Altium Limited closing up shop - Altium Designer discontinued

I read that they were moving their main offices to China, but isn't that a big step from saying they are "closing up shop" and that Altium Designer will be discontinued? They are a very big player in the EDA tools market, and have just brought out a significantly big new version of their main tools. I can't see that mesh well with news that they are just about to collapse.

But maybe you have more information, or have read something I haven't seen?

Indeed, methinks there's some unnecessary scaremongering happening. You're a bit late for April 1st !

t

een?

They've consistently lost money for many years:

They are obviously trying to reduce their overheads.

Or maybe more.. their China sales office was converted into a regional business a little over a year ago (after they appointed Roger Shen) and they set up a preliminary development team some time before last September, so they've been moving down that road a while.

One of their press releases makes mention of perpetual licenses not being viable long-term. If they move away from that, we shall move away from them, without question. Subscriptions are stupid from a business point of view, and clouds are for the birds when that much is invested.

Best regards, Spehro Pefhany

--
"it's the network..."                          "The Journey is the reward"
speff@interlog.com             Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog  Info for designers:  http://www.speff.com

Their shares, worth next to nothing anyway, lost 20% following the announcement.

Since all the bad news is on the street it should be priced in. You know what that means... this could be a buying opportunity!

Rick

(sigh) I've oven pondered what the EDA folks could do to "fix" their business model. Or, if it is a field in which there *is* no realistic fix.

[e.g., if a potter throws a "vase", he'll have a hard time selling it -- especially if it is particularly "nice". OTOH, if he adds an *ear* to it -- magically transforming it into a *mug* -- it is much more salable. However, as a *mug*, most folks won't pay much for it. I.e., stay away from vases -- and don't waste much time on a "mug" since it won't command much in the market]

Consider: they want/need a steady income stream. So, either need to reach new users at a regular rate, derive income from "support", add new features to make folks *want* to upgrade *or* "re-sell" to their existing customer base.

I suspect the size of that market isn't very big -- especially when you consider all the folks nipping at the low hanging fruit. So, I don't imagine they see much growth (number of seats) there -- at least nowhere near enough to carry their needs!

"Support" can only *practically* be "hand-holding". I.e., effort spent fixing bugs is counterproductive, here -- who wants a buggy EDA tool? Charging folks for bug-fixes isn't a viable income option.

Adding features usually means increased development costs and more possibilities to introduce bugs into a product. People using EDA tools (at least the ones with deep enough pockets) are not the sort who want to "play with MythTV" (etc.) and tolerant of "not ready for prime time". Besides, how many features can you add that are truly *worth* the cost/risk of an upgrade?

And, all but the very wealthiest users tend to grumble about "subscriptions" -- it feels like "renting" software! (if you are content witht he current incarnation of the product, a subscription leaves you vulnerable to vendor going belly up... something that "worked good enough" for you now "works not at all")

And we know tying the tool to proprietary hardware is a losing proposition (Mentor, CV, etc.)

Maybe better off making mugs?

A big problem for Altium is their sales model. For a recent project the layout guy was using Altium and other projects for the client were done in Altium, so I needed to do all the schematic entry in Altium. Not a big issue for me since the layout guy was doing all the symbols and foot prints.

I'm not about to jump into the deep end with Altium, so I needed a simple 1 year license. $650 bucks for schematic only. I had to jump through hoops to get it. No simple order on the web, download and be done. I had to deal with the "sales rep". Several emails and phone calls with them trying to sell me the full package. Finally got it through to them all I was going to buy was schematic for 1 year.

Now they want a purchase order from me. Then they will send me a invoice and I can pay with check and then 30 days later I'll get the product. No I'll just pay by credit card to speed things up. They send me a email with CC auth info. I'm supposed to fax it back. Sorry - no fax here. So I have to fill it out, scan it and email it to them. Their suggestion was to snail mail it to them.

Finally after 3 weeks of this I get a package with the Altium DVD (I already downloaded it) and *another* JTAG pod - like I need another one on the shelf.

All of this for a $650 order - no wonder they are loosing money. Companies with sales models locked in the '80s need to go away.

--
Joe Chisolm
Marble Falls, Tx.

[...]

That's economically short-sighted to the point of being silly. What would you prefer: paying for the tool in proportion to how long you actually use it, or have the tool vendor charge you the _entire_ cost of a lifetime's support up-front, because a "strictly no subscriptions" pricing scheme cannot work any other way?

Because, face it: you *are* paying for support and maintenance, and yes, that means a large portion of the overall asking price will be proportional to how much time you're going spend with the tool. The only question is whether you pay in proportion, or get fined according to a worst-case scenario.

That's a MIGHTY big if to bet your company's entire legacy of work products on for the foreseeable future.

