OpenBSD Journal

OpenBSD 4.4 released

Contributed by johan on from the we-owe-you-UCB dept.

The OpenBSD team is pleased to announce the release of OpenBSD 4.4. See the announcement for more information.

Place an order worldwide or order in Europe, or if you only download from FTP then make a donation. If you do FTP your release, be sure to use a local mirror and not the main ftp server:

o spargel.kd85.com  Austria       o ftp.eu.openbsd.org  Sweden
o anga.funkfeuer.at  Austria       o openbsd...uni-erlangen.de  Germany
o obsd.cec.mtu.edu  MI, USA       o ftp3.usa.openbsd.org  CO, USA
o ftp5.usa.openbsd.org  CA, USA       o rt.fm  IL, USA

Highlights of OpenBSD 4.4 are listed below.

  • New/extended platforms:
    • OpenBSD/sparc64
      Much more platform support. Machines using the UltraSPARC IV/T1/T2 and Fujitsu SPARC64-V/VI/VII are now supported.
    • OpenBSD/socppc
      For machines based on the Freescale MPC8349E System-on-Chip (SoC) platform that use Das U-Boot as a boot loader.
    • OpenBSD/landisk
      Added shared libraries support.

  • Improved hardware support, including:
    • Several new/improved drivers for sensors, including:
      • New fins(4) driver for Fintek F71805F LPC Super I/O Hardware Monitor.
      • New andl(4) driver for Andigilog aSC7611 Hardware Monitor.
      • New kate(4) driver for AMD K8 embedded temperature sensor.
      • New sdtemp(4) driver for JEDEC JC-42.4 memory module thermal sensors.
      • New adtfsm(4) driver for Analog Devices ADT7462.
      • New km(4) driver for AMD K10 (Phenom, Opteron Barcelona et alia) embedded temperature sensor.
      • The it(4) driver now has better probing procedure.
      • The lmtemp(4) driver now support LM76.
    • Support for the Intel G33 and G35 chipsets in agp(4).
    • New lii(4) driver for Attansic L2 10/100 Ethernet devices.
    • Preliminary support for UVC USB webcams: uvideo(4) and video(4).
    • The openchrome(4) driver replaces the via X.Org driver for VIA video cards.
    • AMD Geode video card driver for X.Org.
    • New vmt(4) driver which implements VMware Tools.
    • New auglx(4) driver for AMD Geode LX CS5536 integrated AC'97 audio.
    • New ix(4) driver for Intel 82598 PCI Express 10Gb Ethernet.
    • New acpithinkpad(4) driver provides additional ACPI support for IBM/Lenovo ThinkPad laptops.
    • New acpiasus(4) driver provides additional ACPI support for ASUS laptops, including the EeePC.
    • New gecko(4) driver supporting the GeckoBOA BC GSC+ port found on some hppa systems.
    • New tsec(4) driver supporting the Freescale Triple Speed Ethernet Controller.
    • The re(4) driver now supports RTL8102E and RTL8168 devices.
    • The cas(4) driver now supports National Semiconductor Saturn devices.
    • The pccom(4) driver has been removed; all platforms use com(4) now.
    • cardbus(4) and pcmcia(4) now work on most sparc64 machines.
    • The udcf(4) driver now supports mouseCLOCK USB II devices.
    • The msk(4) driver now supports 88E8040T devices.
    • The ath(4) driver now supports many more Atheros wireless devices.
    • The ciss(4) driver now supports HP Smart Array P212, P410, P411, P411i and P812 devices.
    • The uftdi(4) driver now supports ELV Elektronik and FTDI 2232L devices.
    • The umsm(4) driver now supports Option GlobeTrotter 3G+, Huawei E220 and more HSDPA MSM devices.
    • The ubsa(4) driver now supports ZTE CMDMA MSM devices.
    • The axe(4) driver now supports Apple USB A1277 devices.
    • The puc(4) driver now supports more Netmos devices.
    • The mgx(4) driver now supports 2D acceleration on selected boards.
    • The isp(4) driver firmware for some controllers has been updated.
    • The isp(4) driver no longer hangs during probe on some machines.
    • The bge(4) driver has better support for BCM5704 chipsets in fiber mode which helps with some blade servers.
    • The bge(4) driver has better support for the BCM5906 chipset on some systems.
    • The bge(4) driver has much better support for PCI Express chipsets resulting in much faster transmit performance.
    • The bge(4) driver has support for the BCM5714/5715/5780 chipsets using fiber interfaces.
    • The bnx(4) driver has support for the BCM5706/5708 chipsets using fiber interfaces.
    • The ral(4) driver now supports Ralink Technology RT2700 devices.
    • Serial ports other than com0 can now be used for console on amd64.
    • The serial console on i386 and amd64 has improved compatibility with server management cards.

  • New tools:
    • rpc.statd(8), the host status monitoring daemon for use with the NFS file locking daemon.
    • Initial import of ypldap(8), a drop-in replacement for ypserv to glue in an LDAP directory for get{pw,gr}ent family of functions.
    • Deprecated slattach(8) and nmeaattach(8) in favor of ldattach(8).
    • Import of tcpbench(1), a small TCP benchmarking tool.

  • New functionality:
    • aucat(1) is now able to play and record audio in full-duplex, mix unlimited number of streams, handle up to 16 channels, resample streams on the fly, support various 24-bit and 32-bit encodings and does format conversions on the fly.
    • httpd(8) now supports IPv6.
    • dhcpd(8) now supports basic synchronization of the /var/db/dhcpd.leases file to allow running multiple instances for redundancy.
    • WPA/WPA2-PSK support for several models of wireless cards.
    • rpc.lockd(8) now supports NLMv4 and does actually do locking.
    • ftp(1) now supports recursive mget transfers.
    • ftp(1) now uses keep alive packets by default.
    • ftp(1) now accepts empty passwords in URLs.
    • locate(1) now supports -b flag to perform search only on the last component of the path.
    • Allow cdio(1) in TAO mode to set the write speed.
    • cdio(1) no longer blanks media twice.
    • Add ability in cdio(1) to determine media capabilities and make it figure out if media supports TAO or blanking.
    • Initial version of softraid(4) crypto support.
    • dhcpd(8) now groks options tftp-config-file and auto-proxy-script in dhcpd.conf(5).
    • dhclient(8) option handling much more resistant to abuse.
    • dhclient(8) now aware of interface link state and reacts to changes.
    • DIOCRLDINFO, DIOCGPDINFO, and DIOCGPART support added to block devices previously lacking it.
    • disklabel(8) no longer supports the '-r' option, and obtains all disklabel information via ioctl's.
    • disklabel(8) no longer suggests offsets and sizes that would result in partitions starting or ending outside the OpenBSD section of the disk.
    • disklabel(8) now correctly reads back the 'vendor' field from text disklabels.
    • disklabel(8) editor mode '?' and 'p' commands are more compact and the 'l' command has been added to produce previous verbose output.
    • I/O's outside the bounds of the RAW_PART are now prevented, allowing proper detection of invalid I/O's.
    • USB floppies now have a valid cylinder count calculated, rather than 0.
    • newfs(8) can now create filesystems on devices with sector sizes other than 512, although such filesystems cannot yet be read.
    • rtadvd(8) now revokes its privileges and runs as its own user _rtadvd.
    • scsi(4) probing displays less useless verbiage and fewer spurious error messages.
    • st(4) devices can now be detached.
    • ATAPI devices are now identified as such, rather than as SCSIn devices.
    • SATA tape drives now work.
    • scsi(4) probing now displays the ID of the initiator on the bus.
    • scsi(4) debug capabilities improved to show commands and input or output data as appropriate.
    • scsi(4) probing makes better use of the TEST UNIT READY command to clear errors and allow successful attachments.
    • scsi(4) probing can now find more fibre channel attached devices.
    • Several mbuf pool cache corruption issues were fixed.
    • identd(8) now supports IPv6 in standalone mode.
    • cal(1) now shows week numbers too.
    • In pf(4), implement a sloppy tcpstate tracker which does not look at sequence numbers at all.
    • pf(4) rule accounting now has a counter to record how many states in total have been created by a rule.
    • The kill states feature in pfctl(8) now supports two additional match targets: Kill by rule label or state ID.
    • Make relayd(8) use sloppy pf(4) state keeping for routed sessions (Direct Server Return).
    • Added support in relayd(8) for transparent L7 forwarding in relays.
    • Added support for dynamic IPv6-to-IPv4 or IPv4-to-IPv6 TCP relays in relayd(8).
    • Improved the DNS mode and use OpenBSD's Id shuffle code from named(8). in relayd(8).
    • Extend awk(1) with bitwise operations.
    • Updated the display code for systat(1) which adds views for pf(4) states, rules and queues.
    • Imported initial support for IEEE 802.3ad/LACP in trunk(4).

