Talk:Comparison of open-source operating systems

Grey cells for row/column headings

So I noticed that while most of the row/column headings were grey, there were a few row headings that weren't. (Note: They are grey for me, but depending on your CSS stylesheet, they may be different. They are the cells started with a ! rather than a | .) So anyways, I changed it everywhere except for "Important programs", which was the only place where it looked like it might possibly have been on purpose. Personally, I'd rather they were all grey, but was there a reason for this? Armedblowfish 21:19, 26 May 2006 (UTC)[reply]

Yes it was. i meant that: x86 is a main category and the x86 smp support is a sub, as they are the same architecture. --preceding unsigned comment by Balihb 23:51, 26 May 2006 (UTC)[reply]

Maybe there's another way to make this clearer... perhaps a template that makes the cell a lighter grey? (Btw - I hope you don't mind that I indented your comment and credited you/added a date. To sign your comments on talk pages, just use ~~~~ or ~~~ if you don't want a date.) Armedblowfish 00:46, 27 May 2006 (UTC)[reply]

linux and other arch

i can't find the following arch's. they from the linux kernel arch dir:

cris m32r um v850 xtensa

and also a few from ecos. --BaliHB 22:16, 28 May 2006 (UTC)[reply]

Wikipedia has articles on M32R, V850, and Xtensa. At a glance, M32R seems to be supported by Linux, V850 runs on uClinux, and I'm not sure if Xtensa runs on any open source operating systems at all. Armedblowfish 23:38, 28 May 2006 (UTC)[reply]

hmmm...

> cd /usr/src/linux/arch/

> ls

alpha cris i386 m68k parisc s390 sparc v850

arm frv ia64 m68knommu powerpc sh sparc64 x86_64

arm26 h8300 m32r mips ppc sh64 um xtensa

i think that cris is maybe the clamrisc, that eCos support.

--BaliHB 08:20, 29 May 2006 (UTC)[reply]

more operating system

thinking of includeing these:
Contiki, K42

Suggestions

A nice addition to the matrices would a comparison of native (kernel level) threading support. Rather than just a "yes" or "no", it would be interesting to say whether N:M or 1:1 implementations are supported.

Possible additions to the network matrix could be XNS and OSI support. I'm not sure if these two stacks are of interest to many people these days, but several open source operating systems support them.

Finally, Im not sure how appropriate the ATI and Nvidia driver support comparisons are. These drivers are closed source, and usually tied to a specific patch level of the Linux kernel. I feel it is arguable that including these drivers makes the inclusion of other closed source additions to operating systems appropriate. For instance, Wasabi Systems market a journaling filesystem for NetBSD (WasabiJFS). —The preceding unsigned comment was added by Chriswareham (talk • contribs) .

I've added a new row to the "other information and technologies table" for kernel level thread support. —The preceding unsigned comment was added by Chriswareham (talk • contribs) .

Given that some OSes support both 1:1 and M:N with different userland libraries, what does it mean to say that the kernel-level threading support is 1:1 or M:N? Would a kernel that makes M:N difficult or impossible be one that supports 1:1, and a kernel that makes M:N reasonably possible be M:N even if there exist 1:1 threading libraries for it? Guy Harris 18:33, 12 June 2006 (UTC)[reply]

I'd say either whatever the primary threading mechanism is (do any OSs have M:N kernel threads but have 1:1 as preferred threading method?)... Or just list both: "M:N and 1:1". NicM 21:42, 12 June 2006 (UTC).[reply]

Number of entries

There are far far too many columns in these tables, it makes them almost unreadable. How about losing some of the less popular/well-known OSs, like Menuet, NewOS, Haiku, the last seven columns that have virtually no data. In fact, what about just limiting it to:

Linux
FreeBSD, NetBSD, OpenBSD, DFly BSD
OpenSolaris
OpenDarwin
Minix
FreeDOS
GNU HURD
Plan9
possibly Syllable and Inferno and ReactOS

The others could be linked to in see also section. Let's pick a limited number of OSs carefully, fill the columns properly, and have a useful page, rather than the half-empty unreadable mess we have now, eh? NicM 22:00, 12 June 2006 (UTC).[reply]

No. It's other articles that compare them. I think that it should be all of the actively developed os-es here. there won't be much more of them.--BaliHB 00:25, 13 June 2006 (UTC)[reply]

the firs os-es are the popular ones. thats why they are not in alphabetic order.--BaliHB 00:31, 13 June 2006 (UTC)[reply]

