software suppliers (notably Rowley, and our
Altium CAD reseller) give us amazing software support. What type of business
are you in which makes you think this?
--
Tim Mitchell
Reply by ●April 2, 20122012-04-02
> In my experience in NemeriX I saw this many times,
with huge delays in
> product development caused by customers asking too many technical
> questions but not having the production volume to justify the effort
> answering them, this always involved the development team getting
> involved.
NemeriX folded, IIRC. NemeriX purchased a number of CrossWorks for ARM
licenses.
> That's why companies like nVidia for example simply produce 1 reference
> design that everybody copies, including all the physical design criteria
> and usually even gerber output and all SW, they simply cannot afford to
> support people with hundreds of questions.
The unfortunate thing is that even some microcontrollers on the merchant
market, offered for sale to customers, are just not documented adequately
enough. So, as a silicon vendor you have a choice: produce excellent
documentation to fend off questions, or produce sketchy documentation and
push the problem down to the customer or FAEs. Or produce equally crap
examples and say "well, here you are, this is a free example. If it doesn't
work for you, well, it's an example and it's free, what more do you want?"
It's a hard one. Given the rate of introduction of new silicon, possibly
the only way to run is by producing some documentation that is mostly right,
with a chip that's mostly right, and write up the rest as errata once
discovered, or put together some app notes/examples.
With big chips come big errata. The Freescale iMX5 has so many errata,
without any planned fixes, there must have come a time when they said "I'm
sick of this part, it's good enough, just ship it..."
--
Paul Curtis, Rowley Associates Ltd http://www.rowley.co.uk
SolderCore running Defender... http://www.vimeo.com/25709426
> product development caused by customers asking too many technical
> questions but not having the production volume to justify the effort
> answering them, this always involved the development team getting
> involved.
NemeriX folded, IIRC. NemeriX purchased a number of CrossWorks for ARM
licenses.
> That's why companies like nVidia for example simply produce 1 reference
> design that everybody copies, including all the physical design criteria
> and usually even gerber output and all SW, they simply cannot afford to
> support people with hundreds of questions.
The unfortunate thing is that even some microcontrollers on the merchant
market, offered for sale to customers, are just not documented adequately
enough. So, as a silicon vendor you have a choice: produce excellent
documentation to fend off questions, or produce sketchy documentation and
push the problem down to the customer or FAEs. Or produce equally crap
examples and say "well, here you are, this is a free example. If it doesn't
work for you, well, it's an example and it's free, what more do you want?"
It's a hard one. Given the rate of introduction of new silicon, possibly
the only way to run is by producing some documentation that is mostly right,
with a chip that's mostly right, and write up the rest as errata once
discovered, or put together some app notes/examples.
With big chips come big errata. The Freescale iMX5 has so many errata,
without any planned fixes, there must have come a time when they said "I'm
sick of this part, it's good enough, just ship it..."
--
Paul Curtis, Rowley Associates Ltd http://www.rowley.co.uk
SolderCore running Defender... http://www.vimeo.com/25709426
Reply by ●April 2, 20122012-04-02
OK thanks to Paul and Phil for their interesting perspective from the other side
of the parapet!!
I won't say too much about Altium, suffice to say their UK reseller is extremely helpful when negotiating some of the more "unusual features" of the software.
--
Tim Mitchell
I won't say too much about Altium, suffice to say their UK reseller is extremely helpful when negotiating some of the more "unusual features" of the software.
--
Tim Mitchell
Reply by ●April 2, 20122012-04-02
--- In l..., "Phil Young" wrote:
> That's why companies like nVidia for example simply produce 1 reference
> design that everybody copies, including all the physical design criteria and
> usually even gerber output and all SW, they simply cannot afford to support
> people with hundreds of questions.
>
Perhaps this is why Microchip is investing a ton of money in developing and giving away code for their processors. First, it eliminates a lot of questions if there is working code to copy. Second, everybody can try before they buy (silicon, in this case).
