LinuxQuestions.org - PC-BSD - where is it going?

- *BSD (https://www.linuxquestions.org/questions/%2Absd-17/)

- - PC-BSD - where is it going? (https://www.linuxquestions.org/questions/%2Absd-17/pc-bsd-where-is-it-going-755483/)

IMO, the PC-BSD project is a compromise variant, intended to attract
new customers. It is the same old FreeBSD wrapped in a nice GUI + PBI.
I think it's meant to add more popularity to FreeBSD, which is less
popular than Linux distributions.
As for me, I wouldn't like BSD be more popular like Linux.
FreeBSD is better than Linux distros and I hope it will always
be like that.

>FreeBSD is better than Linux distros and I hope it will always be like that.

What's better? I'm using FreeBSD and Slackware - Slackware since the early 90s, FreeBSD since 5.0 because of quality. I don't have any penalties while using the Slack and I'm really sick of this fanboyism. I don't like any single Linux distro, but at least Slack and to some degree Gentoo. Vice versa PCBSD is in my opinion redundant, especially PBI. So what? Who cares? It's open source, if you want to compete do it with facts not fiction.

Quote:

Originally Posted by Oliver_H (Post 3795862)

Well, I think "better" arguments are usually pretty silly.
I happen to really like the PBI system - it's a nice take on the pre-compiled packages that Debian uses (and Redhat et al copied).

I happen to like the interface of BSD better, but that's because I have a really long background with the actual classical Unix systems (System V, IRIX, AIX, BSD, etc) so I am more likely to type ps aux than ps -ef.

The only really significant differences, so far as I know are:
1) GPL vs BSD licenses
2) Top-down Unix vs bottom-up Linux

I don't think an end user of either system (on a properly configured box) would ever really be aware of it. KDE is KDE, Gnome is Gnome, WindowMaker is WindowMaker.

Quote:

Originally Posted by Oliver_H (Post 3795862)

facts not fiction.

The output of top

Code:

last pid:  2331;  load averages:  0.47,  0.55,  0.55  up 0+00:27:16    14:06:30

44 processes:  3 running, 41 sleeping



Mem: 113M Active, 115M Inact, 119M Wired, 1800K Cache, 111M Buf, 643M Free

Swap: 512M Total, 512M Free





  PID USERNAME    THR PRI NICE  SIZE    RES STATE    TIME  WCPU COMMAND

 2276 wolf        1  58    0  116M 28700K RUN      8:01 19.38% operapluginwrapper.

 2238  wolf        1  46    0 52764K 40812K select  0:34  4.69% Xorg

 2254  wolf        3  76    0 82876K 72292K ucond    1:06  4.49% opera

 2321  wolf        2  44    0 43044K 20724K ucond    0:02  0.00% gedit

 2242  wolf        1  44    0 10744K  8364K select  0:01  0.00% wmaker

  352  root          1  44    0  3408K  1168K select  0:01  0.00% moused

  819  root          1  50    0  5608K  2516K ttyin    0:00  0.00% csh

 2252  wolf        1  44    0  5608K  2668K RUN      0:00  0.00% csh

 2246  wolf      1  76    0 27572K  8716K select  0:00  0.00% aterm

 2328  wolf        1  44    0 21096K  5412K select  0:00  0.00% gconfd-2

  821  root          1  44    0  5608K  2576K ttyin    0:00  0.00% csh

 2245  wolf      1  44    0  9884K  6908K select  0:00  0.00% xterm

 2247  wolf      1  44    0  7908K  5328K select  0:00  0.00% xclock

 2330  wolf      1  44    0 19912K  4044K select  0:00  0.00% gam_server

 2249  wolf      1  76    0  5608K  2428K ttyin    0:00  0.00% csh

  832    wolf      1  44    0  5608K  2468K pause    0:00  0.00% csh

  809    root          1  45    0  3764K  1724K wait    0:00  0.00% login

  811    root          1  44    0  3764K  1764K wait    0:00  0.00% login

The output of kldstat

Code:

Id Refs Address    Size    Name

 1    8 0xc0400000 4ef5e8  kernel

 3    1 0xc4666000 e000    fuse.ko

 4    1 0xc56a2000 490000  nvidia.ko

The kernel configuration file:

Code:

cpu                I686_CPU

ident                CUSTOM



# To statically compile in device wiring instead of /boot/device.hints

#hints                "GENERIC.hints"                # Default places to look for devices.



