OpenBSD Journal

Contributing to OpenBSD

Contributed by deanna on from the cure-for-boredom dept.

A handful of requests and project ideas has passed through the lists recently. Any of these would make a great way to help OpenBSD or spend a cold winter/hot summer weekend indoors. Here's a short list; feel free to add more as you find them.
  • write a web interface to the ports system
  • write a driver for broadcom 43xx wireless chipsets
  • donate hardware
  • sponsor a developer?

Write a web application to interface with sqlports

Marc Espie recently wrote to the ports list suggesting that someone write a web application to interface with sqlports, complete with permission to use one's language of choice. sqlports is a database interface to the OpenBSD ports system; more information can be found here.

Write a driver for Broadcom wireless

Theo de Raadt wrote to the misc list explaining that no OpenBSD developers are currently working on support for Broadcom bcm43xx wireless chipsets. So here's your chance to join in on that whole campaign. Message with links to unofficial docs and more information can be found here. Who's up for a challenge from the man himself? :-)

Donate Hardware

Not everyone has the time or skills to code. If you're one of these, and would still like to contribute, why not donate some hardware? Theo recently wrote to tech@ that the list of necessary hardware has been updated. Have a look at his message here, and the list here. Remember that you don't need to actually own the hardware in order to donate it; many places will let you buy it and then ship straight to the developer.

Sponsor a developer?

bert wrote into undeadly suggesting that users sponsor a developer for just pennies a day. I assume he meant by donating hardware, but who knows? Maybe some of them can be convinced, for the right price, to send you a picture for your fridge and an email about their lives every 3 months. :-)

(Comments are closed)


Comments
  1. By Landry (82.245.140.102) on

    >Write a web application to interface with sqlports
    Maybe it's a little OT, but http://ports.openbsd.nu is a nice web interface to browse ports...

    Comments
    1. By Anonymous Coward (64.231.232.11) on

      > >Write a web application to interface with sqlports
      > Maybe it's a little OT, but http://ports.openbsd.nu is a nice web interface to browse ports...

      It is okay for browsing, but one based on sqlports would be a lot more powerful. You'd be able to do things like find all ports that depend on port x, find all ports marked NOT_FOR_ARCHS=somearch, etc. Take a look at all the variables in bsd.port.mk(5) then, take a look at the 3 search options over there.

      Be nice to have this hooked up with cvsweb too. Thanks. :-)

      Comments
      1. By Anonymous Coward (193.11.250.202) on

        > > >Write a web application to interface with sqlports
        > > Maybe it's a little OT, but http://ports.openbsd.nu is a nice web interface to browse ports...
        >
        > It is okay for browsing, but one based on sqlports would be a lot more powerful. You'd be able to do things like find all ports that depend on port x, find all ports marked NOT_FOR_ARCHS=somearch, etc. Take a look at all the variables in bsd.port.mk(5) then, take a look at the 3 search options over there.
        >
        > Be nice to have this hooked up with cvsweb too. Thanks. :-)
        >
        >

        You can now search the depend fields and a lot more areas, if you contact us direct instead of complaining here the search options would have been implemented long time ago ;).

        The NOT_FOR_ARCHS is no problem, but it will take some time to import.

        Comments
        1. By Anonymous Coward (64.231.232.11) on

          > You can now search the depend fields and a lot more areas, if you contact us direct instead of complaining here the search options would have been implemented long time ago ;).

          Okay, tell the developers to stop complaining, too. Read the mail: they want something that they control, that uses their tools, that will be hosted on undeadly, not a clone of some pkgsrc thing. ;)

          Yours is great for the RSS/mail notification stuff and I think it will always have its place.

          Comments
          1. By Fredrik (193.11.250.202) on

            > > You can now search the depend fields and a lot more areas, if you contact us direct instead of complaining here the search options would have been implemented long time ago ;).
            >
            > Okay, tell the developers to stop complaining, too. Read the mail: they want something that they control, that uses their tools, that will be hosted on undeadly, not a clone of some pkgsrc thing. ;)
            >
            > Yours is great for the RSS/mail notification stuff and I think it will always have its place.
            >

            The reason it is a clone of pkgsrc.se i because we could reuse a lot of the code. To connect port.openbsd.nu to undeadly would be a trivial thing to do.