Microchip isn't the only one doing this, of course. But the scale seems larger than others I have seen. I suspect that this will become the norm over time. If all of the protocol stacks are provided (serial, I2C, SPI, USB, ethernet) and sample code is available for all the other peripherals, all that remains is the application layer. This reduces development costs and shortens time-to-market.
Often times, time-to-market means everything. There is a window of opportunity that soon closes when others enter the market. You absolutely have to be there first.
Richard
> That's why companies like nVidia for example simply produce 1 reference
> design that everybody copies, including all the physical design criteria and
> usually even gerber output and all SW, they simply cannot afford to support
> people with hundreds of questions.
>
Perhaps this is why Microchip is investing a ton of money in developing and giving away code for their processors. First, it eliminates a lot of questions if there is working code to copy. Second, everybody can try before they buy (silicon, in this case).
Microchip isn't the only one doing this, of course. But the scale seems larger than others I have seen. I suspect that this will become the norm over time. If all of the protocol stacks are provided (serial, I2C, SPI, USB, ethernet) and sample code is available for all the other peripherals, all that remains is the application layer. This reduces development costs and shortens time-to-market.
Often times, time-to-market means everything. There is a window of opportunity that soon closes when others enter the market. You absolutely have to be there first.
Richard
Reply by ●April 2, 20122012-04-02
Correct, Nemerix Folded when it ran out of cash and had serious production
problems.
The problems turned out to be a subcontract packaging house in Asia using
faulty underfill, but it's never good to unable unable to ship product,
particularly when going through a funding round.
NemeriX used ARC as the embedded core processor with a GHS / Metaware hybrid
toolchain, ARM development tools were for integrating the application
software into customers systems and various toolchains selected to suit
customer requirements.
ARC won out over Tensilica for the core because the Tensilica environment
was GCC / Eclipse based.
The IP was eventually acquired for a fraction of the cost of developing it
and investors lost a lot of money, this shows how important focussing on KEY
customers is.
Had NemeriX not wasted so much time supporting worthless customers for the
earlier chips the cashflow would have been much different and it could have
survived this problem, but it was probably 6-9 months behind where it could
have been with better focus.
From: l... [mailto:l...] On Behalf Of
Paul Curtis
Sent: 02 April 2012 16:45
To: l...
Subject: RE: [lpc2000] Re: ARM Development Tools
> In my experience in NemeriX I saw this many times, with huge delays in
> product development caused by customers asking too many technical
> questions but not having the production volume to justify the effort
> answering them, this always involved the development team getting
> involved.
NemeriX folded, IIRC. NemeriX purchased a number of CrossWorks for ARM
licenses.
> That's why companies like nVidia for example simply produce 1 reference
> design that everybody copies, including all the physical design criteria
> and usually even gerber output and all SW, they simply cannot afford to
> support people with hundreds of questions.
The unfortunate thing is that even some microcontrollers on the merchant
market, offered for sale to customers, are just not documented adequately
enough. So, as a silicon vendor you have a choice: produce excellent
documentation to fend off questions, or produce sketchy documentation and
push the problem down to the customer or FAEs. Or produce equally crap
examples and say "well, here you are, this is a free example. If it doesn't
work for you, well, it's an example and it's free, what more do you want?"
It's a hard one. Given the rate of introduction of new silicon, possibly
the only way to run is by producing some documentation that is mostly right,
with a chip that's mostly right, and write up the rest as errata once
discovered, or put together some app notes/examples.
With big chips come big errata. The Freescale iMX5 has so many errata,
without any planned fixes, there must have come a time when they said "I'm
sick of this part, it's good enough, just ship it..."
--
Paul Curtis, Rowley Associates Ltd http://www.rowley.co.uk
SolderCore running Defender... http://www.vimeo.com/25709426
problems.
The problems turned out to be a subcontract packaging house in Asia using
faulty underfill, but it's never good to unable unable to ship product,
particularly when going through a funding round.