  • Assorted improvements and code cleanup:
    • A greatly changed buffer cache subsystem which maps cache pages only when in use, resulting in improved filesystem performance, and allowing for the effective use of a much larger buffer cache.
    • A greatly improved implementation of malloc(3), the general purpose memory allocator, which catches more mistakes, reduces address space fragmentation, and is faster.
    • The statfs(2) system call has been enhanced to support large filesystems.
    • The strtof(3) function has been added to libc.
    • A lot of work has been done on libm to add several functions towards more C99 compliance.
    • Lots of features have been implemented in OpenCVS, which can now be used to do some real work.
    • New APIs for arc4random(3), one to fill a buffer with random numbers and the other to return a uniformly distributed random number without bias.

  • Install/Upgrade process changes:
    • A new tool sysmerge(8), derived from the old mergemaster port, makes it easier to merge configuration files changes during an upgrade.
    • Fully support OpenBSD inside extended partitions on i386 and amd64.
    • During installation 'dhcp' is now the initial default answer during network configuration.
    • Fetching sets via FTP more reliable due to automatic use of keep alive.
    • Fetching sets via NFS no longer hangs retrying a non-functional mount.
    • Installation ensures hostname.* files are installed with mode 600.
    • Serial console configuration now automatically detects speed.
    • Serial console support extended to all architectures.
    • Partition size display no longer limited to 32 bit sizes.
    • Partition sizes now scaled and formatted to use human readable units.
    • ntpd(8) configuration questions improved.
    • Sparc miniroot root disk detection fixed.
    • Invocations of disklabel(8) by the scripts are now less verbose.

  • OpenSSH 5.1:
    • New experimental fingerprint ASCII art visualisation system for easier verification of remote keys.
    • Added chroot(2) support for sshd(8).
    • Added an extended test mode (-T) to sshd(8).
    • Make ssh(1) support negation of groups in a "Match group" block.
    • Increased the ephemeral key size in protocol1 from 768 to 1024 bits.
    • Better tests of primes in /etc/moduli
    • Refuse to read .shosts or authorized_keys files that are not regular files.
    • Enable ~ escapes for multiplex slave sessions.
    • Support CIDR address matching in Match blocks and authorized_keys from="..." stanzas.
    • Make port forwarding code try additional addresses when connecting to a destination whose DNS name resolves to more than one address.
    • Make the maximum number of ssh(1) sessions run-time controllable via MaxSessions in ssh_config(5).
    • ssh-keyscan(1) now defaults to RSA protocol 2 keys, instead of RSA1.
    • Added an extension to sftp protocol to implement statvfs(3)-like operations and add a df command to sftp(1).
    • Disable execution of ~/.ssh/rc for sessions where a command has been forced by the sshd_config ForceCommand directive.
    • And several bug fixes and performance enhancements.

  • Over 5205 ports, minor robustness improvements in package tools.

  • Many pre-built packages for each architecture:
    • i386: 5033
    • sparc64: 4862
    • alpha: 4852
    • sh: 1285
    • amd64: 4940
    • powerpc: 4466
    • sparc: 3381
    • arm: 4018
    • hppa: 1595
    • vax: 1954
    • mips64: 3099

  • Some highlights:
    • Gnome 2.20.3
    • GNUstep 1.14.2
    • KDE 3.5.9
    • Mozilla Firefox 2.0.0.16 and 3.0.1
    • Mozilla Thunderbird 2.0.0.16
    • MySQL 5.0.51a
    • OpenMotif 2.3.0
    • OpenOffice.org 2.4.1
    • PostgreSQL 8.3.3
    • Xfce 4.4.2

  • As usual, steady improvements in manual pages and other documentation.

  • The system includes the following major components from outside suppliers:
    • Xenocara (based on X.Org 7.3 + patches, freetype 2.3.5, fontconfig 2.4.2, Mesa 7.0.3, xterm 234 and more)
    • Gcc 2.95.3 (+ patches) and 3.3.5 (+ patches)
    • Perl 5.8.8 (+ patches)
    • Our improved and secured version of Apache 1.3, with SSL/TLS and DSO support
    • OpenSSL 0.9.7j (+ patches)
    • Groff 1.15
    • Sendmail 8.14.3, with libmilter
    • Bind 9.4.2-P2 (+ patches)
    • Lynx 2.8.5rel.4 with HTTPS and IPv6 support (+ patches)
    • Sudo 1.6.9p17
    • Ncurses 5.2
    • Latest KAME IPv6
    • Heimdal 0.7.2 (+ patches)
    • Arla 0.35.7
    • Binutils 2.15 (+ patches)
    • Gdb 6.3 (+ patches)

If you'd like to see a list of what has changed between OpenBSD 4.3 and 4.4, look at plus44.html.

Thank you to all of the developers who make OpenBSD possible. Please be sure to make a donation to continue to make OpenBSD releases possible.

(Comments are closed)


Comments
  1. By Anonymous Coward (169.244.70.146) on

    W00t!

    Thank you OpenBDS team. As usual, there are a lot of nifty new features that I can't wait to try out.

  2. By Noryungi (Noryungi) noryungi@yahoo.com on

    As usual, keep up the good work.

    I love OpenBSD and I am proud to be a user and a donator to the project.

    Still waiting for my CDs though... But I ordered at the last minute, so I have only myself to blame ;-)

    Comments
    1. By Anonymous Coward (70.141.212.164) on

      > As usual, keep up the good work.
      >
      > I love OpenBSD and I am proud to be a user and a donator to the project.
      >
      > Still waiting for my CDs though... But I ordered at the last minute, so I have only myself to blame ;-)

      Well, if you already ordered it then you should have no guilt what so ever when you download it.

    2. By Anonymous Coward (85.19.213.88) on

      > Still waiting for my CDs though... But I ordered at the last minute, so
      > I have only myself to blame ;-)

      I ordered as soon as the pre-orders came up, but I ordered from a local provider and I'm still waiting for the CD-set. They've been quick to deliver before, but this is stretching it. I'm going to order one of Wim's "bloated" CD-sets next year ;-)

    3. By Paul Barnier (122.106.86.61) on

      > Still waiting for my CDs though... But I ordered at the last minute, so I have only myself to blame ;-)

      My CD & T-Shirt just arrived to-day, and I ordered kind of late-ish. So cheer up, there's hope!
      (+Thanks to the guys for timely shipping to Oz)

  3. By Markus Peloquin (incripshin) markpeloquin@gmail.com on http://cs.wisc.edu/~markus

    The announcement link doesn't work (first paragraph).

    Comments
    1. By Markus Peloquin (incripshin) on http://cs.wisc.edu/~markus

      Also, is there a reason you removed the mention of a JDK package from the highlights? I think it's rather noteworthy. As much as I dislike Java, I'm happy to see it getting opened up.

  4. By Anonymous Coward (67.69.227.99) on

    I think the syncing dhcpd feature sounds ultra awesome, just like CARP is.

    What I'm not clear on though (not clearly specified in the man page) is, is the dhcpd run in a load-balanced sense on say two (or more) servers and each syncs it's leases to the peer server(s) dhcpd.leases file?

    Or is it a case of master/backup dhcpd where the master syncs to the slave(s) server(s)?

    TIA!

    Comments
    1. By Anonymous Coward (96.21.15.58) on

      > I think the syncing dhcpd feature sounds ultra awesome, just like CARP is.
      >
      > What I'm not clear on though (not clearly specified in the man page) is, is the dhcpd run in a load-balanced sense on say two (or more) servers and each syncs it's leases to the peer server(s) dhcpd.leases file?
      >
      > Or is it a case of master/backup dhcpd where the master syncs to the slave(s) server(s)?
      >
      > TIA!