  2. By Anonymous Coward (70.179.123.124) on

    I'm told kettenis@ needs one of these.
    And who wouldn't want jolan@ to get wireless broadband going strong with one of these puppies?

    There are a number of other relatively low-cost items on the hardware list. Some of it you may even have lying around.
    Hell, I'm shipping a G3 my old roomate left in his closet.

  3. By Chris Snell (166.70.206.242) on http://chrissnell.com

    I'd love to see patches for dhcpd(8) that allowed it to either:

    - Run on VLAN interfaces without having to assign an IP address to those interfaces

    - Run on CARP interfaces

    I'd also love to be able to run dhcpd(8) on two CARP'ed servers and have the leases synced between the two machines, just like pfsync. Right now, you can only run the stock dhcpd on one machine at a time and have to start it up manually on the secondary server, should the primary fail.

    Comments
    1. By Bryan Vyhmeister (166.214.118.96) bryan AT bsdjournal ! net on

      I have to second the request for dhcpd(8) on vlan(4) and carp(4) interfaces! Syncing leases between two servers running dhcpd would be great as well!

      Bryan

    2. By Arnaud (82.229.89.88) on


      > I'd also love to be able to run dhcpd(8) on two CARP'ed servers and have the leases synced between the two machines, just like pfsync. Right now, you can only run the stock dhcpd on one machine at a time and have to start it up manually on the secondary server, should the primary fail.

      Correct me if I'm wrong, but the ISC DHCP V3 as the failover ability built in ?

      Regards

      Comments
      1. By Chris Snell (166.70.206.242) on http://chrissnell.com

        > Correct me if I'm wrong, but the ISC DHCP V3 as the failover ability built in ?

        Indeed it does, but ISC DHCP v3 is not included with OpenBSD. For my firewalls, I prefer to use only "official" OpenBSD software.

        Also, v3 still does not play very nice with IP-less VLAN interfaces or CARP.

    3. By pixelfairy (82.233.69.3) on

      > I'd love to see patches for dhcpd(8) that allowed it to either:
      >
      > - Run on VLAN interfaces without having to assign an IP address to those interfaces
      >
      > - Run on CARP interfaces
      >
      > I'd also love to be able to run dhcpd(8) on two CARP'ed servers and have the leases synced between the two machines, just like pfsync. Right now, you can only run the stock dhcpd on one machine at a time and have to start it up manually on the secondary server, should the primary fail.

      or just run them both with differnt ranges. (at least for now)

  4. By Daniel Ouellet (66.63.10.94) daniel@presscom.net on

    I would definitely be more then happy to sponsor a developer, or more then one for starting of VoIP (SIP) project in OpenBSD. I express the idea a few times on misc@ and this is definitely a serious project for long term and if any developer(s) are serious about it, I would definitely work something out to make that happen somehow.

    A few ideas and needs are as follow, some smaller then others obviously.

    - NAT traversal built into PF.
    - Conference bridge for SIP VoIP.
    - Media server for stream of music on hold and personalize music on hold for virtual PBX.
    - cnam SIP database for CLID. Already have PERL scripts doing this, but should really be C and fully daemon in BSD license as well.
    - SIP library for building upon full virtual PBX.
    - Full SIP network server for Virtual PBX hosting.
    - Full SIP access server for users control, etc.
    - PRI library support to replace Cisco 5350 type ISDN PRI to Telco gateways with something that would use common hardware and provide good quality in G711.
    - FAX to Email server and library for standard G711 access from Telco.
    - Addition to PF to allow preset of DSCP value of the packets header for proper QoS propagation on enable QoS networks.
    - Addition to PF to also process on the fly the DSCP value of packets as well.

    There is a few more things that would be nice, but as you can see, there is a good lists and some small, some bigger and all are modular could be very interesting.

    If any are actually interested to do this, or part of and would like to be sponsor to do this, I would be happy to exchange more emails directly.

    I already do this for a few years already and I hate the being stuck in proprietary product that never do what you need it to do and get stuck for very long time when bugs show up!