# Use the following to compile in values accessible to the kernel

# through getenv() (or kenv(1) in userland). The format of the file

# is 'variable=value', see kenv(1)

#

# env                "GENERIC.env"



makeoptions        NO_MODULES        



options        SCHED_ULE                # ULE scheduler

options        PREEMPTION                # Enable kernel thread preemption

options        INET                        # InterNETworking

options        FFS                        # Berkeley Fast Filesystem

options        SOFTUPDATES                # Enable FFS soft updates support

options        UFS_ACL                        # Support for access control lists

options        UFS_DIRHASH                # Improve performance on big directories

options        UFS_GJOURNAL                # Enable gjournal-based UFS journaling

options        MSDOSFS                        # MSDOS Filesystem

options                NTFS

options        CD9660                        # ISO 9660 Filesystem

options        PROCFS                        # Process filesystem (requires PSEUDOFS)

options                LINSYSFS

options                LINPROCFS

options        PSEUDOFS                # Pseudo-filesystem framework

options        GEOM_PART_GPT                # GUID Partition Tables.

options        GEOM_LABEL                # Provides labelization

options                COMPAT_LINUX                # Linux compatibility

options        COMPAT_43TTY                # BSD 4.3 TTY compat (sgtty)

options        COMPAT_FREEBSD4                # Compatible with FreeBSD4

options        COMPAT_FREEBSD5                # Compatible with FreeBSD5

options        COMPAT_FREEBSD6                # Compatible with FreeBSD6

options        COMPAT_FREEBSD7                # Compatible with FreeBSD7

options        SYSVSHM                        # SYSV-style shared memory

options        SYSVMSG                        # SYSV-style message queues

options        SYSVSEM                        # SYSV-style semaphores



# Bus support.

device                acpi

device                eisa

device                pci



# Floppy drives

device                fdc



# ATA and ATAPI devices

device                ata

device                atadisk                # ATA disk drives

device                atapicd                # ATAPI CDROM drives

device                atapifd                # ATAPI floppy drives

device                atapicam        # SCSI support for ATAPI devices



# SCSI peripherals

device                scbus                # SCSI bus (required for SCSI)

device                da                # Direct Access (disks)

device                cd                # CD

device                pass                # Passthrough device (direct SCSI access)





# atkbdc0 controls both the keyboard and the PS/2 mouse

device                atkbdc                # AT keyboard controller

device                atkbd                # AT keyboard

device                psm                # PS/2 mouse

device                kbdmux                # keyboard multiplexer

device                vga                # VGA video card driver



# syscons is the default console driver, resembling an SCO console

device                sc





# Parallel port

device                ppc

device                ppbus                # Parallel port bus (required)

#device                lpt                # Printer

device                ppi                # Parallel port interface device





# PCI Ethernet NICs that use the common MII bus controller code.

# NOTE: Be sure to keep the 'device miibus' line in order to use these NICs!

device                miibus                # MII bus support

device                fxp                # Intel EtherExpress PRO/100B (82557, 82558)



# Pseudo devices.

device                loop                # Network loopback

device                random                # Entropy device

device                ether                # Ethernet support

device                pty                # BSD-style compatibility pseudo ttys

device                md                # Memory "disks"



# The `bpf' device enables the Berkeley Packet Filter.

# Be aware of the administrative consequences of enabling this!

# Note that 'bpf' is required for DHCP.

device                bpf                # Berkeley packet filter



# USB support

device                uhci                # UHCI PCI->USB interface

device                ohci                # OHCI PCI->USB interface

device                ehci                # EHCI PCI->USB interface (USB 2.0)

device                usb                # USB Bus (required)

device                umass                # Disks/Mass storage - Requires scbus    #and da

device                ums                # Mouse

# Sound support

device                sound                # Generic sound support

device                snd_ich                # ICH sound

It is logical, simple and stable.
Do you want more facts?

Quote:

Originally Posted by mobydick (Post 3796201)

It is logical, simple and stable.
Do you want more facts?