  5. By Steffen Wendzel (217.238.62.114) on

    "Some highlights: * Gnome 2.20.3 * GNUstep 1.14.2 * KDE 3.5.8 * Mozilla Firefox 2.0.0.16 and 3.0.1 ..."

    Are these versions highlights? Maybe they were it months before but they aren't _now_. These software versions include many bugs and maybe even security problems.

    Don't get me wrong; I think OBSD is a nice server O.S. but why should someone use it as a desktop O.S. if one can get KDE 4.1.x, GNOME 2.24 and the like with Ubuntu or similar current systems?

    best regards
    Steffen Wendzel

    Comments
    1. By Anonymous Coward (143.166.226.61) on

      Difference is that it works _NOW_ and not in the future. You can chase the latest and greatest bleeding edge stuff and the ramification is that you never ship anything. Settling on a version is great because it gets hashed out and "just works" unlike your eternal-daily-patching-linux-crap.

      Comments
      1. By Jeroen Janssen (J.Janssen) on http://opennsd.net

        > Difference is that it works _NOW_ and not in the future. You can chase the latest and greatest bleeding edge stuff and the ramification is that you never ship anything. Settling on a version is great because it gets hashed out and "just works" unlike your eternal-daily-patching-linux-crap.

        That, and OpenBSD is FREE, FUNCTIONAL, and SECURE. Something you can hardly say of Linux (eg. Ubuntu). Except the functional thingy maybe.

        -J

      2. By Steffen Wendzel (217.238.62.114) on

        > Difference is that it works _NOW_ and not in the future. You can chase the latest and greatest bleeding edge stuff and the ramification is that you never ship anything. Settling on a version is great because it gets hashed out and "just works" unlike your eternal-daily-patching-linux-crap.

        hm... I thought a few times about the idea of a reply to your comment since I dont' expect you to reply on a real objective way. But I will try it nevertheless ...

        If you -- for example -- mean that KDE 3.5.8 works "_NOW_", why do you not expect 3.5.10 (most linux distributions) or 4.1.2 (very new linux releases like ubuntu 8.10) to be more stable? The bugfix list between 3.5.8 and 3.5.10 for example is very very long. This means: It is more stable.

        And firefox 3.0.3 should be more stable than OpenBSDs 3.0.1 version.

        I don't expect you to agree by replying but please think about it nevertheless.

        Comments
        1. By Anonymous Coward (67.69.227.99) on

          > > Difference is that it works _NOW_ and not in the future. You can chase the latest and greatest bleeding edge stuff and the ramification is that you never ship anything. Settling on a version is great because it gets hashed out and "just works" unlike your eternal-daily-patching-linux-crap.
          >
          > hm... I thought a few times about the idea of a reply to your comment since I dont' expect you to reply on a real objective way. But I will try it nevertheless ...
          >
          > If you -- for example -- mean that KDE 3.5.8 works "_NOW_", why do you not expect 3.5.10 (most linux distributions) or 4.1.2 (very new linux releases like ubuntu 8.10) to be more stable? The bugfix list between 3.5.8 and 3.5.10 for example is very very long. This means: It is more stable.
          >
          > And firefox 3.0.3 should be more stable than OpenBSDs 3.0.1 version.
          >
          > I don't expect you to agree by replying but please think about it nevertheless.

          Are you helping with the effort involved in doing such work? Just curious to know your position.

          Comments
          1. By Anonymous Coward (71.239.81.9) on

            > > > Difference is that it works _NOW_ and not in the future. You can chase the latest and greatest bleeding edge stuff and the ramification is that you never ship anything. Settling on a version is great because it gets hashed out and "just works" unlike your eternal-daily-patching-linux-crap.
            > >
            > > hm... I thought a few times about the idea of a reply to your comment since I dont' expect you to reply on a real objective way. But I will try it nevertheless ...
            > >
            > > If you -- for example -- mean that KDE 3.5.8 works "_NOW_", why do you not expect 3.5.10 (most linux distributions) or 4.1.2 (very new linux releases like ubuntu 8.10) to be more stable? The bugfix list between 3.5.8 and 3.5.10 for example is very very long. This means: It is more stable.
            > >
            > > And firefox 3.0.3 should be more stable than OpenBSDs 3.0.1 version.
            > >
            > > I don't expect you to agree by replying but please think about it nevertheless.
            >
            > Are you helping with the effort involved in doing such work? Just curious to know your position.


            Boy...didn't see that response coming from an OpenBSD'er! That's straight out of the standard playbook, isn't it?

            The guy was asking a legitimate question and you can't just give him a straight answer. It's very simple:

            Yes or no? Is there a reason that OpenBSD selects older versions of applications? There has to be a reason. Maybe it's technical in nature....does the OpenBSD team take the extra time to audit the code and ensure security? Is it an ideological thing? Is it that you would prefer to release newer versions but lack the human resources?

            Just give him a straight answer without being a jerk. Grow up.

            Comments
            1. By Anonymous Coward (2001:470:8802:3:20f:b5ff:fe45:7cfe) on

              Stop being such a complete moron.

            2. By Anonymous Coward (80.171.13.115) on

              > > Are you helping with the effort involved in doing such work? Just curious to know your position.
              >
              >
              > Boy...didn't see that response coming from an OpenBSD'er! That's straight out of the standard playbook, isn't it?
              >

              You mean the "are you offering to shoulder some of the burden you ask others undertake on your behalf" playbook?

              Because that's what it's about. If this person is unwilling to do so, then their request is, ultimately, just more whining about not getting everything they want for free.

              > The guy was asking a legitimate question and you can't just give him a straight answer. It's very simple:
              >

              Yes, and it's something that maybe a good, long, minute's thinking would reveal to you: with limited resources, people do the best they can with what they have.

              > Yes or no? Is there a reason that OpenBSD selects older versions of applications? There has to be a reason. Maybe it's technical in nature....does the OpenBSD team take the extra time to audit the code and ensure security? Is it an ideological thing? Is it that you would prefer to release newer versions but lack the human resources?
              >

              That was a straight answer: "are you willing to do the work this would require?" is a legitimate response. It may not be the answer the original poster wanted, but it *is* an answer.

              > Just give him a straight answer without being a jerk. Grow up.

              Just because you act all butt-hurt on another's behalf doesn't mean you're not a douchebag.

            3. By phessler (2001:470:1f0a:8b2::2) on

              > Is there a reason that OpenBSD selects older versions of applications? There has to be a reason.

              The reason is that it was the most recent version when we froze the ports tree, for the release cycle. Testing the release, and getting the CDs printed simply takes time.

        2. By TeXitoi (86.68.39.30) on

          > If you -- for example -- mean that KDE 3.5.8 works "_NOW_", why do you not expect 3.5.10 (most linux distributions) or 4.1.2 (very new linux releases like ubuntu 8.10) to be more stable? The bugfix list between 3.5.8 and 3.5.10 for example is very very long. This means: It is more stable.
          >
          > And firefox 3.0.3 should be more stable than OpenBSDs 3.0.1 version.
          >
          > I don't expect you to agree by replying but please think about it nevertheless.

          First, KDE 3.5.9 is in OpenBSD 4.4.

          Then, the version of KDE and Firefox correspond to the version when the ports have been frozen.

          In current, KDE version is 3.5.10 and firefox version is 3.0.3p0.

          I do not know if there will be ports updates in stable. Ports in release will not change.

        3. By Anonymous Coward (128.171.90.200) on

          > And firefox 3.0.3 should be more stable than OpenBSDs 3.0.1 version.

          It might be, it might not, it might be exactly the same stability-wise.

          It may have more bug fixes, it may have exposed new bugs.

          Chasing the bleeding edge often means less stable, compared with older releases that have been in the field longer and have known weaknesses.

          It's really hard to say without specific details.

        4. By Marc Espie (213.41.185.88) espie@openbsd.org on

          Actually, by the time 4.4 is announced, we've moved on to other things.

          There's about a two months delay (a bit more...) between the release being frozen and tested and the announcement.

          Pushing newer stuff has never worked... we end up scrambling to fix last minute bugs. We prefer to have stuff working.

          That said, this part of the announcement is probably a bit false. I expect that 4.4 ships with kde 3.5.9, actually.