NemeriX used ARC as the embedded core processor with a GHS / Metaware hybrid
toolchain, ARM development tools were for integrating the application
software into customers systems and various toolchains selected to suit
customer requirements.
ARC won out over Tensilica for the core because the Tensilica environment
was GCC / Eclipse based.
The IP was eventually acquired for a fraction of the cost of developing it
and investors lost a lot of money, this shows how important focussing on KEY
customers is.
Had NemeriX not wasted so much time supporting worthless customers for the
earlier chips the cashflow would have been much different and it could have
survived this problem, but it was probably 6-9 months behind where it could
have been with better focus.
From: l... [mailto:l...] On Behalf Of
Paul Curtis
Sent: 02 April 2012 16:45
To: l...
Subject: RE: [lpc2000] Re: ARM Development Tools
> In my experience in NemeriX I saw this many times, with huge delays in
> product development caused by customers asking too many technical
> questions but not having the production volume to justify the effort
> answering them, this always involved the development team getting
> involved.
NemeriX folded, IIRC. NemeriX purchased a number of CrossWorks for ARM
licenses.
> That's why companies like nVidia for example simply produce 1 reference
> design that everybody copies, including all the physical design criteria
> and usually even gerber output and all SW, they simply cannot afford to
> support people with hundreds of questions.
The unfortunate thing is that even some microcontrollers on the merchant
market, offered for sale to customers, are just not documented adequately
enough. So, as a silicon vendor you have a choice: produce excellent
documentation to fend off questions, or produce sketchy documentation and
push the problem down to the customer or FAEs. Or produce equally crap
examples and say "well, here you are, this is a free example. If it doesn't
work for you, well, it's an example and it's free, what more do you want?"
It's a hard one. Given the rate of introduction of new silicon, possibly
the only way to run is by producing some documentation that is mostly right,
with a chip that's mostly right, and write up the rest as errata once
discovered, or put together some app notes/examples.
With big chips come big errata. The Freescale iMX5 has so many errata,
without any planned fixes, there must have come a time when they said "I'm
sick of this part, it's good enough, just ship it..."
--
Paul Curtis, Rowley Associates Ltd http://www.rowley.co.uk
SolderCore running Defender... http://www.vimeo.com/25709426
Reply by ●April 2, 20122012-04-02
Il 02/04/2012 17:53, rtstofer ha scritto:
>
>
> --- In l... ,
> "Phil Young" wrote:
>
> > That's why companies like nVidia for example simply produce 1 reference
> > design that everybody copies, including all the physical design
> criteria and
> > usually even gerber output and all SW, they simply cannot afford to
> support
> > people with hundreds of questions.
> > Perhaps this is why Microchip is investing a ton of money in
> developing and giving away code for their processors. First, it
> eliminates a lot of questions if there is working code to copy.
> Second, everybody can try before they buy (silicon, in this case).
>
> Microchip isn't the only one doing this, of course. But the scale
> seems larger than others I have seen. I suspect that this will become
> the norm over time. If all of the protocol stacks are provided
> (serial, I2C, SPI, USB, ethernet) and sample code is available for all
> the other peripherals, all that remains is the application layer. This
> reduces development costs and shortens time-to-market.
>
> Often times, time-to-market means everything. There is a window of
> opportunity that soon closes when others enter the market. You
> absolutely have to be there first.
>
> Richard
>
Richard,
I asked you if you think to pay $500 for a networking library like the
Microchip. So I mean with similar documentation and examples. I asked
this because if you read carefully their license you should see that you
can use that library only with microchip products and with the same
library but without any license restriction I think you should pay for
it right?
More or less $500 are about 5 to 10 hours of cost for a developer.
A part that I think you should consider:
1. how much time do you need to start to use the library?
2. Who helps you in case of problems?
Microchip and other vendors, are getting some free libraries because
they develop or pay for them to introduce their silicon so we can say
that who uses their silicon pays some money also to this development.
