From stas at FreeBSD.org Wed Apr 8 15:09:57 2009 From: stas at FreeBSD.org (Stanislav Sedov) Date: Wed Apr 8 15:10:08 2009 Subject: [rubygems-]: Request for Review and -exp Run In-Reply-To: <49DCF1F2.3040501@p6m7g8.com> References: <49DCCC93.2010209@p6m7g8.com> <1239216349.50237.2.camel@hood.oook.cz> <49DCF1F2.3040501@p6m7g8.com> Message-ID: <20090409002918.e48eecde.stas@FreeBSD.org> -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On Wed, 8 Apr 2009 14:50:26 -0400 "Philip M. Gollucci" mentioned: > Pav Lucistnik wrote: > >> Diff: > >> ----- > >> http://people.freebsd.org/~pgollucci/rubygems.diff > > > > If only the individual ports are changed, we does not need to waste > > resources on exp run. Just run these seven ports in Tinderbox... > > Would have helped if you looked at the diff or did you and that still applies ? > > Those seven ports have the deltas listed in *ADDITION* to the first 3 > which are global across all remaining rubygem-* ports (except maybe mongrel > and merb) > > > $ grep -c Index: rubygems.diff > 111 I don't think Exprun is really required. We can run this on tinderbox - ruby ports builds pretty fast. The change seems to be straightforward and limited to the affected ports only. - -- Stanislav Sedov ST4096-RIPE -----BEGIN PGP SIGNATURE----- iEYEARECAAYFAkndCSoACgkQK/VZk+smlYEPNQCfaZJD/LzuQ2ckl+qx8qV/Udjb orYAn2h5syUNYJet8/nBH58oUTfO2NpO =Xplo -----END PGP SIGNATURE----- !DSPAM:49dd1d31967001990318312! From pgollucci at p6m7g8.com Wed Apr 8 15:35:15 2009 From: pgollucci at p6m7g8.com (Philip M. Gollucci) Date: Wed Apr 8 15:35:26 2009 Subject: [rubygems-]: Request for Review and -exp Run In-Reply-To: <20090409002918.e48eecde.stas@FreeBSD.org> References: <49DCCC93.2010209@p6m7g8.com> <1239216349.50237.2.camel@hood.oook.cz> <49DCF1F2.3040501@p6m7g8.com> <20090409002918.e48eecde.stas@FreeBSD.org> Message-ID: <49DD2449.4080509@p6m7g8.com> Stanislav Sedov wrote: > I don't think Exprun is really required. We can run this on tinderbox - ruby > ports builds pretty fast. The change seems to be straightforward and limited > to the affected ports only. I've already run it on a tinderbox at $work (RideCharge Inc) unfortunately its not public. -- ------------------------------------------------------------------------ 1024D/DB9B8C1C B90B FBC3 A3A1 C71A 8E70 3F8C 75B8 8FFB DB9B 8C1C Philip M. Gollucci (pgollucci@p6m7g8.com) c: 703.336.9354 Consultant - P6M7G8 Inc. http://p6m7g8.net Senior Sys Admin - RideCharge, Inc. http://ridecharge.com Contractor - PositiveEnergyUSA http://positiveenergyusa.com ASF Member - Apache Software Foundation http://apache.org FreeBSD Committer - FreeBSD Foundation http://freebsd.org Work like you don't need the money, love like you'll never get hurt, and dance like nobody's watching. From stas at FreeBSD.org Wed Apr 8 15:39:18 2009 From: stas at FreeBSD.org (Stanislav Sedov) Date: Wed Apr 8 15:39:30 2009 Subject: [rubygems-]: Request for Review and -exp Run In-Reply-To: <49DD2449.4080509@p6m7g8.com> References: <49DCCC93.2010209@p6m7g8.com> <1239216349.50237.2.camel@hood.oook.cz> <49DCF1F2.3040501@p6m7g8.com> <20090409002918.e48eecde.stas@FreeBSD.org> <49DD2449.4080509@p6m7g8.com> Message-ID: <20090409023917.223c7d1e.stas@FreeBSD.org> On