blog nauseam » Thoughts

Cloud benefits – constant updates

Steve Johnson — Mon, 12 Dec 2011 02:54:18 +0000

One promoted benefit Software as a Service is that you are always up to date. There are no local applications to install and maintain. You don’t need to go through expensive and disruptive annual updates and/or install service packs or hotfixes; all this is taken care of for you. The latest and greatest software is always automatically available to you, and because everybody is always using the same version, there will be no compatibility issues. You won’t need to worry about your OS being compatible with the latest release, either. Bugs, if not exactly a thing of the past, will be quickly taken care of without you even being aware of them.

This is something you all want, right? What could possibly be wrong with this picture?

Cloud concerns – security again

Steve Johnson — Fri, 18 Nov 2011 04:02:26 +0000

It’s probably worth pointing out that if you you have no problem emailing your designs around the place without some form of protection or encryption, there’s little point in getting all worked up about Cloud security. Email isn’t remotely secure. FTP isn’t exactly watertight, either. If you’re still interested in Cloud security issues, this post includes some relevant links you might like to peruse.

First, here’s what Autodesk’s Scott Sheppard had to say about Project Photofly (now 123D Catch Beta) security last month: Project Photofly FAQ: What about the security of my data? This covers some of the same kind of stuff I’ve already discussed, but from an Autodesk point of view (albeit a pretty transparent and honest one, as you might expect from Scott). Here are some selected quotes:

In essence, we don’t want to accept liability when we don’t take money…

We intend to have a reasonably secure service, better than email, but less secure than a bank account.

We store your files on Amazon’s S3 service, and they maintain their own physical and data security policy that is considered robust.

Next, here are the 123D Terms of service, which raise many of the same alarm bells I mentioned before. Selected quotes:

We reserve the right to change all or any part of these Terms, or to change the Site, including by eliminating or discontinuing the Site (or any feature thereof) or any product, service, Content or other materials, and to charge and/or change any fees, prices, costs or charges on or for using the Site (or any feature thereof).

By uploading, posting, publishing, transmitting, displaying, distributing or otherwise making available Shared Content to us and/or any Users of or through the Site you automatically grant to us and our sub-licensees…the worldwide, perpetual, royalty-free, fully paid-up, irrevocable, non-exclusive, sublicensable (through multiple tiers) right and license to have access to, store, display, reproduce, use, disclose, transmit, view, reproduce, modify, adapt, translate, publish, broadcast, perform and display (whether publicly or otherwise), distribute, re-distribute and exploit your Shared Content (in whole or in part) for any reason and/or purpose (whether commercial or non-commercial) by any and all means in any and all media, forms, formats, platforms and technologies now known or hereafter devised, invented, developed or improved.

Please note that with respect to Non-public Content, we will not authorize your Non-public Content to be made available to others on a public section of the Site, although we cannot guarantee complete security (e.g., of cloud servers).

Moving on to another Cloud security-related issue, something that Owen Wengerd raised on Twitter was the idea that:

…once data is on the cloud, it can never be deleted.

Deelip Menezes thought this whole idea somewhat loopy:

Actually I’m implying that it is ridiculous to even start thinking along those lines.

However, I see Owen’s point. Once your data is on someone else’s server, you have no control over it. You have no idea where it lives, how often it is backed up, what happens to those backups, and so on. Let’s say you place some highly sensitive design data on the Cloud. It might be commercially sensitive, or about something that represents a possible terrorist target, or just something you don’t want certain parties to know about, ever. A week later, you delete the design data. Now, is it really gone? Any responsible Cloud infrastructure vendor must regularly take multiple backups and store them securely. So you now have multiple copies of your “deleted” data floating around, who knows where? What happens to old servers when they die? Where do backup hard drives, tapes, etc. go? If backups are stored off-site, how are your files going to be permanently removed from the media?

While there may be policies, procedures and ISO standards in place, we’re dealing with humans here. If one backup copy of your data ended up in a country where a rogue employee decided to better feed his family by selling off old hard drives, your nuclear power plant plans could end up not safely deleted at all, but instead delivered into the hands of some people you’d really prefer not to have it.

This may sound like paranoid nonsense, but risk from non-deleted data is real. There was a local case where a company was illegally siphoned of funds and went bust. The company’s old internal email servers were supposedly wiped and sold off. Somebody bought them, undeleted the data and was able to pass on incriminating emails to the police. While that ended up being a good thing in terms of natural justice and it’s not even a Cloud issue, it illustrates that making sure your stuff is properly deleted can be very important. This is related to something that Ralph Grabowski mentioned on Twitter; the “right to be forgotten”. Here is a Google search that includes various links that touch on some of the struggles related to this issue.

Finally, here’s something related to the possibility of the data being accessed illegally while it’s up. You put it up there, somebody copies it, you delete it, it’s not really gone and you are none the wiser. Is that something that only tin foil hat wearers need worry about? Have a read of this article before answering that one: Cloud Services Credentials Easily Stolen Via Google Code Search. Selected quotes:

The access codes and secret keys of thousands of public cloud services users can be easily found with a simple Google code search, a team of security researchers says.

Now the team is offering one word of advice to companies that are considering storing critical information on the public cloud: Don’t.

…an attacker who knows Google and some simple facts about cloud services authentication can easily find the access codes, passwords, and secret keys needed to unlock data stored in public cloud services environments such as Amazon’s EC3.

We found literally thousands of keys stored this way, any one of which could be used to take control of computers in the cloud, shut them down, or used to launch attacks on other computers on the same service.

Here’s a PDF of the presentation, if you’re interested.

Cloud concerns – trust

Steve Johnson — Tue, 15 Nov 2011 03:20:31 +0000

Using any software involves some degree of trust in the vendor. Using the Cloud requires a much higher level of trust.

Autodesk boss Carl Bass is a maker of carefully crafted things, so I’ll use that as an analogy. Using standalone software requires the sort of trust that a maker has in a tool manufacturer. Will the tools work properly and last a long time? Or will they break, potentially damaging the materials or even the user?

Using SaaS requires that same kind of trust, plus others. Will the tool manufacturer keep making that tool? If not, will spare parts continue to be available? Will the manufacturer change the tool design so it doesn’t suit your hand any more, or doesn’t work as well on the materials you use? Beyond that, there are some aspects of the relationship that stretch this analogy somewhat. For example, a SaaS vendor resembles a manufacturer that won’t allow you to buy tools, only lease them. Except the manufacturer can change the lease terms or end it any time it likes, and then come into your workshop and take all your tools away. Oh, and this take-your-tools-away right also applies to the company that delivers the tools to your door.

Using Cloud storage requires yet further levels of trust. It’s not tool manufacturer trust, it’s bank safety deposit trust. Will your carefully crafted creations be kept safe? Or will they be stolen or damaged? If they are, will you be compensated? If you can’t afford to pay the bank fees or want to use another banker because the teller was rude to you, will the bank politely return your valuables to your safe keeping or transfer them to the new bank? Or will they end up in the dumpster at the back of the bank?

Trust is vital. I’m convinced that a CAD on the Cloud takeover will live or die based on trust, more than any other factor. Potential Cloud customers must be able to trust that the vendor is going to do the right thing by them. Without trust, any vendor that expects to win its customers over to the Cloud has absolutely no hope. None. Forget it. Pack up and go home now, and save us all a lot of bother.

