SoylentNews is people
SoylentNews
Effective immediately, the new release of Ubuntu, 17.10, aka 'Artful Aardvark' has been released!
This release will be supported for 9 months (until 2018) for Long Term Support, stick with release 16.04, instead.
Official flavors (e.g. Kubuntu) are also released.
See the above release notes for a full list of changes and where you can get a copy.
[Full disclosure: the majority of SoylentNews' servers run Ubuntu 16.04 LTS though we have taken steps towards moving to Gentoo.]
Also:
The customized version of GNOME that Ubuntu 17.10 uses is very much in the mould of the (now defunct) Unity desktop, so it won't be to everyone's tastes.
This discussion has been archived.
No new comments can be posted.
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
Krogt, n. (chemical symbol: Kr):
The metallic silver coating found on fast-food game cards.
-- Rich Hall, "Sniglets"
(Score: 4, Insightful) by crafoo on Sunday October 22 2017, @12:07AM (16 children)
Linux doesn't run the vast majority of useful software for industry and engineering. We are talking about massively complex systems. And no, it's not the fault of the people writing these pieces of software, for the most part. Linux is a massive pile of shit to support a binary distribution on. Part of this is by design - breaking ABIs and a dumb-as-fuck dynamic linking system. So Linux gets garage-tier CAD/CAM/Simulation/Analysis software. I'm not against the free software movement, in fact I agree in principal. But there are very real reasons why Linux won't be taken seriously for anything but coding IDEs anytime soon.
(Score: 2) by rylyeh on Sunday October 22 2017, @02:01AM (1 child)
Bullshit. Those developers want the proprietary lock-in black box they get from windows.
They see 'other' OS's as a risky waste of time and money.
"a vast crenulate shell wherein rode the grey and awful form of primal Nodens, Lord of the Great Abyss."
(Score: 2) by mhajicek on Sunday October 22 2017, @05:58AM
They don't want to spend millions of dollars on cross platform development and support for a tiny niche market. There's a bit more to it than clicking "compile for Linux."
The spacelike surfaces of time foliations can have a cusp at the surface of discontinuity. - P. Hajicek
(Score: 0) by Anonymous Coward on Sunday October 22 2017, @02:40AM (10 children)
The moment before the vendor clicked "Compile", it wasn't a binary.
What's keeping them from selecting "Compile as a Linux binary"?
As rylyeh notes, this isn't a choice made to expand the company's user base--which your typical company would like to do.
This choice is about lock-in.
...and the idea of giving money to lazy people doesn't excite me.
Apparently, there are some who feel that subsidizing sloth is something other than a negative incentive.
-- OriginalOwner_ [soylentnews.org]
(Score: 2) by mhajicek on Sunday October 22 2017, @05:54AM (9 children)
If cross platform graphics support is so easy, why wouldn't everyone do it? The fact is that you can't just recompile Windows software to run on Linux. Do you know anything about how professional grade software is developed? Especially software that's been continually added onto since 1980, and still contains code that was written for DOS? They would have to start over from scratch writing with cross platform in mind, which is not something they'll spend tens of millions of dollars doing just to appease a tiny niche market.
The spacelike surfaces of time foliations can have a cusp at the surface of discontinuity. - P. Hajicek
(Score: 0) by Anonymous Coward on Sunday October 22 2017, @08:23PM (6 children)
Now, see.
You started from "Windoze software" and I started from "software".
cross platform graphics support
Games have already been mentioned.
Your apps are more graphics-intensive than games??
I'm not buying that for 1 second.
Again: This is about sloth and lock-in.
Your excusing your lazy, greedy vendors shows how easily you allow yourself to be led around by the nose.
Again: I doubt that you have even broached the subject with your vendors.
-- OriginalOwner_ [soylentnews.org]
(Score: 2) by mhajicek on Sunday October 22 2017, @08:39PM
I have asked, and there is no vendor of professional grade five axis CAM who is willing to consider developing for Linux. Not one. Why is this my fault again?
The spacelike surfaces of time foliations can have a cusp at the surface of discontinuity. - P. Hajicek
(Score: 2) by mhajicek on Sunday October 22 2017, @08:58PM (4 children)
I forgot to address your other points. Yes it is more graphics intensive, just in a different way. It's not worried about fancy visual effects, but representing a model which may be several feet across with a facet every .002" or less is pretty intensive. And all the software that currently exists in the field was written to use the Windows API, with the exception of the older DOS code that they're still using. Unless someone is willing to pony up a few tens of millions to start over from scratch this is what there is.
The spacelike surfaces of time foliations can have a cusp at the surface of discontinuity. - P. Hajicek
(Score: 0) by Anonymous Coward on Sunday October 22 2017, @10:09PM (3 children)
If you're trying to impress me, you're failing.
Going in a different direction, but on a related theme, I remember DJ Delorie (a developer of the gEDA suite) mentioning how for laughs he had demonstrated that, by tweaking a parameter in the setup file, the software would (assuming that you have enough RAM/virtual memory) lay out a printed circuit board that was 64 feet by 64 feet.
There's nothing special about software for which you have to pay a license fee again every year.
I remember Terry Porter regularly posting a link to photos of his latest FOSS-created PWBs at sci.electronics.design back around the turn of the century.
