Auto-suggest helps you quickly narrow down your search results by suggesting possible matches as you type.
Showing results for
Show only
|
Search instead for
Did you mean:
This page has been translated for your convenience with an automatic translation service. This is not an official translation and may contain errors and inaccurate translations. Autodesk does not warrant, either expressly or implied, the accuracy, reliability or completeness of the information translated by the machine translation service and will not be liable for damages or losses caused by the trust placed in the translation service.Translate
I'm a new CAD user trying to pick which CAD software to use, but because a large portion of my work involves software development, I use Linux on all my machines, and it's not feasible to switch. From what I've seen, Fusion 360 is the best tool for the kind of work I'll be doing, and I even purchased a few Windows licenses to run 360 inside VirtualBox. Unfortunately, nothing I've tried seems to eliminate the graphical glitches inside a VM which make it very difficult to use. After spending weeks and hundreds of dollars (Windows licenses) to try to get 360 to run decently inside a host Linux system, I've reached a wall and am about to give up and find a different solution. Dual-booting or maintaining separate machines is just too great an interruption to my workflow to be feasible.
I can understand why the port from the Windows version to the Mac version was costly and quite difficult, but once you already have a Mac version built, the port to Linux should be much simpler.
At this point, I can't wait around long enough for a Linux version to be built, but every day I'm sure you're losing additional people like me. I'm sure that someone has run the numbers and concluded that a Linux port wouldn't be cost-effective to make, but I hope it's worthwhile giving it a serious second look. It's easy to underestimate the number of additional users it could bring. Many users only ever learn one CAD program and stick with it, and as Linux is continually increasing in popularity, you're likely missing out on quite a few lifetime users. Additionally, as the feature gap between existing free/low-cost Linux CAD programs and Fusion 360 (slowly) closes, it's entirely possible that the more portable programs will start to displace Fusion 360 in environments where maintaining a dedicated Windows/Mac machine is an inconvenience: Although entire Linux-only environments may not be very common today, many people do use at least one Linux machine during the course of their work, and for many of those people it will be advantageous to switch to software that will work on all of their computers. For me, as a software developer myself, I'm much easier to add the features I need to something like FreeCAD than to fudge around with Windows VMs trying to get 360 to work.
I see a lot of people trying to get more platform independent and I wouldn't be the only one who would appreciate a Ubuntu compatible Autodesk Fusion 360.
I will certainly admit that there is a huge amount of extra work needed when trying to support multiple OS. That said, if you supported Linux (probably Ubuntu) a ton of people would ditch Microsoft and Apple. I've already switched to Fusion 360 from SolidWorks simply because of the CAM. However, I'm still stranded in windows because fusion 360 is windows based. Check out how crazy this is. I have a mac (because I like the hardware) that I run exclusively in Windows so that I can use autodesk fusion and at times solidworks (sheet metal). I have a VM on the machine for running Ubuntu because it's such a better programming environment.
I couldn't understand how those people are thinking?
For small businessmen, a 100$ cut for an operating system on machine is important. For big ones, just consider 1500$ on 15 machines. So if I have to worry only about the software I need instead of investing in operating system I would not think for a second to buy a software if its available on Linux.
No software company is considering a huge crowd they can gather just to by supporting Linux.
Yes, we are watching this thread. Our current Linux support plan is to ensure that the Jaguar Browser Client supports Webgl browsers and thereby those who prefer Linux.
This situation remains that we are not in a position to support linux with the installed client. This is is due to several factors, 3rd party components and QA effort to test across distributions and hardware are the primary obstacles. There is a project on our server side components to run on a very specific linux configuration. This is a precursor to us re-evaluating the costs on the client. This is not out of disrespect to the community it is simply a question of effort. We hope the browser client will fill this need once it is generally available.
Whops this is a case of Bad code names. Our Browser Client Code name is "Jaguar." Chrome/Chromium, Firefox would be browsers we would support on Linux.
Sad fact about the webgl based CAD's they can never replace a installed Client mainly because they will not be able to perform on a large assemblies and do alot of parallell calculation. Thats mainly because of the overhead the browser like Chromium would carry on the Gfx cards.
Unless if you did something like tmpvar (a node.js coder) did with livecad and parsed all models and vector data in the backend of a server application and passed it onto the client through websockets or RTC or similar.
That would come down to that your browser client would need to carry all the end users of a browser based product not only linux users.
You could prob. take the tour over doing C++ cross compiled into ASM.js or something like that, but it would still not be able to perform on large model sets and be suitable for simpler single objects and not large assemblies.
So basicly.. a toy.
But well thats just my 2 cents i guess.. after been doing webgl and node.js/asm.js programming for the past.. oh so 5 years as my main profession.