With that in mind, a few days ago I added a poll that asks Do you trust Autodesk to do the right thing by its customers?. I deliberately didn’t mention it, just to see what would happen. The initial results are interesting, with only 25% trust so far. If you haven’t already voted, I encourage you to do so.

I also encourage you to share your thoughts on the subject by commenting here. Although you’re welcome to comment as you see fit, it would be good to hear specific reasons you have for whatever level of trust you may have. Do you trust Autodesk? If so, exactly what has Autodesk done to deserve that trust? If not, just what has Autodesk done to deserve your distrust? I’m concentrating on Autodesk because that’s mostly what this blog’s about, but if you’re not an Autodesk customer, feel free to add your thoughts about any CAD vendor you like.

Poll of evil

Steve Johnson — Mon, 14 Nov 2011 05:53:18 +0000

I have closed the Which of these is most evil? poll, which had been running from 20 February 2009. It attracted 2,351 voters, each of whom could distribute up to three votes among thirteen (yes, that number was deliberate) candidates. Here are the ranked results:

Satan (36%, 846 Votes)
Microsoft (31%, 721 Votes)
Apple (26%, 614 Votes)
RIAA/IFPI/MPAA (26%, 601 Votes)
Miley Cyrus (23%, 546 Votes)
Autodesk (23%, 536 Votes)
Disney (16%, 382 Votes)
Google (10%, 230 Votes)
Dell (7%, 172 Votes)
The Pirate Bay (6%, 147 Votes)
Sony (6%, 140 Votes)
Steve Johnson (4%, 89 Votes)
Gaahl (3%, 82 Votes)

That top three is not going to shock anyone (except perhaps some fanbois), but are some surprises in the list. For example, more than a quarter of voters were aware enough of the evils of Big Content to be able to decipher the alphabet soup RIAA/IFPI/MPAA choice and select it. More than four times as many people think this litigious pack of demons is voteworthy than think the same about arch enemies The Pirate Bay. That’s not so shocking for those of us with our fingers on the pulse of popular opinion, but I was surprised to see so few people choose Big Content arch-villain Sony. Rootkit, anyone?

For Autodesk, this poll is something of a triumph, with less than a quarter of voters putting the company in the top three. Mind you, Autodesk was faced with some very stiff competition, being very narrowly edged out of fifth place by Miley Cyrus.

Only one in ten of you thought Google was worthy of selection. This is Google, a company that knows more about you than you do. Google, which passes out your information whenever it feels it might gain some strategic advantage from doing so, and really doesn’t care when it violates your privacy. Google, which insists on knowing my phone number before it lets me sign up for its Facebook-copy thing, because it obviously feels it doesn’t already have enough information about me. Google is apparently “do no evil” enough to attract far fewer votes than more sinister recipients such as, say, Disney.

Dell has been on my personal brown list for some years now, since repeatedly sending out fax spam to me and many other Australian businesses. It forced me to deal with its abysmal “customer service” [sic] Indian call centre in order to try to get it stopped. After making me wait for ridiculously long times while passing me round between various clueless, indecipherable people, a manager finally lied to me to get me off the phone. He assured me I would be taken off the list. The Dell fax spam continued until I finally gave up and threw the machine away; rather that than attempt to deal with Dell again.

Prior to this, I had no dealings with Dell and had just assumed it was a reasonably respectable company. It was only after this episode that I learned that Dell is utterly without ethics; my experience was perfectly normal. Indeed, victims of its shonkier practices (illegal bait-and-switch marketing, lying about stock and deliveries, repeatedly sending out “repaired” units that are totally non-functional, etc.) will probably think that I got off very lightly indeed. Dell has never seen a cent from me and never will. I’ve been very happy to pass on my feelings about the company to everyone who has ever asked for my hardware advice, as happens from time to time. 7% or not, Dell can go to Hell.

Finally, it’s official, I am more evil than Gaahl. Who? Gaahl is a Satanic death-grunt vocalist from black metal band Gorgoroth. He has performed in corpse paint on a stage decorated with sheep’s heads on spikes, and blood-splattered naked women hung up on crosses. Gaahl has been convicted of viscious violent assault multiple times, including one occasion where he was alleged to have threatened to drink his victim’s blood. I’m sure my metal friends will be very impressed by me being considered more evil than that. \m/

CAD on the Cloud according to Autodesk’s Jim Quanci

Steve Johnson — Thu, 10 Nov 2011 02:15:25 +0000

In all of the Cad on the Cloud discussion so far, both here and elsewhere, there have been a lot of anti-Cloud comments and very little in the way of response from the pro-Cloud crowd. Participation in the debate from Autodesk people has been minimal. In one way I can understand that, because given the current atmosphere, who would want to stick their head above the parapet? On the other hand, Autodesk wants to position itself as a Cloud leader and obviously needs to bring its customers with it. It is unlikely that many hearts and minds will be won over with press releases and other forms of corporate self-praise. Therefore, it makes sense for someone to get their hands dirty and engage with the plebs.

Step forward Jim Quanci, director of the Autodesk Developer Network. In the last edition of upFront.eZine, Jim was brave enough to enter the fray with a “letter to the editor” response to Ralph’s The Cloud is Dead position. I appreciate that Jim went to some length to compose his epistle and I thought he deserved more of a reply than Ralph’s one-liner. There will probably be a few letters in reply in next week’s upFront.eZine, but I’m placing my own response to Jim’s arguments here. It’s quite a lengthy tome and it would be unreasonable to expect Ralph to publish the whole thing unedited. I have quoted parts of Jim’s letter for the purposes of comment and criticism; for the full context you should read the original in upFront.eZine.

You’ve been at this CAD thing a long time; use more of that long term perspective you have. Think past that time you and I have retired. I think of my two kids in college. My younger son (in engineering school) believes having valuable data (like a mid-term paper) on his PC is an accident waiting to happen. Why would any sane person want to do that?

This is similar to several pro-Cloud arguments I have seen that I find unconvincing. Just because kids do a lot of stuff on the Cloud these days doesn’t mean much of anything. I have a couple of very smart kids myself. Like all kids, not everything they do makes sense, and I’m not about to start copying their behaviour. As people grow up, they start doing different, more mature things. That will, hopefully, include the appropriate use of technology. It may involve storing data locally, on the Cloud, or both. Yes, storing one copy of your work on your PC is indeed an accident waiting to happen. This is something I have learned through difficult experience. No, storing one copy of your work on the Cloud isn’t any more sane, particularly if your ISP is down when you really, really need to get at it to meet a deadline. This is something that Jim’s son will hopefully not need to experience in order to learn.

Jim then gives a potted and somewhat debatable history of CAD on the PC, with the implication that CAD on the Cloud in 2010 is just the same as CAD on the PC in 1982, with the implication that the same kind of takeover will inevitably happen. He concludes that part of his argument as follows:

The naysayers on the cloud could be the same naysayers we saw with the PC, just ‘find and replace’ a few words and the reasoning is identical (control, trust, capability, performance, productivity, etc). ‘Sure PC’s are good for word processing and spreadsheets -– but not CAD. They are just toys.’