That's no different from the tool company going belly-up even faster because once it sold a certain number of licenses, its business dries up for good. No cash flow --> no company. It's as brutally simple as that. So as soon as your last machine running that tool you used to be content with breaks down (taking your last working installation of, say, NT3 with it), you're toast.

In the end of it all, _somebody_ has to put food on the table of the people who make the tools, and do so for as long as the world at large, including you, wants those tools. The only questions are who pays how much, and when --- and what happens once there is no longer enough wide-spread interest to keep the original tool maker financially afloat.

Yes, some kinds of tools really had better be open-source, for exactly the reasons above.

A subscription for support/updates and a subscription to even allow you to continue running the software are two very different things. The former means that, if you need additional functionality, you may need to migrate to a new solution. The latter means that you may suddenly lose the ability to work with your old, existing designs because the tool vendor's gone to the code repo in the sky and now you can't get past FlexLM anymore.

We've got a decade-plus lifecycle on products, and have certainly seen plenty of tools come and go in that time. And I even use a few tools here, such as Aldec Active-HDL, where I have to pay an annual license in order to keep it running. But that's just my simulation package; nothing in it directly impacts code shipping on products. If it's code that's shipping, I absolutely HAVE TO know that I can get the toolchain that built it back up and running, regardless of whose license server may or may not be a distant relic of the past. And that requirement has massively shaped which tools I will and won't use.

--
Rob Gaddi, Highland Technology
Email address is currently out of order

You are focusing tool much on the tools here - the big issue is /data/. For the tool itself, it can well make sense to pay only for what you use, when you use it. But you have to have access to your data. When the only way to access your data is using a particular tool from a particular company, then that company controls your data. If you can buy a perpetual license as an insurance here, then it is probably worth it.

So while I am all in support of open source tools - for some sorts of tools, they are definitely the best choice - I see open standards for data formats as much more important. And by "open standards", I don't mean the kind of "open as long as you pay us lots of money every year for a copy of a badly written spec" standards that are often seen in embedded development. I mean a standard format for EDA files that is specified and documented, available to everyone, and implemented by any serious tool vendor (and open source authors, of course). That would make the tools /much/ more valuable, and let vendors compete on features, support, etc., rather than vendor lock-in on people's data.

It doesn't matter what *I* think but, rather, what the *market* thinks! If you look back on EDA tools over the last 30 years (anyone want a copy of DASH-PCB? STRIDES? etc. -- complete with genuine FutureNet mouse and security PALs...), this is how the market has dealt with these vendors.

Huh? If I have and expect I can continue to live with Version X's limitations, feature set, bugs, etc. -- i.e., no *need* to upgrade to Version Y (SINCE Y MAY NEVER COME!), then what am I risking? I know the product worked "so far". I can recreate what I have for all of the projects completed *with* it. Chances are, I can probably use "it" to make whatever small changes a design might need, going forward (sure, there are cases where this could push the design over the hairy edge... but, that risk also exists with Version Y, Version Z, etc. -- especially if those versions never materialize

*or* prove to be incompatible with Version X, introduce new bugs that your design "tickles", etc.)

Exactly -- as was intended to be my point, above. I am very slow to update my tools. First, it costs a fair bit of coin to update everything I *have* used each time an update is available (especially since I may not *need* that tool for one, two or three releases!).

Second, each update incurs a fair bit of non-monetary cost: time to install it (never as straightforward as they want you to believe), time learning new features, UI changes, etc. Remember, you have to recover this "spent time" (and money) due to "increased productivity" in that tool (if it costs you a day or two, are you going to recover that on a 4 week project?)

Third, each update runs a very real risk of *breaking* something that you had previously used "successfully" -- now you end up either abandoning any projects that were created with the "earlier" release

*or* end up having to actively maintain *two* releases (ever try to get a vendor to support release N-1 or N-2? "Oh, you should upgrade to version N!" "I *have* upgraded to Version N and it is incompatible with my Version N-2 designs. I am just asking you to fix this problem *in* Version N-2... without dragging in all the Version N cruft that BREAKS my project!")

Previously, I would image my system (and preserve the image) at the end of each project just so I could recreate the development environment in effect *for* that project. This has now been complicated by having to preserve the actual *hardware* on which the image ran ("virtualization" wasn't as feasible in the past as it is now).

Exactly. Or, you end up running a separate machine with the clock set at some bogus date *hoping* the license manager hasn't "seen" any timestamps "too near" to license expiration. The tool is just making itself hard to use. (imagine if you were *forced* to use a hammer with your non-dominant hand... for no *real* reason other than "you've run out of your allotted N dominant hand days")

I have products at twice that. I wonder how the hell these people even find *parts* to maintain them!! :<

If you have any sort of certification process involved with your products, chances are the lifecycle tends to be considerably longer than the "90 day wonders" you find with consumer kit.