Besides WebGL only existing on the market as stable for the last 2-3 years or so..
Wouldn't it be possible to load the sprites and resources from a local source?
Also if content was cache defined on the server end the client would only need to download them once before getting a 304 no?
I'm not entering a pissing contest, I'm actually wondering if WebGL would allow it. Have never thought of it before.
And sure, WebGL wouldn't have direct access to some GL calls such as physics, but there is hardware acceleration support and it should be speedy enough to actually use Fusion for at least the very basic stuff until more development gets put into WebGL (One more major application would actually benify the whole open source process)
Again, no offense but I'd rather take a "a toy" over nothing at this point. It might bump stuff in the right direction.
This reason has been thrown out there before. But if I may ask: what does the Linux distribution have to do with anything? From a software development perspective, all distributions are pretty much the same, except for some superficial things like packaging. When developing for Linux, you don't target distributions; you target APIs like GTK and OpenGL.
My point is: If you release this for any popular distro, it will work as is for many other distros too, and the Linux community usually takes care of the rest, repackaging as needed.
(I have no idea how to quote stuff.. So a simple name drop will have to do)
I could answer the Linux part at least.
I've been a Linux developer for at least 10 years now, I've stoped counting at this point.
There is a huge difference between distros even tho many refuse to accept it.
But there's at least two branches, Rolling releases and LTS releases.
The two differs in that they might change underlaying core functions all of a sudden breaking a lot of non-modular products.
For instance, Arch, Gentoo and other rolling releases changed the file system a year or so back removing /usr "completely" which broke a few things (even their own package managers). When you're working with heavy applications (rendering etc) you might rely in certain kernel parameters to be present or you would even need to develop some basic drivers on your own.. These will become distro dependant.
Also if you try to package your application for user convenience, you're also down in the rabbit hole. I would argue that we here don't give two hoots about that and a simple .tar.gz would do... But some might argue here.
There's also the whole systemd/initd battle, you'll most likely create background services and sure, in this case that wouldn't be much hard work in creating something for both worlds.. But the point still stands, there are differences you'd might not think about until you get down there in the mud.. So sticking to one is better.
I would really not give two hoots about distribution packaging.. A tar ball and some library dependencies and even maybe a kernel diff could be used nd it would all get figure out quickly during compilation. If you target the debian basen distributions and maybe the Fedora one you would hit like 90% of all the desktop linux users that arent sys devs or tinkerers.
Regarding the webgl part, yes you could load sprites and texture from local storage no problem. There is even working game conversions into WebGL, make not tho, we are talking about Quake 1 and similar.. Its really hard to do proper memory management and heap inspection within a webrowser
And as is, fusion is struggling with assmeblies in the size of 4-6mb as it is. So taxing the system with extra load from a webbrowser with a diffuse to say the least integration into the OpenGL.
Making it work in WebGL is a pipe dream, and can easily be illustrated with the existing in browser cad alernatives out there today, its a joke.. would maybe be useful as a sketchingpad mainly if even that.
That said, using something like electron or other chromium based desktop runtimes with some minor tweaks and patches could prob. sustain a desktop client without the extra interference a full blown web browser would have. (Basicly removing the user customization in form of badly coded plugins, cache, and behind the scenes processes running amok checking your latest facebook updates and such.)
Yeah, Arch and Gentoo are bleeding edge distros. They're meant for tinkerers, and each of them fully expects their system to break now and then because of an update (especially in the case of Gentoo). I've used both myself, and it was fun while it lasted. I have recently switched to Mint for my main systems, which has greatly reduced my headaches. 🙂 In other words: They don't count. Their package trees are maintained mostly if not fully by their respective communities. If a software manufacturer wants to support Linux, they will target LTS.
You also mention the kernel. I know that some distros provide custom kernels, but that usually comes down to small tweaks. I've never heard of this breaking anything big.
The fact remains that Autodesk can just target a single distro, and the rest would (mostly) take care of itself.
I definitely have my doubts about the browser client. I would love to see it. Remind me how long have we been waiting for that again? I know I applied for the beta many times but gave up like 6 months ago after never hearing back.
As for the installed client, I think some of these concerns must show some lack of understanding of the system. Don't you guys have any interns or employees who are LInux geeks over there? If you can make something that can work on one distro I don't think it shouldn't be hard for you to make something that can work on any distro. But hey if you guys develop an internal distro that is say Ubuntu, Fedora or Suse based that had everything needed to make your software work till you got to the next phase I'd give it a go. That way you could control repos, updates and all that. LinuxCNC was it's own distro for the longest time. Not ideal but acceptable. Then they got it dialed in to where you can install it on just about anything with RTK. As has been mentioned many times. If F360 can run on Mac you shouldn't have too far to go to get it to run on most Nix systems.
