blog nauseam Begs Adesk: No Acad for Mac, Please

Steve Johnson this morning makes very good arguments against porting AutoCAD to the Macintosh (at Why AutoCAD for Mac is a bad idea), but I disagree with him.

Mr Johnson argues that Windows has its roots infecting (my word) too much of AutoCAD; in any case, Autodesk's $250 million worth of personal and other cuts should prevent it from spending the manpower and dollars on coding a Mac version. 

He summarizes, "Reversing or working around that [Windows->Mac] process is a very substantial undertaking."

I disagree. Creating new CAD programs is trivial these days. Even one-man programmers do it, like the touchscreen-oriented MoI. 'Nother example: small CAD company IMSI/design produced the better-than-LT DoubleCAD in under two years (I think that was the timeframe). An example from Autodesk Labs: the new Project Dragonfly online "CAD" program.

I suspect that Autodesk would not port AutoCAD/Win to the Mac. They would start from scratch, linking code modules to create something fresh that happens to use AutoCAD's command structure/APIs and reads/saves in DWG format -- all wrapped in the OS X interface.

They would... they would be reading a Mac version of Deelip Menezes' book, "OpenCAD: A Step by Step Guide to Developing a Professional CAD Application." But instead of using ODA's APIs, Autodesk'd use their own. 

(On the other point: Autodesk is not just firing employees; Autodesk is also hiring in some areas.) 

The tough one is whether Acad/Mac would sell, and here I agree with Mr Johnson, who writes, "Any Mac user with any sense wouldn’t touch the first new Mac release with a bargepole." While today's Mac market is hundreds of times larger than when Autodesk made its original attempt 20 years ago, a brand-new Acad/Mac today must compete against established vendors -- VectorWorks, Graphisoft, Ashlar Vellum, and so on -- who spent the last 20 years improving their Mac-based software.

- - -

At time of writing this, the Macheads must still be asleep; 67% of the comments are begging for AutoCAD/Linux. 

Comments

[Brian Benton]

I disagree, to a certain degree, with the statement that "Creating new CAD programs is trivial these days." One (or a small group) programmer can make a CAD program relatively easy, but the programs you mentioned are limited in ability as to what AutoCAD, and other long serving CAD programs can do. In time they can potentially progress and have abilities that can fully compete. But it will take time and effort. Do you think it's a good idea for Autodesk to make a MAC version? I take it from your post that you don't because you share Mr. Johnson's perspective that it won't sell enough to justify the cost. Also, how much more shelf life does AutoCAD have? Is it worth spending the money on developing a MAC version only to shelve the program in 5 or 10 years?

[Royal Farros]

We appreciate the nod here... but, truth is, DoubleCAD XT and DoubleCAD XT PRO were not created in a few years... it took us a few years just to do the interface and usability work.

The issue isn't that you can't create the bones of a new CAD system from scratch in a short timeframe... it's being able to make that CAD app robust enough.

CAD users say they want simplicity and speed... but what they pay for is reliability and features, more features, and even more features.

Reliability and all those features take many years to do... more than just a few.

We were fortunate to have a great CAD code base to work from... TurboCAD (which, as you know, is on Version 16).

There's no way we could create a reliable, industrial-strength AutoCAD LT work-alike like DoubleCAD XT and DoubleCAD XT PRO -- in the time that we did -- without having the headstart we had.

So, I would substitute the word "trivial" with something like "really hard"... or maybe even more apt, "proceed at your own peril."

[JT]

Of course CAD for Linux must come first, as there aren't yet professional CAD programs there. And the first one who does it (seems BricsCAD) will have an interesting market advantage (a real one, as Linux users are more used to support programs than to crack them)... for the moment.

Linux offers quite an appeal for those users and can only grow for now, but in some years we will have open source CAD programs working (being developed right now), and porting AutoCAD won't matter anymore, as it happens now with Mac.

By then, the real battle will be around multiplattform BIM, I think.

[R.K. McSwain]

I couldn't care less if they make a MAC version as long as resources are not stolen from the regular WIN AutoCAD effort. The 12-month release cycle is too short as it is, we don't need to hurt that any more.

[Darren Young]

Not sure where this MAC hype is coming from all of the sudden. We heard people calling for a Linux port a while back but aside from JT's comments above, that seems to all but have died down, you rarely hear any commentary on it any more.

I'd take AutoCAD on a Linux platform over MAC anyday.

[Tony Tanzillo]

I've seen a few comments here that seem to be blind to Autodesk's effective-monopoly, and the variety of ways that monopoly market power is abused.

First, Autodesk doesn't compete with other vendors, or at least, not in the traditional sense of the word.

Competition based largely on the technical merits and overall quality of a product is irrelevant.

Ask any organization what their criteria is for choosing CADD software. Most will cite factors such as the cost of interoperabiity, availability of skilled users, and the extent of third-party supporting resources of all kinds, as the most prominent factors.

What does any of that have to do with the technical merits of a product itself, or how productive one can be with it, out of the box? Absolutely nothing!

So, what Autodesk sees in the Mac, is the same thing it sees in the PC, which is a way to offer technically-challenged if not downright inferior products it can have developed offshore on the cheap, that must only meet one and only one criteria, that being DWG interoperability.

Has anyone ever pondered what percentage of the annual USD $15.8 billion in estimated interoperability costs cited by the NIST report, Autodesk can directly take credit for?

If your guess is half, you're low.

So, how long Ashlar, and other Mac vendors have been hard at work building quality Mac-based products isn't terribly important in terms of Autodesk's core marketing strategy.

If most CADD users were to base purchasing decisions largely on the technical merits of the products they chose, relatively few of us would be using Autodesk products.

[Matt Stachoni]

In today's day and age, I find it dumbfounding to still read reports that say nonsensical off-the-cuff things like this:

"...Today's Mac market is hundreds of times larger than when Autodesk made its original attempt 20 years ago"

Really? Aside from the obvious (How does a percentage number grow hundreds of times larger when the largest number possible is 100?), this statement is not even remotely true. Apple's OS X market share is tiny compared to Windows, and small even compared to Linux/Unix.

While it's hard to pin down exact numbers (even if they were meaningful), in 1989 Apple's U.S market share was well under 8%. Their global market share was under 2%.

At the time Apple's market share did well in certain targeted industries, such as in desktop publishing, music production and image editing. They even did somewhat well in architecture thanks to ArchiCAD.

But the market had not yet dramatically shifted direction to the cheap PC in the '90s, as non-crappy versions of Windows came out, along with the subsequent Windows versions of Apple-only apps like Photoshop.

However, as of Q1 2009, Apple's U.S market share is (depending on who you talk to) hovering around 7.8%.

So, Apple may be selling a lot more Macs nowadays, but taken in total, their market share has barely moved if at all.

If Autodesk did develop a successful, complete, non-sucky Mac version, it would not be as much competing against other CAD developers as it would be competing against itself, as existing AutoCAD users move sideways to a different platform.

And while Autodesk would certainly charge high dollars for the privilege to do so, I doubt seriously it would result in any brand new licenses (and their required Subscription fees), which is where the serious money is made.