    The platform I am using now for 5 years+ already and that is getting very popular is Broadsoft but I can't stand them! It's however the one that still give you the most flexibility.

    Anyway, if that was a real question on sponsor developers for interesting projects, then here is a list of possible start and will see if that's of any interests.

    Comments
    1. By Daniel Ouellet (66.63.10.94) daniel@presscom.net on

      I should also have added that I would even go as far as sponsor a SIP VoIP hackathon should there be a real interest in doing so!

      If I recall properly, Theo said that it cost usually around 30K to do one for the week. Obviously that doesn't include everything for sure, but that's around what the cost was so far for the last few.

      I really don't know how much can actually be done in a full week and if anything final could be done, but that should for sure provide a good start and might allow to see this become a reality in the near future.

      May be I wouldn't be allow to sponsor this as I sure think many would benefit form it as well, and sure would provide for long term viability for this idea, but if not, I would find a way to make it happen!

      What I have is a small business only with limited resources, but OpenBSD allow me to be in business and be able to actually make this happen as a selfish interest I must admit, would also be a nice way for me to contribute back to the project in way that I believe I could do.

      I am sure in the end, I wouldn't be the only one enjoying it for sure!

      It's hard to compete on the same footage then the Verizon, AT&T, MCI, WorldCom or the world as there is no leverage for the small business and being lock in in proprietary platform where you pay the same licenses cost, but sure don't get the same results as yours is much smaller then the big guys!

      Would be nice to be able to compete on quality and service oppose to wallet size with no one can really win against these guys!

      In any case, I would be more then welling to sponsor more then one if it really triger nice interests in the idea.

      Comments
      1. By Daniel Ouellet (66.63.10.94) daniel@presscom.net on

        Sorry!

        This part "May be I wouldn't be allow to sponsor..."

        Should have been "May be I wouldn't be allone to sponsor..."

        Meaning may be others would like to see this happen as well and contribute to it as well, but if not, so be it.

        Sorry, even proof reading doesn't always coem out how it should.

    2. By mvanbaak (213.154.226.2) on

      All the SIP server and PBX parts are already out there. Have a look at http://www.asterisk.org
      It would be nice tho if the zaptel parts will be ported to OpenBSD. The PF additions is a nice feature to have too.

      Comments
      1. By Wijnand (85.144.78.187) on http://nedbsd.nl

        > All the SIP server and PBX parts are already out there. Have a look at http://www.asterisk.org

        This is GPL and thus not free.

        > It would be nice tho if the zaptel parts will be ported to OpenBSD. The PF additions is a nice feature to have too.

        True.

      2. By Anonymous Coward (66.63.10.94) Daniel Ouellet on

        > All the SIP server and PBX parts are already out there. Have a look at http://www.asterisk.org

        It's not free, plus look again. Asterisk DO NOT support virtual PBX hosting, meaning multiple isolated company in one server configuration. Asterisk can work well, or relatively well for a company PBX, but not when virtual hosting in needed.

        Sorry, no go for that and no go for not free.

    3. By mcbride (210.138.35.53) mcbride@openbsd.org on

      > - NAT traversal built into PF.

      Not going to happen; PF is not the place to do layer 7 mangling - you're just asking for security problems (not to mention developer nightmares and kernel bloat).

      A PF-aware transparent proxy might be nice, though.

  5. By SH (82.182.103.172) on

    One may donate cash as well.

  6. By Anonymous Coward (88.198.180.62) on

    OpenHTTP would be a neat point on the list too!

    Comments
    1. By Daniel Ouellet (66.63.10.94) daniel@presscom.net on

      > OpenHTTP would be a neat point on the list too!

      Yes it would be and it would remove one more GNU software from the base system, but that's not a small undertaking for sure.

      Same would apply to OpenName and OpenSMTP, but I am not sure any would find it interesting to do, but I could be wrong.

      What would be nice is all three are in the base system as is, so it would make the default install even more OpenBSD like if I can say that.

      But I think developers are more interested in new things, oppose to redo/replace already working one.

      Would be nice however.