          3.5.10 came out too late for our release process.

          As far as KDE 4 goes, I did a large part of the porting work for KDE 4.0. We never activated it because it was not really useful (kde 4.0 was missing a large amount of functionality).

          I haven't had enough time to get kde 4.1 to work yet. There are still some issues, the most annoying one being that it might require us to build a lot of things with g++4 (I hope to avoid that) and hence to recompile a lot of shit...

          That's for the reasonable answer.

          Now, you still sound like a standard whiner. Let me remind you that OpenBSD developpers write the system *for themselves* first, and choose what they work on.

          You can help in various ways:
          - you can bribe us. Give money to the project. Heck, give money directly to me, and maybe I will spend more time working on kde4.
          - you can port stuff yourself (that's often non practical).
          - YOU CAN TEST STUFF !!!!


          Each release, it's the same story: we freeze things, we ask for testers, and we don't get anything. Almost no test report. MOST OF THE BUGS FOUND BEFORE THE RELEASE ARE FOUND BY DEVELOPERS.

          Yet, users still feel entitled to gripe after the release, that thing don't work as they should, or that so-and-so linux distro has more recent shit.

          Well, fuck you. Help testing. If users find bugs in a more timely maneer, then maybe, just maybe, the release process won't be the hell it currently is, and you might have a chance to see more current software in a release, BECAUSE IT WILL HAVE BEEN TESTED.

          Clear enough ?

          Comments
          1. By Anonymous Coward (217.238.24.102) on

            > As far as KDE 4 goes, I did a large part of the porting work for KDE 4.0. We never activated it because it was not really useful (kde 4.0 was missing a large amount of functionality).
            >
            > I haven't had enough time to get kde 4.1 to work yet. There are still some issues, the most annoying one being that it might require us to build a lot of things with g++4 (I hope to avoid that) and hence to recompile a lot of shit...

            Okay, I can understand this but ...

            > That's for the reasonable answer.
            >
            > Now, you still sound like a standard whiner. Let me remind you that OpenBSD developpers write the system *for themselves* first, and choose what they work on.

            That is why there are releases, install images for users and donations: because you develop it for yourself.

            > You can help in various ways:
            > - you can bribe us. Give money to the project. Heck, give money directly to me, and maybe I will spend more time working on kde4.
            > - you can port stuff yourself (that's often non practical).
            > - YOU CAN TEST STUFF !!!!

            Didn't you just call me a whiner? What do you think do you sound like? BTW. I submitted different patches within the past. I just told that OBSD is far away from providing the most stable and/or most current 3rd party software. And what did you do? You act like a child I told that your superhero isn't the best one in all affairs. Does your ego really depend on your O.S. so much?

            > Each release, it's the same story: we freeze things, we ask for testers, and we don't get anything. Almost no test report. MOST OF THE BUGS FOUND BEFORE THE RELEASE ARE FOUND BY DEVELOPERS.

            Didn't you just said that you develop the O.S. for yourself?

            > Yet, users still feel entitled to gripe after the release, that thing don't work as they should, or that so-and-so linux distro has more recent shit.

            That is why I use it. Just a simple reason. And it is pretty stable. I already added the comment that I like(d) OBSD as a server OS because of its security and clear design.

            > Well, fuck you. Help testing. If users find bugs in a more timely maneer, then maybe, just maybe, the release process won't be the hell it currently is, and you might have a chance to see more current software in a release, BECAUSE IT WILL HAVE BEEN TESTED.
            >
            > Clear enough ?

            an evidence of incapacity in an reasonable discussion. clear, yup. Talking with you is like talking with a fundamentalist. This discussion is finished for me.

            Comments
            1. By Steffen Wendzel (217.238.24.102) on

              PS. Just forgot to enter my name. The posting was not sent by some anonymous user (like the most replies here) -- it was sent by me.

              Comments
              1. By Marc Espie (213.41.185.88) espie@openbsd.org on

                > PS. Just forgot to enter my name. The posting was not sent by some anonymous user (like the most replies here) -- it was sent by me.

                Okay, so you're one of the guys who does a bit of testing, and not your average whiner.

                Well, we still don't get enough tests. So for now, the release process can't change.

                AND we could still use more manpower. Having newer software isn't free, it takes a lot of effort.

            2. By Jeroen Janssen (J.Janssen) on http://openbsd.cc

              > > As far as KDE 4 goes, I did a large part of the porting work for KDE 4.0. We never activated it because it was not really useful (kde 4.0 was missing a large amount of functionality).
              > >
              > > I haven't had enough time to get kde 4.1 to work yet. There are still some issues, the most annoying one being that it might require us to build a lot of things with g++4 (I hope to avoid that) and hence to recompile a lot of shit...
              >
              > Okay, I can understand this but ...
              >
              > > That's for the reasonable answer.
              > >
              > > Now, you still sound like a standard whiner. Let me remind you that OpenBSD developpers write the system *for themselves* first, and choose what they work on.
              >
              > That is why there are releases, install images for users and donations: because you develop it for yourself.
              >
              > > You can help in various ways:
              > > - you can bribe us. Give money to the project. Heck, give money directly to me, and maybe I will spend more time working on kde4.
              > > - you can port stuff yourself (that's often non practical).
              > > - YOU CAN TEST STUFF !!!!
              >
              > Didn't you just call me a whiner? What do you think do you sound like? BTW. I submitted different patches within the past. I just told that OBSD is far away from providing the most stable and/or most current 3rd party software. And what did you do? You act like a child I told that your superhero isn't the best one in all affairs. Does your ego really depend on your O.S. so much?
              >
              > > Each release, it's the same story: we freeze things, we ask for testers, and we don't get anything. Almost no test report. MOST OF THE BUGS FOUND BEFORE THE RELEASE ARE FOUND BY DEVELOPERS.
              >
              > Didn't you just said that you develop the O.S. for yourself?
              >
              > > Yet, users still feel entitled to gripe after the release, that thing don't work as they should, or that so-and-so linux distro has more recent shit.
              >
              > That is why I use it. Just a simple reason. And it is pretty stable. I already added the comment that I like(d) OBSD as a server OS because of its security and clear design.
              >
              > > Well, fuck you. Help testing. If users find bugs in a more timely maneer, then maybe, just maybe, the release process won't be the hell it currently is, and you might have a chance to see more current software in a release, BECAUSE IT WILL HAVE BEEN TESTED.
              > >
              > > Clear enough ?
              >
              > an evidence of incapacity in an reasonable discussion. clear, yup. Talking with you is like talking with a fundamentalist. This discussion is finished for me.

              Thanks for your not so understanding view on this. The developers made OpenBSD primarily for themself, and in SECOND for the users. So, they are thinking about us, the users as well.

              I completely understand Marc's ranting about this. You (in general, not personal) want something of the developers. Is it so "retarted" of them to ask a favor in return?

              If no one else would be using OpenBSD (besides the developers), there would be no one to test it out. But since there are users, who are actively using the operating system, which is given away for free, they have the right to ask something in return. You totally have the right to ignore them, but that would be kinda contradictional/irrational, since you are using the system. What's the problem, with sending an email when you encounter a bug? You are running -current, right? (since you call yourself a "security specialist", "open source developer" and whatsoever.

              Marc could explain himself more subtile, but since it's out of our hands how other people react...

              Anyway, you have your answer, porting takes A LOT of time, and they have to do the process all over with a new version, like Firefox 3. They can't "just" port it, it has to be secure as well.

            3. By Chris (142.132.30.47) on

              >
              > an evidence of incapacity in an reasonable discussion. clear, yup. Talking with you is like talking with a fundamentalist. This discussion is finished for me.

              This discussion was finished for you before it even started. You came here with some chip on your shoulder (who knows why), and have been taking exactly what you want from every reply, regardless of their actual content.

              A couple of minutes of trivial thinking (or barring that, a little bit of reading right here on the site) would allow you to come up with the answers you seek all on your own, but somehow you've decided you definitely *must* be more clever than everyone else here, and the profound, insightful questions you thought to ask must be brought to the people. The truth can't be suppressed!

              It's amazing to me that you accuse others of being jerks, but can't seem to see that you're being a tremendous asshole yourself. I suspect you got just the reaction you expected to get, and somehow feel clever as a result. Good work!

