Development Direction

Here you can discuss ReactOS related topics.

Moderator: Moderator Team

Post Reply
Walt
Posts: 42
Joined: Tue Sep 04, 2018 1:39 am

Development Direction

Post by Walt »

I've used Windows since 3.1 and still have that machine. Another machine runs a '98 system. I spend perhaps 6 hours daily using XP and in the last week, maybe 30 hours getting ROS 0.4.11 up on a Dell D610 laptop after trying and failing to make it work on a (desktop) Dimension 3000. I spent several hours reading various forums here. In my working years I did IBM System 360/370 development, mostly in assembler language.

ROS is impressive. Yeah a lot of stuff doesn't work but an awful lot does, and what's visible is very nice indeed. That said, I want to offer a suggestion: Focus urgently on the potential user base for an XP replacement and on getting those potential customers using it at least part time as soon as possible.

That potential ROS customer base is going away and once they're gone they won't come back. Only those who either now use XP or who recently did so would be likely to run ROS. Once they're on Win 8 for a couple of years, nearly all of them will move next to Win 10 regardless of how great ROS is at that time.

Furthermore ROS needs XP-like drivers. Where will those come from when all the machines out there are two or three generations post-XP? Ten years from now almost nobody will have a machine that was ever supplied with XP installed and for which drivers are thus available.

XP has about 5% of the U.S. computer OS market now I believe. I think it's much bigger abroad but the trend there too is to later systems for most of the same reasons that apply here.

ROS is wonderful but it's not practical for anything as of today. If it's going to take the place it deserves that has to change and change quickly.

I suggest focusing on two things: Expand the base of machines on which it can be installed, focusing on increasing the fraction of real hardware installs that go 'right the first time.' And aim to get it on the Internet 'out of the box' with at least Firefox 52 Esr.

The first is because very few people have several spare computers or will go find one that might be a good bet. (And in fact my failed Dimension 3000 was a good bet -- it's in the list of 'I got ROS running on ...' machines.) When someone goes to the trouble to install ROS it should come up right now on virtually all of the most popular machines.

A competent browser is urgent because that's the most common application: Nearly everyone using a real computer uses a browser regularly. Sure other applications are important but a machine without a browser that will render nearly all web sites decently is a toy and there's not time to do what Mozilla has done from scratch.

A working browser requires working internet. Ethernet after you find and install the driver for your NIC is barely a short term solution -- okay for a year perhaps but WiFi needs to come along promptly.

Third on the list would USB support.

Focusing on these goals would require pruning other efforts and I understand the issue of volunteer software development. But if it is mainly a hobby for the next five years or so, then it's unlikely ever to be more than that.

ROS has a 20 or 22 year history depending on who you read. That amount of effort and the slick design that's visible now deserves real success -- not just that a bunch of people had fun doing it.

No doubt these thoughts rub some the wrong way but that's how it looks to this ROS newbie.
User avatar
dizt3mp3r
Posts: 1874
Joined: Mon Jun 14, 2010 5:54 pm

Re: Development Direction

Post by dizt3mp3r »

That stuff is all obvious - you are simply stating the obvious. You want ReactOS to work and be feature complete.

ReactOS is not ready to use, it is not feature complete. We all know that. It is in Alpha. It won't be ready to use for ages yet. Everything you ask for will be there but not yet. Certainly not now.

Your advice can only rebound back on yourself and others like you, people that want ReactOS to be something now. Your recommendations can serve only to inspire you and others like you to DO something to contribute to ReactOS through code, documentation or a donation, or something. Don't complain because that is what you are really doing - instead DO.
Skillset: VMS,DOS,Windows Sysadmin from 1985, fault-tolerance, VaxCluster, Alpha,Sparc. DCL,QB,VBDOS- VB6,.NET, PHP,NODE.JS, Graphic Design, Project Manager, CMS, Quad Electronics. classic cars & m'bikes. Artist in water & oils. Historian.
oldman
Posts: 1179
Joined: Sun Dec 20, 2009 1:23 pm

Re: Development Direction

Post by oldman »

You are unkind dizt3mp3r to Walt. He is not making demands, just putting forth his views, observations (a lot of which are very complimentary) and making some suggestions. I for one enjoyed reading it, but not your aggressive reply.

