see http://rt2x00.serialmonkey.com/phpBB2/viewtopic.php?t=1350&start=0&postdays=0&postorder=asc&highlight= i tried emerging net-wireless/rt2x00-9999 but failed two times: first it doesnt support MAKEOPTS="-j3" (fails with message 'file not found') and it still doesnt emerge: # MAKEOPTS="" emerge -va =rt2x00-9999 These are the packages that I would merge, in order: Calculating dependencies ...done! [ebuild N ] net-wireless/rt2x00-9999 -debug 0 kB Total size of downloads: 0 kB Do you want me to merge these packages? [Yes/No] y >>> emerge (1 of 1) net-wireless/rt2x00-9999 to / >>> md5 files ;-) rt2x00-9999.ebuild >>> md5 files ;-) rt2x00-2.0.0_beta3.ebuild >>> md5 files ;-) files/digest-rt2x00-2.0.0_beta3 >>> md5 files ;-) files/digest-rt2x00-9999 * Determining the location of the kernel source code * Found kernel source directory: * /usr/src/linux * Found sources for kernel version: * 2.6.16-gentoo-r6 * Checking for suitable kernel configuration options... [ ok ]>>> Unpacking source... * Fetching CVS module source/rt2x00 into /usr/portage/distfiles/cvs-src ... * Running cvs -q -f -z3 -d ":pserver:anonymous:@cvs.sourceforge.net:/cvsroot/rt2400" login Logging in to :pserver:anonymous@cvs.sourceforge.net:2401/cvsroot/rt2400 * Running cvs -q -f -z3 -d ":pserver:anonymous@cvs.sourceforge.net:/cvsroot/rt2400" update -dP rt2x00-9999 * Copying source/rt2x00 from /usr/portage/distfiles/cvs-src ... * CVS module source/rt2x00 is now in /var/tmp/portage/rt2x00-9999/work >>> Source unpacked. * Preparing ieee80211/80211 module CC [M] /var/tmp/portage/rt2x00-9999/work/rt2x00-9999/ieee80211/ieee80211_module.o CC [M] /var/tmp/portage/rt2x00-9999/work/rt2x00-9999/ieee80211/ieee80211_tx.o CC [M] /var/tmp/portage/rt2x00-9999/work/rt2x00-9999/ieee80211/ieee80211_rx.o CC [M] /var/tmp/portage/rt2x00-9999/work/rt2x00-9999/ieee80211/ieee80211_wx.o CC [M] /var/tmp/portage/rt2x00-9999/work/rt2x00-9999/ieee80211/ieee80211_geo.o LD [M] /var/tmp/portage/rt2x00-9999/work/rt2x00-9999/ieee80211/ieee80211.o CC [M] /var/tmp/portage/rt2x00-9999/work/rt2x00-9999/ieee80211/ieee80211_crypt.o CC [M] /var/tmp/portage/rt2x00-9999/work/rt2x00-9999/ieee80211/ieee80211_crypt_wep.o CC [M] /var/tmp/portage/rt2x00-9999/work/rt2x00-9999/ieee80211/ieee80211_crypt_ccmp.o CC [M] /var/tmp/portage/rt2x00-9999/work/rt2x00-9999/ieee80211/ieee80211_crypt_tkip.o Building modules, stage 2. MODPOST CC /var/tmp/portage/rt2x00-9999/work/rt2x00-9999/ieee80211/ieee80211.mod.o LD [M] /var/tmp/portage/rt2x00-9999/work/rt2x00-9999/ieee80211/ieee80211.ko CC /var/tmp/portage/rt2x00-9999/work/rt2x00-9999/ieee80211/ieee80211_crypt.mod.o LD [M] /var/tmp/portage/rt2x00-9999/work/rt2x00-9999/ieee80211/ieee80211_crypt.ko CC /var/tmp/portage/rt2x00-9999/work/rt2x00-9999/ieee80211/ieee80211_crypt_ccmp.mod.o LD [M] /var/tmp/portage/rt2x00-9999/work/rt2x00-9999/ieee80211/ieee80211_crypt_ccmp.ko CC /var/tmp/portage/rt2x00-9999/work/rt2x00-9999/ieee80211/ieee80211_crypt_tkip.mod.o LD [M] /var/tmp/portage/rt2x00-9999/work/rt2x00-9999/ieee80211/ieee80211_crypt_tkip.ko CC /var/tmp/portage/rt2x00-9999/work/rt2x00-9999/ieee80211/ieee80211_crypt_wep.mod.o LD [M] /var/tmp/portage/rt2x00-9999/work/rt2x00-9999/ieee80211/ieee80211_crypt_wep.ko CC [M] /var/tmp/portage/rt2x00-9999/work/rt2x00-9999/rt2x00core.o CC [M] /var/tmp/portage/rt2x00-9999/work/rt2x00-9999/rt2400pci.o CC [M] /var/tmp/portage/rt2x00-9999/work/rt2x00-9999/rt2500pci.o CC [M] /var/tmp/portage/rt2x00-9999/work/rt2x00-9999/rt2500usb.o /var/tmp/portage/rt2x00-9999/work/rt2x00-9999/rt2500usb.c:1285: Fehler: unbekanntes Feld