The problem is that they know that customer service is another cost and
they would avoid it so releasing the source code and making a blog or a
mailing list they create a self organizing support.
>
>
>
> --- In l... ,
> "Phil Young" wrote:
>
> > That's why companies like nVidia for example simply produce 1 reference
> > design that everybody copies, including all the physical design
> criteria and
> > usually even gerber output and all SW, they simply cannot afford to
> support
> > people with hundreds of questions.
> > Perhaps this is why Microchip is investing a ton of money in
> developing and giving away code for their processors. First, it
> eliminates a lot of questions if there is working code to copy.
> Second, everybody can try before they buy (silicon, in this case).
>
> Microchip isn't the only one doing this, of course. But the scale
> seems larger than others I have seen. I suspect that this will become
> the norm over time. If all of the protocol stacks are provided
> (serial, I2C, SPI, USB, ethernet) and sample code is available for all
> the other peripherals, all that remains is the application layer. This
> reduces development costs and shortens time-to-market.
>
> Often times, time-to-market means everything. There is a window of
> opportunity that soon closes when others enter the market. You
> absolutely have to be there first.
>
> Richard
>
Richard,
I asked you if you think to pay $500 for a networking library like the
Microchip. So I mean with similar documentation and examples. I asked
this because if you read carefully their license you should see that you
can use that library only with microchip products and with the same
library but without any license restriction I think you should pay for
it right?
More or less $500 are about 5 to 10 hours of cost for a developer.
A part that I think you should consider:
1. how much time do you need to start to use the library?
2. Who helps you in case of problems?
Microchip and other vendors, are getting some free libraries because
they develop or pay for them to introduce their silicon so we can say
that who uses their silicon pays some money also to this development.
The problem is that they know that customer service is another cost and
they would avoid it so releasing the source code and making a blog or a
mailing list they create a self organizing support.
>
Reply by ●April 2, 20122012-04-02
--- In l..., "M. Manca" wrote:
> Richard,
> I asked you if you think to pay $500 for a networking library like the
> Microchip. So I mean with similar documentation and examples. I asked
> this because if you read carefully their license you should see that you
> can use that library only with microchip products and with the same
> library but without any license restriction I think you should pay for
> it right?
Well, yes, the Microchip stack is limited to Microchip devices. That seems fair to me.
By the way, that mythical $500 stack probably involves per-unit royalties so the life-cycle cost will be much higher. I have seen C libraries without per-unit royalties but the more esoteric stuff usually does require royalties.
C libraries almost have to come with the IDE/compiler toolchain to get out from under the licensing restrictions of the open source libraries.
> More or less $500 are about 5 to 10 hours of cost for a developer.
> A part that I think you should consider:
> 1. how much time do you need to start to use the library?
I think that will be the same regardless of the vendor. That there are copious examples is helpful.
In fact, since I wouldn't necessarily get source code for a purchased library, the startup time for the Microchip stack might be orders of magnitude less than a commercial stack. If I have a problem, I can just put in some debug statements and track it down. With an object file and a handbook, I might be stuck for a long time.
As you point out, time is money.
> 2. Who helps you in case of problems?
For a $500 stack, how much help are you going to get from any vendor? Not very much! Unless, of course, there are royalties involved to increase the vendor's interest.
> Microchip and other vendors, are getting some free libraries because
> they develop or pay for them to introduce their silicon so we can say
> that who uses their silicon pays some money also to this development.
Absolutely! There's no free lunch unless you are a hobbyist and doing one-off projects.
> The problem is that they know that customer service is another cost and
> they would avoid it so releasing the source code and making a blog or a
> mailing list they create a self organizing support.
Yes and that's far better than an 800 number that transfers to the middle east help center.
