Several years ago, I wanted to get into autocad again but not pay for the latest and greatest. It was just for my own personal use of playing around and designing things so I had no worries of format compatibilities with other companies, etc. I had gotten a stack of autocad lt 2002 books for free from a professor at the college I worked at, so I set out on ebay to find a copy. Got one complete in box!
When it arrived, I fired up the installer, used the serial, and everything installed without a hitch. I ran the program and it asks to activate online. "No problem," I think, I used a legit serial from a legit copy, I should be fine. What I didn't consider was that the activation server was taken offline (I'm sure the official reason was "cost savings" but "force people to upgrade" fits in there pretty well in my opinion).
So now I had a perfectly good box of software and a stack of manuals for a product I couldn't use for no technical reason other than the company didn't want me to use it.
Damn, not even a decade of keeping their activation servers online by Autocad for their own software? That is ridiculous, it'd be like Microsoft shutting down their infrastructure for Windows 7 activations today.
I feel at a certain point, if you've sold software that calls home and you are unwilling to maintain the servers and there are still people using it, you should hand it off to the Internet Archive or a similar org so they can run it. Leaving users stranded with no hope of using software they're familiar with, have paid for, and have supported is just a bad move.
Moves like what Autocad did to you are why large companies are using libre software, if upstream dies for some reason, your business won't click off tomorrow.
Being free to choose someone else to pay for maintenance work seems like a pretty good feature. Rejecting Free software on the basis that the original maintainer might stop seems like a veiled statement of unwillingness to actually pay for maintenance, not a statement about the viability of supporting Free software.
There is software where a single yearly license costs more than someone's salary.
The problems with these servers though is typically isolated to hobby users who don't have the money to hire people to keep the ancient software working for them.
Which is funny, because in the worst case they can continue the development of
the software by themselves, which is virtually impossible with commercial
software.
No, this isn't. If that was your worry you especially wouldn't use proprietary software where the issue is much more prevalent. As others point out, you could pay FOSS maintainers if you wanted to.
Instead, people who don't use X make up excuses as to why not, rather than simply saying they like Y. If you like AutoCAD, use it.
In taking customer feedback it's important to only ask potential customers, not people who'd never use your product anyways. Even if their feedback is honest it won't help with what you're doing.
That sounds right, and your experience obviously gives you a negative experience of the manufacturer. But as an engineer in this situation I can relate. If we had phone-home activation I'd probably try to argue that it should be expiring re-enabled every year in the product so that if we abandoned it then users wouldn't be stranded.
The question I'd have to answer is - would this be a net positive? Would it lead to more or less sales? Would we only lose customers who wouldn't buy a new license anyway? What would the cost be of my proposed feature that helps customers in the future?
In our situation there is zero risk of e.g. losing business to open source, so management would likely be deligthted with the idea that users would not be running our old software. In reality we avoid the problem entirely by just selling subscriptions like so many others these days. The software has a kill switch that makes it unusable after the planned release date of version N+2.
The managers of this hypothetical software company would surely complain if their cars were remotely disabled because the manufacturer wanted to sell a new model. Similarly their customers should complain (via the courts) when the product they bought is remotely disabled to force them to upgrade.
As a sibling poster points out, running a license server is a trivial expense and if you require it for your product you should budget to keep it running. It's roughly $400 for a perpetuity paying $20/y to run the server indefinitely.
I completely agree - but hopefully if the management knows what they are doing in 2002 they don't sell a license that gives the buyer perpetual right to use the software (they can't guarantee that so long as the maintenance of license servers is required. Having a perpetual license is basically an infinite future expense even if the cost is trivial.)
> Having a perpetual license is basically an infinite future expense even if the cost is trivial.
It's a one-time expense if budgeted properly. Buy a perpetuity to cover it. If you ever discontinue the service you can release an unlimited copy of the software and recover the value of the perpetuity.
I bought a copy of Office 2010 Home and Student a few years ago from a brick and mortar store. I tried to install it a few weeks ago and got the message that I installed it too often (which perhaps I did because I only need it once a year and I use Linux, so I always install it in a throwaway virtual machine).
The installer told me to call a number but it looks like they disabled that number and they are currently in the process of setting up a new system. Meanwhile I can't use my legit copy...
There is pretty much no free CAD software that can compete with AutoCAD. The problem starts with the DWG format is based on proprietary RealDWG that they license to vendors. Free software will not be given a license.