        5. By diw (diw) on

          > hm... I thought a few times about the idea of a reply to your comment >since I dont' expect you to reply on a real objective way. But I will >try it nevertheless ...
          >
          > If you -- for example -- mean that KDE 3.5.8 works "_NOW_", why do >you not expect 3.5.10 (most linux distributions) or 4.1.2 (very new >linux releases like ubuntu 8.10) to be more stable? The bugfix list >between 3.5.8 and 3.5.10 for example is very very long. This means: It >is more stable.
          >
          > And firefox 3.0.3 should be more stable than OpenBSDs 3.0.1 version.
          >
          > I don't expect you to agree by replying but please think about it >nevertheless.

          No worries. Here's a stab.
          There is a flaw in your reasoning.

          You say:
          >The bugfix list between 3.5.8 and 3.5.10 for example is very very >long. This means: It is more stable.

          This is inaccurate. It may be true. It doesn't have to be true.
          In my experience it is usually not true.

          Consider:
          New versions of software usually correlate with new features. Certainly with significant portions of new code. Although many bugs from previous versions may have fixes included in the latest version, time tells how many new bugs are found that did not previously exist.

          Don't take my word for it - have a look at Secunia.
          Here's an apt sample - Vulnerability Report: Linux Kernel 2.6.x - http://secunia.com/advisories/product/2719/?task=advisories
          Many of the recently discovered vulnerabilities affect the 2.6.x kernel only.

          Down to number 10 - "Linux Kernel ASN.1 BER Decoding Vulnerability".
          http://secunia.com/advisories/30580/
          Read carefully the products involved.
          "The vulnerability is reported in versions prior to 2.4.36.6 of the 2.4 branch and prior to version 2.6.25.5 of the 2.6 branch."

          In other words, many bugs are not found in older versions of the kernel. They have been introduced with new code.
          Solution to bugs? Upgrade to a newer version with bugfixes incorporated.
          Side effect? Newer versons contain new code with other bugs.

          The simple deal is that we have a high degree of confidence in what we are getting. We may not have blinkenlights, etcetera.
          Give me reliability over surprises any day of the week. :]
          It may not be perfect but it is a responsible approach to releasing software. Compare the insanity that is the large portion of the OS marketplace.

          Anecdotally, I would much rather run stable, lean, secure software any day of the week for just about everything I do. Give me an old app any day. Or better yet, a new app very carefully crafted ("we craft the code - we make the code better").
          I guess, that care takes patience.

          You said in your previous post:
          > Don't get me wrong; I think OBSD is a nice server O.S.
          Have you used OpenBSD?

          In my limited experience, users of OpenBSD understand the deal.
          The points you raise calling the legitimacy of OpenBSD into question are exactly the reasons why OpenBSD is a "nice server OS".
          What other possible reason could you have for thinking that "OBSD" is a nice server OS?
          They are also the reasons why Ubuntu is for fun.

          To recap. You said:
          > The bugfix list >between 3.5.8 and 3.5.10 for example is very very >long. This means: It >is more stable.
          Guess which part of that you got right?
          You also said:
          > And firefox 3.0.3 should be more stable than OpenBSDs 3.0.1 version.
          'Should' being the operative word. :]

          In your words:
          > I don't expect you to agree by replying but please think about it >nevertheless.

          > Are these versions highlights? Maybe they were it months before but >they aren't _now_.
          They are to me buddy. You're raining on my parade.

          Best wishes.

          Comments
          1. By Anonymous Coward (217.238.24.102) on

            > > Are these versions highlights? Maybe they were it months before but >they aren't _now_.
            > They are to me buddy. You're raining on my parade.

            In this case you shouldn't care about the rain. Do your parade nevertheless, OpenBSD 4.4 is great -- I just said that the "Highlights" aren't such great highlights in my point of view.

        6. By Marc Espie (213.41.185.88) espie@openbsd.org on


          > If you -- for example -- mean that KDE 3.5.8 works "_NOW_", why do you not expect 3.5.10 (most linux distributions) or 4.1.2 (very new linux releases like ubuntu 8.10) to be more stable? The bugfix list between 3.5.8 and 3.5.10 for example is very very long. This means: It is more stable.

          > And firefox 3.0.3 should be more stable than OpenBSDs 3.0.1 version.

          Sarcasm: wow, we didn't think of THAT one, thanks for enlightening us.

          Seriously. Sadly enough, this software is developed on linux, and tested on linux. Other oses hope it's clean enough, and run it. There are lots of dynamic aspects in kde, stuff that doesn't work when you get outside linux. FreeBSD fixes some of them (in fact, some of their fixes are wrong, or not portable at all, I've seen my share of them).

          So, yeah, you get more bug-fixes upstream, some of them introduce new issues on OpenBSD.

          Trust me, if it was THAT easy to port software to OpenBSD, we'd be faster in doing it.

          Unfortunately, a lot of people upstream don't have a clue about portability (and it gets worse, the FSF brainwashing is so thorough that a lot of developers don't believe in regular RELEASES any more, or in manpages... score -1 for stability). So you get the chance to spend a lot of time to fix portability issues.

          Comments
          1. By Steffen Wendzel (217.238.24.102) on

            > Seriously. Sadly enough, this software is developed on linux, and tested on linux. Other oses hope it's clean enough, and run it. There are lots of dynamic aspects in kde, stuff that doesn't work when you get outside linux. FreeBSD fixes some of them (in fact, some of their fixes are wrong, or not portable at all, I've seen my share of them).

            I agree in this point. It is bad that the developers don't port all their software directly to *BSD too what makes the port maintaining more difficult.

        7. By Brian Hershey (208.38.87.217) on

          > > Difference is that it works _NOW_ and not in the future. You can chase the latest and greatest bleeding edge stuff and the ramification is that you never ship anything. Settling on a version is great because it gets hashed out and "just works" unlike your eternal-daily-patching-linux-crap.
          >
          > hm... I thought a few times about the idea of a reply to your comment since I dont' expect you to reply on a real objective way. But I will try it nevertheless ...
          >
          > If you -- for example -- mean that KDE 3.5.8 works "_NOW_", why do you not expect 3.5.10 (most linux distributions) or 4.1.2 (very new linux releases like ubuntu 8.10) to be more stable? The bugfix list between 3.5.8 and 3.5.10 for example is very very long. This means: It is more stable.
          >
          > And firefox 3.0.3 should be more stable than OpenBSDs 3.0.1 version.
          >
          > I don't expect you to agree by replying but please think about it nevertheless.

          Hi Stephen,

          I think that the underlying security issues in older, unpatched versions of KDE, GNOME, and Firefox are mostly issues with the host OS itself. Because OpenBSD has mitigated most types of stack overflows with things like W^X, privilege separation, and encrypted swap, the security issues inherent in KDE-3.5.8 running on Linux do not necessarily exist when running on OpenBSD. Sure, they're older, and not as pretty, and don't burn the window off of your screen, making you all l33t, but they work. They're rock-solid stable 99.9% of the time. Maybe you should just run it, and see. I'm pretty sure that you'll love it.

          Comments
          1. By Brian Hershey (208.38.87.217) on

            > > > Difference is that it works _NOW_ and not in the future. You can chase the latest and greatest bleeding edge stuff and the ramification is that you never ship anything. Settling on a version is great because it gets hashed out and "just works" unlike your eternal-daily-patching-linux-crap.
            > >
            > > hm... I thought a few times about the idea of a reply to your comment since I dont' expect you to reply on a real objective way. But I will try it nevertheless ...
            > >
            > > If you -- for example -- mean that KDE 3.5.8 works "_NOW_", why do you not expect 3.5.10 (most linux distributions) or 4.1.2 (very new linux releases like ubuntu 8.10) to be more stable? The bugfix list between 3.5.8 and 3.5.10 for example is very very long. This means: It is more stable.
            > >
            > > And firefox 3.0.3 should be more stable than OpenBSDs 3.0.1 version.
            > >
            > > I don't expect you to agree by replying but please think about it nevertheless.
            >
            > Hi Stephen,
            >
            > I think that the underlying security issues in older, unpatched versions of KDE, GNOME, and Firefox are mostly issues with the host OS itself. Because OpenBSD has mitigated most types of stack overflows with things like W^X, privilege separation, and encrypted swap, the security issues inherent in KDE-3.5.8 running on Linux do not necessarily exist when running on OpenBSD. Sure, they're older, and not as pretty, and don't burn the window off of your screen, making you all l33t, but they work. They're rock-solid stable 99.9% of the time. Maybe you should just run it, and see. I'm pretty sure that you'll love it.

