0.3.11 September

[nute]

There seem to be vague indications that the bugs plaguing networking and firefox need to be
fixed before 0.4.0 or even 0.3.11 can come out, but I don't know that that is going to happen.
One has to use firefox 2 still and even then for some reason, on my site at least, I have
trouble getting web browsing to work at all.

The interface needs a lot of work, but I don't know if that has anything to do with the delay
getting 0.3.11 out. ReactOS explorer especially is not where it needs to be yet.

There is no log in implementation yet even though a password is requested at install.

There has been a fat implementation for a long time, but there has never been a defragmenter.
There is the open source defragmenter ultradefrag, but it doesn't work in ReactOS. It is meant
for NTFS filesystems anyways.

I assume there is a major hardware effort under way and that may be partially why there is a delay.
I for one hope that there will be USB2 support soon.

Clearly there are major memory management problems still or ReactOS wouldn't randomly crash
with illegal memory access errors. It is hard to pin down what causes those errors to happen.

I would like to see ReactOS's TCP/IP implementation get fixed so that it works as well as any other
implementation in any other environment before the other problems get much attention. I suggest
focusing on this area over other areas because networking is essential to get updates installed and
speed development. While ReactOS may not be stable enough yet to worry about updates, I'm hoping
it will be stable in the not too distant future. Whether the networking problems are because of buffering
issues that can be tracked to memory management issues or poor driver support, I'm not sure. I don't
know when networking is going to be smooth in ReactOS, but the sooner the better.

The question is, can ReactOS development speed up enough for the OS to be relevant? Running the
demo of Diablo II or firefox 2 isn't going to solve the Joe blow needs Windows once in a while to run
xyz program issue. There is open source tax software now for example, but it doesn't cover every
jurisdiction and I believe it is still in beta. People are still using Intuit's TurboTax which is only
available for Microsoft Windows as far as I'm aware. If ReactOS could run TurboTax, that would
be great.

Actually, I suppose ReactOS explorer and networking need to be fixed simultaneously so that you can
get around ReactOS better in general.

[EmuandCo]

Some of your Suggestions remind me of a new Roadmap....

[cuttysark]

And how would you propose we decide whether to increment the progress bar?

For instance, the ReactOS 0.3.10 Release Newsletter:
1 - Fixed inability to work with partitions bigger than 8Gb
2 - Added possibility of installing into any of four primary partitions
3 - Initial support for SerialATA controllers along with enhanced ATA support
4 - Initial USB keyboards and mice support
5 - Greatly improved network cards support (20 different NICs were tested successfully)
6 - Increased stability in networking
7 - A clone of MS paint application has been introduced
8 - Initial support for MSVC compilation
9 - Better cleanup of system resource usage
10 - Synchronization of most of the Wine usermode DLLs and some Win32 subsystem code shared with Wine

10 Important changes, 10% per change in the progress bar...

[AlexEagar]

I like cuttysark's measurement. I personally think as long as ReactOS is in alpha it shouldn't have to be more stable than a previous release. I say 10 notable improvements or two months, whichever comes later. Then, before major versions are released focus more on regression and stability. I think most ReactOS fans just want to hear about improvements being made and see the version number increase. As long as you can do those two things you'll continuously get more supporters and more donations.

[vicmarcal]

For instance, the ReactOS 0.3.10 Release Newsletter:
1 - Fixed inability to work with partitions bigger than 8Gb
2 - Added possibility of installing into any of four primary partitions
3 - Initial support for SerialATA controllers along with enhanced ATA support
4 - Initial USB keyboards and mice support
5 - Greatly improved network cards support (20 different NICs were tested successfully)
6 - Increased stability in networking
7 - A clone of MS paint application has been introduced
8 - Initial support for MSVC compilation
9 - Better cleanup of system resource usage
10 - Synchronization of most of the Wine usermode DLLs and some Win32 subsystem code shared with Wine

10 Important changes, 10% per change in the progress bar...

That´s not even a 3% of the total project.

[swight]

Maybe the progress bar would be for a version number change(vs total project).

[Aeneas]

I have been watching the ReactOS project already for a long time (last time tried 0.2.7) and I must say I find its progress rather impressive. What I find not particularly impressive is the "if you do not like it, go away"-attitude. On this attitude I would like now to comment:

For those having this attitude, please note, we ARE most probably "away" right now, because using native ReactOS is not common. I would like to remark that there is a difference between a hobby and an open-source-project in my eyes. An open-source-project is rooted in idealism. Let's say, the kind that Gandhi had when he made India independent. Noone actually hired him to do that, and yet, he did not pursue India's independence as a "hobby".

The Linux kernel, I read somewhere, was at one point 8 million lines of code, while the whole Debian project - with all packages - 260 million lines. ReactOS, so to say, has just to target the 8 million, as the 260 million here already exist - in a LOT of Windows software, both commercial and open source. If 0.3.11 represents less than 3% of the general effort left, then at such a speed (about 1% per month), you will be "done" in 2018! I think this is regrettable, in particular in view of the achievements so far. Because till then, Microsoft will have brought out 2-3 other OS' already. You have a chance to catch up, as MS has reached a point where they cannot make much progress on their current path as, to put it bluntly, "they have already invented everything". A jump from WinXP to Win7 is not like a jump from Win3.11 to WinNT4. But if MS somehow radically changes something, and you are NOT ready to fill the gap, this project can be lost completely... or get a relevance like, say, FreeDOS. A really nice DOS, by the way. But in the 2000s it simply came too late. FreeDOS does NOT have the relevance of, say, Linux.