      Comments
      1. By sng (67.171.149.18) on

        > > OpenHTTP would be a neat point on the list too!
        >
        > Yes it would be and it would remove one more GNU software from the base system, but that's not a small undertaking for sure.

        Apache isn't GNU and the version in base is for al intents and purposes a fork.
        >
        > Same would apply to OpenName and OpenSMTP, but I am not sure any would find it interesting to do, but I could be wrong.
        Why? BIND works fine and also isn't GNU. And SMTP just plain sucks. No good reason for the project to get involved in something that broken from the outset.
        >
        > What would be nice is all three are in the base system as is, so it would make the default install even more OpenBSD like if I can say that.
        BIND and Apache already are. And are forks in all but name. So the point of renaming would be?
        >
        > But I think developers are more interested in new things, oppose to redo/replace already working one.
        >
        > Would be nice however.

        Comments
        1. By Anonymous Coward (68.104.220.48) on

          > > > OpenHTTP would be a neat point on the list too!
          > >
          > > Yes it would be and it would remove one more GNU software from the base system, but that's not a small undertaking for sure.
          >
          > Apache isn't GNU and the version in base is for al intents and purposes a fork.
          > >
          > > Same would apply to OpenName and OpenSMTP, but I am not sure any would find it interesting to do, but I could be wrong.
          > Why? BIND works fine and also isn't GNU. And SMTP just plain sucks. No good reason for the project to get involved in something that broken from the outset.

          Uh... do you have a different idea for interfacing with the rest of the world on an email infrastructure? Or do you just send smoke signals?

          An open, simple but functional alternative to Sendmail would be a welcome thing, especially since most of the other alternatives don't have adequate licensing freedom. By "simple but functional" I mean as in the difference between openntpd and ntp.org's implementation. It doesn't have to be fancy.

          > >
          > > What would be nice is all three are in the base system as is, so it would make the default install even more OpenBSD like if I can say that.
          > BIND and Apache already are. And are forks in all but name. So the point of renaming would be?
          > >
          > > But I think developers are more interested in new things, oppose to redo/replace already working one.

          openntpd, opencvs, openbpgd, openospfd; all alternatives to something that was "already working". They were in need of alternatives, not something new, so there we have it.

          DS

  7. By Amir S Mesry (66.23.227.241) on

    A Mips port would be very helpful as it would enable OpenBSD to be ran on the WRT54G platform. I would love it, and I think many others would as well.

    I can't write c code nor assembly, otherwise I would try!

    Comments
    1. By Anonymous Coward (71.231.77.111) on

      > A Mips port would be very helpful as it would enable OpenBSD to be ran on the WRT54G platform. I would love it, and I think many others would as well.
      >
      > I can't write c code nor assembly, otherwise I would try!

      No way. This is the single best idea I've ever heard. I currently run 2 WAP54G and if those things could run OpenBSD on them I'd be the happiest geek around. Of course the only task on this list that I'm up to is the sqlite ports gui one.


      Comments
      1. By Anonymous Coward (64.231.232.11) on

        It would be nice if people could keep to the topic of concrete things that actual developers have asked for, instead of some ridiculous wish list. Linksys support? Rewriting "GNU" Sendmal and BIND.. to get rid of the GPL? Christ, think first, then post. :)

        Comments
        1. By Daniel Ouellet (66.63.10.94) daniel@presscom.net on

          > It would be nice if people could keep to the topic of concrete things that actual developers have asked for, instead of some ridiculous wish list. Linksys support? Rewriting "GNU" Sendmal and BIND.. to get rid of the GPL? Christ, think first, then post. :)

          Yes you are right and that's why it said "But I think developers are more interested in new things, oppose to redo/replace already working one"

          Plus the use of might be interesting, didn't imply it should be done.

          That was a reply to present that redoing things already working is not the first interest of many.

          So, as you said "Christ, think first, then post", where is yours? (;>

          I put some of my ideas before that and it was thought of and not the full list presented as it would be much longer of things not available already and that I think are interesting, but that's just me. (:>

          So, "Christ, think first, then post." it was thought of before posting.

          Best,

          Daniel

          PS; The OpenName and OpenSMTP, were cheap one I agree! That's why it said that it most likely it wouldn't triger interest oppose to new ideas. Hope this make it more clear for you?