I do hope that you come along and edit your post. There is a need to encourage people, not discourage them. I know that some come here and make demands and deserve to be told off, but that is not the case with Walt's post.
Please keep the Windows classic 9x/2000 look and feel.
The layman's guides - debugging - bug reporting - compiling - ISO remaster.
They may help you with a problem, so do have a look at them.
Zombiedeth
Posts: 119
Joined: Fri Jun 27, 2014 9:01 am

Re: Development Direction

Post by Zombiedeth »

I think many people would like ReactOS to be a drop in replacement for windows right now or even 2 years from now when windows 7 reaches end of life.
Even if by some chance ReactOS could reach beta by then it still would not be in a position to be used by the masses. Maybe more people could use it in niche scenarios in a VM or on older hardware. But even with that all being the case i know there are many talented programmers working on ReactOS that will continue to embrace and extend the good things about windows. Many technologies that microsoft could have supported will find there way into ReactOS and all that is enough for me to help where i can. I enjoy watching the developments in ReactOS and i think i will continue to. And some day i think i will get to use ReactOS as a mainstream OS. :)
User avatar
dizt3mp3r
Posts: 1874
Joined: Mon Jun 14, 2010 5:54 pm

Re: Development Direction

Post by dizt3mp3r »

oldman wrote: Thu Sep 20, 2018 1:53 pm You are unkind dizt3mp3r to Walt.
No, I wasn't unkind, I just told the truth. So many come here with observations about what the team should do and those comments are really implied criticism through dissatisfaction with the team's progress and what it has achieved.

What most don't realise is the team is YOU and I and includes the OP if he chooses to be a part of it.

The posters here generally think there is a team that co-ordinates, plans and works to a schedule that can be influenced. In reality, my perception over the years is that all the developers simply work on what inspires them personally. So, the suggestion is if you want something to happen, make it happen by doing it (if you have the skill) or donate or contribute in some small alternative way (if I could code in C++ competently I would be doing more).
Skillset: VMS,DOS,Windows Sysadmin from 1985, fault-tolerance, VaxCluster, Alpha,Sparc. DCL,QB,VBDOS- VB6,.NET, PHP,NODE.JS, Graphic Design, Project Manager, CMS, Quad Electronics. classic cars & m'bikes. Artist in water & oils. Historian.
PurpleGurl
Posts: 1790
Joined: Fri Aug 07, 2009 5:11 am
Location: USA

Re: Development Direction

Post by PurpleGurl »

Walt wrote: Thu Sep 20, 2018 6:11 am ...

That potential ROS customer base is going away and once they're gone they won't come back. Only those who either now use XP or who recently did so would be likely to run ROS. Once they're on Win 8 for a couple of years, nearly all of them will move next to Win 10 regardless of how great ROS is at that time.

Furthermore ROS needs XP-like drivers. Where will those come from when all the machines out there are two or three generations post-XP? Ten years from now almost nobody will have a machine that was ever supplied with XP installed and for which drivers are thus available.

...

ROS is wonderful but it's not practical for anything as of today. If it's going to take the place it deserves that has to change and change quickly.
Well, nothing you said was particularly offensive. It's quite obvious, if anything, and most of this is already being attempted to the best of our abilities. If you can help close up the gaps, that would be appreciated.

Loyal XP users don't really go away. They just keep using XP, even if that limits their hardware. So if they find us, then they might migrate if we give a better solution than what they currently have. That has to do more with them, not us. If XP is working for them, they won't change unless we can provide better.

ROS is designed to use real XP/2003 drivers, though that needs more work to be more stable. Yes, I agree that newer XP drivers would come in handy. However, that is not really ReactOS' job. I mean, if someone here wants to code a driver for something, and they know how, nobody is stopping them. It would be nice if ROS would run well on Core machines and have XP/2K drivers for them. Now, using the Microsoft model, beyond default drivers, drivers are the responsibility of the hardware manufacturers. However, that is not an option since most manufacturers have moved beyond XP and even 7 in a lot of cases.

Intel has went a step further and said that they now take the OS design into account when they design their chips, and thus Core i9 is designed for Windows 10 (and maybe somewhat higher) only. They claim that even if there were drivers for older OSs, it might still be unstable. I don't see why that has to be the case, though making stable drivers for older versions might be tricky, though I imagine not impossible. That could be a worthwhile project, even if someone also has to write a custom HAL (if the problem is the BIOS/EFI, then it seems replacement routines could be put in RAM, with some mappings changed).

