UAC has already seen quite some changes since it's initial introduction, and we can benefit from seeing all those changes and not start at the same place as where it originally started. (ask everything)dizt3mp3r wrote:I'm not convinced by this one line though I'd accept the rest of what you said.they came up with the best compromise they could between average user interaction and letting programs even do anything in the first place.
Should we just accept that we've all had negative experiences with Windows 'features' and that all that negative experience is relevant in the context. We all know that Microsoft could have done it better and that is why were are here? If ReactOS could improve on the implementation as well as replicate it then we would be in a better place...
UAC will be the outcome I am certain but with some minor tweaks to make it very slightly more relevant, I suppose that is what we should be considering rather than jousting over an opinion here or there.
Version compatibility
Moderator: Moderator Team
-
- Developer
- Posts: 246
- Joined: Fri Dec 19, 2014 10:00 pm
Re: Version compatibility
Re: Version compatibility
It's not as if one guy in a cubicle decided on UAC and implemented it in a weekend. They had several board meetings with seasoned software engineers deciding on and designing it over several months. They weighed the pros and cons and considered the alternatives. There isn't much you can do when you already have decades of software being used to doing whatever it wants.dizt3mp3r wrote:I'm not convinced by this one line though I'd accept the rest of what you said.they came up with the best compromise they could between average user interaction and letting programs even do anything in the first place.
Should we just accept that we've all had negative experiences with Windows 'features' and that all that negative experience is relevant in the context. We all know that Microsoft could have done it better and that is why were are here? If ReactOS could improve on the implementation as well as replicate it then we would be in a better place...
UAC will be the outcome I am certain but with some minor tweaks to make it very slightly more relevant, I suppose that is what we should be considering rather than jousting over an opinion here or there.
Microsoft is a wizard of software design, they hire people incredibly brilliant to develop their products. Do you know what makes them brilliant? It's not that they can throw away what they already have and make something "better". Any idiot can do that. What makes them brilliant is that they can work with the standards already given to them and work on them, keeping compatibility, not breaking anything, while making the necessary changes to allow the product to flourish. This is why Dave Cutler won a 2007 National Medal of Technology and Innovation award. That's why MS hired Dave Cutler to begin with. Nobody else could do what he did.
The fact, you can't secure Windows down any further without simply disallowing applications to run, or otherwise causing them to crash due to having no access to a resource they always expected to have access to. You could sandbox them, but what if they need to access another application? What about all the edge cases of making changes to the OS and core files? Do you think you can come up with a solution that would make all the existing software happy while not designing something that would be confusing and difficult to use, or just cause bugs? If you say yes, you should send in your application at MS, they would love to hear from someone even smarter than the guys they hired to design UAC.
That doesn't make UAC bad. It's quite good for what it does. But you can't expect it to be all or none. It's pretty much doing the best it can, anything else and things will break.
I'm honestly amused when people say they can do better than MS. Please, go on, tell us what you would do differently. You lose if you cause a breaking change. Comments like that remind me of github issues like this.
But, honestly, nobody ever comments past "well they could do better", just, kind of expecting the ReactOS team to develop the idea for them. I don't think you realize that it isn't the ReactOS team's job to design anything, they don't want to deviate from the NT design. Doing that would both require tons of designing and testing, and risk breaking applications, which would make the entire project moot, as the whole goal is to be completely compatible with Windows. Be compatible first, then maybe design a better replacement. At least if it fails, they always have the existing design which is proven to work.
Actually, forget my github issue comparison. People who think they could do better than MS just remind me of those teenagers on facebook who say "why don't all the scientists just cure ebola." Don't say that unless you actually have a proposal that won't break anything or make customers complain. Because that is not an easy task, actually that's probably Microsoft's most difficult job. It honestly insults me.
Who is online
Users browsing this forum: Bing [Bot] and 49 guests