The problem migrates to AutoCAD being like C++ and having a long list of legacy objects that are never broken.
A competing format was headed by Bentley with OpenDWG but they abandoned it somewhat after RealDWG gave in to giving them the doormat for free.
Bentley has abandoned OpenDWG but the software that is built on it has continued to improve. IntelliCAD is not free, but there are enough companies building on it, that it would be difficult to kill.
My employer is a member of the Open Design Alliance [0] working on IntelliCAD. I work a layer up, building plugins for both AutoCAD and IntelliCAD. Free trials are not hard to find, so you can decide if it will work for you.
I am no expert in CAD software, my only experience is with FreeCAD - and I found it is pretty good for things I needed. Project evolves fast. There are PPAs with up-to-date binaries too.
The UI is really nice. Version 3.0 whenever it comes out is going to be awesome. You may also try FreeCad for things this can not model, but that has its own problems - particularly UI.
The OSS world still needs a full-featured NURBS library that supports trimming. Solvespace does, but the performance suffers as complexity increases.
Reminds me of that ballroom dancing software program's page:
what happens when the Douglasses are no longer around?*
...
Here is how the plan works. Immediately upon learning of our deaths the executors of Dick's estate (his two highly computer literate kids) will post two files on our www.compmngr.com web site and will send out a broadcast email advising our customers how to download the files. The first file is a small standalone computer program called RegisterEvent.exe, which allows you to create your own registration files. So you won't have to register with Douglass Associates and you won't have to pay a registration fee. You can read more about RegisterEvent and how to use it below. The second file is a ZIP file containing all the source code for COMPMNGR and its supporting programs. This file will only be of interest to those few users who want to continue COMPMNGR development and who either know C++ programming or or willing to hire a C++ programmer.*
> what happens when the Douglasses are no longer around?
> ...
> Here is how the plan works. Immediately upon learning of our deaths the executors of Dick's estate (his two highly computer literate kids) will post two files on our www.compmngr.com web site and will send out a broadcast email advising our customers how to download the files. The first file is a small standalone computer program called RegisterEvent.exe, which allows you to create your own registration files. So you won't have to register with Douglass Associates and you won't have to pay a registration fee. You can read more about RegisterEvent and how to use it below. The second file is a ZIP file containing all the source code for COMPMNGR and its supporting programs. This file will only be of interest to those few users who want to continue COMPMNGR development and who either know C++ programming or or willing to hire a C++ programmer.
You guys know, it just dawned on me that beyond the practicalities of the question, we're looking at something incredibly significant here. That is, if this story is in fact genuine and not just a fictious commentary on subscription models.
So first of all... a software activation model from the '80s? Twenty years before Microsoft "invented" it? Back when even ARPANet was only a curiosity? That's something incredibly awesome. Key disks, manual code entry, and serial or parallel hardware keys were a thing - remote activation, now that's something else entirely.
It's like finding an ancient Egyptian D&D-like rulebook and twenty sided dice dating from the age of pharaoh Ramses. Not "impossible" as such, just extremely remarkable and improbable.
Then, the other wonder here is not that the activation server stopped, it's that it stopped in twenty-effing-seventeen. I'd have expected that BBS to have been scrapped around '95 at the latest. Also, phone numbers - where I grew up, phone number formats and area codes have changed three times since the '80s, a BBS call coded in the '80s would have stopped working before '91.
Also, I cannot imagine the change-averseness of someone who keeps using a CAD suite from the '80s (probably on some ancient XT or DEC or something weird and irreplacable) well into the 2010s. I mean my dad is the image of conservativeness, and yet even he migrated all his CAD work from the old DOS tools over to Windows-based, modern stuff when Windows NT 4.0 came out...
I'm not saying "it's not true", just that if it is, this should probably be on some Youtube computing history show as the curiosity of the year.
> It's like finding an ancient Egyptian D&D-like rulebook and twenty sided dice dating from the age of pharaoh Ramses. Not "impossible" as such, just extremely remarkable and improbable.
D20: Rome. Close, but not that close. (Found in Egypt though)
Yes, originally bought in the 80s, but like a lot of one-man-band hobbyist software, could well have had a "lifetime upgrades" option that at some point introduced the DRM.
I had a similar situation for a company that purchased it's accounting software way back in the first 90ies, which tied the software to an hw dongle.