FOSS has been making money for pros for a long long time while preserving their freedom.
-- OriginalOwner_ [soylentnews.org]
(Score: 2) by mhajicek on Monday October 23 2017, @08:50AM (2 children)
There's absolutely nothing special about it, except that there are zero viable alternatives in the real world. If it's so easy, why is the best free CAM only passible for 3 axis hobby use? Why can't it be on par with the software that's had hundreds of millions of dollars poured into it over almost fourty years of continuous development and improvement?
The spacelike surfaces of time foliations can have a cusp at the surface of discontinuity. - P. Hajicek
(Score: 0) by Anonymous Coward on Monday October 23 2017, @09:39PM (1 child)
I've pointed to other folks with needs that they filled from within their own ranks.
Someone saw a lack of Free(dom) Software for a particular set of tasks and set about filling the void.
(In the FOSS world, we call it "scratching an itch".)
At $60k per individual, how many of you in that field would it take to commission a cross-platform FOSS app?
The narrowness of your vision is disappointing.
Techies are supposed to be problem solvers and you guys instead just bend over and take it.
...and in my field, as alluded to, several proprietary products have had changes to their licenses.
Others have been bought by a competitor and abandoned, crippled, made spyware, and/or increased in price.
Keep your fingers crossed that none of that happens to your single-platform proprietary apps and you then have to start spending buckets of money all over again and learn an entirely new UI.
-- OriginalOwner_ [soylentnews.org]
(Score: 2) by mhajicek on Monday October 23 2017, @10:41PM
My business is machining parts, not writing software. You're proposing that instead of buying software that I can use right away, I set aside all of that capital for a decade while someone tries to remake something equivalent? That's horrible business sense. How about this: your car breaks and you need an immediate replacement. Take the money that you could use to buy said replacement, and instead give it to a group of people who may or may not develop a usable replacement in a decade. In the mean time you walk.
No. If Linux wants to replace Windows it will need to run software that was written for Windows. It's this "No, you come to me" attitude that's keeping Linux from succeeding.
The spacelike surfaces of time foliations can have a cusp at the surface of discontinuity. - P. Hajicek
(Score: 2) by Whoever on Monday October 23 2017, @03:48AM (1 child)
If you start with a Windows-only codebase, yes.
If you start and use something like Qt, then porting to Linux becomes easy.
(Score: 2) by mhajicek on Monday October 23 2017, @08:40AM
That's the thing. ALL PROFESSIONAL GRADE 5 AXIS CAM IS WINDOWS CODE BASE!!! All of it. All. Of. It.
The spacelike surfaces of time foliations can have a cusp at the surface of discontinuity. - P. Hajicek
(Score: 0) by Anonymous Coward on Sunday October 22 2017, @12:21PM (1 child)
Bullshit. If you are making a binary-only distribution, you just have to ship all your libraries with your binary and set your library load path accordingly. That's already the standard Windows way of doing things for the most part.
(Score: 0) by Anonymous Coward on Sunday October 22 2017, @08:48PM
There's fat binaries as well: Compile in all the dependencies.
This makes it possible to e.g. run very old apps under newer versions of Linux.
It also gets rid of the .deb|.rpm|many distros "problem".
(...and newer packaging paradigms have already made mention of this an anachronism.)
You can also put the fat-binary app on a thumbdrive and take it with you from Linux box to Linux box.
This sort of thing is done for Windoze-compatible (FOSS) apps over at PortableApps.
Those apps don't require installing DLLs or modifying the Windoze Registry.
Again: Having the source code available makes this possible for those devs.
Again: Closed-source proprietary apps always come in last when it comes to freedom for the user.
People who hold up their closed-source single-platform apps as some kind of shining example are the kind of people who would point to the chains their slave masters put on them and call that jewelry.
-- OriginalOwner_ [soylentnews.org]
(Score: 0) by Anonymous Coward on Monday October 23 2017, @01:26PM
I think you're grossly overstating the difficulty of writing cross-platform software that runs on Linux. Considering the fact that a loosely affiliated group of volunteers has tens of thousands of working software packages that build from source on Linux, and most of the fucking internet runs off it, I would say supporting it isn't especially difficult.
The real problem is simply legacy code. If some company made the completely reasonable decision to use Visual Studio 6 or similar to start writing their colossal piece of software on Windows, and fifteen years later it's a colossal Frankenstein of unbelievably complicated code, then there is simply no business case to port it. Most companies still use Microsoft Windows on the desktop. No sane Chief Financial Officer is going to greenlight a project to do a clean rewrite in cross-platform tools when the Return On Investment (in the form of new customers who will run the new application from Linux) is negative. And to be clear, I'm not criticizing companies, Microsoft, or Visual Studio for having huge and complicated legacy apps. If they had started out writing the software for OS/2, OS 9, OS X, or Red Hat Enterprise Linux 1 the codebase would be every bit as enormous and complex today and a clean cross-platform rewrite would be every bit as impractical and financially unsound.
Your industry is stuck on Windows for legacy reasons. It's not good or bad, it just is. The best I can do as an FSF member is support the development of fully free software alternatives to the software you're using. But it's awfully difficult for the free software community to put together something that is probably quite literally almost an order of magnitude more complicated than the Chromium project.