Now, if you want XP/2K drivers for newer hardware, it would be nice if you'd be able to help code those. Or you can wait until the design target changes. Thus you might like it better if/when Vista/7 support is available. If someone is impatient for changing the kernel target, they could fork it now, so long as they follow the licensing terms and stay "clean" in their development. Hopefully they will not just put out another vanity junk release where they only change the names and graphics without making any actual progress.
Ancient
Posts: 82
Joined: Tue Mar 27, 2018 11:32 pm

Re: Development Direction

Post by Ancient »

Walt,

We may have similar backgrounds / experience. The developers here are convinced they need to implement the 32 bit version of Windows first, on a single CPU within 4 GB of RAM. This isn't unreasonable, but it taking a long time.

As an experiment, I tried to install and run RoS on a USB drive without success. I then tried a Linux Mint system. The Mint installed, and was able to read NTFS without issue. Installing the Proton software from Steam / Valve resulted in an expanded WINE comparability. Mint has a few Windows themes, which can help make it look similar to Windows.

DX11 (DRM generally) is an issue for WINE and Proton (the Steam variant of WINE), also some programs that require root kit like functions (Black Desert Online) are a problem, and should be a problem for any Linux, but even for Windows.

The developers here must know their window is closing, but shouldn't feel they've failed. If USB support, and X86 support, with multi processors can't be completed in the next few years, there will be no significant use for this product as things like Mint / Proton will have surpassed most any goal this product will have. Also AMD is working better with the Linux community. This helps a lot with things like drivers, particularly video drivers.

Goals here are admirable, but it is tough get a functional version out in a reasonable time. I feel sorry for these folks, they are working so hard for so long. Lets hope they can complete a functional X86 system in a reasonable time, which supports most single core capability. Then lets hope adding muti core, better plug and play, USB boot, and whatever don't take forever. Currently nobody runs a new single core less than 4 GB RAM PC. Most are at least dual or quad core (or effective core) computers, maybe at the smallest end with 2 GB, but usually with a minimum of 4 GB of RAM. These would be the $100 to $200 SoC PC's at Amazon, EBay and elsewhere. Which come with Windows 10 Home installed.

In 2 years, it is likely 8 effective core 8 to 16 GB RAM systems will be default. Heck 500 GB drives which cost nearly $500 2 years ago are a lot less than $100 today. The hardware is moving faster with AMD and Zen. It was slower with only Intel driving development. Somehow RoS will have to not only attain it's immediate goal, but must also easily boot off USB (maybe USB C) and support multiple cores. Lets hope RoS is around, or the only alternative will always be Linux with WINE.
Walt
Posts: 42
Joined: Tue Sep 04, 2018 1:39 am

Re: Development Direction

Post by Walt »

Wow, what a lot of great comments. First, I hope it's clear how much I admire what's been done. If I had a spare million dollars I'd land it on this project. Unfortunately my wife and I are several zeroes short of that. I'll make a small donation in a couple of months when we can afford it, but it won't be what the project deserves. More of 'deserves' later.

Ditz3mp3r:
That stuff is all obvious - you are simply stating the obvious. You want ReactOS to work and be feature complete.
Not exactly. I want it to be USABLE AS SOON AS POSSIBLE for practical work by people who don't mind a few rough edges. Rough edges are things like you cannot set folder options so you can see system files, 'Uninstall' isn't usable from the Application Manager, the boxes to click to minimize or exit a window aren't displayed (though they work), and your icons will be randomly rearranged. At this stage such things are to be expected. But in the ~five hours use since I got 0.4.11 installed on the Dell D610 I've had to REinstall it twice because of different unexpected bad events. And before that I spent about 30 hours trying to install it on a Dell Dimension 3000 -- which is listed as 'hardware that has been found to work.' This seemed to be disk error that prevented the second stage of install from working -- but when I put the same disk (having installed from a boot CD on the 3000) in the D610, it took right off.

It SEEMS very slow but I don't have it on an SSD as I do my 'daily use' systems; that's on order and I'll be able to make a real comparison in a week or so.

The 'rough edges' wouldn't prevent someone interested in the ROS idea from using it regularly for browsing or perhaps running a favorite game. The crashes (so reinstall, yes, including the needed supporting files, your browser, its add-ons, bookmarks ...) filter out all but the most fanatical and we are not a viable client base.

I'm talking ONLY about priorities -- NOT about some expectation that it run as well by next month (or year) as XP SP3 does now, 15+ years on.