In spite of the old system, the software works absolutely fine and there were no reasons to upgrade or change. In fact, the hardware it was running on was upgraded several times. Until one day, the dongle broke (a bump by the cleaning staff cracked it apparently). The company behind the software went bankrupt years ago, and there was no way to either obtain a new dongle or extract the data from the software (custom db as well).
The software was easy to crack by just nop'ing a couple of bytes, but the owner of the company was absolutely shocked by the experience as they reverted to paper archives for about a week with no easy way out.
I was quite young at the time and it served me as a life lesson. Since then I never used or considered anything with DRM, remote licensing or SaaS for exactly that reason. I never had trouble finding alternatives either.
> Until one day, the dongle broke (a bump by the cleaning staff cracked it apparently). The company behind the software went bankrupt years ago, and there was no way to either obtain a new dongle or extract the data from the software (custom db as well).
I have seen similar things happen, some proprietary data analysis framework from the 90s relying on hw dongle (from a company busted years ago) used in research lab. Talk about reproducible results, when you wouldn't even be able to reproduce it yourself when it breaks, I wonder how these people think. I guess their cancer research isn't very important so...
I am starting to use Kicad and FreeCAD in more and more projects for this very reason: I want to be able to open my projects 30 years from now.
Sure, the commercial tools are slightly more competent and have a nicer interface (although the FOSS tools are improving with every release). But there is no guarantee AutoCAD doesn't go bankrupt next week or changes business next year or is bought by Apple and becomes the next Logic.
I have also been moving towards Kicad. (Amusingly enough, I was also bitten by the Apple eMagic/Logic buyout).
Unfortunately, the way technology moves on, the only way to really ensure this is to commit to it as an active process.
Much like backing up your data, it has to be an ongoing and active maintenance process. This takes time and effort and incurs risk and disruption (changing formats and software).
I understand how the person in the original post got themselves into this problem, and I don't blame them. But with the benefit of hindsight one can see the lesson. For data to be durable over time, it has to be in the most compatible formats or the most general formats which allow them to be economically (wrt time/effort/money) moved to other systems.
With OSS and open formats you have a fighting chance, but even then the ultimate cost can be having to maintain the software yourself which for most people who are using design software isn't likely their main goal.
We once checked the activation of a peace of software which was very crappy because the server was down a lot. Turned out in case of a valid code it was returning a HTTP 200 OK.
The software never contacted it's home server again, but instead a hosts file redirected the request to a random server.
I have a strong policy of not using tools that prioritize themselves over me. Everyone’s time is valuable, and I have better things to do than sit on the phone screwing with activation problems or being forced to discontinue my work because I lack an Internet connection.
First thing your app does is present a login screen instead of creating a usable window? Won’t work without an Internet connection? Error messages that default to “accusing me” instead of “helping me”? Upgrades that mysteriously fail to support older file formats? If so then sorry, I am completely reworking my environment to never use your stuff again.
When I was younger there was a similar problem with an online game that had the official servers shut down after a few years of operation. The community at the time resorted to using a Kali Net hack to play. Luckily the server software was written by Dan Kegel and was open source http://kegel.com/anet/
After a little coaxing to get it running on 64 bit I managed to get it running and the community could finally play for another few years..until that server died and this time it was me who didn't have the time/interest to fix it again. Then it sort of died for good.
I wonder if there is an opportunity to provide a service/servers for legacy games so that every time the nostalgia hits a gamer he can come play online.
>I wonder if there is an opportunity to provide a service/servers for legacy games so that every time the nostalgia hits a gamer he can come play online.
I think it is what GameRanger does. I played Age Of Empires games and Rise Of Nations with it, it works quite well.
This is the perfect response to the question of why being permissive about DRM is a bad thing, something which is being asked in this other thread about Google Chrome:
This reminds me of a text editor I've seen linked here a while ago; it was written by one guy, and he'd been selling it and maintaining it for something like 30 years (and he still got new customers!). I don't remember the name of the program or the website.
Also, Bob Staake, a New Yorker illustrator, has been doing all his work in Photoshop 3.0 until 2008, and he's probably still using it. He was running it in the Classic environment on a PPC mac in 2008, I remember reading that he switched to a VM recently but can't find any info on it.
> This reminds me of a text editor I've seen linked here a while ago; it was written by one guy, and he'd been selling it and maintaining it for something like 30 years (and he still got new customers!). I don't remember the name of the program or the website.
