A Modest Proposal to Element14

Posted by: Dave Vandenbout 1 year, 11 months ago

(0 comments)

Hi, Element14. First off, let  me tell you how much I like Eagle. I bought version 4 back in 2001 for $1212. (Schematic editor, pcb layout and auto-router - woo hoo! Times were good.) Then in 2004, I upgraded to V4.1 for $315. (I don't know why I had to pay for a minor release, but there you go.) Finally, I went to V5 in 2009 for $400. I've never regretted my choice - over the years, Eagle has gotten better and better and my total outlay is still less than $2000. I can hardly get a schematic editor in one of the major CAD packages for that.

But I haven't made the move to V6. The price tag for me is around $600 and for that I get two new features: differential trace routing and design files stored as XML. Diff traces would be nice, and I guess XML files will be great once people create some cool tools for processing them. But to me, it's not a lot for $600. (Not that I begrudge you the right to profit from providing an excellent tool.) But more than the price, the main reason I'm not upgrading is because Eagle is isolating me from the community.

Over the past few years, I've transitioned XESS into an open source hardware (OSHW) company. (You can see my OSHW designs here and here.) Part of the reason for that (well, probably most of it) is I want to share my designs with others so they can modify and use them for their own projects. But your Eagle pricing is getting in the way of that. People can buy Eagle Lite for $69, but it's limited to a single schematic sheet and a 3" x 4", two-layer PCB (that's so limited that maybe you should call it Eagle Aerogel). Lite would choke on many of my designs. Going to Eagle Standard gets you up to 99 pages of schematics and six PCB layers, but the PCB size is still limited to 3" x 4". And that costs you $820 for the full package (although it does go down to $630 if you dump the autorouter). That's pocket change to an engineering company, but it's huge in the OSHW community. So while I can make my board designs open, it's really just an academic exercise if nobody else can afford the software to work with them.

Your change to XML files makes this problem even worse. If I release V5 files, at least there are a substantial number of people out there with the V5 software already. But if I move to V6, those XML files are not backward compatible with the V5 software so even fewer people can use my designs.

Recently, I've been seeing people choose CAD packages other than Eagle. (Diptrace and Pulsonix come to mind.) Maybe the pricing is a lot better for these tools, but the sharing issues still remain. It seems like the same set of monkeys, just in a different tree. I'm not really interested in that.

But I am interested in the open-source CAD packages. In particular, KiCAD is gaining a lot of users and the developers are making a lot of improvements. Sure, the user interface can be a bit clunky, but you guys aren't exactly in a position to throw stones. And it's not feature complete (I would really miss my Eagle ULPs), but neither were OpenOffice and Inkscape until, one day, they were and then I stopped using Word and Xara. Perhaps I'll be able to dump Eagle for KiCAD one of these days...

But that would be a shame. I really like Eagle and it would be a pain to move everything to a new tool. And you've provided a reasonably-priced package that's good enough for a large portion of PCB designs and built up a large user base as a result. But it's just not reasonably-priced enough for the OSHW community. If it were, then I think Eagle would benefit from a large influx of new OSHW people who would contribute design examples, part libraries, tutorials, new ULPs and other software tools (maybe even some for manipulating those XML files). In turn, the OSHW community would get access to a great tool and the talent and knowledge of the large population of existing users.

Of course, it all comes down to money and how you protect your existing market while serving the OSHW community. Here's a couple of ideas I had about that:

  • Increase the capabilities of Eagle Lite to something like ten schematic pages, four PCB layers, and 12" x 12" boards and charge $99 for it (call it Eagle Svelte). That would satisfy a lot (most?) of the OSHW people. You could probably even leave the auto-router out of it (nobody really uses it much, any way). You might cannibalize some of your Eagle Standard market, but make the revenue back on sales to the OSHW market.
  • Offer an Eagle Svelte license for free to anyone who creates an OSHW project. This shouldn't hurt your commercial market because few companies are going to let their engineers release open-source projects just to get a free CAD package that saves them $600. A bigger problem might be your administrative costs of validating the OSHW projects and doling out the licenses.
  • Offer a free Eagle Svelte license that can open any Eagle design but which embeds an OSHW license statement in any design files it saves. That would deter most companies from using it and they would go with the paid version for their commercial, proprietary designs. (Of course, they could remove the OSHW license statement from the XML files once they completed a design, but it would be restored whenever they made modifications. I'm sure their legal department would have something to say about the dangers of doing stuff like that.)

Those are my ideas, but I'm sure you can think of others. I would just hope you stay away from things like adding restrictions on the component families that can be used in OSHW designs or tieing the output design files to a specific PCB fabrication service. I've seen that done before and it just blunts the impact and makes the whole thing into a wasted effort.

So if anyone from Element14 has taken the time to read this, thank you!  I hope we can find a solution that preserves (and even grows) your revenue from Eagle while giving more of the OSHW community access to your software.

Currently unrated


Comments

  • There are currently no comments

New Comment

required
required (not published)
optional