Windows 7 soon to be no longer functional for 6.0

Software
63

I’m going to express a minority opinion here as well. I dislike that new versions will not work on older operating systems. Please note that I am not complaining about it, the programmers and developers can do as they like, and there is no basis for complaint unless you are either part of that effort or are putting up your own cash to fund it. I am simply offering an opinion and perspective that has not been expressed here yet.

I see the whole “upgrades” and “updates” thing as yet another planned obsolescence scam. It’s more of getting you to buy something new, to discard and replace an existing something that works perfectly well.

Built a new computer in 2006 when the motherboard died on my 486 machine, and was forced to discard windoze 98 and run XP. A few years back, two things happened at about the same time. AVG started demanding my email, and it became impossible to find another decent free antivirus that worked on XP. So XP became exponentially more risky for general online use, and that is now how people are being forced out of XP. I had also gotten a laptop that had come loaded with this windoze 10. Garbage. Mandatory, automatic “updates” that couldn’t be turned off. “Telemetry”. It didn’t take me long to wipe it and get windoze 7, which didn’t have any of that awfulness, and it was still possible to get a decent free antivirus that didn’t demand your email.

Linux? I had previously tried it, and still have it on another hard drive in the same computer that I run XP on (so I can boot into either OS). The concept behind linux is GREAT (free and open source, privacy and security built in). But it fails by not being user friendly. I’m not a programmer type, and everything in linux is command line gobbledegook. “Do” this, and “sudo” that. Yuck. And no firewire support! Having several pieces of firewire hardware, this is an essential feature to me. Sure, the developers see firewire support as a potential security issue, so at least make it so firewire support can be enabled or disabled at will, for those who want or don’t want it. That should have been a no-brainer from the start.

Unfortunately, this comment is ultimately the thread winner. Lack of support by decent free antivirus programs for the older operating systems is going to eventually force one to choose between the latest windoze privacy invasions and mandatory automatic updates, or the not user friendly linux. Being able to still run windoze 7 (or even 8) is only a stopgap measure that will soon end up going the way of XP and 98.

Because of this, and ONLY because of this, it doesn’t make much sense for further development of KiCAD to continue supporting the older operating systems. We will only eventually be forced out of them anyways. There is no logical reason to direct limited resources towards ensuring backwards compatibility with an OS that is about to fall victim to the planned obsolescence scam. In a perfect world, I’d be able to run the latest version on my XP machine, and backwards compatibility would always continue to be maintained. But the practical reality is unfortunately otherwise. I don’t like it, but I see it as a decision that is forced more by the hardware manufacturers and the antivirus makers (probably with a good amount of encouragement from bill gates), than by the developers of KiCAD.

64

Planned obsolescence is true in some sense, every commercial device and software must be abandoned in some phase because of commercial pressures and capitalism. Capitalism in some for is good but it has its downsides.

But it’s not a scam even if you don’t like it. Many companies would no doubt like to continue supporting their old customers, but it becomes a burden. It’s not practically possible to support several codebases or keep piling up know-how and equipment for all possible old device versions and models actively.

65

I always preferred the older business model of buying quality that was designed to last a lifetime. Only problem with that is repair parts eventually are no longer produced, and technology eventually makes it cheaper to replace something rather than repair it. And WHY that is a problem is because many times the new product does not do exactly what the old one did, performance is not as good, or some undesirable “feature” has been added.

Nikon has done an absolutely wonderful job of this! I absolutely love the superior build quality of the older manual focus lenses, and being able to use them with a modern digital camera. Best of both worlds. Brand loyalty well-earned.

66

You say that, but we don’t support Linux 2.6 era operating systems either. Is the Linux community running a planned obsolescence scam too?

Windows 7 absolutely has telemetry, they just didn’t document it nor provide a tool to let you see it. Under Windows 10 they wanted to comply with certain regulation among other things so it is documented and visible.

67

Yes and no.
Interchangeable camera lenses of quality last a lifetime or three and have an industrial standard connector to fit the camera body. By continuing to manufacture lenses with the standard connector, companies can sell their products to anyone with any brand of camera body that has the standard connector. If a company changes its connector, it will lose most of its market, hence, the standard remains.
Now the camera body is a different story…
How many bodies into which can you still stuff a roll of 35mm film?
Can you still buy 35mm film?
Have you tried to get the busted Charge Coupled Device replaced in your 10 year old Nikon camera body?
Have you tried getting replacement rechargeable battery packs or a replacement charger for that 10 year old digital camera body?

The usual answer is “sorry, we no longer service that model… go buy the new bigger, better, faster model”.

I didn’t think a great deal of your example, GoldenAge, but I do like your comments.
The utter waste of energy and resources with the world wide scam of planned obsolescence is killing this planet and electronics seem to be a major part of the problem.

Hmmm, I think I’ve had too many beers tonight; or maybe too few. Perhaps I should go and contemplate my navel :slightly_frowning_face:

PS.
I still have a protest button from my university days back in the early '70s.
It reads:

consume
be silent
die

1 Like
68

That is why in (I think) 2007 when buying digital camera body I selected Pentax - it is powered by 4x AA rechargable batteries. At first I had a problem that whenever it was needed the batteries were close to empty. But since Eneloops were introduced that problem disappeared.

69

Even car makers are only supposed to make spare parts available for 10 years, and this often does not happen in reality. A far more expensive purchase and huge materials/energy investment

70

Old software that is vulnerable to attacks does not “work well”. Cybersecurity is a never ending arms race. You lose the moment you stop.