I like their business model. I remember the bad old days when development tools were multi-thousand dollar investments (and for high end FPGA tools, they still are) and only companies could work with the newest chips. Today it's a free-for-all with software and toolchains all over the place. I like it! Olimex should be a lot of credit for making various chips available in a usable format.
I think the chip vendors are forced to provide no-cost toolchains because, if they don't, their competitors will. The days of "Here's my latest chip, you figure it out!" are long gone.
Why does Xilinx make the WebPack ISE toolchain available for free? It's a huge expense with little return. It's simple! They want college kids to play with their silicon and when they finally get a job, perhaps they recommend Xilinx to their bosses. Yes, Xilinx makes a very costly toolchain for the commercial environment and yes, they guard their IP cores very carefully but even that is beginning to change. Altera has released their NIOS core (competitive advantage over Xilinx) complete with an Eclipse/GNU toolchain and pretty soon Xilinx will release their MicroBlaze project (I think they already have but I'm not sure). It will happen...
Have you ever looked into the video game "Unreal Tournament" (and the sequels)? Did you know that the entire level development software is available for free? Every smart kid on the planet can build a REAL video game using this stuff. Maybe they do and then maybe they go on to work for one of the big game companies and make a bazillion dollars. Good for them!
http://udk.com/
Unreal isn't the only company making their platforms available. But, BTW, if that kid does make money (over $50,000), Unreal wants a big piece of the action. In the meantime, the kid can experiment with a REAL development tool.
All the tools in the world would be worth nothing if their wasn't a fun summer camp for the kids.
http://www.internaldrive.com/
Richard
> Richard,
> I asked you if you think to pay $500 for a networking library like the
> Microchip. So I mean with similar documentation and examples. I asked
> this because if you read carefully their license you should see that you
> can use that library only with microchip products and with the same
> library but without any license restriction I think you should pay for
> it right?
Well, yes, the Microchip stack is limited to Microchip devices. That seems fair to me.
By the way, that mythical $500 stack probably involves per-unit royalties so the life-cycle cost will be much higher. I have seen C libraries without per-unit royalties but the more esoteric stuff usually does require royalties.
C libraries almost have to come with the IDE/compiler toolchain to get out from under the licensing restrictions of the open source libraries.
> More or less $500 are about 5 to 10 hours of cost for a developer.
> A part that I think you should consider:
> 1. how much time do you need to start to use the library?
I think that will be the same regardless of the vendor. That there are copious examples is helpful.
In fact, since I wouldn't necessarily get source code for a purchased library, the startup time for the Microchip stack might be orders of magnitude less than a commercial stack. If I have a problem, I can just put in some debug statements and track it down. With an object file and a handbook, I might be stuck for a long time.
As you point out, time is money.
> 2. Who helps you in case of problems?
For a $500 stack, how much help are you going to get from any vendor? Not very much! Unless, of course, there are royalties involved to increase the vendor's interest.
> Microchip and other vendors, are getting some free libraries because
> they develop or pay for them to introduce their silicon so we can say
> that who uses their silicon pays some money also to this development.
Absolutely! There's no free lunch unless you are a hobbyist and doing one-off projects.
> The problem is that they know that customer service is another cost and
> they would avoid it so releasing the source code and making a blog or a
> mailing list they create a self organizing support.
Yes and that's far better than an 800 number that transfers to the middle east help center.
I like their business model. I remember the bad old days when development tools were multi-thousand dollar investments (and for high end FPGA tools, they still are) and only companies could work with the newest chips. Today it's a free-for-all with software and toolchains all over the place. I like it! Olimex should be a lot of credit for making various chips available in a usable format.
I think the chip vendors are forced to provide no-cost toolchains because, if they don't, their competitors will. The days of "Here's my latest chip, you figure it out!" are long gone.