Wed, 8 Apr 2009 18:25:13 -0400 "Philip M. Gollucci" mentioned: > Stanislav Sedov wrote: > > I don't think Exprun is really required. We can run this on tinderbox - ruby > > ports builds pretty fast. The change seems to be straightforward and limited > > to the affected ports only. > I've already run it on a tinderbox at $work (RideCharge Inc) > > unfortunately its not public. > Then it should be safe to commit if the run was successfull. -- Stanislav Sedov ST4096-RIPE !DSPAM:49dd2794967001105412841! From pgollucci at ridecharge.com Fri Apr 10 15:31:57 2009 From: pgollucci at ridecharge.com (Philip M. Gollucci) Date: Fri Apr 10 15:32:03 2009 Subject: [RFC]: use --no-rdoc to implement NOPORTDOCS for rubygem-* Message-ID: <49DFC680.9000104@ridecharge.com> http://people.freebsd.org/~pgollucci/no_rdoc.diff The only rubygem-* port this breaks is www/rubygem-merb mysql> select count(1) from build_ports bp, ports p where bp.port_id = p.port_id and p.port_directory like '%/rubygem-%' and bp.build_id = 27; +----------+ | count(1) | +----------+ | 168 | +----------+ -- ------------------------------------------------------------------------ Philip M. Gollucci (pgollucci@ridecharge.com) p: 703.549.2050x206, did: 703.579.6947 Senior System Admin - RideCharge, Inc. http://ridecharge.com 1024D/DB9B8C1C B90B FBC3 A3A1 C71A 8E70 3F8C 75B8 8FFB DB9B 8C1C Work like you don't need the money, love like you'll never get hurt, and dance like nobody's watching. From pgollucci at p6m7g8.com Fri Apr 10 15:40:45 2009 From: pgollucci at p6m7g8.com (Philip M. Gollucci) Date: Fri Apr 10 15:40:52 2009 Subject: [RFC]: RUBYGEM_AUTOPLIST implied by USE_RUBYGEMS=yes In-Reply-To: <49DFC680.9000104@ridecharge.com> References: <49DFC680.9000104@ridecharge.com> Message-ID: <49DFCAEB.60505@p6m7g8.com> Philip M. Gollucci wrote: > http://people.freebsd.org/~pgollucci/no_rdoc.diff > > The only rubygem-* port this breaks is www/rubygem-merb > > Now that all rubygem-* ports use RUBYGEM_AUTOPLIST except www/rubygem-merb and www/rubygem-mongrel I'd like to see Mk/bsd.ruby.mk have USE_RUBYGEMS= yes imply RUBYGEM_AUTOPLIST=yes Then tweaking the previous no_rdoc.diff patch we can set RUBYGEM_AUTOPLIST=no for these 2 'unique' ports. I'd provide a diff, but I'm so far ahead of whats in CVS right now it would be useless. -- ------------------------------------------------------------------------ 1024D/DB9B8C1C B90B FBC3 A3A1 C71A 8E70 3F8C 75B8 8FFB DB9B 8C1C Philip M. Gollucci (pgollucci@p6m7g8.com) c: 703.336.9354 Consultant - P6M7G8 Inc. http://p6m7g8.net Senior Sys Admin - RideCharge, Inc. http://ridecharge.com Contractor - PositiveEnergyUSA http://positiveenergyusa.com ASF Member - Apache Software Foundation http://apache.org FreeBSD Committer - FreeBSD Foundation http://freebsd.org Work like you don't need the money, love like you'll never get hurt, and dance like nobody's watching. From pgollucci at p6m7g8.com Fri Apr 10 16:24:39 2009 From: pgollucci at p6m7g8.com (Philip M. Gollucci) Date: Fri Apr 10 16:24:49 2009 Subject: */rubygem-* => MAKE_JOBS_SAFE=yes Message-ID: <49DFD535.2000107@p6m7g8.com> Is it worth setting this ? They all have NO_BUILD= yes from Mk/bsd.ruby.mk (at any rate all 168 of them work on quad core