If ROS has enough installs, the customer base will pull it along -- more developers would be likely to appear, maybe even more money. As a year after year 'maybe someday' people will go where they must -- Win 10, Linux, Chrome ...
Your advice can only rebound back on yourself and others like you, people that want ReactOS to be something now. Your recommendations can serve only to inspire you and others like you to DO something to contribute to ReactOS through code, documentation or a donation, or something. Don't complain because that is what you are really doing - instead DO.
The critical path is the coding and debugging. Money, of course, would allow buying more of that. Unfortunately VERY few of the people who need or admire ROS can do much to contribute in either department. I can do some lower level things -- there are a couple of articles that seem to be needed here that I can draft -- and I will. I intend running a new build every week or so on a real machine and using it for whatever it will do; perhaps I can make inputs to the debugging process once I figure out how that works.

Looking at this as a total outsider -- a month ago I'd never heard of it -- and commenting on priorities is an immediate effort to contribute.

Oldman:
You are unkind dizt3mp3r to Walt. He is not making demands ...
I didn't adequately thank you, Oldman, for all the help on the Dell 3000 part of my install efforts. So -- THANKS!

I don't consider dizt3mp3r's comment particularly unkind: He sees the issue from another side and expressed his views. And because he did, I am able to -- hopefully -- clarify mine.

Zombiedeth:
I think many people would like ReactOS to be a drop in replacement for windows right now or even 2 years from now
Sure. But most of us probably live in the real world.

What I would hate to see is the project fading to black from lack of people who see it as valuable and machines on which it will run. Say 5-10 years out.

dizt3mp3r:
The posters here generally think there is a team that co-ordinates, plans and works to a schedule that can be influenced. In reality, my perception over the years is that all the developers simply work on what inspires them personally. So, the suggestion is if you want something to happen, make it happen by doing it (if you have the skill) or donate or contribute in some small alternative way (if I could code in C++ competently I would be doing more).
I've done programming and unless there's a deadline and a boss who can fire you if you're not helping get there, what you say is correct. However, most volunteer programmers want to see their work have an impact in the real world and focusing now on fixing things that make practical use essentially impossible (rather than on feature completeness) would help get ROS there.

I've toyed with the idea of learning C++ and the necessary MS OS basics just to get in the action. But at age 80 in the life I've chosen, no. Too much sitting down, pretty soon you can't stand up. We don't even OWN a TV.

PurpleGurl:
[What you said is] quite obvious, if anything, and most of this is already being attempted to the best of our abilities.
Well ... having now actually RUN ROS for a few hours I think this IS already being attempted. Were I writing my original comment today I'd state it as a need for more emphasis (on expanding the market as fast as possible) rather than a change of direction. There's probably a font or smiley for 'slightly embarrassed' and if so, please assume it.

That said:
ROS is designed to use real XP/2003 drivers, though that needs more work to be more stable. Yes, I agree that newer XP drivers would come in handy. However, that is not really ReactOS' job. I mean, if someone here wants to code a driver for something, and they know how, nobody is stopping them. It would be nice if ROS would run well on Core machines and have XP/2K drivers for them. Now, using the Microsoft model, beyond default drivers, drivers are the responsibility of the hardware manufacturers. However, that is not an option since most manufacturers have moved beyond XP and even 7 in a lot of cases.
And:
... you might like it better if/when Vista/7 support is available.
And Ancient:
If USB support, and X86 support, with multi processors can't be completed in the next few years, there will be no significant use for this product as things like Mint / Proton will have surpassed most any goal this product will have. Also AMD is working better with the Linux community. This helps a lot with things like drivers ...
These are important development direction issues. A strict XP replacement would only run on machines that originally ran XP unless/until someone else writes drivers for later machines, ditto for device support. But ROS can hardly afford to be that strict, because in ten years single-processor machines will be only a memory for those of us who are still alive. Similarly FAT32 was a good place to start but with HDD (and soon SSD) moving past 1 TB NTFS will become non-negotiable. And even the market for computer OS's that can be user installed and upgraded is going downhill as the younger generations adopt mobile devices and packaged solutions with a manufacturer-controlled supply of 'apps.'

At the same time the plan of having ROS progress from an XP replacement to a Vista/7 replacement is probably impossible in the circumstances. Maybe ROS needs the main added functions, but the totality, no.