There are too many already, never mind any more. It either needs some to be removed, or the tables to be rotated so the OSs are the rows rather than columns. In that case, at least the columns could be restricted by adding more tables for more features. NicM 00:36, 13 June 2006 (UTC).[reply]

Look at the way Comparison of operating systems does it, for instance. NicM 00:38, 13 June 2006 (UTC).[reply]

I'we thinked about this a lot. Dunno what would be the best way. With rotation you can't realy manage newly added thinks, but the os management is easyer. But in this case I think that it'll be more features added, than os... --BaliHB 07:22, 13 June 2006 (UTC)[reply]

It isn't a case of which is easier to add, it is which is easier to read... if the OSs are in the rows, we can have a million OSs and it won't matter much: it is also much much easier to split features to make it a sane number of columns, eg, if the hardware/driver table had two many columns, it could be split into two tables "Network drivers" and "Other drivers", or three with "Graphics drivers", etc. I think empty rows also look much better than empty columns. NicM 17:20, 13 June 2006 (UTC).[reply]

I agree that this would be better. Do you know of an easy way to rotate the tables? (If not, I recommend the use of {{inuse}} whenever someone gets around to changing a section.) Armedblowfish (talk|contribs) 17:31, 13 June 2006 (UTC)[reply]

I don't know of an easy way, I was rather hoping somebody else did :-). NicM 18:36, 13 June 2006 (UTC).[reply]

I'm feeling my self guilty as I started the page in the wrong way... Maybe a new editing interface could be nice. Or just a script wich makes the tings easyer... --BaliHB 20:59, 13 June 2006 (UTC)[reply]

Don't feel bad. It will just take time to change. And the Wikitable format is actually much easier than coding in raw CSS or HTML tables. Armedblowfish (talk|mail|contribs) 21:30, 13 June 2006 (UTC)[reply]

I'm almost finished with the script, that does the rotating automaticli. It'has some problems now, but almost finished. Here it is. I hope I'll finish it in a few hours.--BaliHB 22:21, 13 June 2006 (UTC)[reply]

Awesome! Just don't rotate the "Other information and technologies" section, I did that manually (with nice, helpful comments for empty cells.) See this version of the page with both the old table and the new table to compare to make sure there's no mistakes. Armedblowfish (talk|mail|contribs) 22:37, 13 June 2006 (UTC)[reply]

OK. Finihs! :))) Crappy solution, but works and was fast :) --BaliHB 23:06, 13 June 2006 (UTC) ,, Headers should be inserted after 5 lines... --BaliHB 23:17, 13 June 2006 (UTC)[reply]

ati & nvidia

What to do with ati and nvidia. Make an "official dirvers"? Dunno. It's not realy a looking great. :) --BaliHB 00:27, 13 June 2006 (UTC)[reply]

Page is Useless

This page offers no useful information to anyone and should be removed. 202.6.138.34 04:45, 13 June 2006 (UTC)[reply]

You wanna remove the discussion page? why? >:)))--BaliHB 07:20, 13 June 2006 (UTC)[reply]

Yes, yes; jolly good joke old chap. The page is basically just a list of features found in Linux.

It reads like "Linux has X, what other systems have X too?". Its hardly an unbiased comparison. 202.6.138.34 14:24, 13 June 2006 (UTC)[reply]

It is a young page. Many articles start off biased, and become less biased as more editors with different points of view contribute. Your contributions to help make it less biased would be appreciated, of course. Armedblowfish (talk|contribs) 15:59, 13 June 2006 (UTC)[reply]

That's why i called for the help of the os users and developers. Maybe they can give in more.--BaliHB 21:03, 13 June 2006 (UTC)[reply]

old-style refs vs. footnotes

The oldstyle references (as opposed to the new, inline ones) are confusing next to the new inline ones. Perhaps convert to footnote style instead (since they are being used as footnotes anyways), like in Comparison of BSD operating systems? Armedblowfish (talk|contribs) 17:44, 13 June 2006 (UTC)[reply]

E/OS

e/os info:

Hi! 

Thanks for the mention of eos in the articule, and i
will intent to add the information of eoslx.