Well, if we're comparing based on stuff like that

How about topas (AIX):

Code:

Topas Monitor for host:    dev12345            EVENTS/QUEUES    FILE/TTY

Fri Dec 18 08:12:10 2009  Interval:  2        Cswitch  61802  Readch    17.9M

                                                Syscall  589.4K  Writech  28.9M

CPU  User%  Kern%  Wait%  Idle%                Reads    49222  Rawin        0

ALL  36.4  14.6  17.1  32.0                Writes    17632  Ttyout      714

                                                Forks      265  Igets        0

Network  KBPS  I-Pack  O-Pack  KB-In  KB-Out  Execs      383  Namei    118.8K

en4      30.8K    20.6K  16.3K 9032.3    22.0K Runqueue  14.0  Dirblk        0

en0    6071.7    645.6  360.0    74.0  5997.7  Waitqueue  0.0

lo0    1298.3    946.2  946.2  649.1  649.1

en5      2.9    10.0    11.0    1.3    1.6  PAGING          MEMORY

                                                Faults    71241  Real,MB  220671

Disk    Busy%    KBPS    TPS KB-Read KB-Writ  Steals        0  % Comp    87.7

hdisk159 41.9    24.5K  33.5    24.5K    0.0  PgspIn        0  % Noncomp  12.2

hdisk52  40.4    24.2K  34.0    24.2K    0.0  PgspOut      0  % Client  12.2

hdisk266 37.0    21.4K  28.5    21.4K    0.0  PageIn      655

hdisk373 35.0    20.5K  28.0    20.5K    0.0  PageOut    609  PAGING SPACE

                                                Sios      1331  Size,MB  128000

Name            PID  CPU%  PgSp Owner                            % Used      2.2

pss_rx      1200794  3.3 697.4 symai3          NFS (calls/sec)  % Free    98.8

oraexp      8196984  3.2  1.8 root            ServerV2      0

oracle      3654514  3.1  9.2 oradev          ClientV2      0  Press:

psi_os      1507552  2.7  14.8 symai3          ServerV3      4  "h" for help

sftp-ser    2920556  1.5  0.5 devsftp        ClientV3      0  "q" to quit

There's no rival for SMIT/SMITTY anywhere:

Code:

                                  System Management



Move cursor to desired item and press Enter.



  Software Installation and Maintenance

  Software License Management

  Devices

  System Storage Management (Physical & Logical Storage)

  Security & Users

  Communications Applications and Services

  Print Spooling

  Advanced Accounting

  Problem Determination

  Performance & Resource Scheduling

  System Environments

  Processes & Subsystems

  Applications

  Installation Assistant

  Cluster Systems Management

  Using SMIT (information only)









F1=Help              F2=Refresh            F3=Cancel            F8=Image

F9=Shell            F10=Exit              Enter=Do

And compared to the clean /etc/filesystems layout, /etc/fstab is a freaking mess:

Code:

/tmp:

        dev            = /dev/hd3

        vfs            = jfs2

        log            = /dev/hd8

        mount          = automatic

        check          = false

        vol            = /tmp

        free            = false

        quota          = no



/proc:

        dev            = /proc

        vol            = "/proc"

        mount          = true

        check          = false

        free            = false

        vfs            = procfs

/etc/fstab:

Code:

# device name  mount point    fs-type      options                dump-freq pass-num

LABEL=/        /              ext3        defaults                1 1

/dev/hda6      swap            swap        defaults                0 0

none            /dev/pts        devpts      gid=5,mode=620          0 0

none            /proc          proc        defaults                0 0

none            /dev/shm        tmpfs        defaults                0 0

 

# Removable media

/dev/cdrom      /mount/cdrom    udf,iso9660  noauto,owner,kudzu,ro  0 0

/dev/fd0        /mount/floppy  auto        noauto,owner,kudzu      0 0

My point being - this is not the basis for a system comparison.
You need to compare on the architecture of the system, the stability of the kernel, the accessibility of the support, and the range of systems supported.
If we're simply going by the beauty and elegance of some utilities and tools, AIX and VMS are lightyears ahead of any free project... deservedly so.

@mobydick

Facts! Those are no facts. Top and FreeBSD kernel config. So what? DO you want to see some Irix or Solaris? Most driver options in Linux are in kernel config, in FreeBSD most driver options are hidden in sysctl's for example. Do you know any sysctl? Even most of the developers just know a bunch of those. It's a mess, especially if you have to cope with a problem. How many drivers got FreeBSD compared to Linux? So Linux kernel config is a mess, missing most of the drivers and technologies of Linux it's no miracle to see FreeBSD shine in this area.