There are some important positives:

1. There's a substantial base of XP die-hards. It's small (5%) in the U.S. but MASSIVE world wide -- Did I hear 50% just a couple of years ago? Do a search for an XP error message and many of the hits will be by people for whom English is a second language. In total numbers I'm sure its still growing: MS doesn't supply it, but pirates do.

2. Linux has its own set of disadvantages coming from its history in universities. Even choosing a version to run is painful. Then you trash all the programs you have to solve YOUR problems and start at the beginning to learn the new problems and discover the solutions. And what of obsolete hardware and software that does your specific job -- but only runs on Windows? Because there's no manufacturer behind it the support for Linux isn't great. The larger open source projects -- Mozilla for example -- support it but there's nothing like the richness of the Microsoft world.

3. Manufacturer-controlled computing is moving in user-unfriendly directions, increasingly in monopoly and GOVERNMENT controlled directions. If the politics of the last ten years doesn't make you wary of government control -- I probably should not mention specifics -- I don't know what it would take.

4. Among the manufacturer controlled unfriendliness-es is the accelerating movement of everything into the cloud. Where -- as long as your cloud provider makes no mistakes, the power grid and Internet are working, and everybody with authority over your provider likes you -- it will always be available. To you and ... gee, who knows who else?

In sum, as the niche for XP itself closes, the niche for a modernized XP is opening. But ROS needs to focus on catching that brass ring in the next five years. To do that its development needs to aim NOT for sound production release 'someday' but for betas good enough (crash-free enough) to begin to expand the user base. If (I'm guessing) 100 people actually run it regularly now can become 10,000 in a year, who knows what might happen? More developers, maybe money, maybe interest from a company that would distribute it one day ...

The USERS can make ROS a success and if it can pull them in while there are still lots of XP machines out there, they'll do so. It DESERVES 100 times as many users, SOON.

This website is a terrific job. There are items I haven't found yet, but the SITE is fine. And again, I'm VERY impressed with ROS itself.
PurpleGurl
Posts: 1790
Joined: Fri Aug 07, 2009 5:11 am
Location: USA

Re: Development Direction

Post by PurpleGurl »

That is a great post!

I think Win7 compatibility is on the horizon. That has to do with driver compatibility to run on more modern systems. Maybe they could do that with X64. It also has to do with the state of the 3rd party API code. For instance, they've had to use Vista/7 APIs in places already because of getting support from Wine, and Wine tries to run everything. Thus it will use the newer APIs.

On SSD, I likely won't get another SSD drive until Optane-based drives become more mainstream, unless someone else develops another NVRAM standard that isn't currently patent-encumbered. The 3D xpoint memory (Optane) is interesting in that it doesn't require 12v to blow the previous junctions to format it for reuse. Since there is no need to take time to build up a charge pump, the writes are faster than with NAND-based NVRAM. (And, I imagine after some maturity time, they might be able to create tweaked controllers for them that could occasionally write more than a bit at a time on the same module. I mean, there are 3 sets of access lines, and what if they could have an additional line that could address another bit on a single plane at the same time if that is applicable to whatever data? Or even hold an entire plane low or high and bulk write and entire byte or more at once. But I don't know if that is even possible or worthwhile for them to attempt.)

I agree on the USB and multicore. I'd love to see those working soon, but I don't know what I could do to help with that.

I agree that we cannot trust either corporations or governments. While I vote conservative on a number of things, I'm not going to support those who are hostile to the Internet or technology. I don't support the current FCC chairman (who pretty much openly admitted to there being a conflict of interest), and I don't think I'm for the current Supreme Court nominee either. I don't like the whole push to cloud nor the industry pressure to be totalitarian censors and limit whatever content that they or their advertisers disagree with. While I tend to believe that certain categories of content does not belong on the web, I also take issue with a lot of attempts to ban those things from the web as a whole, since people can redefine those categories at any time. Or they can straight out lie on others and use that as an excuse to keep the victims of the lies off the web. Once they start restricting speech, there becomes a sliding slope. So, to stay on topic, I don't think most corporations or governments can be trusted in this -- and I was only speaking from a US perspective. Other nations have their own reasons why their leaders cannot be trusted. So it looks like we can only trust the open source community.