> This reminds me of a text editor I've seen linked here a while ago; it was written by one guy, and he'd been selling it and maintaining it for something like 30 years (and he still got new customers!). I don't remember the name of the program or the website.
My father has a very similar problem. He's using a CAD software which cost a couple thousand bucks decades ago (I don't know how old exactly it is, but I think mid 80s). It's using a LPT-dongle. LPT ports are hard to find nowadays, but all of the USB to LPT adapters I tested didn't work with this dongle. For his PC I settled with a PCIe LPT extension card, but now he got himself a Laptop and I'm back to field one.
These old dongles are a joke. The Sentinel Rainbow parallel port dongle for example is merely a "secure" key store that is unlocked using a key with a whooping length of 16 bits and it uses the "ultra secure" XOR encryption algorithm. The advertised code encryption essentially boils down to having parts of the binary being XOR-ed with some key and to decrypt it, that part of the program would be sent to the dongle and the decrypted text sent back and the program image in memory modified. To make things more difficult the same memory region would be re-used for different texts (who needs re-entrancy anyway).
Long story short: This thing does XOR and all caveats of XOR apply. Like XOR-ing the ciphertext with the plaintext will yield the key. So cracking these things boils down to extracting the key and either emulating the dongle or decrypting the "protected" parts of the program and placing them in the binary in the clear.
I'd be surprised if your fathers program used something more sophisticated than that.
I had this exact problem when I was trying to connect my 3020 CNC to my computer. Most of these LPT to USB dongles was only designed for using it for LPT rather than as a parallel port. This pretty much screws up all the signal timing or analogue reading which a lot of old school dongle used to use.
Though I cannot vouch for this since I ended up ripping my control system out and replacing it with something modern. You should talk with a startech representative to make sure before you buy.
I think this thread is a fascinating living example of how short term 'our' digital world can be viewed, a mistake in my belief, infrastructure projects can easily take 25 years from beginning to end, I don't imagine that organisation would be keen to switch platforms many times during that period. Are we really all working on digital artefacts confident that they will expire their usefulness in such a short time frame?
Perhaps that is worth considering, compare tidying a room every day instead of putting up some storage...
I'm kind of inspired by this story. I love hearing about software with a really long lifespan. It makes you feel like it's worth building something decent.
Admittedly the DRM aspect is a bit naff, but it's hardly evil empire at that scale.
I work on CAE software that started in the 90's but some of our biggest competitors started in the 70's or even 60's. It's actually a bit depressing how little has changed in all that time. Things move very slowly, perhaps because the software is so enormously complicated and has thousands of important features that are all vital to somebody. It's hard to abstract that into something smoother for the user without breaking many well established use cases or wasting computer resources (It's CPU and memory hungry, and no amount of Moore's law has made a dent in that). Now there are startups doing cloud based versions, which is great, but it's still basically the same thing except in the cloud. It's still a steep learning curve. It's still trivially easy for users to get things wrong without even knowing they're wrong. You still pretty much need a university course as training on how to use the software safely.
But, you're right, it's satisfying knowing that it's like building a house. You can put a lot of effort into designing it solidly with a long future in mind.
I worked on a team in 1993 that made point of sale software that "phoned home" and required customers to purchase a modem with the software. We were no cutting edge software company, so I can easily imagine others doing this a few years earlier.
I suppose it could have literally "called home" by using a modem to dial a phone number and connect to a dial-up modem on the authorization server. Since it only does this "every few months", that wouldn't be a prohibitively expensive form of communication.
I really would love to learn how to crack an old piece of software like this. Any ideas on where to start if you don't know much about the software side of things? I know JS and a little bit of C#, but I've never found a great path to learn stuff like this.
Fravia is a name that brings back lots of good memories... him, +ORC, and the +HCU tried to popularise the idea of "cracking tutorials" that were more than "just change these bytes", and were actually quite detailed essays on the various protections around at the time.
As the content was largely user-submitted there was a lot of variation in quality and style. But there were some excellent essays and introductions out there:
Attach Wireshark to the program and capture the network traffic. This might already give you some insight if it's a very simple authentication mechanism - maybe you're even able to edit your hosts.txt file and make a fake server yourself. If that fails, get comfortable with x86 assembly and dive into the code. If you're lucky, they left some symbols for the function names, which makes your life much easier.
WireShark and OllyDbg are the usual starting places for me, both have tutorials around for them. OllyDbg in particular has guides around for by passing serial numbers, etc.