I believe a more courageous approach in the development would be actually very beneficial to the project. No, I cannot contribute code. I would, however, donate. But please understand me correctly: I would donate for an idealistic cause, and not just for the support of someone's hobby. Otherwise I would rather finance my own hobbies. I would indeed rather invest in ReactOS, but I would like to see that it would have a sense.

I DO congratulate heartily for what has been achieved so far, it is nothing less than impressive. But I also hope future development will accelerate.

[Alahndro]

A progress bar could be made from the buglist for the target revision:
http://www.reactos.org/bugzilla/buglis ... alue0-0-0=

Or just add a link on the frontpage to the target buglist. Then everyone not only sees the progress but also what is holding up the release.

And looking at it it seems that there still is a long way to 0.3.11

[Angelus]

Aeneas:

Maybe you don't know what a free (as in beer) open source project means...
Maybe you don't know the great challenge tha involves developing a whole Operating System (not only the kernel, as Linux is)...
Maybe you don't know the difficulties of developing a full Operating System 100% compatible with an existing closed source operating system without infringing any patent...
Maybe you don't know how many people is necessary to carry it out...

But right now, with current resources, this is the most it can be done.

Beginnings are always difficult. But as long as ReactOS becomes more stable and functional, new developers will enjoy the project (sure). Even as a hobby, because they like to code.

[Haos]

I have been watching the ReactOS project already for a long time (last time tried 0.2.7) and I must say I find its progress rather impressive. What I find not particularly impressive is the "if you do not like it, go away"-attitude. On this attitude I would like now to comment:

Such attitude is necessary when a person is trying to convince us that he knows best what he thinks that should be done, but refrains from helping out or even proving his words.

[cruonit]

There has been a fat implementation for a long time, but there has never been a defragmenter.
There is the open source defragmenter ultradefrag, but it doesn't work in ReactOS. It is meant
for NTFS filesystems anyways.

http://www.flexomizer.com/PermaLink,gui ... c1957.aspx

based on:

http://technet.microsoft.com/en-us/sysi ... 97427.aspx

maybe ask Dave to release the app as open source:
"Given that Windows "7" has substanitally improved built-in defragmentation, I won't be doing any work to ensure full compatability with that version of Windows."

[cruonit]

also
http://kessels.biz/JkDefrag/

JkDefrag is completely free, no cost, no time limit. If you've payed for your copy then you've been scammed! The executables are released under the GNU General Public License, and the sources are released under the GNU Lesser General Public License.

[greenie]

Just a quick note I think alot of the time people are told "x, don't like it then leave". I remember this happening to nearly every post in the design forum for a while. I almost thought that the design forum should have been closed completely. Might have been linked to 0.3.9 release.
Most of the time it's called for sometimes it's not. I remember in some posts people say "i mean Y and is NOT related to X". With an instant response " We are not doing X, don't like it leave". Though this is generally the exception not the rule.

[RaptorEmperor]

The Linux kernel, I read somewhere, was at one point 8 million lines of code, while the whole Debian project - with all packages - 260 million lines. ReactOS, so to say, has just to target the 8 million, as the 260 million here already exist - in a LOT of Windows software, both commercial and open source. If 0.3.11 represents less than 3% of the general effort left, then at such a speed (about 1% per month), you will be "done" in 2018! I think this is regrettable, in particular in view of the achievements so far. Because till then, Microsoft will have brought out 2-3 other OS' already. You have a chance to catch up, as MS has reached a point where they cannot make much progress on their current path as, to put it bluntly, "they have already invented everything". A jump from WinXP to Win7 is not like a jump from Win3.11 to WinNT4. But if MS somehow radically changes something, and you are NOT ready to fill the gap, this project can be lost completely... or get a relevance like, say, FreeDOS. A really nice DOS, by the way. But in the 2000s it simply came too late. FreeDOS does NOT have the relevance of, say, Linux.

Aeneas, I understand your logic, and it does make sense if you don't understand the difference between how Windows and Linux work.

The problem with that logic is that Windows isn't built much like Linux. Linux is based on Unix, which is very modular. You can run Linux with just the kernel in command line and everything's good. Or you can add a GUI of your own choice, and pretty much select all kinds of system components at will. Ultimately, though, Unix is a command-line operating system, while Windows NT was designed from the ground up as a GUI-based operating system.

The GUI is tightly integrated with the Windows kernel, as are many other components, and Windows is functionally useless without them. For example, a lot of software that I test on ReactOS has issues with ReactOS Explorer, such as drawing windows and such. Linux devs would never have to deal with that, because they can just focus on the kernel and let the GNOME, KDE, and X11 devs handle those issues. The difference between developing ReactOS and when Linux was released it was based on the much smaller, simpler Unix operating system from the early 1990s than the Windows ReactOS is trying to duplicate here in 2009.

[Z98]

Umm, the GUI is not tightly coupled with the kernel. In the original design it never was. MS made the decision later on to push the Win32 subsystem into the kernel for strategic/performance reasons. The result is that it's tightly coupled, but nothing about the NT kernel requires it to be.