I, personally, have used a number of the low-end tools: Eagle, PCB123, and have tried out a few more. I've never used one of the big iron packges though, so I can't really comment on the differencs.
What I do know though, is that a significant difference in capabilities can still keep people from leaving a very expensive package, even for a free tool. However, if the free tool does what you need, there's not a lot of reason to look anywhere else.
I've also used a few low end / low cost tools (right now I'm using DipTrace) for my own designs. At work we are using the Mentor high end tools.
The high end tools do have significant additional features and capabilities. The big ones that I can see is the integration and control of part libraries and much more capable PCB routing tools. They also have a signal integrety tool that is very good.
These additional features require a large up-front learning curve to the point that we have people in the company who specilize in PCB layout and library part creation and mantainence. Since the learning curve is so steep, these tools don't make as much sense for a small company. The company I work for is large enough that it make sense to have a PCB layout group that knows how to use the tools.
What I think might make sense for the big ECAD companies is to have a low end tool with an easy migration path to their high end tools. Something with the same "look and feel" and file format...
The history of all of the mainline PCB tools is similar: PADS, Orcad, Altium all started as inexpensive, affordable tools. Then they were bought up by large companies who had legacy tools that were behind on development. Then the price went up by 10X and the originators made good money.
This is good for the tool vendor who gets rich, but it is not so good for the customers who gave him a lot of help in debugging and refining the product.
Is DesignSpark a different animal? Hard to say, it is owned by RS Components, not by the members who are contributing. What guarantee do we have that the tools will be available in the future?
I'd like to know more about the licensing model for DesignSpark before I commit to it.
This is an interesting point. The gap has closed in capabilities of the free tools if and only if you compare them to the state of schematic and PCB CAD software in the mid 1990's. DesignSpark is certainly the best of the "free" tools, but as you stated they're not really free.
This is all a bit like govt. bail-outs. In the end it's more a case that everybody pays, whether they know it or not.
There's still a strong future for powerful and professional commercially-backed design tools. One thing is certain: unless the component vendors actually make enough money from the sales generated from the "free tool" users, the whole thing is unlikely to continue. But what does this mean to the user? In the same way govt. bail-outs and money-printing activities tax the population through inflation, the cost of developing and supporting the software is amortized across the cost of the components being sold.
If you look at that as a cost of advertising, no-one's offended. But if you look at it as a price increase on a component which everybody has to pay regardless, it's totally unfair - it's a part-vendor's "software tax".
I propose a fairer solution to the part vendors, then (which I am certain they would not heed): charge a small additional percentage of the BOM cost for the software.
Another way to look at the problem is this: No part vendor is going to glibly invest in the size and ongoing cost of development team required to make a truly world-class design tool. By world class I mean capable of the design complexities and time-saving features that serious PCB designers need.
If your PCBs are simple enough to use free or low-cost tools: more power to ya, but it's unlikely that your product is really new and truly innovative or unique. What enables new and unique products/inventions, is new and unique tools to make them.
The web based tools will shake up the industry. A few $Mil from distributors will not keep the traditional CAD company going. That is very true.
It will be plenty for a 15-30 person web company running on Google or Amazon rented servers. There will be no license headaches and no dongles. The old timers will laugh at these new tools, and that is a good sign.
The big CAD companies should take note of Facebook. Any new disruptor that seems to get real traction is purchased. Purchased for a premium to close the deal quick. So, can't post now, I have to start my online CAD company. (Sound of desperately fast typing.)
Ben wrote: "No part vendor is going to glibly invest in the size and ongoing cost of development team required to make a truly world-class design tool."
Not sure I agree with that point. The real question is what a parts distributor has to gain by maintaining a develoment team? Now I'm not about to suggest that Altium Designer could be built from scratch for a small amount of money (or time), but to focus on your point, Altium's public figures state that its R&D costs last year were $10.5M. As a comparison, RS Components posted revenues last year of around $US1.7B. That makes the 'ongoing cost of the development team required to make a truly world-class design tool' just 0.6% of their revenue.
And more to my point, these big component suppliers continue to position themselves as adjunct marketing departments for device vendors, so the important question is how much additional marketing revenue could RS components generate through direct access to the design community through a tool such as Altium Designer?
Though Impulse is more on the design side of this tools question we find that extensive support from concierge vendors such as ourselves is valued when a tool falls into the critical path of a given project. In these cases, the cost of a traditional tool is rapidly "paid for" by lowering the wasted time on the actual project. In prior times, training (e.g. how to compile software to FPGA) would often precede tool use. Now it's reversed. Developers jump into a free or loaned tool and ask for help when stuck. The impetus is on us tool makers to offer meaningful free (even if time limited) tools that can be used with the assumption that no one will ever read the manual. It's just the changing nature of things.
I have a feeling that the future will bring about many changes to the PCB Design world in terms of cost of the tools we use to do desgin with and what there capabilities are goingforward. Bershire Hathaway is all hands on deck by aquring Mouser a few short years ago. Ok so 2007 is more than just a few short years but with Warren Buffet at the helm I can see the opportunity to team up Mouser with a Mentor type of tool to incorporate the ability to have a design tool directly linked to that database. The trend is for these higher end tools to have a direct link to the parts and databases of component suppliers through Digikey, Farnell, ect. It could be and interesting 5 yrs or so but by then I'll be one of those extinct board designers.
Since the Free ECAD tools are quite functional, the expensive commercial tools must justify their cost by providing exceptional value---either better technology (RF modeling, etc) or better business integration (linked to distributor catalogs, easy BOM/ordering). If you don't need those badly enough to justify the cost, Free ECAD is the right choice---especially in the current job market where skill portability is very important.
The problem with Free tools is that there's so many of them. I think that just like GCC won the compiler war (although it's being challenged by LLVM right now, we'll see how that will turn out), there will be a free ECAD package that will dominate by a virtuous feedback of more developer momentum, and resulting continuous improvement. Today it looks that KiCAD is winning. Note that KiCAD has a roadmap with quite a few interesting advanced technical capabilities. I would love to see KiCAD integration with distributor website catalogs, too, and it doesn't look very difficult to do, in principle, so I bet it'll appear sometime in the future.