Why does Xilinx make the WebPack ISE toolchain available for free? It's a huge expense with little return. It's simple! They want college kids to play with their silicon and when they finally get a job, perhaps they recommend Xilinx to their bosses. Yes, Xilinx makes a very costly toolchain for the commercial environment and yes, they guard their IP cores very carefully but even that is beginning to change. Altera has released their NIOS core (competitive advantage over Xilinx) complete with an Eclipse/GNU toolchain and pretty soon Xilinx will release their MicroBlaze project (I think they already have but I'm not sure). It will happen...
Have you ever looked into the video game "Unreal Tournament" (and the sequels)? Did you know that the entire level development software is available for free? Every smart kid on the planet can build a REAL video game using this stuff. Maybe they do and then maybe they go on to work for one of the big game companies and make a bazillion dollars. Good for them!
http://udk.com/
Unreal isn't the only company making their platforms available. But, BTW, if that kid does make money (over $50,000), Unreal wants a big piece of the action. In the meantime, the kid can experiment with a REAL development tool.
All the tools in the world would be worth nothing if their wasn't a fun summer camp for the kids.
http://www.internaldrive.com/
Richard
Reply by ●April 2, 20122012-04-02
Interesting. Are these the number of licenses "ever" sold or yearly? Are they
for their Tasking IDE or their PCB design SW?
Most customers will be on a subscription (new PCB design licences will be about $5..7k and substription maybe 1k a year). Altium Designer is a wonderful product and they need a big team to maintain and further develop it (as they do all the time). Didn't expect that they would have financial difficulties (maybe the management is the big expense as in many companies now-a-days?)
Regards
Mark
--- In l..., Leon Heller wrote:
>
> On 02/04/2012 15:46, Tim Mitchell wrote:
> > ----Original Message----
> > From: l...
> > [mailto:l... ] On
> > Behalf Of Phil Young
> > Sent: 02 April 2012 15:33 To: l...
> >
> > Subject: RE: [lpc2000] Re: ARM Development Tools
> >
> > > On the subject of how much support for $500, I would
> > > answer NONE.
> > >
> > > Companies that fail to realize this don't tend to last
> > > that long.
> >
> > This is completely not my experience as the owner of a small company
> > that develops embedded products. Our software suppliers (notably Rowley,
> > and our Altium CAD reseller) give us amazing software support. What type
> > of business are you in which makes you think this?
> I don't know about the Altium resellers, but Altium themselves have had
> severe financial problems for many years:
>
> http://www.google.com/finance?q=ASX%3AALU&hl=en
>
> They claim to have sold about 18,000 licenses. With 285 staff to pay I
> can't see how they manage to keep trading.
>
> Leon
> --
> Leon Heller
> G1HSM
>
Most customers will be on a subscription (new PCB design licences will be about $5..7k and substription maybe 1k a year). Altium Designer is a wonderful product and they need a big team to maintain and further develop it (as they do all the time). Didn't expect that they would have financial difficulties (maybe the management is the big expense as in many companies now-a-days?)
Regards
Mark
--- In l..., Leon Heller wrote:
>
> On 02/04/2012 15:46, Tim Mitchell wrote:
> > ----Original Message----
> > From: l...
> > [mailto:l... ] On
> > Behalf Of Phil Young
> > Sent: 02 April 2012 15:33 To: l...
> >
> > Subject: RE: [lpc2000] Re: ARM Development Tools
> >
> > > On the subject of how much support for $500, I would
> > > answer NONE.
> > >
> > > Companies that fail to realize this don't tend to last
> > > that long.
> >
> > This is completely not my experience as the owner of a small company
> > that develops embedded products. Our software suppliers (notably Rowley,
> > and our Altium CAD reseller) give us amazing software support. What type
> > of business are you in which makes you think this?
> I don't know about the Altium resellers, but Altium themselves have had
> severe financial problems for many years:
>
> http://www.google.com/finance?q=ASX%3AALU&hl=en
>
> They claim to have sold about 18,000 licenses. With 285 staff to pay I
> can't see how they manage to keep trading.
>
> Leon
> --
> Leon Heller
> G1HSM
>