Using terms like ‘naysayers’ for anti-Cloud people doesn’t add much to the debate, any more than calling Cloud supporters ‘mistyheads’. That aside, there are several ways in which this argument is flawed. First, as Ralph pointed out, ‘past performance is no guarantee of future returns’. Second, if you do wish to use history as a guide to the future, it is fair to say that the Cloud appears to be an aberration in the overall trend away from the bad old days of centralised computing towards putting control into the hands of individuals. Third, the ‘naysayers’ on the Cloud are generally not the same people who were ‘naysayers’ during the rise of the PC. On the contrary, they are typically those people who supported and actively participated in the PC revolution. They are those who have watched that history evolving and who have learned hard lessons from it along the way, instilling a stubborn resistance to giving up their hard-won control and freedom.

The cost-based naysayers. What makes them think the cost is going to go up?

I have to admit I literally LOL’d at this one. Jim, they think the cost is going to go up because they weren’t born yesterday and they’re not totally clueless. OK, hands up all those people who think that Autodesk and various others are investing hundreds of millions of dollars in the Cloud as part of a cunning plan to ultimately take less money from their customers? Anyone? Nobody? Hang on, there’s one at the back. Thanks, Jim, you can put your hand down now.

Then there are the ‘Unique to the Cloud’ benefits of increased productivity through mobility, collaboration and for all practical purposes unlimited computing power.

It cannot be denied that the Cloud has several unique benefits. Neither can it be denied that the Cloud has several unique drawbacks. The question comes down to where the balance of pros and cons falls for a particular application and a particular customer. The jury is very much out on that one. I have already discussed several of the Cloud benefits, but not many people appear to be that impressed by them. The drawbacks, however, appear to be dealbreakers for many.

Everyone needs to think past today, this year and even the next five years. As you and I know, being of the mature sort with children, five years is the blink of an eye and ten years goes so very fast. The impact of big changes are almost always over sold in the short term (applied to existing problems and processes) and under sold in the long term (it’s hard to envision what the new problems and processes will be in a world we haven’t yet experienced).

I can agree with most of this, particularly the part about it being oversold in the short term. The difficulty of predicting the future also rings true.

Sixteen years ago I bought my first copy of Netscape. The web was oh so slow through dial-up and though thoughtless people with graphics heavy web pages. Back then one could see the web as a marketing, sales and education tool. But no one was predicting Google and Facebook. What will the Cloud enable in a similar period of time?

No idea! But companies that waited till the PC and Web future was clear are themselves mostly in the dustbin of history. Ken Olsen died earlier this year, the PC having ‘done in’ his minicomputer. One might say Compaq was done in by the Web enabled Dell. What software companies will and will not survive the Cloud? One of the biggest software franchises in history, Microsoft Office, may be one of the first victims of a too slow migration to the Cloud.

This attempt to align the Cloud with the winners and the non-Cloud with the losers is specious. The winners and losers haven’t been decided yet, and there may not even be any. In any case, computing history is full of examples of pioneers who did the hard work for little or no reward and relative latecomers who cashed in on it. Also, I remember predictions of doom for Microsoft some years ago when the Internet was ramping up; Microsoft itself was worried by being run over by the Internet. It hasn’t happened yet, though.

Who out there would recommend their children invest most of their time becoming masters of the PC as a great career development investment? How about becoming masters of the web and the Cloud as a good forward looking career development investment?

I’m quite happy for my children to learn to become masters of the PC. Learning web development skills is likely to remain useful, too. Concentrating on one area to the total exclusion of the other is not a wise strategy, because nobody knows what’s going to happen in the computing world by the time they will need those skills. It’s quite likely that many of the skills my kids learn now will be near-useless to them by the time they need to use them, whether those skills relate to standalone or web-based software. Unless they’re learning AutoLISP, of course; that’s a gilt-edged investment. My 25-year-old skills in that arena are still feeding those same kids. Who would have predicted that when VBA was The New Black? And where’s VBA now?

Five years ago when folks like salesforce.com and NetSuite were breaking new ground offering CRM and ERP software as a service, one might have had some doubts. But not anymore. The train has left the station – and folks that missed getting on board better start running hard to catch up (or retire).

I could ask if the clue train stops at Autodesk Station, but that wouldn’t be adding much to the debate either, so I won’t. Instead, I’ll point out that CRM and ERP ain’t CAD. Your smartphone ain’t CAD. Facebook ain’t CAD. CAD on the Cloud is a whole different battle and it needs to be fought on its own merits. CAD is much more than text and a few small raster images, yet Autodesk has stumbled badly even when trying to provide that kind of simple SaaS, for example the poorly received online Help and the abysmal Lithium discussion group software.

Nobody has even proven that CAD on the Cloud can work properly yet. Real, full CAD on the Cloud, I mean. Not a few ultra-niche selected components with a handful of users kicking the tyres. Not a glorified viewer. Proper CAD. With 3D, instant response, full customisation, APIs, that sort of thing. On the Cloud. In bulk, for millions of simultaneous users. Online 3D games like World of Warcraft indicate that it might be possible, but it’s still not exactly CAD, is it? If and when it can be made to work and perform significantly better than standalone CAD, then you’ve got a chance to start selling it, despite various inherent disadvantages, to a bunch of grizzled CAD Managers who have been trained into cynicism by decades of hard knocks. Good luck with that.

How can anybody preach the absolute inevitability of something that might not actually happen at all? That’s not the basis for a rational discussion; it more closely resembles religious dogma. You and the rest of Autodesk management might be convinced, but that really doesn’t matter much at all. What matters is whether your customers are convinced. Have a look at the comments and polls here and elsewhere; do you think they are?

The best feature ever added to AutoCAD is…

Steve Johnson — Mon, 07 Nov 2011 00:24:14 +0000

…LISP. I have now closed the What are the best features ever added to AutoCAD? poll, and the winner is AutoLISP/Visual LISP, by a long, long way. I don’t always agree with the majority view expressed in the polls here, but in this case I wholeheartedly agree. Adding LISP was the biggest and best thing that ever happened to AutoCAD. Autodesk owes an enormous debt of gratitude to John Walker for incorporating the work of David Betz, who was of course standing on the shoulders of John McCarthy. It’s a crying shame that Autodesk has been so terribly neglectful of Visual LISP for over a decade.

Here are your top ten “best ever” AutoCAD features:

AutoLISP / Visual LISP (32%)
Paper / Model Space / Layouts (21%)
Xrefs (20%)
Copy / Paste between drawings (19%)
Dynamic Blocks (16%)
Object Snaps (15%)
Layer Visibility per Viewport (12%)
Undo (12%)
Grips (12%)
AutoSnap (9%)

Something interesting I noticed is the age of these features:

AutoLISP / Visual LISP – 1985 (significantly improved 1999)
Paper / Model Space / Layouts – 1990 (significantly improved 1999)
Xrefs – 1990
Copy / Paste between drawings – 1991
Dynamic Blocks – 2005
Object Snaps – 1984
Layer Visibility per Viewport – 1990 (improved 2008)
Undo – 1986
Grips – 1992
AutoSnap – 1992

The youngest feature here is 6 years old, the oldest is 27. The average top-ten AutoCAD feature is over 20 years old. What does that tell you?

Cloud discussions generating interest

Steve Johnson — Wed, 02 Nov 2011 13:39:19 +0000

This is one of those self-indulgent posts you probably hate, so feel free to skip it and just read the more interesting stuff.