If you have this tight finance why not go with opencad or freecad?
I'm fairly certain importing correctly from a 35 years old format is not impossible, though there could be surprises for sure.
I understand the familiarity vs productivity issue and that learning something new is always painful, but if you want to avoid sw calling home, go for open source.
(I'm impressed by the way that something this old gets not just replies to questions but living servers to answer. I wonder if the answering machine is also from the 80.)
CAE is an area where open source has always been terrible. They just can't do the very complex UIs that are needed. They're good at back-end stuff, but people need to point and click. FreeCAD is nice but endlessly more frustrating and unintuitive than SolidWorks/etc. At least from my experience.
Sure, but at least it's an option for those so inclined, unlike commercial software, where your choice is just to not buy it or convince the developer to fix it for you (which works if you're a big enough user of their software).
I've worked with a lot of small business people and the HN group just doesn't culturally get what the guy is saying...
Say you're not going for billion dollar unicorn or broke, but for a life long small business. You always have non-main line of business systems that are technically obsolete but it doesn't matter to 99.9% of day to day operations because it works and when it doesn't work its only 0.1% of the time you spend anyway. To replace it when it completely breaks, its fully depreciated so to speak, you could pay $5K today or $5K next year, you certainly have $5K, but how exactly do you benefit by throwing down the cash today? I guess you get the tax deduction today rather than in 5 years but who cares? There's no way 0.1% of your day to day operation can have a significant bottom line effect no matter how much the new technology improves things in its little 0.1% corner of operations.
Even worse, when you forklift upgrade its going to be a hit to your productivity for "awhile" and you're busy today so you simply can't afford it. But sometime in the next five years there is certain to be a recession and then you already have the money and then you'll have the spare time to convert. Just not today. Its only 0.1% of your business and you don't want to turn away paying customers today to min/max that last 0.1% today. Play games with new toys when there's 10% unemployment and you have nothing to do for a week. In fact during that future recession you'll get newer technology than today plus it'll probably be cheaper if they're hurting for customers.
When its technology, 100% tech people are blinded by technophillia, but if it was an obscure metalworking tool that only gets used once a year for a few minutes, even tech people understand that when you're a busy one man shop, once that tool is completely depreciated and obsolete, waiting for it to break is the wisest possible choice.
Absolutely. FFS, I sell RS232 devices. RS-frickin-232. In 2017.
There are businesses out there that make a living supporting devices that were designed to run on 5.25" or 3.5" floppies. The machinery still works fine, but now the owner can't load files because replacement drives can't be found, or they can't buy floppies, or... So a business springs up to interface a thumb drive to a DOS HDD controller.
Same does my father with his PCB/EDA software. He ha a lot of libraries and stuff made with was now a old version of ten years ago. And he will not change.
If you're looking for an AutoCAD alternative try Draftsight from Dassault Systemes. There is a free personal version, but even the paid version isn't that expensive and it works pretty well.
No experience with Draftsight but going from SolidWorks -> Dassault's Catia -> Solidworks again was painful. Catia was (and I presume still is) an excellent product.
While patching software you own is illegal in most jurisdictions, making your own authentication server is legal. I am pretty sure that software is vulnerable to replay attack and could be emulated using fiddler's auto-replay feature.
I guess its something to do with quality. even expensive software is cheap so support is short.
I can still find after market parts for cars 20 years on. And the dealer will always do the service. Software isn't like this.
The terrifying implication here is that your CAD files will need to reside on an Internet-connected computer. You don't get an air gap to protect your secrets.
As much as I hate dealing with remote license servers. Not changing for 35 years just sounds ridiculous. I don't want to hear about all your old libraries, and part catalogs. The engineers at the company I work for do this all day, and we've upgraded probably seven major versions since I've started. Every company we purchase, and incorporate, they move their competing product libraries over to ours.
What kind of hardware could that guy possible be using to run that at this point? There is being stuck in your ways, and then there is being STUCK in your ways.
When it arrived, I fired up the installer, used the serial, and everything installed without a hitch. I ran the program and it asks to activate online. "No problem," I think, I used a legit serial from a legit copy, I should be fine. What I didn't consider was that the activation server was taken offline (I'm sure the official reason was "cost savings" but "force people to upgrade" fits in there pretty well in my opinion).
So now I had a perfectly good box of software and a stack of manuals for a product I couldn't use for no technical reason other than the company didn't want me to use it.