Hi Folks on 14.11.2010 12:34, davidMbrooke wrote: > Hi Erich, > > On Sat, 2010-11-13 at 12:55 +0100, Erich Titl wrote: >> >> I have no clue how you guys get to compile everything, you must have >> local differences to CVS. > > I think all the developers see the same build failures that you do (as > per Andrew's earlier response). Certainly, for me, I have never managed > to build: > isdn, fritz, unicorn, lcd4linux, zaptel, bristuff, libpri, asterisk, > openswan, wlan-ng, lirc, irmp3 > > These are all either "legacy" (need to be removed) or not a priority for > Beta1. I agree it would be good to clean them up. Best thing is probably > to flag the "asterisk" group as an Enhancement (in TRAC) and remove them > from the build for now. Same for isdn & fritz, probably.
Sure, but without at least commenting them in sources.cfg you won't succeed. I suggest to comment out the offending packages and make a buildenv which at least compiles. > >> >> I have not looked at all the individual packages but there are huge >> discepancies to actual packages, for example at openswan, which >> according to sources.cf is fetched from cvs-sourceforge. >> >> The version in CVS is 2.4.7, a real old version for kernel 2.4 the >> actual (old) version for 2.4 is 2.4.15 >> >> The current version for OpenSwan is 2.6.31, IIRC the 2.6 versions are >> specifically geared towards kernel 2.6. For BuC4 we shuld not stick to >> the 2.4 series. > > Agreed. Generally what happens is that someone tries to fix a failing > package. If they are successful (e.g. by upgrading to the latest > upstream) the modified code gets checked in to cvs4-sourceforge. If > there is no success things stay as they are (so usually still pointing > to cvs-sourceforge). Sorting out CVS is another item on the TODO list, > so if you would like to help... > >> Another area of concern for me are some kernel options, for example the >> FIPS compliance is enabled. Why is this so? Even the help text suggests >> to say NO unless one really knows what he is doing. Is this setting us >> in some kind of US controlled mode? >> >> Or why is IPCOMP disabled when ESP is enabled? >> Or why is ESP enabled when IPSEC isn't? >> >> Just a few questions..... > > Definitely kernel (and uClibc, and BusyBox?) options needs a review. > This has been flagged up before. Generally people suggest changes on > this list and if nobody disagrees the changes are incorporated. > >> I suggest to first either drop the packages we cannot compile or make >> them fit. Next we need to upgrade the packages to current versions and >> as unfortunate this may seem, we probably need to fork off completely >> from 2.4 based builds (wich is not done right now). >> >> I am missing the wd1100 driver in the geode based kernel. Do you really >> think there are no WRAP boards around anymore? > > The WRAP boards are a key target platform for BuC 4.x. However, I do not > have one so personally I cannot test compatibility. I will port the wd1100 code again. What is the current canonical form of entering it into the mainline? Still a patch against the kernel? cheers Erich ------------------------------------------------------------------------------ Centralized Desktop Delivery: Dell and VMware Reference Architecture Simplifying enterprise desktop deployment and management using Dell EqualLogic storage and VMware View: A highly scalable, end-to-end client virtualization framework. Read more! http://p.sf.net/sfu/dell-eql-dev2dev _______________________________________________ leaf-devel mailing list leaf-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/leaf-devel