Last month, my site statistics went through the roof. Here’s a graph that shows the number of unique visitors and the number of visits per month since I started the blog in February 2008. Page views, hits (a pretty useless statistic) and bandwidth all spiked in a similar fashion.

I remember being very surprised when over 1,500 people visited my blog in the first month, as I would have been very happy with a few hundred readers. I was astonished when more than 5,000 people visited here on the second month. Last month, there were 30,921 unique visitors who visited 58,342 times, viewing 129,206 pages. I’m sure there are other CAD blogs with many times the traffic, but for this blog, October’s numbers were crazy. The mentions on upFront.eZine didn’t hurt, but the daily statistics were already high and didn’t show a huge leap afterwards.

So what’s going on? Well, just posting anything rather than little or nothing (as has happened here from time to time) obviously helps a lot, but I think it’s more than that. I think it’s the Cloud generating interest. While it might be tempting for Cloud proponents to associate interest with excitement, that would be a mistake. Judging from the comments and poll responses here and elsewhere, I’m convinced that many more people are interested in CAD in the Cloud because they are concerned about it, they fear it, they even hate it. Given that atmosphere, I think CAD in the Cloud is going to be a very hard sell.

Fencing in Canberra – video

Steve Johnson — Mon, 31 Oct 2011 23:53:36 +0000

It’s about time I posted about something other than the Cloud, or even CAD.

Every year, there are four national-level fencing competitions in Australia. As they are almost all held on the other side of the continent, I don’t get to compete in them as often as I’d like. However, a couple of months ago I did have the opportunity to compete in the third of these competitions for 2011, held this year in Canberra.

This was very special to me because my mother and sister were in the audience and it was the first time either of them had ever seen me fence. It was also special because my sabre coach, Frank Kocsis, flew out to be with me and his other students. Frank has taken only two years to move me from complete sabre novicehood to being competitive at national level, particularly in the veteran (over-40) events.

This is not an entirely Cloud-free post, because this video of me fencing in the Veteran Men’s Sabre Semi-Final (2:49 long) is hosted on YouTube:

I’m on the right. If a green light goes on, I’ve hit him. If he hits me, it’s a red light. If both lights go on, we’ve both hit each other within 120 milliseconds and the referee awards the hit based on right-of-way rules. Veteran direct elimination bouts (like this semi-final) are fought until one fencer scores ten hits.

If you want to see how the winner of the semi-final did, here is the Final (4:46).

Cloud concerns – tie-in

Steve Johnson — Mon, 31 Oct 2011 04:44:18 +0000

One of the major attractions of the Cloud for vendors is that it ties in customers, providing a reasonably consistent revenue stream. It is an effective anti-competitive strategy. There are various technical and other methods that can be used to ensure that it’s difficult or even impossible for customers to jump ship. While that’s all very nice for vendors, it’s not such a wonderful thing for customers.

Let’s say you’re a CAD Manager who persuades your company to use a great new SaaS service and Cloud storage. Let’s assume it performs well, is secure, has 100% uptime and offers functionality that is not available with standalone software. Your company is pleased with all this and uses it increasingly over several years, eventually moving completely into the Cloud. A good news story, right?

Well, maybe. There are a few things that could go wrong. Very wrong. Wrong enough to get you fired. Most of these things have multiple precedents, some of them quite recent. They are realistic concerns and it’s not really plausible for anyone with any knowledge of the past to argue that they won’t happen in the future. I have grouped these concerns into five categories:

Impermanence. The vendor stops providing the service. There are many possible reasons for this happening. Computing is full of product failures and withdrawals. Autodesk alone has such a long history of dead products and orphaned customers, that it would be a major undertaking just to document them all. If the product’s not making money, it’s unlikely to have a future. The vendor itself could go down the tubes. Computing history is littered with the corpses of once-dominant companies. Because there is a chain of dependencies in a typical Cloud solution, there are several potential points of corporate failure. Maybe Autodesk doesn’t go down, but Amazon does, or Citrix. One day, your software just isn’t there any more. What now?
Price ramping. Once you and enough of your fellow customers are tied in, there is nothing to prevent the vendor from racking up the prices. Autodesk has already done this kind of thing with upgrade pricing and Subscription, so it’s not as if it’s an unlikely scenario. If the boil-a-frog-slowly approach is used and you’re the frog being boiled, you’re better off not being tied down when you decide it’s time to get out of the water.
Unwelcome terms and conditions. The terms and conditions under which you operate are often in the Cloud themselves and can be changed by the vendor without you having any say in the matter. What if one day your company lawyer spots a clause has been added that is totally outrageous (even more than normal, I mean) and there is no way your company can possibly continue to operate under those conditions? Good luck trying to negotiate your way out of that one from a position of weakness.
Unwelcome technical changes. I intend to cover the issue of as-you-go upgrades more fully later, but let’s say the vendor introduces a new feature that seriously impacts your ability to use the software productively. No off switch is provided. Sound familiar? It happens to standalone software. It will happen to your SaaS choice, too.
Ignoreware. Your once-fashionable product stops being The New Black. The vendor decides to concentrate its resources in other areas to attract new customers rather than the ones it has already tied up. While your SaaS product continues to be provided, it is put into maintenance mode and nothing useful is added to it. As the rest of the computing world moves on, your SaaS product does not. Holes start to appear that make your life difficult or impossible. Again, Autodesk history is replete with examples of this kind of thing.

If you’re using standalone software and any of the above occurs, it’s probably a pretty big deal, but you can work around it in the short term by simply continuing to use the product that works. You can keep doing this into the medium term, perhaps for several years. Sure, if Autodesk goes down you’ll find that you can no longer authorise new installations or transfer software from one computer to another. But you’ll have some breathing space. You’ll still have all your data, bang up to date. You’ll be able to continue working productively while you look around for an alternative.

If you’ve deeply committed your company to a Cloud solution and the SaaS hits the fan, it’s more than a big deal. It’s a disaster. It could kill your company. It could kill your career. As a computer once asked me, “Do you really want to do this?”

(so (long (and (thanks (for (all (the (parentheses))))))))

Steve Johnson — Thu, 27 Oct 2011 01:37:35 +0000

A few days ago, John McCarthy died at the age of 84. He didn’t make a fortune selling gadgets, he just profoundly affected the world of computing. He will be remembered mainly as the father of LISP, without which it is quite possible that AutoCAD and Autodesk would not have survived beyond the 80s. However, his original thinking went well beyond the development of a language. For example, 50 years ago he came up with an idea that is very relevant to what we are actively discussing today:

In 1961, he was the first to publicly suggest (in a speech given to celebrate MIT’s centennial) that computer time-sharing technology might lead to a future in which computing power and even specific applications could be sold through the utility business model (like water or electricity). This idea of a computer or information utility was very popular in the late 1960s, but faded by the mid-1990s. However, since 2000, the idea has resurfaced in new forms (see application service provider, grid computing, and cloud computing.)

(Credit: Wikipedia)

Back to LISP, I still use John’s antique language today. It’s still the best language choice for the vast majority of the development I do. Thanks, John.