            Sorry, I meant *Steffen*.

    2. By Anonymous Coward (85.19.213.88) on

      > Don't get me wrong; I think OBSD is a nice server O.S. but why should
      > someone use it as a desktop O.S. if one can get KDE 4.1.x, GNOME 2.24
      > and the like with Ubuntu or similar current systems?

      Because OpenBSD is not Ubuntu and there's more to an OS than the desktop environment it is running?

      Both GNOME and KDE are huge and require a significant porting effort.

      Comments
      1. By jasper (80.60.145.215) on

        > > Don't get me wrong; I think OBSD is a nice server O.S. but why should
        > > someone use it as a desktop O.S. if one can get KDE 4.1.x, GNOME 2.24
        > > and the like with Ubuntu or similar current systems?
        >
        > Because OpenBSD is not Ubuntu and there's more to an OS than the desktop environment it is running?
        >
        > Both GNOME and KDE are huge and require a significant porting effort.

        yep, gnome is indeed a tad older then what ubuntu ships with, but gnome 2.20 received quite some stability and other fixes in our tree. openbsd 4.5 will most likely ship again with a new version of gnome ;-)

    3. By Jacob Meuser (24.22.108.209) jakemsr@sdf.lonestar.org on

      > "Some highlights: * Gnome 2.20.3 * GNUstep 1.14.2 * KDE 3.5.8 * Mozilla Firefox 2.0.0.16 and 3.0.1 ..."
      >
      > Are these versions highlights? Maybe they were it months before but they aren't _now_. These software versions include many bugs and maybe even security problems.
      >
      > Don't get me wrong; I think OBSD is a nice server O.S. but why should someone use it as a desktop O.S. if one can get KDE 4.1.x, GNOME 2.24 and the like with Ubuntu or similar current systems?

      isn't that sorta like saying "OpenBSD comes with apache 1.3.x. why would someone want to use OpenBSD as a server?"

      why does "desktop" mean bleeding edge KDE and Gnome???

  6. By Howard (171.159.192.10) on

    Received my CD set last Saturday, already have OpenBSD 4.4 installed on two of my systems, and was just waiting for the "official" release to put it on the rest.

    Thanks, and keep up the great work!

  7. By Jeroen Janssen (J.Janssen) on http://openbsd.cc

    The install images mirrored @ Amsterdam:

    i386: http://nachtmahr.nl/install44_i386.iso
    amd64: http://nachtmahr.nl/install44_amd64.iso
    MD5: http://nachtmahr.nl/MD5

    Cheers to the developers!

  8. By Wim Vandeputte (wvdputte) wim@kd85.com on https://kd85.com/notforsale.html

    Puffy is bloating:

    20.2G   3.8
    23.2G   3.9
    26.0G   4.0
    33.9G   4.1
    35.2G   4.2
    47.8G   4.3
    53.1G   4.4
    55.4G   4.4
    

    Comments
    1. By Anonymous Coward (85.19.213.88) on

      > Puffy is bloating:
      >
      >
      > 20.2G 3.8
      > 23.2G 3.9
      > 26.0G 4.0
      > 33.9G 4.1
      > 35.2G 4.2
      > 47.8G 4.3
      > 53.1G 4.4
      > 55.4G 4.4
      >

      Good thing you don't charge per GB then.

    2. By Brad (2001:470:8802:3:20f:b5ff:fe45:7cfe) brad at comstyle dot com on

      > Puffy is bloating:
      >
      >
      > 20.2G 3.8
      > 23.2G 3.9
      > 26.0G 4.0
      > 33.9G 4.1
      > 35.2G 4.2
      > 47.8G 4.3
      > 53.1G 4.4
      > 55.4G 4.4

      Big difference between "bloat" and more packages with each release.

      Comments
      1. By Wim Vandeputte (wvdputte) wim@kd85.com on https://kd85.com/notforsale.html

        > > Puffy is bloating:
        > > 
        > > 
        > > 20.2G   3.8
        > > 23.2G   3.9
        > > 26.0G   4.0
        > > 33.9G   4.1
        > > 35.2G   4.2
        > > 47.8G   4.3
        > > 55.4G   4.4
        > 
        > Big difference between "bloat" and more packages with each release.
        

        Wel, I'm looking at it from the FTP mirror's perspective. The load on those has increased dramaticly (judging from my personal experience).

        The culprits are:

        -rw-r--r--  1 root  122   236M Aug 13 02:06 ./amd64/install44.iso
        -rw-r--r--  1 root  122   226M Aug 11 22:18 ./alpha/install44.iso
        -rw-r--r--  1 root  122   232M Aug 11 22:34 ./macppc/install44.iso
        -rw-r--r--  1 root  122   216M Aug 13 02:13 ./i386/install44.iso
        -rw-r--r--  1 root  122   185M Aug 11 22:31 ./hppa/install44.iso
        -rw-r--r--  1 root  122   267M Aug 12 01:15 ./vax/install44.iso
        -rw-r--r--  1 root  122   222M Aug 12 04:34 ./sparc64/install44.iso
        -rw-r--r--  1 root  122   173M Aug 12 00:44 ./sparc/install44.iso
        
        and the rest is packages, mostly texlive, openoffice and wesnoth

        -rw-r--r--  1 root  wheel   253M Aug 21 03:25 ./packages/i386/texlive_texmf-docs-2007p0.tgz
        -rw-r--r--  1 root  wheel   119M Aug 21 03:25 ./packages/i386/texlive_texmf-full-2007p1.tgz
        -rw-r--r--  1 root  wheel   111M Aug 21 03:25 ./packages/i386/texlive_texmf-minimal-2007p2.tgz
        -rw-r--r--  1 root  wheel   253M Aug 25 23:25 ./packages/amd64/texlive_texmf-docs-2007p0.tgz
        -rw-r--r--  1 root  wheel   119M Aug 25 23:25 ./packages/amd64/texlive_texmf-full-2007p1.tgz
        -rw-r--r--  1 root  wheel   111M Aug 25 23:25 ./packages/amd64/texlive_texmf-minimal-2007p2.tgz
        -rw-r--r--  1 root  wheel   253M Sep 29 19:24 ./packages/arm/texlive_texmf-docs-2007p0.tgz
        -rw-r--r--  1 root  wheel   119M Sep 29 19:23 ./packages/arm/texlive_texmf-full-2007p1.tgz
        -rw-r--r--  1 root  wheel   111M Sep 29 19:23 ./packages/arm/texlive_texmf-minimal-2007p2.tgz
        -rw-r--r--  1 root  wheel   253M Oct  2 22:09 ./packages/vax/texlive_texmf-docs-2007p0.tgz
        -rw-r--r--  1 root  wheel   119M Oct  2 22:09 ./packages/vax/texlive_texmf-full-2007p1.tgz
        -rw-r--r--  1 root  wheel   111M Oct  2 22:09 ./packages/vax/texlive_texmf-minimal-2007p2.tgz
        -rw-r--r--  1 root  wheel   253M Aug 21 03:37 ./packages/sparc64/texlive_texmf-docs-2007p0.tgz
        -rw-r--r--  1 root  wheel   119M Aug 21 03:37 ./packages/sparc64/texlive_texmf-full-2007p1.tgz
        -rw-r--r--  1 root  wheel   111M Aug 21 03:38 ./packages/sparc64/texlive_texmf-minimal-2007p2.tgz
        -rw-r--r--  1 root  wheel   253M Sep  2 19:16 ./packages/alpha/texlive_texmf-docs-2007p0.tgz
        -rw-r--r--  1 root  wheel   119M Sep  2 19:16 ./packages/alpha/texlive_texmf-full-2007p1.tgz
        -rw-r--r--  1 root  wheel   111M Sep  2 19:16 ./packages/alpha/texlive_texmf-minimal-2007p2.tgz
        -rw-r--r--  1 root  wheel   253M Aug 31 06:16 ./packages/powerpc/texlive_texmf-docs-2007p0.tgz
        -rw-r--r--  1 root  wheel   119M Aug 31 06:17 ./packages/powerpc/texlive_texmf-full-2007p1.tgz
        -rw-r--r--  1 root  wheel   111M Aug 31 06:17 ./packages/powerpc/texlive_texmf-minimal-2007p2.tgz
        -rw-r--r--  1 root  wheel   253M Sep 24 02:37 ./packages/sh/texlive_texmf-docs-2007p0.tgz
        -rw-r--r--  1 root  wheel   119M Sep 24 02:37 ./packages/sh/texlive_texmf-full-2007p1.tgz
        -rw-r--r--  1 root  wheel   111M Sep 24 02:37 ./packages/sh/texlive_texmf-minimal-2007p2.tgz
        -rw-r--r--  1 root  wheel   253M Sep 24 00:40 ./packages/sparc/texlive_texmf-docs-2007p0.tgz
        -rw-r--r--  1 root  wheel   111M Sep 24 00:40 ./packages/sparc/texlive_texmf-minimal-2007p2.tgz
        -rw-r--r--  1 root  wheel   253M Sep  2 19:07 ./packages/mips64/texlive_texmf-docs-2007p0.tgz
        -rw-r--r--  1 root  wheel   119M Sep  2 19:08 ./packages/mips64/texlive_texmf-full-2007p1.tgz
        -rw-r--r--  1 root  wheel   111M Sep  2 19:08 ./packages/mips64/texlive_texmf-minimal-2007p2.tgz
        