      2. By jsg (210.15.216.215) on

        > > A Mips port would be very helpful as it would enable OpenBSD to be ran on the WRT54G platform. I would love it, and I think many others would as well.
        > >
        > > I can't write c code nor assembly, otherwise I would try!
        >
        > No way. This is the single best idea I've ever heard. I currently run 2 WAP54G and if those things could run OpenBSD on them I'd be the happiest geek around. Of course the only task on this list that I'm up to is the sqlite ports gui one.
        >

        The Broadcom MIPS based NAS systems that have disks in them would perhaps make an interesting target were Broadcom to release some documentation.

    2. By Anonymous Coward (66.39.191.42) on

      > A Mips port would be very helpful as it would enable OpenBSD to be ran on the WRT54G platform. I would love it, and I think many others would as well.
      >
      > I can't write c code nor assembly, otherwise I would try!

      hunh? openbsd already supports several MIPS platforms

      not all of the platforms are still there in current, but the code is still in the CVS attic

      Comments
      1. By Anonymous Coward (66.23.227.241) on

        > > A Mips port would be very helpful as it would enable OpenBSD to be ran on the WRT54G platform. I would love it, and I think many others would as well.
        > >
        > > I can't write c code nor assembly, otherwise I would try!
        >
        > hunh? openbsd already supports several MIPS platforms
        >
        > not all of the platforms are still there in current, but the code is still in the CVS attic

        Updated I should say, I knew it did support it in previous releases.

  8. By Sam Fourman Jr. (69.179.12.207) sfourman@gmail.com on

    I think The Gnome Desktop 2.16.x Would be great to have in the ports tree

    I am New(ish) to OpenBSD and do not *YET* have the understanding necessary to port such a complex app.
    I have made a few comments to the Ports list with mixed results

    I do have a partial version of Gnome 2.14.2 running on -current


    Sam Fourman Jr.

    Comments
    1. By Anonymous Coward (82.181.60.23) on

      If you could get in touch with some committer and get your 2.14 port in tree that would be a great improvement to the 2.10 version which we currently have.

  9. By Janne Johansson (82.182.176.20) jj@inet6.se on www.inet6.se

    Espie asked for people tocode tracing code for m4.
    http://www.security-express.com/archives/openbsd/2006-03/1537.html

  10. By Simon 'simmel' Lundström (83.140.211.6) on

    While WPA (or even WPA2) is adds very little to wireless security, it still is superior to home users. Setting up VPN on friends laptops when they come over once or twice a year is just too much.

    Having WPA on my WRAP would replace my 3Com Travel router and that would be nice.

  11. By Anonymous Coward (82.134.90.244) on

    ppp IMPROVEMENTS so that you don´t have the damn choice between the damn userland PPD wich does not get any REAL speed at fast lines but PF keeps ya NAT-Settings and co OR the kernel PPP wich is faster but after a reconnect you`ve to do pfctl -F all -f /etc/pf.conf otherwise NAT and powertforwarding is fucked.

    So PPP improvements would make me more interested into OpenBSD. :)

    Comments
    1. By Anonymous Coward (87.78.70.152) on

      > the kernel PPP wich is faster but after a reconnect you`ve to do pfctl -F all -f /etc/pf.conf otherwise NAT and powertforwarding is fucked.

      huh? refering to dynamic ip changes?
      use ($interface) in pf.conf. that way pf will use the new/actual ip. no need to do anything on reconnect. it just works!

      Comments
      1. By Anonymous Coward (87.106.20.213) on

        > > the kernel PPP wich is faster but after a reconnect you`ve to do pfctl -F all -f /etc/pf.conf otherwise NAT and powertforwarding is fucked.
        >
        > huh? refering to dynamic ip changes?
        > use ($interface) in pf.conf. that way pf will use the new/actual ip. no need to do anything on reconnect. it just works!

        You`re correct but pf wont notice if the IP did changed until you relaoded the configurations (at least with Kernel pppoe aka pppoe0, with tun0 (userland) you`ve no problems at all).

        If you know any solution for this: please let me know. I`m not aware of any.. so you`ve to make the choice...