“The Cloud is Dead” is not what I said

Steve Johnson — Wed, 26 Oct 2011 05:57:54 +0000

I guess most of this blog’s readers also read WorldCAD Access and upFront.eZine, so it probably hasn’t escaped your notice that in the latest upFront.eZine, Ralph had procalimed “The Cloud is Dead” and referred to some of my recent posts here as supporting evidence. I’d just like to point out that it’s Ralph announcing the death of the Cloud, and not me. Personally, while I agree with some of Ralph’s points, I think there’s life in the Cloud yet and the obituary is somewhat premature.

My own attitude toward the cloud matches that of most of you, judging by the poll results here. I see pros and cons, and have strong concerns about many of the cons. However, I intend to cover both sides objectively. Look out for more coverage soon.

Autodesk Cloud – don’t panic, business as usual

Steve Johnson — Fri, 30 Sep 2011 05:00:27 +0000

Autodesk recently made a big announcement about its Cloud initiatives, and reactions have been all over the place. Some people can barely contain their breathless excitement while others are outraged to the point of passing out the pitchforks. Why? It’s pretty much business as usual.

It’s nothing like Dassault’s disastrous we’re-moving-you-to-the-Cloud FUD campaign against its own product, SolidWorks. There’s no hint here of AutoCAD (real AutoCAD, I mean, not “AutoCAD” WS) being moved to the Cloud, or anything as radical as that. (Yes, I know there’s a limited experiment along those lines but that’s nothing to do with this announcement). It’s just a collection of relatively minor changes to Autodesk’s existing on-line services, collected together to make a newsworthy press release.

(As an aside, I must say this was a much more worthwhile announcement than the ridiculously over-hyped DE8.16N thing. So I was supposed to get excited about a routine upgrade of a product I have already been using for months, on an OS I don’t use, when the upgraded product is still half-baked just like the first underwhelming effort? Fortunately, I didn’t get sucked in by the pre-announcement build-up so I wasn’t disappointed, just amused when the truth was revealed. Autodesk PR, please don’t cry wolf so often; keep the hype in reserve for the hypeworthy stuff.)

Back to the Cloud thing, and putting aside hype and horror, here’s the stuff that has just happened:

Autodesk Cloud documents lets anybody store up to 1 GB documents on-line, or 3 GB if you’re a Subscription customer. This isn’t new, but until recently it was an Autodesk Labs project called Nitrous. The infrastructure is provided via Amazon and Citrix.
AutoCAD WS has been updated to integrate its storage with Autodesk Cloud documents. Remember, WS isn’t anything like real AutoCAD, but rather a limited on-line DWG editing tool. There’s a WS iPhone app, but that’s not new.
There’s an Autodesk Design Review iPhone app for reviewing DWF files you’ve stored in Autodesk Cloud. It won’t do DWG; use WS for that.
There are several cloud-based services that are available “free” to Subscription-paying users of a small subset of Autodesk software, mostly Revit and Inventor-based suites. They are:
- Inventor Optimization
- Cloud Rendering
- Green Building Studio
- Conceptual Energy Analysis
- Buzzsaw (now bundled with Vault Subscription)
AutoCAD users need not apply for any of these services.

So some of Autodesk’s on-line services are now being provided only to Subscription customers, and one is offered in improved form for Subscription customers. There are two obvious reasons for this: tie-in and revenue.

First, Autodesk wants its customers tied to the Subscription gravy train, if you’ll excuse a fairly awful mix of metaphors. Offering Subscription benefits like this is preferable to some of the much less pleasant arm-twisting that has been happening recently (e.g. trebling upgrade prices). Is it too much to hope that Autodesk has learned that offering carrots to its customers is a better strategy than threatening them with sticks?

Second, Autodesk needs to start making money out of this stuff somehow. For some years, it has spent several fortunes on buying and developing on-line services and then given them away for nothing, usually as Labs projects. This obviously can’t go on for ever, but just slapping a charge on these services wasn’t going to fly. Bundling Cloud services up with Subscription is a way of easing people into paying for them, and this is something I expect to be expanded in future, for example with AutoCAD WS. Once that’s been established for a few years, it wouldn’t surprise me to then see Subscription for at least some of the services split off, so you’re paying for Cloud services explicitly. By then, enough customers may consider them to be worth paying for and they may therefore survive beyond the short term.

Will it work? I’m not sure. Time will tell which of these services will thrive and which will die, and such uncertainty is one of the many reasons real-world customers aren’t excited about getting their heads in the Cloud. I don’t intend to make use of these services (I’m not even allowed to), so I’m not too bothered what happens to them. Like the vast majority of Autodesk customers, I will just carry on using conventional software in that old-fashioned 20th century way that just happens to work very well. Autodesk will go on providing its software in that way, because that’s what most customers will want for at least a while yet, and Autodesk can’t survive on wisps of Cloudy revenue.

Move along, people, nothing to see here.

Will Autodesk have to explain itself to the SEC?

Steve Johnson — Mon, 12 Sep 2011 05:06:02 +0000

The observant among you may have noticed that for many years, Autodesk’s free patches, service packs and updates haven’t added any new functionality. Bugs may get fixed, severe performance issues may be addressed, but design errors generally have to wait for the next release (at the earliest), and new features definitely don’t get added.

The last time new functionality was added to AutoCAD in a free maintenance release was Release 13′s c4 update which shipped on 12 February 1996. (There was a public beta available some months earlier; I picked up a copy at Autodesk University 1995). That free update contained not only a host of bug fixes, but also more useful new features than some later full-price upgrades (e.g. AutoCAD 2000i). In an outbreak of outstanding customer service, a c4 CD was shipped free to all registered users. Maybe Autodesk was trying to recover from disastrously shipping Release 13 prematurely, but issuing such a comprehensive update free of charge was still highly commendable.

Why did Autodesk stop providing new functionality in free updates? While it involves more work for Autodesk and hardly encourages paid upgrades or Subscription, the reason we’ve been given over the years is that there are accounting regulations that prevent Autodesk from providing new functionality in free updates. This does not apply to benefits from paid Subscription, and various new features for Subscription users have indeed appeared (albeit in fits and starts) over the intervening years.

I have to admit that I have always thought that this accounting thing was a pretty unlikely-sounding excuse for Autodesk’s inactivity. This attitude was reinforced by a lack of Autodesk response to my requests for further information about the alleged regulations. Until recently, I didn’t care enough about this matter to bother finding out for myself, but something extraordinary just happened that piqued my curiosity.

What happened? Autodesk released a free Service Pack that included new functionality for the first time in over 15 years. I was particularly interested in this, because part of what’s new is a new command providing a subset of the functionality of my ClassicArray™ plug-in. When I put in the time and effort to develop this product to fill a hole of Autodesk’s making, I did so on the assumption that Autodesk wasn’t going to provide an Array dialog box until at least AutoCAD 2013. It turns out that this assumption was wrong.

So what’s all this about accounting regulations preventing new functionality being provided free between releases? Was I right to be vaguely cynical about that? After some research, it would appear that I was wrong about that, too. There is an FASB (responsible to the SEC) accounting standard called SOP 97-2, which covers software revenue recognition. I’m no accountant and the regulations are large and complex, but here is my layman’s understanding of the basics.