see http://rt2x00.serialmonkey.com/phpBB2/viewtopic.php?t=1350&start=0&postdays=0&postorder=asc&highlight= i tried emerging net-wireless/rt2x00-9999 but failed two times: first it doesnt support MAKEOPTS="-j3" (fails with message 'file not found') and it still doesnt emerge: # MAKEOPTS="" emerge -va =rt2x00-9999 These are the packages that I would merge, in order: Calculating dependencies ...done! [ebuild N ] net-wireless/rt2x00-9999 -debug 0 kB Total size of downloads: 0 kB Do you want me to merge these packages? [Yes/No] y >>> emerge (1 of 1) net-wireless/rt2x00-9999 to / >>> md5 files ;-) rt2x00-9999.ebuild >>> md5 files ;-) rt2x00-2.0.0_beta3.ebuild >>> md5 files ;-) files/digest-rt2x00-2.0.0_beta3 >>> md5 files ;-) files/digest-rt2x00-9999 * Determining the location of the kernel source code * Found kernel source directory: * /usr/src/linux * Found sources for kernel version: * 2.6.16-gentoo-r6 * Checking for suitable kernel configuration options... [ ok ]>>> Unpacking source... * Fetching CVS module source/rt2x00 into /usr/portage/distfiles/cvs-src ... * Running cvs -q -f -z3 -d ":pserver:anonymous:@cvs.sourceforge.net:/cvsroot/rt2400" login Logging in to :pserver:anonymous@cvs.sourceforge.net:2401/cvsroot/rt2400 * Running cvs -q -f -z3 -d ":pserver:anonymous@cvs.sourceforge.net:/cvsroot/rt2400" update -dP rt2x00-9999 * Copying source/rt2x00 from /usr/portage/distfiles/cvs-src ... * CVS module source/rt2x00 is now in /var/tmp/portage/rt2x00-9999/work >>> Source unpacked. * Preparing ieee80211/80211 module CC [M] /var/tmp/portage/rt2x00-9999/work/rt2x00-9999/ieee80211/ieee80211_module.o CC [M] /var/tmp/portage/rt2x00-9999/work/rt2x00-9999/ieee80211/ieee80211_tx.o CC [M] /var/tmp/portage/rt2x00-9999/work/rt2x00-9999/ieee80211/ieee80211_rx.o CC [M] /var/tmp/portage/rt2x00-9999/work/rt2x00-9999/ieee80211/ieee80211_wx.o CC [M] /var/tmp/portage/rt2x00-9999/work/rt2x00-9999/ieee80211/ieee80211_geo.o LD [M] /var/tmp/portage/rt2x00-9999/work/rt2x00-9999/ieee80211/ieee80211.o CC [M] /var/tmp/portage/rt2x00-9999/work/rt2x00-9999/ieee80211/ieee80211_crypt.o CC [M] /var/tmp/portage/rt2x00-9999/work/rt2x00-9999/ieee80211/ieee80211_crypt_wep.o CC [M] /var/tmp/portage/rt2x00-9999/work/rt2x00-9999/ieee80211/ieee80211_crypt_ccmp.o CC [M] /var/tmp/portage/rt2x00-9999/work/rt2x00-9999/ieee80211/ieee80211_crypt_tkip.o Building modules, stage 2. MODPOST CC /var/tmp/portage/rt2x00-9999/work/rt2x00-9999/ieee80211/ieee80211.mod.o LD [M] /var/tmp/portage/rt2x00-9999/work/rt2x00-9999/ieee80211/ieee80211.ko CC /var/tmp/portage/rt2x00-9999/work/rt2x00-9999/ieee80211/ieee80211_crypt.mod.o LD [M] /var/tmp/portage/rt2x00-9999/work/rt2x00-9999/ieee80211/ieee80211_crypt.ko CC /var/tmp/portage/rt2x00-9999/work/rt2x00-9999/ieee80211/ieee80211_crypt_ccmp.mod.o LD [M] /var/tmp/portage/rt2x00-9999/work/rt2x00-9999/ieee80211/ieee80211_crypt_ccmp.ko CC /var/tmp/portage/rt2x00-9999/work/rt2x00-9999/ieee80211/ieee80211_crypt_tkip.mod.o LD [M] /var/tmp/portage/rt2x00-9999/work/rt2x00-9999/ieee80211/ieee80211_crypt_tkip.ko CC /var/tmp/portage/rt2x00-9999/work/rt2x00-9999/ieee80211/ieee80211_crypt_wep.mod.o LD [M] /var/tmp/portage/rt2x00-9999/work/rt2x00-9999/ieee80211/ieee80211_crypt_wep.ko CC [M] /var/tmp/portage/rt2x00-9999/work/rt2x00-9999/rt2x00core.o CC [M] /var/tmp/portage/rt2x00-9999/work/rt2x00-9999/rt2400pci.o CC [M] /var/tmp/portage/rt2x00-9999/work/rt2x00-9999/rt2500pci.o CC [M] /var/tmp/portage/rt2x00-9999/work/rt2x00-9999/rt2500usb.o /var/tmp/portage/rt2x00-9999/work/rt2x00-9999/rt2500usb.c:1285: Fehler: unbekanntes Feld »owner« in Initialisierung angegeben /var/tmp/portage/rt2x00-9999/work/rt2x00-9999/rt2500usb.c:1285: Warnung: Initialisierung von inkompatiblem Zeigertyp make[3]: *** [/var/tmp/portage/rt2x00-9999/work/rt2x00-9999/rt2500usb.o] Fehler 1 make[2]: *** [_module_/var/tmp/portage/rt2x00-9999/work/rt2x00-9999] Fehler 2 make[1]: *** [modules] Fehler 2 make: *** [rt2x00-nodebug] Fehler 2 !!! ERROR: net-wireless/rt2x00-9999 failed. !!! Function linux-mod_src_compile, Line 512, Exitcode 2 !!! Unable to make KERNDIR=/usr/src/linux KERNOUT=/usr/src/linux nodebug. !!! If you need support, post the topmost build error, NOT this status message. psytux psycho # uname -a Linux psytux 2.6.16-gentoo-r6 #2 Thu May 4 19:39:14 CEST 2006 i686 Intel(R) Pentium(R) 4 CPU 3.40GHz GNU/Linux
Please provide errors in english. I suspect this is an upstream problem. Please remember that -9999 ebuilds are sourced from CVS / SVN repositories and as such can't be guaranteed to build.
(In reply to comment #1) > Please provide errors in english. > > I suspect this is an upstream problem. Please remember that -9999 ebuilds are > sourced from CVS / SVN repositories and as such can't be guaranteed to build. > sorry for that, here is the english message /var/tmp/portage/rt2x00-9999/work/rt2x00-9999/rt2500usb.c:1285: Error unknown member
(In reply to comment #1) > Please provide errors in english. > > I suspect this is an upstream problem. Please remember that -9999 ebuilds are > sourced from CVS / SVN repositories and as such can't be guaranteed to build. > sorry for that, here is the english message /var/tmp/portage/rt2x00-9999/work/rt2x00-9999/rt2500usb.c:1285: Error unknown member »owner« mentioned in initialisation /var/tmp/portage/rt2x00-9999/work/rt2x00-9999/rt2500usb.c:1285: Warning Initialisating with incompatible pointer type im sure this translation is wrong, but as you can see this bug is fixed in upstream cvs but the ebuild doesnt use current cvs: http://rt2x00.serialmonkey.com/phpBB2/viewtopic.php?t=1350&start=0&postdays=0&postorder=asc&highlight
Yeah, SF CVS is crappy and isn't always sync'd to the devleoeprs CVS properly. I wrote an ebuild that uses the daily CVS tarballs that you might like to try - you can find it as the rt2x00-cvs-9999.ebuild attached to http://bugs.gentoo.org/show_bug.cgi?id=132240
Firsly, the fact that the rt2x00 public cvs tree is out of sync by a magnitude of days/weeks with their private cvs tree is not a Gentoo bug. Secondly we're not in the habit of patching cvs ebuilds. Last, but definitely not least, the ebuild warns you not to report bugs about the cvs here.
Actually, there's something to fix here, probably by introducing another masked ebuild, Allen Brooker's one (if it works ok) would do it. "Daily" tarballs (hourly actually, afaik) on the project site are more reliable than SF CVS and should then be available directly in portage (http://rt2x00.serialmonkey.com/rt2x00-cvs-daily.tar.gz). Why leave the portage situation as it is: we have the choice between outdated sources and (most often) unavailable ones...