        Comments
        1. By Anonymous Coward (87.106.20.213) on

          I forgot: I´m using port-redirections and NAT
          And I do have problems with the kernel pppoe0 after reconnects (if I don´t realod the rules like descriped above).

          Comments
          1. By Anonymous Coward (87.78.70.152) on

            adsl with dynamic ip, using in-kernel pppoe.
            7 rdr-rules and 3 nat'ed subnets.
            at least 2 reconnects a day.
            no problems with pf not picking up ip-changes.

            this is way offtopic. contact the pf-mailinglist with a full bug/problem-report. (pf-benzedrine_cx)

            Comments
            1. By Amir S Mesry (66.23.227.241) on

              > adsl with dynamic ip, using in-kernel pppoe.
              > 7 rdr-rules and 3 nat'ed subnets.
              > at least 2 reconnects a day.
              > no problems with pf not picking up ip-changes.
              >
              > this is way offtopic. contact the pf-mailinglist with a full bug/problem-report. (pf-benzedrine_cx)

              This is a known issue. I read it would be fixed in 4.0.

              I can duplicate it like crazy.

  12. By Anonymous Coward (88.191.33.169) on

    I just thought about oBSD during the day and this is what come sinto my mind:

    - improvements to ppoe (descriped into another answer)
    - OpenHTTP -> Configuration should be propably compatible to Apache (no
    need to at all)
    - Build in ssl support, no need for modules f.e. but modules should be
    possible too (for PHP as example)
    - Better encryption of SVNDs (different Algos, also mentioned in posts
    at mailinglists)
    - Propably a setup during the install?
    - More GNU replacements
    - ccd improvements so that raidframe can get removed
    - BSD licensed LZMA implementation wich would SAVE a lot memory (space
    on drive) and would enable the Kernel (f.e. on floppies) to keep more
    Drivers because LZMA compresses better
    - Also packages/ports (*.tar.gz) could use it because decompressing
    those files are a common scenario even compressing them once needs
    more time
    - Saves Bandwith + LZMA is faster then bzip2 during decompression
    - Better SMP handling and propably more SMP for different Archs
    - ACPI (much process was submited recently so AMD64-Systems may propably
    power off if I use halt -p on 4.1 or later)
    - Improvements ot "login" to enable better checks for passwords (adding
    complexity checks wich ma ycould get defined) + allowing logins just
    on special dates+times (f.e. during work time only!)


    I know NEW stuff is more interesting but the "it works"-phillosophy sounds kinda unusual for a BSD project (or are you realy statisfied with all the GPL stuff?)

    I also know that many users would love "more" drivers even there`s (in my oppinion) a lot other stuff to do except the Routing-Daemons or the WLAN-Drivers or RAID-Support wich may get special attention.

  13. By Anonymous Coward (68.12.154.246) on

    I would love to see a PPTP and/or IPSec passthrough type of proxy to work with PF. If a $40 linksys can do this, why not PF?

    Comments
    1. By Anonymous Coward (217.112.179.74) on

      > I would love to see a PPTP and/or IPSec passthrough type of proxy to work with PF. If a $40 linksys can do this, why not PF?

      I'd really like to see this too. It has already been discussed in http://undeadly.org/cgi?action=article&sid=20041009000521 but frickin is not a part of OpenBSD.

  14. By Otto Moerbeek (87.210.142.234) otto@drijf.net on http://www.drijf.net

    I guess some developer has to say it....

    This article is NOT about requesting new features or whatever. This article is about supporting development, by doing things yourself or by sponsering developers.

    Trust me, developers have plenty of ideas and as a result have pretty long TODO lists.

    So we're not very interested in wish lists, feature requests and such, UNLESS you offer to support a developer to actually get something done.

    Given the amount of time I spend on OpenBSD, I would love to get some compensation for that. I get some hardware now and then, for which I'm thankful, but the value of the hardware does not compare to the hours spent. Almost all of this is done in my spare time. Of course I love to do this, but please bear in mind all this when you're asking for features... And ordering CDs just helps to keep the general project infrastructure going, it is not enough to fund developers.