Advantages would be:

1. lots of manpages and actually readable ones
2. developed as a whole system

I didn't mention stability, because stability depends on the hardware context. If you have got better support for your hardware, especially for some quirks FreeBSD isn't top, it's Linux that shines in this area.

FreeBSD has got more and better development cycles in contrast to Linux, that's a fact. But again this is no miracle, with only a "few" testers you have to do it that way. With 1000 times more testers you can do it the messy way like Linux.

But Linux is a massive showstopper if it comes to documentation or continuity (especially in terms of how to configure it from release to release).

I like *BSD, but it's more a matter of personal bias, because it's from release to release familiar terrain. Vice versa Volkerding achieves similar qualities with Slackware.

I compared FreeBSD to Linux distros, not other UNIX flavours.

Quote:

Originally Posted by Oliver_H (Post 3797704)

1. lots of manpages and actually readable ones
2. developed as a whole system

These I forgot to mention.

Quote:

Originally Posted by Oliver_H (Post 3797704)

I like *BSD, but it's more a matter of personal bias, because it's from release to release familiar terrain. Vice versa Volkerding achieves similar qualities with Slackware.

This is my favourite Linux distribution.
It is designed like FreeBSD, to be simple.
I have been running FreeBSD for more than 2 years (started with 6.0-RELEASE).
One cannot call me a fanboy, because I do all my day-to-day
work on my FreeBSD box.
And I got used to it. Now I find Linux distributions awkward and strange.
Tried a few times OpenSolaris, but I found it mysterious and
gave up the idea of learning it.

Quote:

Originally Posted by MBybee (Post 3796786)

Well, if we're comparing based on stuff like that

How about topas (AIX):

Code:

Topas Monitor for host:    dev12345            EVENTS/QUEUES    FILE/TTY

Fri Dec 18 08:12:10 2009  Interval:  2        Cswitch  61802  Readch    17.9M

                                                Syscall  589.4K  Writech  28.9M

CPU  User%  Kern%  Wait%  Idle%                Reads    49222  Rawin        0

ALL  36.4  14.6  17.1  32.0                Writes    17632  Ttyout      714

                                                Forks      265  Igets        0

Network  KBPS  I-Pack  O-Pack  KB-In  KB-Out  Execs      383  Namei    118.8K

en4      30.8K    20.6K  16.3K 9032.3    22.0K Runqueue  14.0  Dirblk        0

en0    6071.7    645.6  360.0    74.0  5997.7  Waitqueue  0.0

lo0    1298.3    946.2  946.2  649.1  649.1

en5      2.9    10.0    11.0    1.3    1.6  PAGING          MEMORY

                                                Faults    71241  Real,MB  220671

Disk    Busy%    KBPS    TPS KB-Read KB-Writ  Steals        0  % Comp    87.7

hdisk159 41.9    24.5K  33.5    24.5K    0.0  PgspIn        0  % Noncomp  12.2

hdisk52  40.4    24.2K  34.0    24.2K    0.0  PgspOut      0  % Client  12.2

hdisk266 37.0    21.4K  28.5    21.4K    0.0  PageIn      655

hdisk373 35.0    20.5K  28.0    20.5K    0.0  PageOut    609  PAGING SPACE

                                                Sios      1331  Size,MB  128000

Name            PID  CPU%  PgSp Owner                            % Used      2.2

pss_rx      1200794  3.3 697.4 symai3          NFS (calls/sec)  % Free    98.8

oraexp      8196984  3.2  1.8 root            ServerV2      0

oracle      3654514  3.1  9.2 oradev          ClientV2      0  Press:

psi_os      1507552  2.7  14.8 symai3          ServerV3      4  "h" for help

sftp-ser    2920556  1.5  0.5 devsftp        ClientV3      0  "q" to quit

There's no rival for SMIT/SMITTY anywhere:

Code:

                                  System Management



Move cursor to desired item and press Enter.



