SolidWorks

There is only one problem -- you can't _PURCHASE_ it so you are hooked forever and have to pay for the rest of your active life. And when (not "if" but "when" -- it is ALWAYS "when", no matter how big and powerful the company was in its heydays) the company goes down you lose your entire job you did in all those years you've been using that miracle.

And yearly _RENT_ now cost the same (if not more) that is cost to _PURCHASE_ that software when they were selling it.

The smart move is to stick with the last release BEFORE it becomes cloud-based or rental-only (as applicable).

Ask if the "new features" (and possible bug fixes/additions) are worth chasing -CURRENT.

There are also ways to get capabilities without chasing a particular product/vendor (e.g., I was using 3D models of circuit boards back in the 80's -- but, NOT in an EDA package!)

I think there are some free equivalents.

The viewer is free. I have a guy in the Azores who does our mechanical design. He's great, and fast.

John Larkin Highland Tech Glen Canyon Design Center Lunatic Fringe Electronics

Once again, there is one small problem -- once it went rental you can't buy that last release anymore...

"If I only was as smart yesterday as my wife is today..."

I have Altium Designer with permanent license and grandfathered subscription still valid until mid-2027. Will stick with the latest version available before that subscription expires.

However, there is no way to purchase a permanent license today for any version. You are essentially buying a whatever version is current at the time of purchase and lock yourself to that version for a year. For a price that is about the same as I paid for a permanent license. It was their marketing ploy back in the days when Eagle went rental -- "Unlike other bad guys we give you a PERMANENT license! If you are an Eagle user we will give you a discount!". This is how I purchased my AD permanent license back then.

As we can see their word worth absolutely nothing.

The same thing happened to IDA and I'm lucky enough to have a permanent license with the last version before that happened.

Don't believe a single word they say. There is a simple test to check if they are lying -- if they open their mouth they are lying.

I have a Garmin navigator somewhere that I bought just because they promised "Lifetime maps updates". IN WRITING -- it is in their printed EULA included with that device. It is even on the original box in very big letters. Then I tried to update maps in it after 4 years or so and that was impossible. Their lawyer droids CLARIFIED what "Lifetime" means telling that it assumes the "projected device lifetime" or 3 years, whichever is SHORTER.

A miser pays twice, but a sucker pays always. Don't be that sucker, throw that garbage where it belongs and NEVER EVER buy anything from such smartarses.

The rental model is OK for rarely used tools but not for something like CADs. And it shouldn't be a yearly rental. The right model is e.g. Toyota with their TechStream car diagnostic and control software. You don't CREATE something that could be used later and/or needs to be maintained. You only need it when you are working on your car. And they will sell you a relatively cheap ONE WEEK license that is sufficient for a repair job. You don't CREATE something with it; it is just a tool like e.g. some big iron SDS hammer drill or whatever -- you don't need it on a daily basis and might only need it ONCE in your entire lifetime so it makes sense to rent it SHORT TERM when needed.

CADs are totally different and they are ALL proprietary so you can't rent a different tool to work on something that was created with another tool.

It is not necessary that anything has to be free, it is okay if it is cheap enought. :-D

I know SolidWorks from my company, but since 2006 I have Viacad3D 6.0 ($50) for private use and also now the newest Viacad3D 16 ($100) in my company. Of course there are a view differences to the expensive SolidWorks, but at least I did not notice them. But what I notice, the old Version from 2006 is living in a virtual machine now, will work forever and it is on modern hardware 10 times faster than the fatty Solidworks that need minutes to open a modell! And yes, since we all have 3D printer now, there is now live without

Olaf

Of course not! You have to have been "keeping up" with it as new releases came out. Then, simply STOP when the business model changes.

There are no shortcuts via hindsight.

Exactly. They aren't fools. They need a continuous revenue stream to keep those folks employed.

It's far too easy to pick a version and *stick* with it, ignoring all of their future "enhancements"; if it worked "well enough" last month, why won't it work "well enough" NEXT month?

They can always sell the business and the "new owners" can refuse to honor the promises of the prior owners.

You, of course, don't "own" the software that you've licensed. But, it will keep behaving as it used to if you avoid crossing the threshold to the cloud/subscription world.

"I'm sorry but we no longer sell ("stock") that version of the product. But, as a prior customer, you are entitled to this special UPGRADE for just $X (even if X is 0)..."

Of course, once you migrate your designs/projects forwards, there's no way to step backwards.

Note my recent comment re: a client trying to support a layout I'd done in OrCAD PCB 7 with PCB 9 -- not possible. And, I'm sure that migrating a layout from 9 back to 7 would similarly be impossible.

Hence my emphasis on keeping ALL tools available -- they are "knowns", even if they have warts.

I don't see how it can have value anywhere as you need to keep paying so it is *available* when you later decide you need it. There are no real savings.