with -j8 in my tbs) -- ------------------------------------------------------------------------ 1024D/DB9B8C1C B90B FBC3 A3A1 C71A 8E70 3F8C 75B8 8FFB DB9B 8C1C Philip M. Gollucci (pgollucci@p6m7g8.com) c: 703.336.9354 Consultant - P6M7G8 Inc. http://p6m7g8.net Senior Sys Admin - RideCharge, Inc. http://ridecharge.com Contractor - PositiveEnergyUSA http://positiveenergyusa.com ASF Member - Apache Software Foundation http://apache.org FreeBSD Committer - FreeBSD Foundation http://freebsd.org Work like you don't need the money, love like you'll never get hurt, and dance like nobody's watching. From pav at FreeBSD.org Fri Apr 10 23:32:56 2009 From: pav at FreeBSD.org (Pav Lucistnik) Date: Fri Apr 10 23:33:03 2009 Subject: */rubygem-* => MAKE_JOBS_SAFE=yes In-Reply-To: <49DFD535.2000107@p6m7g8.com> References: <49DFD535.2000107@p6m7g8.com> Message-ID: <1239429944.70926.1.camel@hood.oook.cz> Philip M. Gollucci p??e v p? 10. 04. 2009 v 19:24 -0400: > Is it worth setting this ? > > They all have NO_BUILD= yes from Mk/bsd.ruby.mk So it would be no-op? > (at any rate all 168 of them work on quad core with -j8 in my tbs) > > -- Pav Lucistnik If God is perfect, why did He create discontinuous functions? -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 195 bytes Desc: Toto je =?UTF-8?Q?digit=C3=A1ln=C4=9B?= =?ISO-8859-1?Q?_podepsan=E1?= =?UTF-8?Q?_=C4=8D=C3=A1st?= =?ISO-8859-1?Q?_zpr=E1vy?= Url : http://lists.freebsd.org/pipermail/freebsd-ruby/attachments/20090411/da0b122e/attachment.pgp From stas at FreeBSD.org Sat Apr 11 00:20:31 2009 From: stas at FreeBSD.org (Stanislav Sedov) Date: Sat Apr 11 00:20:39 2009 Subject: */rubygem-* => MAKE_JOBS_SAFE=yes In-Reply-To: <49DFD535.2000107@p6m7g8.com> References: <49DFD535.2000107@p6m7g8.com> Message-ID: <20090411112030.2279adc2.stas@FreeBSD.org> -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On Fri, 10 Apr 2009 19:24:37 -0400 "Philip M. Gollucci" mentioned: > Is it worth setting this ? > > They all have NO_BUILD= yes from Mk/bsd.ruby.mk > > (at any rate all 168 of them work on quad core with -j8 in my tbs) > Have you checked if it have any real benefit? - -- Stanislav Sedov ST4096-RIPE -----BEGIN PGP SIGNATURE----- iEYEARECAAYFAkngRMIACgkQK/VZk+smlYEQUwCfY5DKdedUG/mYkaQlqKQEpsbE bPQAnAm9DADkgp1syomlPFxR1En/DWpk =9RVJ -----END PGP SIGNATURE----- !DSPAM:49e044bd967002020418523! From stas at FreeBSD.org Sat Apr 11 00:32:12 2009 From: stas at FreeBSD.org (Stanislav Sedov) Date: Sat Apr 11 00:32:18 2009 Subject: [RFC]: RUBYGEM_AUTOPLIST implied by USE_RUBYGEMS=yes In-Reply-To: <49DFCAEB.60505@p6m7g8.com> References: <49DFC680.9000104@ridecharge.com> <49DFCAEB.60505@p6m7g8.com> Message-ID: <20090411113216.24170d6f.stas@FreeBSD.org> -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On Fri, 10 Apr 2009 18:40:43 -0400 "Philip M. Gollucci" mentioned: > Philip M. Gollucci wrote: > > http://people.freebsd.org/~pgollucci/no_rdoc.diff > > > > The only rubygem-* port this breaks is www/rubygem-merb > > > > > > Now that all rubygem-* ports use RUBYGEM_AUTOPLIST > except www/rubygem-merb and www/rubygem-mongrel > > I'd like to see Mk/bsd.ruby.mk have > USE_RUBYGEMS= yes > > imply RUBYGEM_AUTOPLIST=yes > > Then tweaking the previous no_rdoc.diff