  Software Installation and Maintenance

  Software License Management

  Devices

  System Storage Management (Physical & Logical Storage)

  Security & Users

  Communications Applications and Services

  Print Spooling

  Advanced Accounting

  Problem Determination

  Performance & Resource Scheduling

  System Environments

  Processes & Subsystems

  Applications

  Installation Assistant

  Cluster Systems Management

  Using SMIT (information only)









F1=Help              F2=Refresh            F3=Cancel            F8=Image

F9=Shell            F10=Exit              Enter=Do

And compared to the clean /etc/filesystems layout, /etc/fstab is a freaking mess:

Code:

/tmp:

        dev            = /dev/hd3

        vfs            = jfs2

        log            = /dev/hd8

        mount          = automatic

        check          = false

        vol            = /tmp

        free            = false

        quota          = no



/proc:

        dev            = /proc

        vol            = "/proc"

        mount          = true

        check          = false

        free            = false

        vfs            = procfs

/etc/fstab:

Code:

# device name  mount point    fs-type      options                dump-freq pass-num

LABEL=/        /              ext3        defaults                1 1

/dev/hda6      swap            swap        defaults                0 0

none            /dev/pts        devpts      gid=5,mode=620          0 0

none            /proc          proc        defaults                0 0

none            /dev/shm        tmpfs        defaults                0 0

 

# Removable media

/dev/cdrom      /mount/cdrom    udf,iso9660  noauto,owner,kudzu,ro  0 0

/dev/fd0        /mount/floppy  auto        noauto,owner,kudzu      0 0

Its funny that you mentioned smit. I am actually looking at using java to build a smit clone for linux. That is one tool that I feel is really missing from linux. It makes AIX administration so much easier for certain tasks.

Quote:

Originally Posted by jstephens84 (Post 3798591)

Good plan - not sure I'd care for it in Java, per se... but yes. I have often wished that other Unix systems (and clones, like Linux) had smit. Especially that 'show command' bit that makes it easy to automate complex or weird menu paths.

Also - comparing based on a few specific generic tools is every bit as silly as comparing Unix systems. Linux is, like it or not, much more deeply entrenched in the big company mindsets. It's the tools like Veritas, Oracle, DB2, SAP, Peoplesoft and such that matter. Not tools like top.

Systems that use BSD in the background like appliances and such are fighting a losing battle for mindshare.

<edit> It's probably worth mentioning that a lot of the 'problems' with Linux come from the pressure applied by companies like IBM pushing their code into Linux... desktop kernels shouldn't really be built the same way as server kernels. </edit>

sysinstall would be probably the closest thing FreeBSD has to smit, and that's a stretch...or a starting point, depending on how you look at it.

Quote:

Originally Posted by rocket357 (Post 3799962)

sysinstall would be probably the closest thing FreeBSD has to smit, and that's a stretch...or a starting point, depending on how you look at it.

I'd call it a starting point - you might be able to tuck some of it under the 'post install config' section, but I think you'd really just want to start over, but make it feel like sysinstall.

BTW - sysinstall is SO missing from Linux. Every Linux distro seems to have grown some method of replacing it, but usually they just pile tools together until they get there. GParted, User Manager, etc.

I'm a firm believer in "wrap your habits around the OS, not the OS around your habits" (i.e. I learn the "native" OS tools moreso than write my own, unless of course the native tools just aren't up to par), but a system configuration manager (even a simple script) would be nice for Linux/BSD.

Quote:

Originally Posted by mobydick (Post 3798577)

Tried a few times OpenSolaris, but I found it mysterious and gave up the idea of learning it.

Same here. I'm trying to get my mind around FreeBSD.

-JJ

Quote:

Originally Posted by MBybee (Post 3684336)

Is it going anywhere? Was my experience just unusual? What experiences have you guys had with it?

I just tried it for the first time yesterday. It bombed. I burned the boot-only CD, then performed a network install over the Internet using that disc. The installation seemed to be going well, although it took about 5 hours. When it finished, I had a brain-dead system, unable to load Xorg, and it got caught in an infinite loop with a 640x480 text screen scrolling the same four lines forever.:( Strange, because the GUI installer looked good and seemed to handle my monitor and video card without problems.

I found a report of an identical experience a couple of months ago on the PCBSD forum, different video card, different monitor, so my experience obviously wasn't a fluke. A seemingly more experienced user suggested booting into single-user mode and editing the Xorg configuration by hand. I can't believe it! As much maligned as Bill Gates is, he never would have suggested that Windows 95 users tweak video drivers by hand in order to get the GUI to come up! I carefully transcribed what I was seeing on my screen with pencil and paper and added my comment to the thread several hours ago, and it has yet to appear while the moderator "approves" my posting. What a joke -- the Ubuntu forums could never operate in this manner.

Face it: PCBSD is going nowhere. Unless they can come up with a Live CD version that lets people prove it out quickly without having to spend half a day, only to face what looks like an insurmountable roadblock to the novice, it won't catch on.