And yes, ROS is the best project of this sort, or at least the only one that seems to matter. I remember how hard it was to find what I was looking for on the web when I first came over here. I was hoping to find alternative Windows files that could cause a performance increase. I had always perceived Microsoft as bloated and hoped to find maybe hobbyists who coded their own faster libraries and files. I didn't find that, but I found ReactOS and figured that was even better. Don't just recode one piece, recode it all.
Last edited by PurpleGurl on Thu Sep 27, 2018 9:13 pm, edited 1 time in total.
Ancient
Posts: 82
Joined: Tue Mar 27, 2018 11:32 pm

Re: Development Direction

Post by Ancient »

Lets assume an average user wants to get away from Windows and it's many issues. The only place to go is POSIX or RoS. Of the various POSIX alternatives, Mint has decent repute, it already has Windows themes (for various versions of Windows). Out of the box the Mint Cinnamon will read NTFS, and will work with most any file system. Installing the Valve variant of WINE (named PROTON or maybe Proton) gets you a very high compatibility with Windows API's while on Linux. It is easy to make a boot USB system, a CD ROM system, or install on a hard or SSD drive, it will even offer to set up dual boot with Windows.

Not certain if the folks here have reviewed Valve Proton, but it seems very nice. It still requires users to be a bit more knowledgeable than many are, but so does RoS. Windows is often turnkey because it's pre-installed on many systems.

So assuming I'm smart enough to figure out how to load an appropriate video driver for Proton on Mint, what will RoS offer me that Mint / Proton doesn't? What will RoS offer in 2 years, in 4 years, in 6 years at anticipated rates of development?

Would RoS development efforts be better spent to augment WINE / Proton? You have a development system here, and a lot of knowledge. Could it be better leveraged accepting Linux, embracing Proton, then improving upon that? A problem with development, is trying to leverage assets to get the best results. Sometimes this requires rethinking. My suggestion to developers here is take a few days off, try something like Mint Cinnamon and Proton, see how it compares to any result you can realistically attain with the current RoS kernel and API interface. Then consider your effort isn't wasted, but can be morphed to improve upon Mint (or any other variant of Linux you like) and Proton (effectively a superior WINE).

If the goal is to beat Microsoft and offer an alternative, is it better to keep plugging away with the current system, or to take what you know, the great stuff you have, the development capability here, and turn it to improve Proton, to get a better more Windows friendly Linux graphics shell. Reinventing every wheel takes a lot of time, your knowledge is extensive, but you're mired by the need to write a lot of code from scratch or to try and retrofit open source code to work within the RoS kernel. Maybe it is possible to focus on end user experience. An RoS 2, based on Linux / Proton, would let you develop improvements on top of something that is already good, and also allow Steam not to lock Proton features up in proprietary binaries.

What is the goal of RoS, can it be better met by changing development direction? RoS has been a great concept and a lot has been learned, a lot is applicable to more modern and evolved code in Linux / Proton. It is unlikely at the current trajectory that RoS will match Mint / Proton features already existing in the next half decade or so. At least that is my opinion. I am not suggesting RoS stop development, but rather that it re-target development to improve a Linux system with Proton as a turnkey product. You can do a lot to help make it a smoother product, and not be Steam OS. While Steam OS is centered on Steam now, it may not always be, and Steam may not always want to be so open with the Linux community.

This group could help keep Proton development moving forward as open source, and could provide an easier turnkey Linux / Proton solution which is easier for many users to set up, install and run. Essentially this group could possibly fulfill it's objective by embracing Linux / Proton rather than continuing on the current development path.
User avatar
dizt3mp3r
Posts: 1874
Joined: Mon Jun 14, 2010 5:54 pm

Re: Development Direction

Post by dizt3mp3r »

ReactOS isn't going to change its direction so most of this thread is pointless. Fact Number ONE.

ReactOS isn't a ship that you can change direction by the mere application of rudder towards something more acceptable to you. There are a lot of individual developers working on things that they want to contribute toward based upon their particular skill and inclination. Moving the current set of devs toward what is really another o/s build doesn't do much to redirect the whole of ReactOS. Most might simply not join the new direction stopping mid flow... Those that are contributing know what and why they do it.

If you want something enough then do it. Fork ReactOS, change its direction, make your direction desirable to the world, create a team and take it in that direction.

This thread is fundamentally about disatisfaction over the progress of ReactOS, get over it. It is where it is. IF you want it to go faster then contribute enough to complete, if you are a multi-millionaire, simply hire some developers to give you the functionality you require. Kernel, memory management, file system, USB And networking should just get it to beta.