        
        -rw-r--r--  1 root  wheel   146M Aug 21 03:22 ./packages/i386/openoffice-2.4.1p0.tgz
        -rw-r--r--  1 root  wheel   148M Aug 25 23:23 ./packages/amd64/openoffice-2.4.1p0.tgz
        
        -rw-r--r--  1 root  wheel  67.1M Aug 21 03:26 ./packages/i386/wesnoth-1.4p1-lite.tgz
        -rw-r--r--  1 root  wheel   144M Aug 21 03:26 ./packages/i386/wesnoth-1.4p1-tinygui.tgz
        -rw-r--r--  1 root  wheel   144M Aug 21 03:26 ./packages/i386/wesnoth-1.4p1.tgz
        -rw-r--r--  1 root  wheel  67.0M Aug 25 23:26 ./packages/amd64/wesnoth-1.4p1-lite.tgz
        -rw-r--r--  1 root  wheel   144M Aug 25 23:26 ./packages/amd64/wesnoth-1.4p1-tinygui.tgz
        -rw-r--r--  1 root  wheel   144M Aug 25 23:26 ./packages/amd64/wesnoth-1.4p1.tgz
        -rw-r--r--  1 root  wheel  67.7M Sep 29 19:23 ./packages/arm/wesnoth-1.4p1-lite.tgz
        -rw-r--r--  1 root  wheel   144M Sep 29 19:23 ./packages/arm/wesnoth-1.4p1-tinygui.tgz
        -rw-r--r--  1 root  wheel   144M Sep 29 19:23 ./packages/arm/wesnoth-1.4p1.tgz
        -rw-r--r--  1 root  wheel  67.1M Aug 21 03:38 ./packages/sparc64/wesnoth-1.4p1-lite.tgz
        -rw-r--r--  1 root  wheel   144M Aug 21 03:38 ./packages/sparc64/wesnoth-1.4p1-tinygui.tgz
        -rw-r--r--  1 root  wheel   144M Aug 21 03:38 ./packages/sparc64/wesnoth-1.4p1.tgz
        -rw-r--r--  1 root  wheel  67.8M Sep  2 19:17 ./packages/alpha/wesnoth-1.4p1-lite.tgz
        -rw-r--r--  1 root  wheel   145M Sep  2 19:17 ./packages/alpha/wesnoth-1.4p1.tgz
        -rw-r--r--  1 root  wheel   145M Sep  2 19:17 ./packages/alpha/wesnoth-1.4p1-tinygui.tgz
        

        Comments
        1. By Anonymous Coward (219.90.220.101) on

          > The culprits are:
          > 
          > 
          > -rw-r--r--  1 root  122   236M Aug 13 02:06 ./amd64/install44.iso
          > -rw-r--r--  1 root  122   226M Aug 11 22:18 ./alpha/install44.iso
          > -rw-r--r--  1 root  122   232M Aug 11 22:34 ./macppc/install44.iso
          > -rw-r--r--  1 root  122   216M Aug 13 02:13 ./i386/install44.iso
          > -rw-r--r--  1 root  122   185M Aug 11 22:31 ./hppa/install44.iso
          > -rw-r--r--  1 root  122   267M Aug 12 01:15 ./vax/install44.iso
          > -rw-r--r--  1 root  122   222M Aug 12 04:34 ./sparc64/install44.iso
          > -rw-r--r--  1 root  122   173M Aug 12 00:44 ./sparc/install44.iso
          
          When I go to grab a snapshot, I'll often ftp to a directory, mget * and wander off for lunch. Recently (last year or so) the big .iso files have been present and I've not noticed them until after I've returned to the terminal. Whoops. Didn't really need the .iso.

          ...And mget * is so much easier than mget *.tgz INSTALL bsd.rd bsd bsd.mp boot (maybe netboot, lif44.fs, etc)

          Comments
          1. By Pierre Riteau (82.249.66.180) on

            >
            > The culprits are:
            >
            >
            > -rw-r--r-- 1 root 122 236M Aug 13 02:06 ./amd64/install44.iso
            > -rw-r--r-- 1 root 122 226M Aug 11 22:18 ./alpha/install44.iso
            > -rw-r--r-- 1 root 122 232M Aug 11 22:34 ./macppc/install44.iso
            > -rw-r--r-- 1 root 122 216M Aug 13 02:13 ./i386/install44.iso
            > -rw-r--r-- 1 root 122 185M Aug 11 22:31 ./hppa/install44.iso
            > -rw-r--r-- 1 root 122 267M Aug 12 01:15 ./vax/install44.iso
            > -rw-r--r-- 1 root 122 222M Aug 12 04:34 ./sparc64/install44.iso
            > -rw-r--r-- 1 root 122 173M Aug 12 00:44 ./sparc/install44.iso
            >
            >
            > When I go to grab a snapshot, I'll often ftp to a directory, mget * and wander off for lunch. Recently (last year or so) the big .iso files have been present and I've not noticed them until after I've returned to the terminal. Whoops. Didn't really need the .iso.
            >
            > ...And mget * is so much easier than mget *.tgz INSTALL bsd.rd bsd bsd.mp boot (maybe netboot, lif44.fs, etc)

            Well, you know that you can write a one-liner script to download only the required files, which would be even easier than using "mget *", don't you?

            Comments
            1. By mho (130.237.216.121) undeadly@mho.nu on

              > > ...And mget * is so much easier than mget *.tgz INSTALL bsd.rd bsd bsd.mp boot (maybe netboot, lif44.fs, etc)
              >
              > Well, you know that you can write a one-liner script to download only the required files, which would be even easier than using "mget *", don't you?

              Not easier than just doing
              ftp ftp://ftp.se.openbsd.org/pub/OpenBSD/4.4/amd64/*

              - mho

              PS
              I cannot seem to manage to persuade firefox-3.01 to let me login with undeadly's current cert...

              - mho

              Comments
              1. By Jeroen Janssen (J.Janssen) on http://openbsd.cc

                > > > ...And mget * is so much easier than mget *.tgz INSTALL bsd.rd bsd bsd.mp boot (maybe netboot, lif44.fs, etc)
                > >
                > > Well, you know that you can write a one-liner script to download only the required files, which would be even easier than using "mget *", don't you?
                >
                > Not easier than just doing
                > ftp ftp://ftp.se.openbsd.org/pub/OpenBSD/4.4/amd64/*
                >
                > - mho
                >
                > PS
                > I cannot seem to manage to persuade firefox-3.01 to let me login with undeadly's current cert...
                >
                > - mho
                >

                Install the CACert root certificate via http://www.cacert.org/index.php?id=3, that fixes the issues. (not only undeadly, but every site using CACert)

                Comments
                1. By Magnus Holmberg (mho) on


                  > > PS
                  > > I cannot seem to manage to persuade firefox-3.01 to let me login with undeadly's current cert...
                  > >
                  > > - mho
                  > >
                  >
                  > Install the CACert root certificate via http://www.cacert.org/index.php?id=3, that fixes the issues. (not only undeadly, but every site using CACert)

                  Thanks, that did the trick!