On top of that, the product is in a state of perpetual flux; do you want to PLAN an effort based on your MEMORY of what a tool PREVIOUSLY was like? Only to discover that it has changed while you were looking the other way?

What do you do when/if <something> has changed that causes your prior investment to be valueless? (e.g., models now in a new format) Do you invest the time to migrate those things *forward*? I.e., do you make an ADDITIONAL investment in the tool (money is easy to quantify; the time required to adapt to something new -- AND UNKNOWN -- can be incredibly high!)

My *car* isn't likely going to change (without a significant evaluation process). So, why can't I just *buy* the tool? Price it so it makes sense for the market into which you are selling it.

I used to buy Workshop Manuals for every car I owned. Even at a few hundred dollars a pop, it was convenient to be able to browse them AT MY OWN PACE before and after problems arose.

Now, I have to *rent* access to a manual "online". Should I just plan on spending a few days downloading every page and archiving it for posterity? What if I didn't KNOW that a certain page existed that could have addressed my problem directly?

That has always been the "a little bit pregnant" hook that all EDA tools had; you invest in developing libraries which means switching to another tool represents forfeiting one investment and taking on another. So, you didn't jump around between products (like you could office suites, compilers, desktops, etc.)

One consolation is you don't see these firms rolling in the dough as their pricing would make you *think*. And, individual shops can just decide to STOP using a tool and let a service bureau do their work -- one less seat generating revenue yet the same number of boards being designed.

I wouldn't want to be in their market...

What a lot of crap, a box like that I can assemble from what I have laying about in less time than playing with yet an other CAD or cat! or whatever .

Now if you had a 4D or where are we now 5D? printer maybe it would make sense. Or ask AI.

BEEP

They will sell it if you want. This is what auto mechanics do who are in the car repair business -- they need it every day. it is not all that cheap but they do sell it.

On the other hand, there are lots of hacked pirated versions coming with Chinese OBD adapters and without. They DO work and can do [almost] everything the legal version does except connecting to the Toyota data server. It is not critical for regular repairs / setups /configuration but it can't reflash (or provision) any ECU as all firmware images are kept on their server. It is not even an attempt to somehow get control over a mere user -- there are simply too many ECUs in all Toyota cars (TechStream supports them all) and different versions of their firmware, specific for particular VINs and so on. It is simply impossible to fit them all in a piece of software and to make a decision what version to pull based on a particular VIN.

But if you, e.g., replaced some ECUs you have to "marry" them to the rest of the vehicle that is not possible without a legal licensed TechStream. This where that one week license comes handy for those who fix their cars themselves.

I do also have 8-volume PRINTED shop manual for my Lexus LS460L but you can do almost nothing without TechStream. Maybe tighten some loose nut/bolt or replace something pure mechanical but even that is not always possible without sending some commands to the system first. TechStream finds 50+ ECUs in my LS460L (don't remember exact number -- it's been a while since I last talked to my vehicle) and each and every of those have settings and configuration. Some of them also have UNIQUE IDs so the system must be instructed to take a new ECU replacing the old ID otherwise it won't work properly and might even refuse to start.

I have an eternal Altium Designer license and stopped the yearly subscription in 2021 when AD crashed 5 seconds after calling it up, repeatedly. I got no help at all when I said that my Win10 machine ran under VMware. They pretended that VMware could not handle networks. But that's VMwares bread and butter.

They probably fear that it really could run forever or, even worse, I could give a copy of the VM to everybody and his dog.

The idea is that I cannot expose customer data to the net, so the win10 world ends at the ext4 /d partition that is mapped to d: by VMware and that contains the actual designs.

I have made a quick design of a 100 * 100 mm board with holes in

Gerhard

Am 19.06.26 um 11:16 schrieb Gerhard Hoffmann:

I found the problem myself. I had renamed an old Protel library from .lib to .pcblib. Altium does not check the contents in any way and crashes.

We designed the extrusions ourselves. The Hammond boxes ARE crap.

We can 3D design the box, the PC boards, the end plates, the light pipes, the connector locations, and be sure everything will fit and can be manufactured repeatably.

Being able to see it all in the 3D viewer, spin it around, measure things, is a huge aid to thinking about the design. Electronics is half packaging.

You couldn't to this

John Larkin Highland Tech Glen Canyon Design Center Lunatic Fringe Electronics

?

Raspberry Pi hat with veroboard and ebay modules, all .1 inch

Have not used El Tea Spice in many years Like that tennis player, live, versus the math guy with his calculator Who wins? LOL

Anyways, they nopw ting how did thse Egyption build the Pyramids...

Soon if anything is left of humanity after trump's nuke wars they will wonder 'how did those humming beans make those cellphones. And what are salettes and GPS, all we see is space junk, in space AND on the houses or what's left of the houses.

There were bigger pyramids in South America.

He's trying to keep Iran from nuking Israel, which they were probably planning to do.