So, if you discover a problem in your implementation and want to fire the tool up "next year" to fix it, you're SOL when the firm goes belly up (or, see below, *refuses* to sell you the license and, instead, insists that you "upgrade" to the current version... "for only $99...")

A firm I worked at deployed several (6 figure) products relying on third party licenses -- that suddenly were not available AT ANY PRICE. Take your development dollars and throw them out the window, your product is now at EOL!

Yeah, but "Pay us more money or you'll never see your data again, hnar hnar, hnar." is certainly an _easier_ model to compete on. Even if it does set the customer thinking of Snidley Whiplash

--
Rob Gaddi, Highland Technology
Email address is currently out of order

Just to set the record straight on the (mis)title of this discussion, Altium are not discontinuing Altium Designer in any way shape or form. Also, we are not shutting up shop in Australia. We are moving our corporate HQ to Shanghai, China, where we already operate a sizeable office with around 60 staff. A number of development and management staff will be relocating to China. We will retain a substantial presence in Australia (we will remain an Australian company), with a number of people, including the key Aus/NZ support and sales teams, manning an office in Sydney. Our Hobart, Tasmania operation remains unchanged.

For the record, I am product manager with Altium Limited, currently based in the Sydney office.

[attributions elided]

I don't think "open source" is *the* natural consequence of this. As I stated in my original post, I was pondering the flaws in the business model that appear to make this "de rigueur" for that industry.

E.g., another solution is to change the *scope* of the product to eliminate those labor intensive aspects that drive the cost of doing business up ("symbols", "footprints", "3D models", etc. no doubt represent a significant and ongoing cost).

Yes.

Note that you can still have "access to your data" -- though not in a form that is worth a damn (e.g., my post re: legacy file formats). If, for example, I need to add a foil jumper to an existing layout, I am very happy to play with a Gerber editor and manually tweak the photoplots (sure beats cutting rubylith! :> ). But, most EDA tools leave you stuck at this point. Doing *anything* else requires access to a FUNCTIONING tool.

E.g., it is *really* annoying to have to trace foils to figure out what this particular "signal" happens to be that is proximate to an area of the board that you want to modify. *Much* easier to click on a trace and have the signal name appear "magically".

*But*, worth a one year subscription just in case you *might* want to do that??

Exactly. Depending on the complexity of the "interface", the quality of that spec could suffer and still prove sufficient for folks to derive useful information from it. E.g., I save all my FrameMaker documents in "portable" form so there is a *chance* of processing them with something other than FM (I learned that decades ago with Ventura -- before they "improved" it by encoding files in some proprietary, non-text format).

Yes. Now, you (the vendor) can entice me to *elect* to use your tool instead of EXTORTING that behavior from me. Isn't this what The Market is supposed to be about?

Without putting you "on the spot", could you comment IN GENERAL on where the costs tend to lie in your business? Yes, I realize getting a firm answer here is folly -- but, perhaps you could shed some light on what drives this business in this direction (where cash flow seems to always be a problem).

Thanks,

--don

I don't usually have much sympathy for overly high priced tools, but you are right in that eda tool development must be one of the most intellectually labour intensive things to do, especially at the top end and you need some of the brightest minds to get it right. Added to that. it's a limited market and the necessarily high prices limit usage to high end companies like aerospace. Gut feeling just tells me they started to have the company equivalent of illusions of grandeur, going too far up market too fast and believing their own hype. If you claim a lot, then you have to deliver a lot etc. It wasn't too bad in the old Protel days. Reasonable prices and they even kept their original Easytrax and Autotrax available as a free download. I did some work at a client recently who were still using Protel 99 and it's still usable, if a little clunky in places.

Same with any high priced "really usefull" software product. People who don't have the financial resources will find workarounds and there are a lot of them. Just as open source has eroded the dev tools market, it will erode the eda market in time :-)...

Regards,

Chris

n

Ah, so well said. My way of taking this into account is probably at the extreme (don't know of another person or company doing it like this) but I do use only tools I have written - not just for software development, for hardware as well. For schematics and PCB I use a graphics editor I wrote in the mid 80-s, less sleek (not necessarily less convenient, though - and certainly very convenient to me) than the popular tools of the day

), but it does take me places all right nmc3top.gif , etc.).

Infinitely more important, I am tempted to say :-). A tool compares to say a written novel, the data standards compare to the alphabet used for the purpose. We are far from having such an EDA alphabet, though.

Dimiter

------------------------------------------------------ Dimiter Popoff Transgalactic Instruments

------------------------------------------------------

u

hen

One of the 3 links got damaged,

is the correct link. To me perhaps the most annoying thing about google groups which I use is this habit to split lines, in this case they split a link (so that the first half still worked but was the wrong link...) And it is not that I do not try to keep my posts reasonably narrow, their idea of where to split is just impredictable.

Dimiter