As an structural engineer and former Autodesk user, I completely agree with the letter. I always saw Autodesk as a company where software goes to die a slowly painful death.
It started with Autodesk Structural Detaling. They bought the software from a small company, renamed the product... and that was it. It had the same bugs release after release... we as customers tried our best to report the bugs, but nobody listened.
We stopped using ASD and moved to Graitec Advance Steel and Advance Concrete, for a couple of years everything was great, quick support, remote sessions in case you struggled with something, new features, bugs fixed... and then... Autodesk bought it. I still remember the morning when I read the news... I was horrified. I remember talking with some guys at Graitec and they were shocked also.
After this... was Autodesk Structural Detailing story all over again. Just a product rename... and that was it. No more bug fixing.. no more new features.. they killed Advance Concrete in favor of Revit (which is one of the dumbest thing ever... because Revit was years behind Advance Concrete as a concrete detailing sw).
For concrete detailing we switched to Allplan. I think Allplan and Graphisoft's Arhicad have a bright future together.
Perhaps i'm being idealistic but if the cycle is that aggressive this screams for architects to start caring about buying FOSS. When you buy software it's an investment - you invest your time in integrating it into your workflow and your employees learning to use it... your investment is tied to the future of that software.
If it's FOSS (funded), no amount of money autodesk throw at it can bury it's source code, buying it would guarantee to some degree a future in the software and thus your investment. Even if autodesk bought the originating company.
But this would take a widespread cultural change in the values of architects combined with the first company willing to gamble on that business model. If a large group of architects are willing to sign a letter like this, perhaps now is the opportune moment to try and spread that value to avoid this problem in the future.
> Small company rises to produce quality software, it sells, responds to bugs, listens to users, adds features etc.
I would say:
Small company invests time and money into creating a decent enough product. They listen to customers to build their roadmap and respond to most bugs in a reasonable time. Small company could grow to a medium or large and still hold true to their roots.
Large company buys a mediocre to decent company instead of doing R&D. New company is integrated into large companies suite of products.
Large company fixes only enough to keep the product running or shuts it down completely. Support loses its personal touch and is transferred to the cheapest FTE location. Large company’s corporate politics causes infighting between departments with similar products.
Large company merges with equally large company to reduce competition. Execs cash out and leave ASAP. That’s where I work now.
Sorry I edited away your quote since it was implicit in parent.
Your elaboration sounds more accurate i agree, but I think the end result is the same (a dead product with buried source code and wasted investment from it's users).
I work for a competitor in this space, and we make a living off of poaching AutoDesk customers. We're owned by a pretty deep-pocketed holding company as well, so I don't think we're going to get bought.
IMO it's not just about "values of architects". They're a population that does cognitively demanding work (so polished UX is of fundamental importance) that tends to have more of a fine arts background than a software background (so right-to-modify does nothing for them).
FOSS is just ... a bad deal for them. They want someone they can call and yell at when they're on a deadline and the software isn't working, and they want tools that get out of their way.
I get where you are coming from, but when I say FOSS i do not necessarily mean a "FOSS project"...
I can imagine when it's very niche with a small user base, you probably do need closed source + for profit company developing it with a high per-license fee to produce anything of useful quality. What I am suggesting is _both_. The FOSS part is purely licensing, as an insurance policy for the users, and to make acquisitions repugnant to the likes of Autodesk (since all that would achieve is funding the software that competes with theirs)... I know that is far from easy to achieve and there is no straight forward business model for that - like I said, i'm probably being too idealistic.
[edit]
I wonder if such a legal mechanism exists which automatically triggers FOSS licensing upon acquisition or bankruptcy (i.e what happened to blender, but as a requirement)... that would prevent takeovers that do not benefit users as a kind of legally binding promise while also sidestepping the issue of being profitable while developing niche FOSS software.
If legally feasible this is a nice promise any currently existing software company can add to their products without changing their business model.
In fact thinking about this, as someone who pretty much never buys software anymore, not because I am unwilling to spend money, but because the experience of having the rug pulled from under me too many times frankly makes closed source unpalatable to me - this would make buying (currently) non-free closed source software a lot more comfortable again.
It also feels like a good strategy to combat anti-competitive monopolies like autodesk from destroying choice in software.
> I wonder if such a legal mechanism exists which automatically triggers FOSS licensing upon acquisition
I can't imagine a VC funding a company with a poison-pill clause like this. Such a clause would discourage acquisition, and don't most VCs (if not most founders) hope for acquisition as a very desirable exit strategy?
I think it's definitely possible. Look at Blender and GRASS GIS, some of the most successful open source projects among non-programmers of all time. Architecture tool is merely the midpoint between those too domains, right? :D
Blender was originally a closed-source product by Not a Number Technologies (NaN). After NaN went bankrupt, the
creditors agreed to let Blender become released as open source for 100,000 €.
True, but not really relevant. (Unless you're implying that you think Autodesk is likely to go bankrupt and might be open to a similar arrangement?)
The Blender we have today is much improved compared to the last version developed by NaN. Most of its major features today did not exist when it first become open source. It was certainly helpful to have a working core at the start—I'm not saying that the 100k € was a poor investment, by any means—but I would be surprised to find that there is much left of the original code by this point.
Agreed. Of course I don't think this is what GP was implying, at all, but for the record it would devalue a ton of work to say that Blender is where it is because it started as a closed source product.
Blender is where it is because the core organizers over the past decade have had a clear vision, great community management, and have worked their butts off to make the software as good as it is today -- in terms of UX, capabilities, performance, marketing... everything across the board.
When I started using Blender, it was not as powerful or usable as other software in the field. It was impressive, and having a good core did help, but the program today has just advanced so far, I'm not sure it's really comparable.
See also Krita, for another Open Source project I think is headed in the same direction as Blender. None of this stuff happens by chance, developing Open Source software that's popular with a general non-programmer audience is really stinking hard, and the teams that can pull it off deserve all of the praise they get.
Architecture tools are way more niche. I think a better place to look is mechanical CAD: how many decent open source mechanical CAD programs are there? Approximately one: SolveSpace. And that is quite basic - even simple stuff like bevels is not supported. Absolutely nobody whose livelihood depended on it would use it.
It's just too niche and too complicated for there to be enough developers willing to writing it for free.
My uneducated guess was that engineering tools have a lot more hard requirements than architecture CAD. That's why architects higher engineers to say whether their buildings will fall down!
Call Siemens and ask what they charge for Parasolid. Hint: you have to ask, so...
Computational Geometry is hard. FOSS CG Kernels are not going to spring from the ether. CGAL exists, but its target audience is CG researchers. OpenCASCADE exists, but it's really a loss-leader for proprietary extensions and consulting contracts.
At a complete wild guess I would say something like $50m.
(Obviously it depends what you mean by "usable" - I use it now, so in a sense it is already usable. I guess you mean to get to the point that businesses would rely on it.)
Blender is programming-adjacent (a lot of Blender users are employed by software businesses to create assets for distribution with software products). In particular, I think that's where a lot of the funding for it comes from.
> Perhaps i'm being idealistic but if the cycle is that aggressive this screams for architects to start caring about buying FOSS.
Look at the city government of Munich. They tried to adopt Linux and FOSS but failed after years of internal issues. Proper support and industry adoption are important properties not easy to emulate for FOSS.
They "failed" because the social party mayor didn't get re-elected and a conservative mayor got elected. The new mayor then blamed LiMux for problems that mainly rooted in the ongoing migration of all departments to LiMux and then ordered the migration back to Windows for a price that would have had LiMux running for decades.
By total coincidence, shortly after announcing the move back to Windows, Microsoft announced to relocate their German headquarters from Hamburg to Munich.
> I always saw Autodesk as a company where software goes to die a slowly painful death.
There's a really nice lightweight vector design app for Mac and iOS called Graphic [0]. It was one of the first fully-featured design tools for iOS, and still the best to use.
It was bought by Autodesk back in 2015, who promptly did nothing with it at all. Pretty much all development ground to a halt. From being updated every few months, the Mac app has only been updated three times in the last five years.
It's since been spun-out to a different company, but development still isn't really happening – the iOS app hasn't seen an update since 2018.
The lack of attention this app has been shown is sad, as it's a good piece of software and I use it all the time. It had the potential to be the vector equivalent of software like Pixelmator or Acorn – lightweight but powerful alternatives to Adobe products created by independent, highly-passionate developers.
Instead, newer apps like Affinity Designer and Vectornator have taken any traction it might have had - and it's probably going to stop working with an upcoming release of macOS if they don't do anything.
Adobe bought Macromedia and killed off fireworks. I still run the last version they made.
I know your pain.
Once you learn how to do everything in an editor and they stop supporting it you need to either reinvest time into learning a new editor or in supporting the obsolete one.
I’ve chosen the latter for now mostly because there doesn’t seem to be a clear winner in the market and I hate to keep investing time in learning new workflows every couple of years to do the same task when I can just keep a VM operating to support an obsolete editor.
Its not just Autodesk, all big software shops tend towards low innovation. Look at Adobe, ESRI, even Microsoft office etc. Core products often haven't changed appreciably in 20 years - the business/money isn't in innovation, it is in having proprietary file formats that are industry standards and bundling big packages of software together that makes it hard to shift off the platform.
Whilst that's true, there's also a question of degree. Revit, which is the source of complaint here, takes long enough to open your file you might as well go make a sandwich while it does so. The developers only found out that CPUs have multiple cores somewhere around 2017, and still haven't really worked out how to actually use more than one core for anything much. They still appear to be largely unclear what a "GPU" is, so don't bother putting anything fancy in your workstation.
It also likes to crash a lot, often taking a bunch of work with it. It's slow in a way that's probably entirely foreign to most readers of HN, and even compared to things like adobe photoshop or ptc creo, the rate of improvements is most kindly described as "glacial". In 2020, revit finally learned what a pdf was for example. Sure, large enterprise software development moves slowly. But there's slow, and then there's Autodesk.
Ha. I work on a competitor and our story is similar.
We're a little more aggressive with multi-threading but you have to understand these code-bases are massive legacy things whose value comes from a ton of business logic accreted over 25 years.
I assure you I've heard of multiple CPU cores (before this I did high-performance physics simulations across thousands of cores) but I still don't have the first idea how to get something to execute on a background thread in this codebase.
Working on an new project in a similar industrial market to most AD products, we've adopted a massively multi-threaded approach to development. But I'm starting to wonder if the product will come to fruition before I'm pushing trolleys 'round at Tesco.
With Adobe I'm extremely worried about their acquisition of Algorithmic, the makes of Substance Painter and Substance Designer. I use these tools in 3D animation/texturing, they are truly fantastic tools. Before Adobe bought them, they offered a purchase option, now it's just subscriptions of course.
I have such a love/hate relationship with Adobe. Their software is so useful. I want to love it. But they basically infiltrate my entire computer in order to install/run it and I’ve been burned by their shady billing practices so many times I now have a prepaid Visa from CVS that’s exclusively for my Adobe subscription. I fundamentally do not understand how a company can simultaneously be so awesome and so awful.
When I was younger, I used to pirate expensive software that I couldn't afford. It's what got me into 3D graphics and graphic design at a young age. Now I'm older and using some of this software professionally, and usually, I'm more than happy to pay for it. But I'll never buy into Adobe's shady subscription model, or any other company that does this. When I pay for a software license, that version of the software should become mine, the same as when I buy any other product. If a company refuses me that option, I steal the software with zero remorse.
The idea of renting software is ridiculous, especially when I have zero need for support services.
The problem with this: updates. You buy a version. They fix the bugs you report and release a new version. Do you pay full price for it? No, it's just an update. You'll pay a reduced upgrade price, but you want it for free. Your price probably doesn't pay for ongoing upgrades and might not pay for the updates. So no one is paying for updates. Every new version has to be sufficiently improved to get everyone to buy a new copy. Or the company could release a version and coast on it until revenue stops coming in.
Renting software is ridiculous, but at least the incentives are kind of pointed in the right direction.
Photoshop is extremely stable software. Other than a couple of new features and UI changes, there's not much difference between Photoshop now and ten years ago.
I agree, but to defend them somewhat long established products like MS Office suffer from 2 competing interests, people that want innovation and people that want stability
For my core users any change at all to MS Office is met with frustration and huge amounts of backlash, even something as simple at changing the icon shading is a problem
Many people are robotic when it comes to using these products, going through the same steps day after day, file after file, and any change to the workflow is cause of extreme frustration,
> Many people are robotic [...] going through the same steps day after day, file after file, and any change to the workflow is cause of extreme frustration,
Work in semi-related process automation, and this. So much!
Software engineers can't begin to understand what happens when you change a hotkey, when someone has been doing something 100 times a day for 10 years.
Using a hot key 100 times a day for 10 years sounds like exactly the kind of thing software engineers SHOULD understand. I heard a lot of anger from Vim users when Apple messed with the escape key.
In my experience (~10 years of doing this sort of work) software engineers typically understand how their users use the software in the abstract, but not in the actual experience.
E.g. It is well understood that users rely on a hotkey "a lot." It is not understood that users are so familiar with a hotkey that, if you relocated it on the keyboard, productivity would decrease 40% for 3 months.
That distinction is fairly important when trying to decide whether {new feature X} is actually worth disrupting {old feature Y}.
UX design is done by designers and product managers. Developer just implemented ability to assign hot key X to function Y. How this ability used was not his/her concern.
Personally, I'd push back on such a change unless there was a very good reason. I feel like this violates the same kind of principle as "we do not break user space EVER". My point above is that a programmer should understand better than most how important muscle memory and shortcuts are to productivity.
I would push back too, and I think most experienced programmers understand the needs of power users[0] - problem is, they don't push back all that often. I encourage developers to try and voice their opinion about UX more often. Who knows, maybe your PM isn't a pointy-haired boss, but will turn out to be a professional willing to listen to rational arguments? I had a privilege working with such PMs; at one place, I ended up having a lot of say about UX (and provided a counterbalance to our designer) - just because I spoke up about the issues. It turned out that my views were shared by the rest of the dev team, it's just that they never bothered to voice it.
--
[0] - Almost everyone who spent most of their day job using a particular set of software tools quickly becomes a power user of these particular tools. The corollary to that is that if your software is the kind to be used at a job, it better be power user friendly, or you're going to wasting your customer's money and the life their employees.
My Experience is software engineers understand how users SHOULD use the software, which rarely matches how users DO use the software
Most of the time I experience a scenario of the engineer or support person saying "Show me what you are doing" only for the user to show some weird and non-logical (to the engineer) way of doing some task, and the engineer or support person replying "Well you should be doing it this other completely different way"
MS Office, or at least Word and PowerPoint have lacked innovation to the user experience on a fundamental level. Years ago they overhauled the UI because people weren't using many of the features on offer. What they failed to do is make those features easy to use. Prime example is numbered lists in Word which is still unnecessarily complicated to use. This isn't a compatibility issue, purely UX. Numbered lists are also buggy, like they were 15 years ago, and are still the best way to corrupt your document.
It's not just lists, just about all word's auto formatting is buggy and unintuitive. Page numbers, tables, headings, tables of contents, paragraph layout on pages with pictures, just putting pictures in a document can be a pain in and of itself. Pretty much anything you do in word above some simple formatting starts to get buggy and unpredictable and the more of these things you have in a document, the more painful editing that document becomes.
The problem for microsoft, it's not a simply "fixing the auto formatting", but "changing how auto formatting behave". Nowadays the buggy auto formatting may be expected by some people and changing (fixing) it may distrupt their workflow (xkcd reference).
Moreover I don't think there are "best" specification for those auto formatter, even a "better" one.
There's a class of software I want rapid improvements to.
And frankly there's a class of software I want to just work and not move the needle needlessly. Even with Adobe, I have NO idea why I have to have 27 processes constantly running to check for updates; why it actually performs updates that grind my system all the time; I can't honestly say that I have personally advanced and started using anything new or different in Photoshop in many many years, let alone MS Word. :-/
(understanding this is a different use case to original-original-poster; most of the software I DO want to advance - our database engine or process scheduler or OS monitoring software can always get better as far as I'm concerned, etc.