I'm designing things to be manufactured. 3D modeling really helps.

We are using the R Pi cpu, the RP2040, mildly overclocked.

John Larkin Highland Tech Glen Canyon Design Center Lunatic Fringe Electronics

To be fair, one can't expect a supplier to support *every* configuration under which their tool/product might operate.

I've had products that were "only" supported on a specific make/model of hard disk (really? will a disk that has the same geometry but is made by a different manufacturer NOT work? will a larger disk from the recommended manufacturer not work?). Conceivably, there COULD be issues "baked into" their implementation that silently rely on certain characteristics of the recommended drive. But, likely not (OTOH, I recall having to patch the BIOS ROMs on my Compaq Portable

I ran DASH-STRIDES hosted on a Chimera II... installed in a SPARC (!) SB2000... under CDE... over a network connection... to another PC. Just *explaining* how all of that magic can work would be too much to expect a support engineer to grok ("But STRIDES is an x86 application; how did the SPARC handle that? And, how did you accommodate the hardware product locks STRIDES required on that hardware??")

Of course, they aren't stupid and likely don't want to encourage people to operate in environments that defeat their product lock-ins as that's what pays their salaries. But, the problem there is with their chosen implementation; expecting no one to sort this out on their own is a bad assumption when your customers are engineers and not basket-weavers!

But you can still do that -- just not with their assistance. And, you'd likely be more inclined to share your results with others if the problem is more challenging.

"Design a product that people WANT to buy, not that they feel COMPELLED/FORCED to buy."

I do that by air-gapping my development environment and sneaker-netting anything "out" (or in) as required. It also mitigates the need for other protections mechanisms that eat into productivity.

What you can do with a given tool -- or, tool generation -- varies. In the mid-80's, doing SMT with tools that were designed for thru-hole technologies was "challenging". My current designs would likely be impossible with those same tools (BGA425's in 13mm packages?

Similarly, expecting the tool to match trace lengths wasn't an issue, back then. Or, 10-layer pad-stacks.

The whole notion of filenames carrying "type" information is just stupid. Especially when names are so mutable! Would Gerhard.girl suddenly change Gerhard.boy's sex?

Another mistake consequential to the use of "plain files" for structured data. Think about all of the code that (should!) be executing to verify the format and content of EVERY file referenced by EVERY program!

Wouldn't it be simpler to ensure the contents are correct before being "admitted" to the file -- and then constraining further changes similarly? As such, any CONSUMER of that content would be assured that it was consistent and in the correct format for their expectations WITHOUT having to verify that each time the contents are referenced?? Not having to remember to take read&write locksWHILE referencing it in an environment with multiple readers/writers?

Oh yes, let's go back to OS/360 and JCL. So much better!

Jeroen Belleman

I don't repair late model cars -- there's no money to be saved; treat those as billable hours and pay for the repair with the monies earned (and far too many things that can go wrong, leaving me at the *mercy* of a shop).

But, I *do* use workshop manuals as references -- what's where, how things fit together, etc. And, for diagnostic information. The "check engine" mentality is just TOO lame for my mindset (can you at least give me a *hint* at what the problem might be??) E.g., an overweight grease monkey happened to dislodge one of the seat sensors in a prior vehicle (due to his size). Much easier to look up the code that was being thrown and discover the opened connector than make a repeat trip to the shop.

Currently, my interests are in sorting out how to extract/inject signals into existing designs without compromising their normal function (or, having to reverse engineer "substantial" designs.)

E.g., how can I access the main display in the dash? How can I tell which evaporators are active in the refrigerator? Or, determine the state of the controls on the oven/stovetop? How can I determine which cooling stage the ACbrrr is operating in?

My automation system is of similar design, of necessity (all interactions are encrypted so a replacement module must be "formally *introduced*" to the system else it won't be seen or acknowledged.

But, I've made that as simple as possible so an end user could perform the task without requiring any special equipment or expertise. *I* dislike being placed in a situation where I *must* rely on another (for no REAL reason) so don't think others should have to rely on *me* for the same arbitrary reasons.

But 3D printing takes a fair bit of time. I can take a model created from an actual design (as a "side effect") and verify that the tall components on board 1 neatly fit over the short components of board 2 (and /vice versa/) just by twiddling the motion controller.

And, verify connectors, indicators, controls, etc. align properly with the holes in the enclosure that will house them.

How can you determine if airflow over and around a set of

CAD isn't just a tool for designing products but, also, for analyzing their characteristics before they have been reified.

No money to be saved is something I can't even find words to characterize...

Lexus dealership, front upper control arms replacement quote, $13K. 4 upper control arms. 8 arms total in front suspension, can be easily found (aftermarket but as good as originals) for less than $300. All 8, not just 4 upper ones. And it is not just 8 control arms but also stearing tie rods (inner and outer), sway bar links, rubber boots: