As the old adage goes, "it's software, anything is possible"! There are certainly no technical restrictions as to why one can't make rio work with a Wayland backend. It might be hard, but not impossible.
Once we got a few basic Plan 9 programs working, it was a combination of the lack of time and motivation that lead to no progress. Also, vx32/9vx was released around the same time which gives you most of what Glendix does. If you haven't looked at 9vx yet, you certainly should: http://swtch.com/9vx/. Rio doesn't have any dependencies other than (a subset of) system calls, /dev/draw, and possibly a few other file systems. The most efficient way to implement those file systems would be for them to directly talk to the backend (in this particular case, that would be Wayland). /net went a step further and implemented the system in kernel space, but it can also be done from user space, the choice is ultimately yours to make. One thing that tripped us up was the per-process namespace that Plan 9 has, which isn't trivial to map to the Linux world (which has namespaces but they're designed differently). Don't hesitate to post to the list with specific issues, the 9fans list can also be very helpful at times. All the best, -Anant On Thu, Oct 18, 2012 at 6:13 PM, Ericson2314 <[email protected]> wrote: > Wow, thanks for the quick reply. I saw the branch with /net in the > repository, that certainly will help. > > Anyways, I am curious what sort of broad issues to look out for. Is there > any sort of hitch or new viewpoint that made lead you to stop working on the > project when you did? Glendix seems like it could bring a huge amount of > energy into Plan9, but If I am going to put some serious effort into it I > want to try to make sure I know what's coming and avoid any nasty surprises. > > I guess as similar questions, if all 51 syscalls were implemented right now, > what would the next challenge be? In native plan9, how to what extent are > the special filesystems like /net implemented on those system calls? does > taking advantage of Linux's existing drivers mean making higher abstractions > beyond the syscalls and/or filesystems? > > Thanks again, > Ericson2314 > > > On Thursday, October 18, 2012 1:03:17 PM UTC-4, Anant Narayanan wrote: >> >> Hey, >> >> We've only implemented the most important system calls for now (read, >> write, open, clone and couple others), in order to get Rio working, I'm >> assuming you will have to implement about a dozen other system calls, and >> then implement /dev/draw with a Wayland backend. We have an early >> implementation of /net that you can use as reference while implementing >> /dev/draw. >> >> Cheers, >> -Anant >> >> On Wednesday, October 17, 2012 11:49:00 PM UTC-7, Ericson2314 wrote: >>> >>> Hello, >>> >>> I don't have the time just yet, but I am interested in working on Glendix >>> in the near future. Exactly how much has yet to be done? >>> >>> Part of my interest is because I am wondering whether rio could be >>> implemented on top of Wayland, he upcoming "replacement" for X. Wayland >>> isn't particularly like rio, but neither is X, and building off wayland >>> should bring in less POSIX or other annoying dependencies. Also Wayland >>> biggest drawback is the lack of networking transparency. The devs have >>> decided to push that problem down the road. Implementing rio on Wayland >>> should at the very least inspire ideas for remote graphics and Wayland, and >>> possibly interest other developers in Plan9 and Glendix. >>> >>> Thanks, >>> >>> On Wednesday, September 5, 2012 12:43:03 PM UTC-4, Anant Narayanan wrote: >>>> >>>> I haven't touched the project in years, but am more than happy to help >>>> anyone who might be interested in working on it! >>>> >>>> Cheers, >>>> -Anant >>>> >>>> On Tue, Aug 28, 2012 at 8:28 AM, EBo <[email protected]> wrote: >>>> > On Tue, 28 Aug 2012 07:50:22 -0700 (PDT), Josh Leverette wrote: >>>> >> >>>> >> Is this dead? >>>> >> Just wondering... I've found traces of life on the IRC channel -- I >>>> >> think, >>>> >> but the website hasn't been updated in 3 or so years and this page >>>> >> hasn't >>>> >> had activity in a couple of years either. It sounds like an >>>> >> interesting >>>> >> project, so I'm curious. >>>> > >>>> > >>>> > speaking for myself I moved on. I have not kept up with anyone else >>>> > that >>>> > might be playing with it. >>>> > >>>> > I did port 9vx to the TinyCoreLinux distro, but stopped banging my >>>> > head >>>> > against that wall because someone kept making changes to the control >>>> > files >>>> > which broke my maintenance tool chain. I never did figure out who was >>>> > trying to "help" maintain the system without ever telling me. I could >>>> > have >>>> > showed them an easier way that would not have wasted weeks of my time >>>> > -- >>>> > till I just gave up. >>>> > >>>> > EBo -- >>>> > >>>> > >>>> > -- >>>> > To post to this group, send email to [email protected] >>>> > To unsubscribe from this group, send email to >>>> > [email protected] >>>> > For more options, visit this group at >>>> > http://groups.google.com/group/glendix?hl=en > > -- > To post to this group, send email to [email protected] > To unsubscribe from this group, send email to > [email protected] > For more options, visit this group at > http://groups.google.com/group/glendix?hl=en -- To post to this group, send email to [email protected] To unsubscribe from this group, send email to [email protected] For more options, visit this group at http://groups.google.com/group/glendix?hl=en