    Every commit takes time. Some only a few minutes, some are the results of days of weeks of work. For almost all commits several developers have spent time to develop the diff, to review and to test. To most valuable resource for developers is TIME.

    Comments
    1. By Anonymous Coward (88.198.180.62) on

      > I guess some developer has to say it....
      >
      > This article is NOT about requesting new features or whatever. This article is about supporting development, by doing things yourself or by sponsering developers.
      >
      > Trust me, developers have plenty of ideas and as a result have pretty long TODO lists.
      >
      > So we're not very interested in wish lists, feature requests and such, UNLESS you offer to support a developer to actually get something done.
      >
      > Given the amount of time I spend on OpenBSD, I would love to get some compensation for that. I get some hardware now and then, for which I'm thankful, but the value of the hardware does not compare to the hours spent. Almost all of this is done in my spare time. Of course I love to do this, but please bear in mind all this when you're asking for features... And ordering CDs just helps to keep the general project infrastructure going, it is not enough to fund developers.
      >
      > Every commit takes time. Some only a few minutes, some are the results of days of weeks of work. For almost all commits several developers have spent time to develop the diff, to review and to test. To most valuable resource for developers is TIME.

      Right but somebody who may can`t do something (or isn`t interested) fromt he original list may could get itnerested to somethign "users" requested. This would help and improve OpenBSD and makes the userbase (wich propably buys the CDs..) much more happy.

      So suggestions SHOULD be welcome...
      Even it`s clear that oBSD can`t do it now (or ever..) because it lacks manpower.

      Comments
      1. By tmclaugh (192.216.27.32) on

        > Right but somebody who may can`t do something (or isn`t interested) fromt he original list may could get itnerested to somethign "users" requested. This would help and improve OpenBSD and makes the userbase (wich propably buys the CDs..) much more happy.
        >
        > So suggestions SHOULD be welcome...
        > Even it`s clear that oBSD can`t do it now (or ever..) because it lacks manpower.

        Just tossing out random things that random people would like to see done is not helpful. These threads in the end just become a lot of noise which have the adverse effect of drowning out the stuff developers are working on. In an ideal world people could toss out suggestions and others would just magically pickup the work. This isn't an ideal world. OBSD is a volunteer project which deals with the reality of how things get done in a volunteer environment. Take the OpenBSD community of users. Almost everyone of us has something or multiple somethings we would like to see. Only a small minority will take the time to actually do something about what they want. The ratio of things users want to what will actually get done is very small. Bookmark this page and look at it 6 months from now, a year from now, etc. See how many ideas on here actually got implemented. Most things only get done in a volunteer environment when someone finally just says, "Ahhh screw it. I'll just take care of it..."

        If you personally want something from this thread then try these steps.
        1) Plan out what you want and how it will work.
        2) Create something mostly functional that others can toy with.
        3) Announce it somewhere.
        4) Pray you piqued the interest of someone else to help you.

        This thread is the equivalent of someone building a house and asking the people who will be living in it (I dunno, you like living with squatters or people from craigslist), "Hey, can you lend me a hand installing this kitchen?" Instead of the help you need you get people arguing over the correct design of 3 different bikesheds which never get built and one doghouse when you own a cat.

        Comments
        1. By J. (194.109.22.148) on

          > [...]
          > If you personally want something from this thread then try these steps.
          > 1) Plan out what you want and how it will work.
          > 2) Create something mostly functional that others can toy with.
          > 3) Announce it somewhere.
          > 4) Pray you piqued the interest of someone else to help you.

          How about a Google Summer of Code or GNOME Bounty construction? If a developer gets X USD for a certain task such is nice to know before you start coding.

          The normal incentive to get a piece of software in the F/OSS community developed is a developer reasoning "I need A therefore I develop A". The mentioned setup would provide an alternative to that.

          Consider you have a piece of hardware which a developer needs. The hardware isn't (well) supported, and you happen to know that if the developer gets their hands on that hardware, it will be supported next OpenBSD release. Here, there are costs/benefits involved too: you pay the price of the hardware to get it supported. I've even seen a collaboration here to buy a USIII machine on eBay. Various people pledged to put money in a USIII machine for a developer.

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.]