CPU/Architecture support

x86/i386/IA-32         
- yes
x86 SMP
- no         
Xen 
- no        
IA-64/AMD64         
- no
AMD64
- no         
PowerPC
- no         
PowerPC SMP         
- no
SPARC32         
- no
SPARC SMP         
- no
Alpha
- no         
MIPS
- no         
ARM 
- yes        
XScale
- no         
M68k
- no         
PA-RISC
- no         
other
- Poket pc under devel
hosted mode
- QEMU is the default mode for run under other systems

Hardware/Driver support

ATA         
- yes
SATA 
- yes        
SCSI 
- yes        
USB 2.0
- no (partitial)         
USB 1.1         
- yes
FireWire
- yes         
PCMCIA/PC card 
- yes        
AGP         
- no
Nvidia official driver IA32         
- no
Nvidia official driver IA64         
- no
Nvidia official driver AMD64         
ATI official driver x86         
- no
ATI official driver x86_64         
- no
sound
- yes         
TV tuner/Video editing/Webcam         
- yes
Networking 
- yes        
NE2000/RTL8029         
- yes
RTL8139
- yes         
Gigabit Ethernet         
- yes
10-gigabit Ethernet         
- yes
Wireless LAN
- partitial         
Bluetooth         
- no
IrDA
- no

Network technologies
        
firewall         
- yes
TCP/IP
- yes         
IPv6 
- yes        
IPX         
- yes
PPP         
- yes
PPPoE
- yes         
DHCP 
- yes        
bridge
- yes         
TUN/TAP 
- yes        
ssh         
- yes
OpenVPN
- no

File system support

FAT16/dosfs, FAT32/vfat         
- yes
NTFS 
- no        
Ext2
- yes         
Ext3 
- no        
XFS 
- no        
ReiserFS         
- no
UFS 
- no        
UFS2
- no         
HFS 
- no        
HFS+
- no         
Minixfs
- yes         
BeFS 
- yes (default in series 028)        
ISO 9660         
- yes
UDF         
- no
NFS 
- yes        
SMBFS
- yes         
RAM disk/tmpfs         
- yes
procfs
- yes         
Virtual memory/swap         
- yes
other special file systems         
- no
RAID
- yes         
quota
- yes         
Resource access control         
- Unix
encryption         
- yes
other special file system features
- no

Important programs

Linux binary compatibility         
- yes
Wine 
- yes (native)        
DOSEMU/DOSBox
- yes         
XOrg 
- yes        
XFree86
- yes         
other gui
- no         
Mozilla Firefox         
- yes
other browser         
- Internet Explorer, Opera and Dillo
OpenOffice.org         
- yes
other office suite         
- Microsoft Office 97
MPlayer         
- no
other media player         
- Xine
Gaim         
- yes
other instant messenger         
- Amsn, Microsoft MSN
Apache         
- yes
other httpd         
- no
QEMU
- yes
other PC emulator         
- native emulation for Darwin, BSD and PowerPc
emulation.
PearPC
- no


Other information and technologies

        
License         
- GNU GPL 2.0
Kernel type         
- Monolithic
Kernel programming language         
- ASM, C, and C++
Kernel Thread support         
- 1:1
OS family:         
- BeOS, Unix
oldest non-EOL version[30]         
- No
Forks
- None


Thanks for all,
Dr. Miguel Chanampa

--BaliHB 09:39, 14 June 2006 (UTC)[reply]

GNU Mach and/or GNU Hurd

PA-RISC is listed as a supported architecture for GNU Mach and/or GNU Hurd. Although work exists to run Mach on PA-RISC <http://www.cs.utah.edu/flux/mach4-parisc/html/pamach.html>, this is based on Mach4 and not GNU Mach. GNU Hurd for non-i386 machines does not exist. Does anyone have any contradictory evidence that I should be aware of? 202.6.138.34 13:53, 22 June 2006 (UTC)[reply]

Contiki

Contiki is not yet listed. --62.104.88.118 18:04, 7 August 2006 (UTC)[reply]

Syllable and Gigabit

Syllable does support gigabit, check out Syllable website news page for more information.

Please update the article

VLC

VLC supports BSD as well. 70.111.224.252 15:28, 3 September 2006 (UTC)[reply]

AMD64, EM64T, and IA-64

I might have it all wrong, but should something which runs on AMD64 and EM64T run on IA-64 as well, and vice versa? Armedblowfish (talk|mail) 14:59, 29 October 2006 (UTC)[reply]