> Many people are robotic when it comes to using these products,
Many people aren't just robotic, they're straight up ritualistic.
I've worked with several people that are absolute wizards at PowerPoint, love to get new features, and quickly integrate them into their workflows.
But those same people have absolutely no mental framework for how Excel works. They'll quite literally use a physical calculator (or their phone's calculator app) in tandem with Excel, treating Excel as a glorified text editor that lets you do cell-by-cell text placement/formatting. They have such a weak grasp of how Excel works that any tiny change whatsoever doesn't just cause extreme frustration, but also extreme anxiety as they scramble to re-establish the thin veneer of usability they previously held.
Having witnessed the above more than once, I've taken to heart the level of naivety that should be presumed when creating UX flows and documentation.
Sounds like economists run the company not engineers - Boeing comes to mind by some reason - where in place of competing with real inventions the money making has sole preference so killing of competition early using big piles of money is a perfectly decent and honorable way to go.
Not something novel that we have never heard of btw.
The gold standard around these parts is purported to be Tekla Structures. No affiliation, just a structural steel guru who’s happy to work with the fabrication drawings it spits out.
I can spot Autodesk fabrication draws a mile away. Either the steel detailers don’t know how to configure it properly, it, as I suspect, it’s a steaming pile of dumpster fires.
I used Autocad for 2D CAD quite a bit, and I don’t see the appeal. There are competing products at a fraction of the cost.
Indeed. Autodesk Advance Steel generated fabrication drawings need a lot of effort to get it right but AS was ahead of Tekla in regard with connection design and modeling/detailing super strange parts/structures.
As for the prices...in 2014, 2 Advance Steel licences were something like 12000 euros and one Tekla license was about 18000 euros.
Is there anything a non-professional can use for the following four usecases (separate software is ok):
1) Reinforced concrete structure design, with support for non-trivial surfaces (e.g. splines/Bezier curves/surfaces).
2) Non-finely-meshed (in comparison to local volume thickness) truss structures, ideally not just classic steel beams + welded/riveted joints,
3) but also aluminium (considering fatigue from non-stationary loads and optimizing for the lifespan/weight/cost pareto frontier (cost would be just a simplistic metric)).
4) The holy grail, a 3d-printed core (technically a large-pore-size, low-density, open-pore foam) that is then covered by fibers via a robot that passes the spool around the core/preform (filament winding).
After hardening the matrix that bonds the fibers together, the 3d-printed core could optionally be removed by melting and/or solvent washing.
This kind of structure is potentially extremely stiff due to combining the truss structure with cylindrical beams and a lightweight fiber-reinforced polymer material.
The issue is just the insane non-triviality in designing such a structure, because there are restrictions to holes and fiber angle shifts due to the winding process, where the fiber bundle has to be accurately woven around the mesh in the designed way. This, combined with the directionality of the fibers makes it necessary to consider much more than normal 3d-printable topology optimization[0].
A concrete-ish example would be a multi-monitor + keyboard/mouse support structure to enable a low-fatigue position where the user leans back about 30~60 degrees to allow the neck to be supported and the user's practical FOV (w.r.t. feasibly eye/head movement) is covered by screen area and the arms can rest on supports.
The constraint that makes a traditional steel truss undesirable would be to have it mobile (thus lightweight), yet stiff enough to handle ~5 m/s² without impairing usability. A practical case would be to overlay travel with getting work done, by mounting this in the back of a van (including straps/seatbelts as required) and being driven.
A different practical example could be amateur airplane design, e.g. devising an autonomous solar glider, where some stiffness is required to not break the solar cells laminated into the wing surface (thinned wafers are lighter and can conform to the wing's curvature, but will then dislike the wing flexing along the span), but weight is extremely important.
Mounting the heavy things (motors, some backup batteries, potential payload) and handling control surface actuation forces won't easily work with a simple fiber-reinforced hollow wing, as a lack of dedicated mounting points (with their own fibers distributing the forces into the outer structural surface) would require a far heavier construction.