                  - mho

  9. By Anthony (2001:470:e828:100:207:e9ff:fe39:24e8) on

    "Make pf(4) counters on table addresses optional, and disabled by default. Saves about 40% memory if counters are not being used."

    This will definitely come in handy for large tables.

  10. Comments
    1. By diw (diw) on

      > I'd like to take a moment to thanks Mr Evers (AKA Theo) for his contributions to humanity!

      Haha.

      http://www.encyclopediadramatica.com/Joseph_Evers

      Best wishes.

    2. By Anonymous Coward (70.173.233.82) on

      > I'd like to take a moment to thanks Mr Evers (AKA Theo) for his contributions to humanity!

      Is this something we would need to read 4chan to understand? Wtf is with this meme?

  11. By Anonymous Coward (2001:470:d0d3:0:5054:ff:fe00:410) on

    I've said this every release for maybe 5 years, on IRC, on forums and even in person to the people selling the CDs:

    May I PLEASE see a file listing from the CDs? I want to confirm that my program is on the official CDs and not just on the FTP.

    If it is on there, I will buy the CDs. If it is not then I will not. If I don't see a listing then I will not buy them.

    This may sound direct and rude, but remember that I've been trying to get this information every release for the last maybe 10 releases and nobody will answer me. (the guy selling the CDs didn't even understand me)

    Comments
    1. By tedu (udet) on

      > I've said this every release for maybe 5 years, on IRC, on forums and even in person to the people selling the CDs:
      >
      > May I PLEASE see a file listing from the CDs? I want to confirm that my program is on the official CDs and not just on the FTP.
      >
      > If it is on there, I will buy the CDs. If it is not then I will not. If I don't see a listing then I will not buy them.
      >
      > This may sound direct and rude, but remember that I've been trying to get this information every release for the last maybe 10 releases and nobody will answer me. (the guy selling the CDs didn't even understand me)

      OpenBSD is on the CDs. Sometimes other software is on them too, but that's just a bonus.

      Really, what's your motivation? You'll pay 50 bucks for a CD with tuxracer on it, but 0 for a CD without, even if it's a free download? Why would you pay even if it is on the CD?

      Comments
      1. By Anthony (2001:470:e828:100:207:e9ff:fe39:24e8) on

        > OpenBSD is on the CDs. Sometimes other software is on them too, but that's just a bonus.
        >
        > Really, what's your motivation? You'll pay 50 bucks for a CD with tuxracer on it, but 0 for a CD without, even if it's a free download? Why would you pay even if it is on the CD?

        It seems easier to just make a donation. :)

        I dunno about anyone else, but all my OpenBSD stuff these days is on hardware that's easier to netboot than to CD boot, and the requirement to be set up to build patches renders having the CDs of the software as it stood at release kinda moot. All I ever end up using is the stickers.

        The flow of money is necessary to keep the project around, but it's the project I want, not CDs.

      2. By Anonymous Coward (195.178.184.70) on

        > Really, what's your motivation? You'll pay 50 bucks for a CD with tuxracer on it, but 0 for a CD without, even if it's a free download?

        Yes. As a souvenir. As I said. *MY* program.

        Why would you pay even if it is on the CD?

    2. By Wim Vandeputte (wvdputte) wim@kd85.com on https://kd85.com/notforsale.html


      > This may sound direct and rude, but remember that I've been trying to get this information every release for the last maybe 10 releases and nobody will answer me. (the guy selling the CDs didn't even understand me)

      I' not sure who 'the guy selling the CDs' is but I don't remember ever have received a request like this

      mail me at wim@kd85.com and I'll see if I can answer your question

      what file name are you looking for? What architecture?

      Comments
      1. By Anonymous Coward (195.178.184.70) on

        > I' not sure who 'the guy selling the CDs' is but I don't remember ever have received a request like this

        Sorry, not *the* guy. I should have said "a" guy selling OpenBSD stuff in the OpenBSD tent at WTH 2005.

        > mail me at wim@kd85.com and I'll see if I can answer your question

        Will do.

    3. By Anonymous Coward (128.171.90.200) on

      > May I PLEASE see a file listing from the CDs? I want to confirm that my program is on the official CDs and not just on the FTP.


      % my
      my: Command not found.

  12. By George Koehler (kernigh) xkernigh@netscape.net on http://kernigh.pbwiki.com/OpenBSD

    Thank you for the new OpenBSD 4.4 release. This time I preordered and received a CD set, but I had to wait for this release to FTP my favorite application packages. I now run OpenBSD 4.4 on my macppc laptop.

    It was very obvious that the OpenBSD developers want users to send their dmesg to dmesg@. In the first mail to root, and also in section 4.9 of the FAQ, they ask them to send the dmesg after the install. When I installed previous versions of OpenBSD, I always forgot and neglected to do this. After I installed 4.4, I mailed my dmesg for the first time.

    OpenBSD now has more ports than ever. If I compare the numbers from 43.html to 44.html, some architectures now have many more pre-built packages! However, a few less important architectures have fewer packages than before.

    The package quantity for sh fell from 2046 to 1285.
    The package quantity for powerpc fell from 4634 to 4466.
    The package quantity for hppa fell from 3971 to 1595.
    The package quantity for m68k and m88k fell to zero.

    My macppc laptop has powerpc architecture. A few of my favorite packages are among the missing. I used ports to build these missing packages and I have installed them. The multimedia/x264 port broke, but I copied a simple fix from -current. The latest snapshot of -current already has about 4753 packages for powerpc.

    In an above post, developer Marc Espie wrote that OpenBSD does not have enough testers. "YOU CAN TEST STUFF ... Each release, it's the same story: we freeze things, we ask for testers, and we don't get anything. Almost no test report. ... Help testing. If users find bugs in a more timely maneer, then maybe, just maybe, the release process won't be the hell it currently is, and you might have a chance to see more current software in a release, BECAUSE IT WILL HAVE BEEN TESTED."

    Next year, maybe I will install a -current snapshot so that I can try to test stuff. I have many excuses to avoid -current. (1) I have not read the OpenBSD FAQ enough times. (2) A -release comes every six months, so waiting is easier than trying -current. (3) If I always forget to email my dmesg, then I am not ready to email my test reports. FAQ section 5 said, "you should think long and hard before committing yourself to using -current."

    However there are some things entering -current now, that I might want to test before 4.5. Meanwhile, I am enjoying my copy of OpenBSD 4.4, the best -release ever!

  13. By Anonymous Coward (213.185.19.190) on

    Torrents for OpenBSD.

    Please seed.

    Comments
  14. By jirib (195.39.62.98) on

    i have a question about dhcp? is it possible to have dhcp server for multiple networks and to offer IPs depending on network physical locations?

    if not, what about to tag incomming dhcp request via pf and check if in dhcpd?

    jirib

    Comments
    1. By phessler (phessler) on http://theapt.org

      > i have a question about dhcp? is it possible to have dhcp server for multiple networks and to offer IPs depending on network physical locations?
      >
      > if not, what about to tag incomming dhcp request via pf and check if in dhcpd?
      >
      > jirib

      yes, dhcpd can serve different networks on different networks. When you configure it, give it different networks to serve ranges from, and it will give them based on which networks it has interfaces on.

      Comments
      1. By jirib (89.102.13.148) on

        > yes, dhcpd can serve different networks on different networks. When you configure it, give it different networks to serve ranges from, and it will give them based on which networks it has interfaces on.

        what about if dhcpd is not in the same networks? if i used dhcprelay?

Latest Articles

Credits

Copyright © - Daniel Hartmeier. All rights reserved. Articles and comments are copyright their respective authors, submission implies license to publish on this web site. Contents of the archive prior to as well as images and HTML templates were copied from the fabulous original deadly.org with Jose's and Jim's kind permission. This journal runs as CGI with httpd(8) on OpenBSD, the source code is BSD licensed. undeadly \Un*dead"ly\, a. Not subject to death; immortal. [Obs.]