In a simple case where a vendor (e.g. Autodesk) sells a complete product (e.g. AutoCAD 2012) at a given date, it records and declares the revenue for that product in the appropriate period as a single unit of accounting. If there are multiple elements of the product, things get more complex. If Autodesk ships part of AutoCAD 2012 (the main product) at one time and part (e.g. a Service Pack with new functionality) at another, then it is required to separate the elements into multiple units of accounting. It is required to make available vendor-specific objective evidence (VSOE) for each element of the product. If Autodesk has not done so (which seems likely), there is probably a problem. My understanding is that without VSOE, Autodesk is required to allocate the revenue for AutoCAD 2012 sales not at the point when it was received, but when all the elements have been delivered (i.e. when SP1 was released).

What about an argument that the new ARRAYCLASSIC command and new SNAPGRIDLEGACY system variable are not new functionality in themselves, but merely mechanisms to restore functionality that was available in previous releases? I don’t think that matters. The functionality is new to those customers who purchased AutoCAD 2012 and thereby provided Autodesk with revenue between March and September 2011. If that revenue has been allocated incorrectly, then Autodesk has some revenue shuffling and explaining to do.

I repeat that I’m not an accountant and this is all a layman’s uninformed opinion. It is quite possible that the regulations have recently changed, or that a relaxed interpretation is now permissible, or that I have the wrong end of the stick entirely. I’ve admitted being wrong in this post twice already and it could well be thrice.

If I’m wrong and Autodesk is in the clear, that’s great. Why? Because it means Autodesk customers can look forward to a lot more functionality being provided in future service packs.

AutoCAD for Mac – what’s missing?

Steve Johnson — Wed, 01 Sep 2010 10:06:57 +0000

According to Autodesk, the forthcoming OS X version of AutoCAD has “many of the powerful AutoCAD features and functionality.” So what doesn’t it have? What are the holes? Autodesk hasn’t bothered to let me know a single thing about this software, so I guess I’ll just indulge in some irresponsible and uninformed speculation, based on what I can glean from marketing materials and various better-informed sources. I could have just asked, but who knows if I would have ever got any real answers? Besides, this way is more fun.

First, here’s a quick list of some things that don’t appear to be missing, but which might have been lost in translation:

Command line (in fact, the Mac one appears to be better than the Windows one).
Xrefs.
3D, including visual styles and rendering.
Some kind of Quick View Layouts and/or Drawings feature.
Navcube.
Constraints.
Dynamic Input.
Selection highlighting.
LISP (at least some form of it).

Now for speculation on things that are possibly missing or not fully functional (based largely on screenshots, which is not a reliable indicator):

I don’t see a Communications Center, but I do see an Online Contact pull-down. Maybe that gives access to the same functionality, maybe it doesn’t.
Navbar.
Coordinate display in status bar.
Layout tabs (there’s a control instead)
The layer palette, in its docked form as shown in screenshots, looks very cut-down and would be tricky to use productively in complex drawings. It’s not clear if the old layer dialogue box is supported, but it needs to be.
Action Recorder? As this is a “brochure feature”, it’s no great loss if it’s absent.
Visual LISP? It’s not mentioned, so maybe it’s missing, or lacking the ActiveX bits. That would be a big problem for many organisations. Edit: I have since seen it confirmed that the Visual LISP environment is missing, as are the COM APIs.
Other APIs? DCL? ActiveARX? Deelip’s developing stuff using something, but the blurb just mentions a “flexible development platform” without giving any indication of what that means. Which leads to…
Add-ons, large and small. Many of us use various third-party utilities for making our AutoCAD lives more productive. Will they work? Probably not. Can the developer make them work? Maybe. But only if they want to, and feel the need to make the investments required. For small developers, that may not be the case.
Will your tablet, image and screen menus work in this environment? I don’t know, but here’s a guess: no.
Profiles? I couldn’t possibly work without being able to store and switch between profiles.
Object enablers. Has it got a full set? Or any at all? Dunno, but Autodesk’s object support for DWGs from its own AutoCAD verticals has been patchy, even on Windows.
Performance. Has it got any? Dunno.
Reliability. Has it got any? Dunno.
Longevity. Has it got any? Dunno. I think we can confidently expect an AutoCAD for Mac 2012, and probably 2013 too. Beyond that, it’s anyone’s guess. Maybe I should run a book on it?

Finally, things that do appear to missing:

Ribbon, QAT and The Big Red A. What, no Next Big Thing in UI Design, embraced wholeheartedly by Autodesk from AutoCAD 2009 onwards and still spreading out to the outer reaches of the product range? How will Mac users be able to live with the terrible loss of productivity when compared with their Windows-using colleagues?
CUI. There is a screenshot of a very cut-down interface customisation thing, but it’s not the CUI interface you’ve grown to love. Seriously, it looks extremely limited.
Express Tools. Last time Autodesk tried to sell an AutoCAD without these was 2000i. Remember that? Maybe not, because the i apparently stood for ignore and upgraders avoided it in droves.
AutoCAD verticals. Civil 3D for Mac, anyone? Mechanical? Architecture? Not yet. Verticals, if they ever arrive, are likely to be years away.
Network licensing. All of your Macs will need individial licenses.

It will be amusing to see how the various omissions are spun or glossed over. My guess is that they will be ignored altogether, or some vague indication being given to them being investigated for possible inclusion in a future release. But maybe you can think of more interesting ways of handling it. How about something like this for the missing Ribbon?

Mac, Windows. Chalk, Cheese. Ribbon, no Ribbon. Oil, Water. Creative, Productive. Cat, Dog. Trendy, Nerdy. Choose one. Be whatever you want to be. Because with AutoCAD®, it’s your choice™.

Have a go with your own Spin Segment. Who knows? Autodesk may even use your ideas.

Why we keep upgrading

Steve Johnson — Fri, 27 Aug 2010 04:42:00 +0000

In a comment in response to a Deelip post yesterday, Brad Holtz pointed to an article he wrote in 1999. It’s interesting to note that while much of the computing world today bears little resemblance to the scene at the end of the last century, this article remains almost completely accurate and relevant. Indeed, it’s so right that you might even be tempted to think, “Duh, isn’t that obvious?”

One section that stood out to me had this to say:

Many software systems never even get beyond the acceptable stage …. vendors of these systems are continually coming out with new versions, never stopping long enough to fix the problems with the existing systems.

It’s fascinating to me that this observation came at the very time that Autodesk was switching from a company that wasn’t exactly like that to one that very much was (and still is today), thanks to the 12-month release cycle.

Studying Autodesk’s productivity study

Steve Johnson — Wed, 04 Aug 2010 08:04:58 +0000

Heidi Hewett just reported the following on her blog, about a productivity study:

According to a recent independent study, AutoCAD® 2011 can help you work up to 44% faster with the latest productivity enhancements.

I have a couple of problems with that sentence. First, it’s not an independent study. It’s a study conducted by long-time respected CAD figure David Cohn, but it was specified and paid for by Autodesk:

This productivity study was performed at the request of Autodesk Inc., which funded this work.

That’s not exactly independent then, is it? Second, the study does not state that AutoCAD 2011 is responsible for a 44% improvement. That’s a figure that combines both the effects of AutoCAD 2011 (over AutoCAD 2008), plus the effects of using a newer, faster PC. Just stating that figure wthout such a disclaimer is misleading.