2 Likes
71

At some point legacy = boat anchor to innovation and progress.

72

KiCad is a good example, like most (all?) of the long established PCB CAD packages, it started off with roots in the 16 bit DOS universe, with all of the limitations of integer sizes that entails, affecting working grids etc

73

The reason a lot of use still use Windows 7 is that our PNP machines don’t work on anything beyond Windows 7 – and that is never going to change until we get new PNP machines which is years and years away.

Having the CAD software on the machine that runs it is very useful if not necessary. And no, VMs don’t work because the software writes directly to the serial or parallel ports.

The argument that Windows 7 machines are easily compromised is nothing more than Microsoft FUD (not to mention stupid). I find it hard to believe electronics people know so little about how computers work. A decent hardware firewall and up-to-date antimalware software is perfectly capable of protecting old systems for decades to come. There’s nothing magical or mysterious about it – at least to those who know something about computers.

This really sounds like developer arrogance and/or laziness and they don’t know or care who their users are or need.

1 Like
74

Why in the world are you using the computer running the PNP machine for design work? Computers driving equipment should only be used for that equipment, especially if we are talking computers embedded into machines by the manufacturer.

You are more than free to yell at the python maintainers for dropping Windows 7 support which is the primary driver of us dropping Windows 7 support.

Developer laziness? Seriously? You realize we do this for free right? In our personal time.
You are more than free to download the source code and fix it for yourself

6 Likes
75

Anti malware is always reactive, new signatures after the malware is detected in the wild.
Your W7 machine had better not be fetching anything from outside that is executable, which includes pdf, Javascript,
Every time W10 gets a security update, the bad guys try to figure out what it was that got fixed and then see if it can be exploited on unpatched W10 and older abandoned OS like W7

1 Like
76

Professional users who need support for older systems can purchase support from KiCad’s dedicated service providers https://www.kicad.org/help/professional-support/

77

The issue here is pedantry on Pythons part, who decided to throw out some fairly minimal hacks to run on W7 simply because Microsoft declared W7 EOL. As far as I can tell, there was no pressing technical need for dropping W7. Discussion with the commits that would need to be rolled back is here: https://bugs.python.org/issue32592

IMO, it is wrong to drop an OS with >15% market share for some code cleanup, especially when that OS is Windows where users suffer proprietary builds for decades and Microsoft itself puts hacks into Windows to stay bug-compatible for 10-20 year old user software.
I also don’t think it is such an out-there use case to run KiCad on the PNP controller, just to verify something that is unclear in the production outputs.

Edit: My point here is that there is a different mindset around backwards-compatability on Windows, and Pythons decision is really going against that. Not taking any more tickets on W7 issues would have been completely understandable. Yanking support completely just because they have a 20 year old policy on the matter, irrespective of the real world market share is just petty. All this will accomplish is fragment Python further, with 3.8 becoming the new cutoff point for years of what can be supported in a slow moving corporate environment.

@mpi None of this is KiCads fault, and we can’t expect KiCad developers to maintain a W7 compatible fork of Python. Maybe you can convince Python developers to restore W7 support behind a compile switch. Once that exists, it shouldn’t be a problem to integrate that into KiCad.

1 Like
78

It’s funny you lamblast propertiary sfotawre here but Microsoft is actually enabling 10-20 year old software to work, without changes. On Linux land, sure you could probably make 10-20 year old software work…if you are an software engineer…and recompiled it with tweaks almost certainly required to match the latest distro environment.
But here we are, talking about an open source project you can compile yourself on Windows to fix any issues, just like on Linux :wink:

It’s unlikely any of us working on KiCad will set that compile switch. Our Windows build environment now uses vcpkg which is maintained by Microsoft.

End of day, it’s the same statement Python devs concluded with and I quote

If we had a dedicated maintainer who was supporting Win7 and making releases for it, then we (i.e. they) could support it. But then, there’s nothing to stop someone doing that already, and even to stop them charging money for it if they want (which they wouldn’t be able to do under the auspices of python-dev). So I suspect nobody is really that motivated :wink:

79

I came close to deleting the post but erred on the side of ‘freedom’. Maybe I was wrong? Nothing done HERE will change the status of W7 or the developers decision to drop support. Can we stick to what benefits the development and user experience of Kicad? Please? :wink:

80

I think leaving the thread is a good idea so that people can find the info and discussion if they need it, but I’m not sure further discussion will be that valuable.

81

Curiosity got the better of me, here is what I found: apparently all that is needed is a 3kb stub DLL called api-ms-win-core-path-l1-1-0.dll that can be easily found online but does not look redistributable. Then remove the check in https://gitlab.com/kicad/code/kicad/-/blob/master/libs/kiplatform/msw/app.cpp#L47
And surprisingly, it just works. I tried to trigger the PathCchCanonicalizeEx / PathCchCombineEx API calls that shouldn’t be available on W7 using various os.path. functions, but I couldn’t get it to crash.

image
image1237×495 43.9 KB

So whoever absolutely needs this, it doesn’t seem too hard.
Edit: @craftyjon I’d say there is an argument to allow bypassing the startup check via advanced config so this can be done without recompiling.

2 Likes
82

I wasn’t going to delete the thread. Just the one post. But, I think you may be right about further discussion. What’s done is done so may as well close the thread. If anyone thinks this is a mistake, PM me or any other mod with your reasoning.