Hi All, Here are the meeting minutes for the Free EDA Users Group meeting held on December 6th 2006:
In attendance: Stuart, DJ, JohnL, Steve Morss, and Ales. * DJ mentioned that the last change/fix in CVS gschem (by Ales) of hardcoding the gettext version to 0.15 broke the build on his Fedora Core 6 box (which has 0.14.6). Normally gettext wouldn't be a problem, but gettext 0.15 broke backwards compatibility which prompted the version change in CVS gschem. Ales spent time fixing this during the meeting. This fix was checked in after the meeting. DJ reported that the alternative fix is working for him. * DJ demoed several prototype and working SMD challenge boards. DJ noted that nobody has reported successfully building a working SMD challenge board. * We had a new member join us at this FreeDog meeting: Steve Morss. Welcome Steve! * DJ showed off his memory and Ethernet expansion board built using the gEDA tools. * DJ talked about PCB and PCB manufacturing vendors. * Stuart and Ales discussed the guile 1.8.1 issues. Ales has identified some of the breakage caused by guile 1.8.1 and will checkin fixes soon. * DJ showed off a whole bunch of home etched boards. DJ built etch test structures into these test boards. The manufacturing accuracy DJ was able to achieve was 7 mil space/trace, which we all thought was pretty good for home etching equipment. * Stuart and Ales showed off and discussed the latest gEDA in print articles (mainly the Stuart's letter to the editor in IEEE Spectrum: http://www.spectrum.ieee.org/dec06/4767 ). * John talked about the SMD tools he uses: Equipment --------- Metcal SP200-11 available new at www.hmcelectronics.com for apx. $240 SSC-754A tip 30deg bend, 0.02" tip 7A tweezers (curved tip) Luxo illuminated magnifier Solder ------ Kester 24-6337-6422 .015" diameter solder, SN63PB37, water soluble flux Kester Flux Pen 2331-ZX Solder Wick * Steve showed off a PCB layout for his single board computer. Steve also had examples of a parser that creates footprints using a description file. Steve mentioned that he does some of his scripting in perl; Ales quipped that John has finally found a fellow perl user. <g> :-) * Ales inquired about details on Steve's single board computer. Steve described that it is a FPGA based design using the LEON core which is an open source implementation of a SPARC V8 core. Very cool! The best (working) website about the LEON core I (Ales) could find is: http://www.leox.org/ * There was a lot discussion about a new venue for the next FreeDog meeting. Everybody agreed that the current Starbucks location is nice because of atmosphere and general ambiance, however, the lack of affordable Wifi Internet access is a bit of a problem for people around the world who want to participate too. Everybody agreed that changing the venue to get Internet access is worth the effort. Everybody agreed NOT to go back to MIT for these meetings, since the MIT classroom is kind of sterile. Several ideas for new venues were discussed include a cafe in Reading, MA (Bear Rock Cafe) and maybe a cafe Davis Square. Both have Wifi access. Stuart and Ales will go investigate soon. Stay tuned for more info... * Paolo Nenzi of ngspice might be visiting Boston in January 2007. We want to hold a FreeDog meeting to meet and talk with Paolo. Possible dates for this meeting include Wednesday the 10th or Tuesday the 9th. These dates need to be confirmed with Paolo. * The developers discussed having another code sprint in February at MIT. Ales has already decided that he won't work on code, but rather will work on the documentation and maybe the symbol library. Possible dates include: Feb 3rd, 10th, or 24th (all Saturdays). * An idea was tossed out to make PCB's default build/run mode be set to Lesstif, as the Lesstif PCB is more configurable at runtime. DJ noted that Lesstif has an active developer (DJ) whereas gtk+ is more in "maintenance" and could use a gtk+ expert to push it forward. Stuart mentioned that newbies expect (without even knowing it) to see the gtk+ interface, and they would be blown away by the Lesstif interface. John suggested that it would be nice if *all* version of PCB were built and installed. This would enable the user to easily evaluate all interfaces for themselves. * The discussion turned to how to accept donations. Everybody agreed that it is okay for Ales to turn on the Sourceforge donation button. We will follow the PCB donation model where the individual contributer will decide how their contribution should be spent. However, we really do not want the money to go to individual developers because: 1) there are so many past and current gEDA/gaf developers, 2) splitting the money fairly is virtually impossible, 3) we do not want one developer's contributions be worth more that somebody else's, 4) We are not talking about large sums of money here. We are talking about pizza money. Everybody at the meeting agreed with "not sending money to developers, unless the contributer explicitly states this". Ales will be responsible for getting the Sourceforge donation system setup. Further discussions will probably be needed on what to do with any contributed funds. * The group then discussed gedaconsulting.com. Everybody agreed that it is a pretty darn cool idea. * Ales brought up the point that the gEDA project may be in need of a new home on the Internet. The current hosting site is stable until May/June and then the future becomes less certain. DJ suggested a few possible alternatives (one being mv.com). The current server's status will have to be tracked very carefully so that the gEDA project has a continuous home on the 'net after May/June 2007. The current bandwidth usage of the gEDA project is much larger than it needs be. Ales will cleanup the website and ftp server to reduce bandwidth needs in case we have to move. Stuart will ping some contacts that might get us a stable hosting solution. If anybody has any contacts in places that might help us out here, please get in touch with Ales. * The Windows port was discussed again. Ales isn't opposed to creating a set of native mingw binaries (along with any supporting infrastructure) as long as: 1) There is some sort of barrier to entry to keep the support burden manageable. Something as simple as shipping the binaries in a zip or .bz2 file instead of .exe. 2) The initial release should be billed as an "experiment". If the developer support burden is too high, then the experiment will be terminated after a fixed amount of time. 3) The Windows port should be supported separately from the main gEDA project initially. Maybe this means creating a different SF project and a separate mailing list. This, too, is to keep the support burden manageable. Ales will take responsibility for doing something here. * Ales gave his blessing for creating a wiki page for ngspice in the gEDA wiki. * Stuart passed out several wish lists to discuss, but a friendly Starbucks employee told us to vacate the premises at this point. This has been tabled until the next meeting. In the mean time, discussion on these wish lists should continue on the mailing lists. End of meeting. -Ales _______________________________________________ geda-dev mailing list [email protected] http://www.seul.org/cgi-bin/mailman/listinfo/geda-dev