That's it, knuckle down or FORK it and do better. Nothing is going to change.
Skillset: VMS,DOS,Windows Sysadmin from 1985, fault-tolerance, VaxCluster, Alpha,Sparc. DCL,QB,VBDOS- VB6,.NET, PHP,NODE.JS, Graphic Design, Project Manager, CMS, Quad Electronics. classic cars & m'bikes. Artist in water & oils. Historian.
PurpleGurl
Posts: 1790
Joined: Fri Aug 07, 2009 5:11 am
Location: USA

Re: Development Direction

Post by PurpleGurl »

I don't think the thread is just to complain. I think the OP is sincere. And I agree with the post immediately above this one. We are not Linux.

[As a tongue in cheek comment, maybe we should create our own Linux distro. Let's remove all that Posix stuff, make it where it only runs WinNT drivers, and make it where it only runs Win32 (and hopefully x64 eventually) software.]

And the goal is not to beat Microsoft. It is to create our own Windows. Period. If we had Linux aspirations, we'd be there instead.
User avatar
dizt3mp3r
Posts: 1874
Joined: Mon Jun 14, 2010 5:54 pm

Re: Development Direction

Post by dizt3mp3r »

If I were a multi-millionaire I would contribute significant funds to help take ReactOS to beta, the fact that I am not a multi-millionaire probably indicates that millionaires are more clever than me and more careful with their money. That is perhaps why we see less donations of the size required to complete the o/s. That and the fact that ReactOS taking as long as it has does tend to deter the sensible money. ReactOS will be completed but that will almost certainly not be ready by the time Win7 leaves support in two years. I look to the future and see ReactOS being used entirely within the virtual sphere. As 'cooking' machines begin to have 16 or more cores by default then one or two cpus can easily be dedicated to virtual instances of another o/s. ReactOS could run there.

Personally, I think the time for ReactOS to run on real hardware is beginning to pass. It has simply taken too long to get to the point where it is usable, if only partly, on real hardware. However, I can see it coming to a point where it will be stable enough to operate discrete VMs to perform a particular task.

Stability in a VM, NTFS support, NT6 compatibility, USB+networking via passthru, 32bit single core. I think that is what is needed and it is achievable in reasonable timescales if the devs persist with the current goal, as they are doing and as they will do. A possibility to achieve that at beta within 3-5 years? Something working within two? That should be the development direction and it would keep ReactOS in the computing world's eye.

Eventually 64bit versions with SMP, for me that would be ReactOS with knobs on - but I think that is ten years away.

IF I am correct then it would mean real hardware support subordinated to a niche for the techies (and us nerds out there) that like being masochistic. I think that my timescales are realistic.

PS. The OP may be genuine, I am sure he is. I think he is going through that 'wake up to reality' that we all have to go through regarding the slow but sure progress of ReactOS. He, like all of us wants it now (or soon) but that ISN'T going to happen unless he is Father Christmas and he is happy to leave a stack of money to make it happen. If he is - then he can have everything he wants and all he has to do is to make it happen.

I'd say the same to anyone.

Too many words now, let's get on and do something else to lighten the mood.
Skillset: VMS,DOS,Windows Sysadmin from 1985, fault-tolerance, VaxCluster, Alpha,Sparc. DCL,QB,VBDOS- VB6,.NET, PHP,NODE.JS, Graphic Design, Project Manager, CMS, Quad Electronics. classic cars & m'bikes. Artist in water & oils. Historian.
Walt
Posts: 42
Joined: Tue Sep 04, 2018 1:39 am

Re: Development Direction

Post by Walt »

Many interesting thoughts.

Ancient:
Out of the box the Mint Cinnamon will read NTFS, and will work with most any file system. Installing the Valve variant of WINE (named PROTON or maybe Proton) gets you a very high compatibility with Windows API's while on Linux. It is easy to make a boot USB system, a CD ROM system, or install on a hard or SSD drive ...
This says that Mint Cinnamon has come a VERY LONG way in the last three years or so. When I last tried it -- second of two attempts, I think -- some distros would boot, some would not, and I got nowhere with Wine as far as running Windows programs. It must have worked for some people or on some machines but it was no help to me.

And:
I am not suggesting RoS stop development, but rather that it re-target development to improve a Linux system with Proton as a turnkey product. You can do a lot to help make it a smoother product, and not be Steam OS.
This seems to be saying that the developers of ROS should ditch the effort and provide a better (more XP like?) interface to Linux. Having learned a lot that can be applied to another system isn't a great outcome for a project on the scale of ROS.