(In reply to comment #5) > Actually, there's something to fix here, probably by introducing another masked > ebuild, Allen Brooker's one (if it works ok) would do it. > "Daily" tarballs (hourly actually, afaik) on the project site are more reliable > than SF CVS and should then be available directly in portage Sorry, that is not going to happen as the digest would be different for each daily tarball. Users are welcome to use that ebuild and roll their own digests though as that is something that portage cannot do. > Why leave the portage situation as it is: we have the choice between outdated > sources and (most often) unavailable ones... Such is the dilema when we offer cvs ebuilds. Maybe I should just remove the ebuild? Or maybe I should wait for SF to fix things their end or if the rt2x00 project moves to a better CVS server.
(In reply to comment #6) > (In reply to comment #5) > > Actually, there's something to fix here, probably by introducing another masked > > ebuild, Allen Brooker's one (if it works ok) would do it. > > "Daily" tarballs (hourly actually, afaik) on the project site are more reliable > > than SF CVS and should then be available directly in portage > > Sorry, that is not going to happen as the digest would be different for each > daily tarball. Users are welcome to use that ebuild and roll their own digests > though as that is something that portage cannot do. True. I forgot about that :) Let people get their ebuild from Bugzilla and use their own overlay then... > > > Why leave the portage situation as it is: we have the choice between outdated > > sources and (most often) unavailable ones... > > Such is the dilema when we offer cvs ebuilds. Maybe I should just remove the > ebuild? Or maybe I should wait for SF to fix things their end or if the rt2x00 > project moves to a better CVS server. > Well, as such this is experimental CVS sources, so people shouldn't be surprised if the ebuild fails. We can hope SF will eventually fix their CVS server. What i'd do is ewarn users about the CVS reliability issue, and inform them of the daily-tarball based ebuild available in bugzilla (http://bugs.gentoo.org/show_bug.cgi?id=132240).
Created attachment 86848 [details] Updated ebuild. Changes: o Updated new SourceForge CVS server address for this project. o Added a message about possible lags in SF CVS checkouts, with a link to the bugzilla daily-tarball based ebuild as an alternative.
Please update portage-tree ebuild with this one.
Why? The only difference between this one and the one in portage is that you have added a redundant warning and you have invalid keywords. Still WONTFIX
I did not change the keywords: basically, it's just portage ebuild ones (from my last sync). However I changed the CVS server address, it does not work with the old one since SF changed their CVS server farm. The warnings are not redundant at all: they're pointing to another better ebuild (not available in portage) cause it doesn't fetch outdated CVS sources. Did you really have a look at this ebuild before flaming Roy?
(In reply to comment #11) > I did not change the keywords: basically, it's just portage ebuild ones (from > my last sync). Your sync is old then > However I changed the CVS server address, it does not work with the old one > since SF changed their CVS server farm. http://www.gentoo.org/cgi-bin/viewcvs.cgi/net-wireless/rt2x00/rt2x00-9999.ebuild?rev=1.5&view=markup Maybe you should sync again as I changed the server address 4 days ago. > The warnings are not redundant at all: they're pointing to another better > ebuild (not available in portage) cause it doesn't fetch outdated CVS sources. You ebuild is not better and cannot be put in portage for reasons already described. The fact that SF public CVS lags behind the private CVS by a little bit isn't enough justification. Our public cvs repo lags about 2 hours behind the public rsync repo's - maybe we should add a similar warning to every ebuild? Nah > Did you really have a look at this ebuild before flaming Roy? Did you really have a look at current portage before bugging me again? I'll be expecting that appology now :)
True, my sync was more than 4 days old, i apologise for that. :) However, the CVS lags we talk about are more like weeks than hours out-of-sync. That deserves a warning according to me. In fact, the daily-tarballs are the only one we recommend in this project: too many noise, reports about outdated bugs from CVS fetching people. We don't have such problems with daily-tarballs. The daily-tarball ebuild cannot be integrated to portage for above mentionned reasons, but informing users it actually exists seems to be relevant. Anyway, do what you like. :)
(In reply to comment #13) > True, my sync was more than 4 days old, i apologise for that. :) > However, the CVS lags we talk about are more like weeks than hours out-of-sync. That is why sourceforge have moved to new servers - why don't you test an hour or two after IvD commits a change? > That deserves a warning according to me. > In fact, the daily-tarballs are the only one we recommend in this project: too > many noise, reports about outdated bugs from CVS fetching people. Maybe I'll add a warning if sourceforge messes up again. But until then - no.
Just to follow up, IvD committed a change to cvs and the instant he posted to the rt2400 mailing list I emerged rt2x00 and it got the change. So the sync time is very low now :)
