older versions of NT as a target for development?
Moderator: Moderator Team
Re: older versions of NT as a target for development?
i have to agree with someguy there are still 16bit games that ran on windows 9x very well. like doom 1/2 it would be nice to be able to play some retro games on reactos. besides the last time i checked windows xp still had 16bit code in it.
Re: older versions of NT as a target for development?
And not having built in support for 16bit applications was one of the places the project explicitly decided to break with in terms of compatibility with Server 2k3. By their very nature, these days basically the only secure way to run such applications on a networked machine is to do so inside a VM and firewall the crap out of it. There's literally no concept of multi-user security or even security in general on the platforms these applications were run in, so compromising the machine is almost trivially easy.
Re: older versions of NT as a target for development?
running in vm wire isn't the same as just running it in the os nativly. if there no 16bit code ros is not complete.
Re: older versions of NT as a target for development?
im not the only one who want 16bit to die in "future" software
¿Why Linux is moving to x64?
¿Why Microsoft is moving to x64?
¿Why ARM and x64 bits processors are being used in recent devices?
¿Why even they exist?
¿Why Linux is moving to x64?
¿Why Microsoft is moving to x64?
¿Why ARM and x64 bits processors are being used in recent devices?
¿Why even they exist?
Re: older versions of NT as a target for development?
In NT based Windows the 16-bit subsystem IS running in a VM. The main security problem is that multiple applications running inside the same VDM can clobber each other. The VMD should have the same permissions to the underlying OS as any other process. You will have the same problem running any multiple programs in a single VM if the guest OS isn't "secure".
But I do recall there were some bad OS hacks to make the VDM run at acceptable speeds on machines in the NT 3.1 days, and I doubt that architecture has been enhanced much since then.
A bigger problem for ReactOS is that the only way to be compatible with Windows 3.1 is essentially to BE Windows 3.1. Windows NT, WABI, OS/2, DESQView X, and even early versions of Wine worked by using actual Windows 3.1 in a VDM with some core files replaced or modified.
Since the 64 bit versions of Windows omit the 16-bit VDM software it seems like a reasonable thing to omit going forwards. But if anyone is insane enough to write one they are welcome to do so!
Oh, right the real purpose is to get everyone to purchase the same applications again, just like Apple does.
But I do recall there were some bad OS hacks to make the VDM run at acceptable speeds on machines in the NT 3.1 days, and I doubt that architecture has been enhanced much since then.
A bigger problem for ReactOS is that the only way to be compatible with Windows 3.1 is essentially to BE Windows 3.1. Windows NT, WABI, OS/2, DESQView X, and even early versions of Wine worked by using actual Windows 3.1 in a VDM with some core files replaced or modified.
Since the 64 bit versions of Windows omit the 16-bit VDM software it seems like a reasonable thing to omit going forwards. But if anyone is insane enough to write one they are welcome to do so!
Good question, the only advantage to 64-bit is that it can access more memory - and 32-bit Linux doesn't even have that problem. There will always be people who need an infinite amount of ram, but unless your are doing molecular modeling of a nuclear explosion, running an enterprise database, or playing a video game that is more realistic than going outside, then you are doing something horribly wrong if you need more than 4 gigabytes of RAM.jonaspm wrote:¿Why Linux is moving to x64?
¿Why Microsoft is moving to x64?
Oh, right the real purpose is to get everyone to purchase the same applications again, just like Apple does.
Re: older versions of NT as a target for development?
are there vms for 16bit? if they are easy to install it would not be a problem. however i agree that security is more important than 16bit.
Re: older versions of NT as a target for development?
It's not that hard to reach 4 GB of used RAM today. If I want to play Minecraft (haha, such lousy graphics! ) with my friends, I start OS, Minecraft server, Minecraft client, Chrome (to look up any info needed) and use 4-5 GB RAM without much hassle. I agree that normal people won't use that much, though.SomeGuy wrote:There will always be people who need an infinite amount of ram, but unless your are doing molecular modeling of a nuclear explosion, running an enterprise database, or playing a video game that is more realistic than going outside, then you are doing something horribly wrong if you need more than 4 gigabytes of RAM.
Re: older versions of NT as a target for development?
Minecraft uses a lot of RAM due to unoptimized handling of game data, plus its JAVA (guess, Jira has the same problem being a JAVA app as well). As for graphics, its worthy to look around:It's not that hard to reach 4 GB of used RAM today. If I want to play Minecraft (haha, such lousy graphics! :-) ) with my friends, I start OS, Minecraft server, Minecraft client, Chrome (to look up any info needed) and use 4-5 GB RAM without much hassle. I agree that normal people won't use that much, though.
http://www.minecraftforum.net/topic/127 ... e-8192012/
http://www.minecraftforum.net/topic/120 ... -and-more/
Re: older versions of NT as a target for development?
I have even better looking shader mod applied to MC than those presented - my comment to graphics was only a joke, sorry about it not being explicit enough. However, there is no up-to-date alternative client (I saw some C/C++ ones some time ago, but not finished) that would have lower RAM requirements. Anyway, I only presented it as an real-world example that is not so unlikely to happen, I have a definite opinion about Java's memory footprint.
Re: older versions of NT as a target for development?
Man! I love it xDPlaying a video game that is more realistic than going outside
Re: older versions of NT as a target for development?
Don't confuse 64bit versus the actual x86-64 ISA. x86-64 was created to allow x86 to address more memory, true, but it has plenty of additional changes that make it far superior to the original x86 ISA/architecture.
And with respect to support for 16bit applications, no, nostalgia does not constitute a technical reason for why supporting it is important.
And with respect to support for 16bit applications, no, nostalgia does not constitute a technical reason for why supporting it is important.
Re: older versions of NT as a target for development?
nostalgia is a good reason to support 16bit code.And with respect to support for 16bit applications, no, nostalgia does not constitute a technical reason for why supporting it is important
Re: older versions of NT as a target for development?
I don't know about virtual machines, but DosBox (a free x86 application) is a 16-bit emulator. I've used it a fair bit for nostalgic gaming. You can even install Windows 3.1 in it.andreas84 wrote:are there vms for 16bit? if they are easy to install it would not be a problem. however i agree that security is more important than 16bit.
You'd sacrifice security and a TON of ROS developer time just to play old DOS games? I wouldn't. Just get DosBox and call it a day.Dave3434 wrote:nostalgia is a good reason to support 16bit code.
Re: older versions of NT as a target for development?
can i take the 16 bit code in w98 install disc i have and just slip into ros? and i can also install dosbox if it does the job?
Re: older versions of NT as a target for development?
btw doesn't the pentium 2 and early pentium 3 rely on 16bit code still just wondering?
Who is online
Users browsing this forum: No registered users and 18 guests