Now to the study itself. Let me make it clear that I have no problem with David Cohn, who is respected, experienced and honest. I do not doubt that his study accurately describes his observations of the time taken to perform the chosen operations on the chosen drawings. The problem is that the study is designed to concentrate purely on a set of AutoCAD operations that benefit from the changes of the last three releases. In other words, the dice are very heavily loaded. To David’s credit, he states that very clearly in the study report:

Each drawing was chosen based on a number of criteria designed to showcase one or more features of the software that did not exist in AutoCAD 2008 but were added in subsequent releases. While each drawing could certainly be produced using the features and functions available in AutoCAD 2008, the advanced capabilities added in subsequent releases would likely enable a typical user to produce the drawing faster using AutoCAD 2011.

Since the premise of the test was to determine how much time could be saved by using a new feature, the test itself was already predisposed to show that using AutoCAD 2011 is more productive than using AutoCAD 2008.

A quick skim-read shows that there are several other problems with the study. For example, it doesn’t attempt to measure the productivity of those operations that are common to both releases, which are much more likely to be used in bulk by typical users. The report states that the Ribbon interface is likely to be more productive, but makes no attempt to justify that by comparing the exact same operations performed using the two interfaces.

In addition, both AutoCAD 2008 and 2011 are measured on a typical middle-age PC using XP, but only 2011 is measured on a modern PC running Windows 7. The report states that the latter tests were performed after the former tests, so the times will also be biased by familiarity with AutoCAD 2011, the drawings and the operations required. That’s where the 44% figure comes from, and it doesn’t mean anything.

What’s the point of studies like this, that are self-evidently designed to produce a good-looking outcome? Who are they supposed to fool? Come on Autodesk, either do these things properly or don’t do them at all. Please.

Censorship on the Autodesk discussion groups

Steve Johnson — Mon, 28 Jun 2010 05:24:53 +0000

The Autodesk discussion groups have quite a few problems at the moment, which I will discuss at length in future. One unnecessary problem that has been added to the mix is censorship. Having praised Autodesk in the past for allowing discussion to go unhindered, it’s only fair to slam heavy-handed moderation when I see it.

Before I get started, let me just say that Autodesk is entitled to moderate its discussion groups as it sees fit. The forum belongs to Autodesk and it can do what it likes with it. But just because Autodesk can censor its forums, that doesn’t mean it’s always a good idea to do so. Neither does that it mean that Autodesk is immune to public criticism of that censorship. There is no First Amendment obligation on Autodesk, but there are many other places that censored viewpoints can be repeated. Here, for example.

In this particular case, a section was deleted from a reply I made in a thread about the educational plot stamp. In that section, I mentioned that the educational plot stamp is very easy to remove with an everyday AutoCAD command. I didn’t name that command or give any details of how to use it to remove the stamp.

Now I understand that Autodesk gets the twitches when people discuss circumvention of its educational stamp “virus”, but I didn’t mention anything that isn’t already public knowledge. I discussed this issue at length in Cadalyst some five years ago, again without giving away the details. If you really want to know the details, please don’t ask me because I won’t reply. Google it, it’s out there. You probably don’t even need to do that. It’s a pretty obvious thing to attempt. It was, in fact, the very first thing I tried when I first saw an example of an infected file. It worked perfectly.

Back to the censorship. My post was edited, and I wasn’t happy. I wasn’t contacted about it, so it was not possible to have a reasoned discussion about it with the moderator (as I have done in the past on the AUGI forums and elsewhere). Annoyed, I made a further post, this one objecting to the censorship. In that post, among other things, I pointed out that the Autodesk position on the plot stamp was fictional. Here is what the Autodesk knowledge base item TS63668 (which I can no longer find) had to say on the subject:

Issue
When you plot a drawing that was created in or that contains drawing data that was created in the Educational (Student and Faculty) version of AutoCAD® or AutoCAD-based software, the following plot stamp or watermark appears in the plot:

For Educational Use Only

Solution

There is no way to circumvent the plot stamp. This is as designed to discourage the commercial use of an educational version of an AutoCAD product. Autodesk sells educational versions of software on the premise that the software will be used for educational purposes only.

The statement above in italics is a blatant lie. Hopefully, the knowledge base item is now missing because somebody sensible at Autodesk decided that it’s not a good look to have such fraudulent nonsense on its site, dishonestly masquerading as technical support. Or maybe it’s not missing but I can’t find it because the search engine is bad. After all, Autodesk really, really sucks at search. Perhaps it should buy a search engine company?

I digress; back to the censorship issue again. My post objecting to the first censorship was deleted. I was not contacted to discuss this deletion. I made another post objecting to the second censorship of my objection to the first censorship. This post made no reference whatsoever to the plot stamp issue itself. This post was deleted, too. In a surprise development, I was not contacted to discuss this deletion. Three levels of censorship to cover up an Autodesk lie. I can’t see a problem with that, can you? Except for this:

The Internet interprets censorship as damage and routes around it. — John Gilmore

Discussion_Admin, you were entirely within your rights to perform this censorship. Your moderation guidelines may even require it. But as a result, my statement about the plot stamp being easily removed has been read by a much larger number of people. So it really wasn’t such a good idea to censor it, was it?

Readers, if you have your own Autodesk censorship tales to tell, feel free to tell them here. It should be a fun read.

Does Autodesk discuss future plans?

Steve Johnson — Wed, 26 May 2010 05:57:44 +0000

According to Shaan, Autodesk does not discuss its future plans. Or does it? In a comment, Ralph reckoned it does. Putting aside technology previews and various NDA-bound circumstances (e.g. Beta testing), can you think of cases where Autodesk has revealed what it intends to do in future? Here are a few off the top of my head:

I’ve been to AU sessions dating back to 1995 that pretty much give away the contents of the next release of AutoCAD, using a vague cover-my-butt session title and a disclaimer at the start of the session. I understand that these days, attendees need to sign an NDA before entering such sessions.
Last year in San Francisco, an international blogger audience was given all sorts of information about Autodesk’s future directions (preceded by a similar disclaimer), with no NDA and nothing off the record. I assume something similar happened at this year’s North American bloggers’ event.
The Subscription (Advantage) Packs of the last couple of years have been a dead giveaway about some of the features that are going to make their way into the next release.
The new 50%-cost upgrade policy was announced a year in advance.
Surveys and other customer feedback mechanisms provide a very big clue about what Autodesk is looking at changing next. Some of these are covered by NDA, others are not.
In the specific case that triggered this discussion, Autodesk has been gradually building up expectation of a Mac AutoCAD for quite a while. Yes, it required a little reading Between the Lines, but for some time it has been pretty obvious where all the Mac love was leading to.

Feel free to add your own examples, but it seems to me that Autodesk is perfectly happy to reveal future plans as and when it sees fit. And that’s fine. In each of the above cases, the revelations have been A Good Thing. Good for Autodesk, good for customers.

Maybe the question should be, “Why doesn’t Autodesk discuss future plans much more often?” Stock market regulations, perhaps? But hang on, there are some very major publicly traded corporations that seem to get away with revealing all sorts of things about their future products. For example, Microsoft regularly conducts very widespread public Beta programs that let people know in great detail what’s very likely to appear in the next release, and seems to have survived the experience so far. There’s surely no reason why Autodesk couldn’t do the same if it wanted to.