It's also difficult to consider aerodynamic forces when computing pre-distortion of the unloaded structure, so that design loads make it take on the desired shape, if there are also ad-hoc mounting brackets glued to the inside surface (which would be necessary if they're not designed into the primary structure).
Sorry for the slightly-OT examples.
[0]: Project video "on using the free version of Fusion360 for shelf brackets printed in PLA on a Prusa i3 MK3": https://www.youtube.com/watch?v=3smr5CEdksc - Thomas Sanladerer: Making STRONG shelves with Topology Optimization
Well, one doesn't need to be professional to the extend where one can argue Tekla Structures to be worth the cost, if the goal is to design structures for hobby/DIY work.
Depending on the purpose, structural failure might be sufficiently low-risk to not need more than a cursory glance by an engineer to check whether the concept is sufficiently-sane and the stresses the software calculated are no reason to worry.
People today build these structures without software-assisted structural analysis. Or anything, really, beyond rough numbers.
I hope there'd be something that be good enough to not fly blind when working on hobby projects that would benefit massively from clever structural design. Finding someone qualified to look over the finished design is often reasonable, but paying them to do clerical work (e.g. re-drawing your blueprint in their CAD, just to get the computer to calculate stresses), would be too expensive.
I just see that these tools all have license fees where any non-full-time usage is directly prohibitive.
Even if a good tool has e.g. 500$ + 20% royalties (of the sale price), that'd be far better than what I see out there from my searching.
I see and understand your point and agree to some extent.
Not real sure what to do there?
I'd maybe consider doing a course at my local Tafe ( https://en.wikipedia.org/wiki/Technical_and_further_educatio... ) and do a design course and get friendly with the people who work there and run the workshop part of the campus so they might let you continue to use the facilities are you've finished.
Not sure if you have anything like that around your parts.
Sounds like you might want to use something like finite element software, which is essentially what other programs do. FEniCS is an example of an open source library that could maybe help. People usually use something like ls-dyna or Abaqus
Yes, of course. But I'm specifically interested in the design side, because the analysis side is already handled by open and/or free software, due to it being so close to what a small university research group might just publish or adapt for unconventional FEM analysis.
Design software is much less easy to find, though.
To be fair this is not new. Autodesk has had this MO for decades.
I remember in 1999, I was interning for a company that was getting into CAD, choice was to get 3D Solidworks that no one had heard of, or get the software that literally everybody used - 2D AutoCAD.
That company still uses Solidworks today and has never touched a single Autodesk product. I can guarantee the opposite would not be true if the other decision was made.
We knew in 1999 that Autodesk was never going to innovate, or allow us to.
RIP to SoftImage. And we're all worried about the future of Inventor. The only reason I use that at all was because of the generous student license...other companies don't seem to do that, which feels like a giant mistake.
You may want to check the offerings from Altair. They have a lot to offer and even support many 3rd party tools. They do make a lot of acquisitions but that's to increase their capability, not to destroy competition.
When you are a software company and you buy another software company or you buy a product from another software company, you do not intend to maintain their software. You are purchasing the users, and adding them to your own userbase. That's the only reason to buy a competitor.
Sometimes there are some token gestures, but it always boils down to "we can't compete to gain these users, so we'll buy them."
It started with Autodesk Structural Detaling. They bought the software from a small company, renamed the product... and that was it. It had the same bugs release after release... we as customers tried our best to report the bugs, but nobody listened.
We stopped using ASD and moved to Graitec Advance Steel and Advance Concrete, for a couple of years everything was great, quick support, remote sessions in case you struggled with something, new features, bugs fixed... and then... Autodesk bought it. I still remember the morning when I read the news... I was horrified. I remember talking with some guys at Graitec and they were shocked also.
After this... was Autodesk Structural Detailing story all over again. Just a product rename... and that was it. No more bug fixing.. no more new features.. they killed Advance Concrete in favor of Revit (which is one of the dumbest thing ever... because Revit was years behind Advance Concrete as a concrete detailing sw).
For concrete detailing we switched to Allplan. I think Allplan and Graphisoft's Arhicad have a bright future together.