PurpleGurl
And the goal is not to beat Microsoft. It is to create our own Windows. Period. If we had Linux aspirations, we'd be there instead.
Exactly. An effort similar in concept to what Firefox did to Internet Explorer.

Many other good and interesting things in your posts, PG, but my thoughts might attract unfavorable attention from a moderator so I'll pass.

dizt3mp3r:
Personally, I think the time for ReactOS to run on real hardware is beginning to pass. It has simply taken too long to get to the point where it is usable, if only partly, on real hardware. However, I can see it coming to a point where it will be stable enough to operate discrete VMs to perform a particular task.
Perhaps I'm missing something but if ROS is to be limited to running in a VM why not just use XP in a VM? Don't get me wrong -- the human factors of ROS are better -- but that alone doesn't really justify the effort OR give the world an alternative. Firefox isn't just another face for IE, it offers much new function and has helped force Microsoft to share control of the browser direction with browser users. ROS seems to me to be an effort to do exactly the same for the OS itself -- PG's comment, quoted above.

To take a specific example: If the major manufacturers decide that the money is in moving computing to the cloud, they will stop furnishing OS's that run on real hardware without the cloud. Either you go cloud or you go unsupported on steadily older machines and there's no new function coming along.

Google's Chrome is a huge step in that direction with its blurring of the line between local and web function.

If, however there's an alternative OS (ReactOS) that's popular and running on the latest real machines then any effort to end such real machines creates an opportunity for other manufacturers. If USERS want local computing power and have the (labor intensive) OS at hand, the machines will be made.

It's also true that the major limitations of ROS at 0.4.11 are NOT things that will be resolved by virtualizing the hardware: They are crashes and missing/incorrect critical functions that are (or probably are) above the HAL. Today I found my D610 install had three copies of every file in the C disk -- three 'Program Files' folders, and so on. Don't know when that happened. So I started reinstalling it. 'Install' of course isn't just the OS -- there are a number of supporting programs before the system can be useful even for browsing. I'm gradually collecting everything needed on a single CD but having to burn every item makes that a chore. With no USB support, though, that's the only way.

This has been a very interesting discussion. Enjoyed it!
PurpleGurl
Posts: 1790
Joined: Fri Aug 07, 2009 5:11 am
Location: USA

Re: Development Direction

Post by PurpleGurl »

Thanks Walt. I wanted to include more in my one post, but I decided to pare some of it back and speak more in generalities. I don't like the current push for certain people to try to seize control of the web and block unpopular opinions. And certain giants hold a bit of a monopoly on certain things and the playing field is not level to be able to start places were unpopular ideas are allowed. (Take Gab, for instance, a social networking alternative, and people have constantly tried to freeze their revenue, DDOS them etc.) And like you, I felt it was wise to keep most of my opinions to myself.

Actually, it isn't so much Chrome that is blurring the line, as it is Chromium and Chrome/OS. So the idea there is for a bare-minimum OS that brings up the PC quickly and then loads the apps off the web. Thus it is applications as a service. I can see how that can become unfair and a security risk. Now, with apps as a service, you have to keep paying a subscription. Photoshop now does this. You subscribe to a license and use it over a cloud. I have so many objections to this. Privacy and intellectual property reasons are high on the list. Maybe they are not supposed to, but how can you trust that they don't breech the cloud to examine what you use the software to create? Some might think that's good for eliminating "fake news," "hate speech," "illegal pornography," etc., but what if they steal your ideas or use your private photos in an extortion attempt? I myself had been a victim of a misuse of corporate records. I had Roadrunner Internet (owned by Time Warner), and Time Warner had acquired AOL (who had a reputation, whether a fair one or not, for dishonest tactics). Anyway, I rarely gave out my Roadrunner address, but was getting loads of spam there. Eventually, I discovered what had happened after a former AOL employee was arrested for stealing account information and selling it. This dishonest employee ended up at my web provider due to the merger. So if they sell your email address, what is to keep them from stealing software you write, family photos, personal documents, creative works, etc., and misusing them somehow?

So yes, a different philosophy from what we are promoting. And as the cloud thing becomes more popular, tablets and smartphones are used more, etc., and the more certain freedoms are stifled, it seems that the open source community will be taking a larger role in supporting the traditional PC users. I just hope the alleged catfight discussed in another thread doesn't impact us negatively.
Post Reply

Who is online

Users browsing this forum: Ahrefs [Bot], Google [Bot] and 42 guests