Ultimately, it comes down to a desire for secrecy; a culture of concealment and control. Of course Autodesk may have legitimate reasons for keeping some of its plans from its competitors, but the culture can be so pervasive as to cause some bizarre side effects. You may find this difficult to comprehend, but there are those in Autodesk who got into a tizzy about me speculating in my launch announcement that Autodesk’s general design product (AutoCAD 2010) was going to be followed by something called AutoCAD 2011. There was something of a surreal drama behind the scenes. There were apparently people within Autodesk who genuinely thought I needed privileged information to work out that 10+1=11. No, I’m not making this up.

I’m not sure Autodesk’s secrecy is doing any good for anyone. It’s surely harder to maintain these days and it’s only going to get harder. I suspect several Autodesk blood vessels were burst when AutoCAD Mac Beta 1 was leaked. On the one hand, I can understand that; somebody broke an NDA, a legitimate agreement was freely entered into and then broken. Some people at Autodesk probably had their carefully planned marketing timelines disrupted.

On the other hand, this provided a whole heap of free and largely positive publicity. Potential AutoCAD for Mac users are now hovering in anticipation, filling the Mac forums, spreading the good news among themselves, putting off the purchase of competitive products, considering entering the official Beta program, and so on. At the same time, the news of performance issues in the early Beta is helping to put a dampener on expectations in that area. Lowered customer expectations may turn out to be very useful when the product is actually released. All considered, a good thing for Autodesk, then.

I’m convinced that Autodesk is opening up. That’s great, but there’s a way to go yet.

The Machine that Won the War

Steve Johnson — Wed, 19 May 2010 00:04:23 +0000

I just wanted to get on the record that I don’t trust claims based on statistical data without being able to review in detail the methods used to obtain and interpret the data. Even with the best intentions, full integrity and honesty, it is not difficult to come to completely the wrong conclusions based on apparently compelling statistical evidence.

This isn’t just theory, I’ve seen it happen. Detailed percentages presented at upper governmental levels, based on huge sample sets, giving a totally false impression because of errors and assumptions that occur at various places in the process. The exact same question asked twice in the same survey, giving very different results depending on the section in which the question appeared, providing an unstated context to the question. The devil is in the details, and the details can be extremely subtle.

I have a “put up or shut up” rule that applies to anybody who makes claims based on unrevealed statistical evidence. It applies to corporations, news outlets, bloggers, government ministers, everybody. Without allowing scrutiny of the full details, all statistical claims are null and void, as far as I’m concerned. “Trust me” doesn’t cut it. Sorry, no exceptions.

What does this have to do with the title? Those familiar with Isaac Asimov’s short story of that name will understand. I’m sure Robin Capper worked it out immediately.

Autodesk’s CIP data – massively biased?

Steve Johnson — Fri, 14 May 2010 08:47:36 +0000

You may have seen Shaan Hurley and I having a discussion (ahem) over the validity of his statement:

I really do use the ribbon now with AutoCAD 2010 along with most users as evidenced by the CIP data we receive daily from thousands of AutoCAD users who choose to send the great data.

So, now you know. Most of you use the Ribbon now, Shaan said so. Shaan, as he always has done in the past, declined my invitation to back up this assertion with more details. He has vast amounts of data collected from huge numbers of users. How could that possibly be wrong?

Here’s how. CIP data is biased.

How can millions of data points be biased? Actually, all samples are biased. Only the degree of bias varies. The polls on this blog are no exception. I do my best to keep the questions and options neutral; the only leading questions you’ll see here in serious polls are the ones I copy and paste from Autodesk blogs. But readers of this blog are one self-selecting small portion of Autodesk customers, and people who vote in my polls represent another self-selecting portion of that portion.

The question is, how biased is Autodesk’s CIP data? Without access to Autodesk’s data (which it won’t provide) and resources for alternative data collection from its customers (ditto), the best I can do is use my own biased sample (that’s you lot out there) as a cross-check.

Let’s examine it in light of Ribbon use among AutoCAD 2010 users. In an earlier comparison of my 2009 poll figures and Shaan’s CIP data, I wrote this:

But Shaan’s CIP users are also a biased sample, comprising those AutoCAD users who have CIP turned on. Are users who go with the flow and have CIP on also more likely to go with the flow and leave the Ribbon on? Possibly, but I would have thought the CIP-on bias would be less significant than the blog-reader bias.

I have recently run a poll to try to determine if that “possibly”, that hunch, has any basis. Let’s examine the results I got.

AutoCAD 2010 users, what are your Ribbon and CIP settings?

Ribbon on, CIP on (24.7%, 65 Votes)
Ribbon on, CIP off (19.4%, 51 Votes)
Ribbon off, CIP on (11%, 29 Votes)
Ribbon off, CIP off (44.9%, 118 Votes)

Total Voters: 263

For the sake of argument, let’s make the assumption that my poll sample is unbiased. It’s not, and the degree of bias is unknown, but let’s see what it would mean if it was. Let’s see what kind of results Autodesk would see from its CIP sample:

CIP-on voters (94):
Ribbon on 69% (65)
Ribbon off 31% (29)

Shaan would see from this result that 69% of AutoCAD 2010 users have the Ribbon on, and would be tempted to say stuff like “use the ribbon now with AutoCAD 2010 along with most users”. Understandable. That’s just CIP users, but non-CIP users can’t be that different, surely? Or can they?

CIP-off voters (169):
Ribbon on 30% (51)
Ribbon off 70% (118)

Wow. That’s a huge discrepancy, and it implies that a sample selection based on CIP use introduces a massive bias. I’ve watched this poll grow over the weeks, half-expecting things to even out as the sample size increased. It didn’t. It has been pretty constant, with non-Ribbon non-CIP users outnumbering Ribbonite non-CIP users by a substantial margin.

Let’s put the groups together, shall we?

All voters (263)
Ribbon on 44% (116)
Ribbon off 56% (147)

So, if the voters in my poll were observed by Autodesk via CIP they would appear to be 69% Ribbon users. In fact, only 44% of these voters are Ribbon users.

How many AutoCAD 2010 users really have the Ribbon on? 69%? 30%? 44%? Some other number? I don’t know, and that’s not the point. The point is, Autodesk doesn’t know either. It can take some smart guesses, but just assuming CIP is accurate isn’t smart, it’s just a guess.

Why does this matter? Because Autodesk makes decisions based on this stuff. Decisions that affect you and me and how we use our tools. Have a look at this statement from Autodesk’s Teresa Anania, Director of Industry Management (taken from her interview with Deelip Menezes about Inventor):

…we had data that suggested that the new ribbon UI was well received and would be absolutely all that customers needed …. And now since we have the CIP data that shows us how our customers are using the software, we can analyze this before we permanently turn anything off.

Comments like this (and others from other Adeskers) seem to indicate that there is an unspoken assumption that CIP users accurately represent a true cross-section of users in general.

I know that Autodesk doesn’t rely solely on CIP; it uses a wide range of research tools to find out what users are up to and what they need. I regularly encourage you to participate in various Autodesk surveys, for example. But there are problems of accuracy inherent in all those methods. It would be natural, when faced with a set of apparently conflicting results from different sources, for Autodesk decision-makers to simply assume that the source with the biggest sample size is the most accurate. That could be a dangerous mistake, for both Autodesk and its customers.

Note: my arithmetic was off in several places when I posted this, and I have edited the post to correct some of the figures. These corrections do not invalidate the arguments; the substantial bias is still evident.