patch > > we can set RUBYGEM_AUTOPLIST=no > for these 2 'unique' ports. > > I'd provide a diff, but I'm so far ahead of whats in CVS right now > it would be useless. > I don't think it is a good idead as historically we prefer static pkg-plists over dynamic ones. For ports where names of the files are perfectly known AUTOPLIST feature is not required. When I introduced this option my intention was to work around rdoc names generation problem and to not eliminate rubygem pkg-plists entirely. - -- Stanislav Sedov ST4096-RIPE -----BEGIN PGP SIGNATURE----- iEYEARECAAYFAkngR4AACgkQK/VZk+smlYH4FQCfRjR1Jk22gtjFOtVqkZu6Lper RCgAn0XTaaZKwS1zaHKo5/rhKkDS78Gt =4QuZ -----END PGP SIGNATURE----- !DSPAM:49e0477a967005749415982! From stas at FreeBSD.org Sat Apr 11 00:42:00 2009 From: stas at FreeBSD.org (Stanislav Sedov) Date: Sat Apr 11 00:42:06 2009 Subject: [RFC]: use --no-rdoc to implement NOPORTDOCS for rubygem-* In-Reply-To: <49DFC680.9000104@ridecharge.com> References: <49DFC680.9000104@ridecharge.com> Message-ID: <20090411114203.b70ab3fd.stas@FreeBSD.org> -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On Fri, 10 Apr 2009 18:21:52 -0400 "Philip M. Gollucci" mentioned: > http://people.freebsd.org/~pgollucci/no_rdoc.diff > > The only rubygem-* port this breaks is www/rubygem-merb > Nice. Thanks for working on it! Can you work on the rubygem-merb as well? Probably, we can implement some temporary hack here... - -- Stanislav Sedov ST4096-RIPE -----BEGIN PGP SIGNATURE----- iEYEARECAAYFAkngScsACgkQK/VZk+smlYFuvwCeL+GKJ+JOwUb4kv6g1C4uMili yFUAn1LBXjXn0FLCi5v2W0RIu9Gu86gy =0eeB -----END PGP SIGNATURE----- !DSPAM:49e049c6967005065714533! From pgollucci at p6m7g8.com Sat Apr 11 07:36:35 2009 From: pgollucci at p6m7g8.com (Philip M. Gollucci) Date: Sat Apr 11 07:37:07 2009 Subject: */rubygem-* => MAKE_JOBS_SAFE=yes In-Reply-To: <1239429944.70926.1.camel@hood.oook.cz> References: <49DFD535.2000107@p6m7g8.com> <1239429944.70926.1.camel@hood.oook.cz> Message-ID: <49E0AAE2.9090803@p6m7g8.com> Pav Lucistnik wrote: > Philip M. Gollucci p??e v p? 10. 04. 2009 v 19:24 -0400: >> Is it worth setting this ? >> >> They all have NO_BUILD= yes from Mk/bsd.ruby.mk > > So it would be no-op? Thats correct, but I forgot if we were adding it to all ports for Tracking. From pgollucci at p6m7g8.com Sat Apr 11 07:38:33 2009 From: pgollucci at p6m7g8.com (Philip M. Gollucci) Date: Sat Apr 11 07:38:39 2009 Subject: [RFC]: RUBYGEM_AUTOPLIST implied by USE_RUBYGEMS=yes In-Reply-To: <20090411113216.24170d6f.stas@FreeBSD.org> References: <49DFC680.9000104@ridecharge.com> <49DFCAEB.60505@p6m7g8.com> <20090411113216.24170d6f.stas@FreeBSD.org> Message-ID: <49E0AB5D.60403@p6m7g8.com> > I don't think it is a good idead as historically we prefer static pkg-plists > over dynamic ones. For ports where names of the files are perfectly known > AUTOPLIST feature is not required. When I introduced this option my intention > was to work around rdoc names generation problem and to not eliminate rubygem > pkg-plists entirely. *sigh*. Wouldn't every rubygem port have to use it to work around that issue (the zfs one?)? That means every rubygem should fail in QAT then b/c of the NOPORTDOCS .... he must be excluding them. From pav at FreeBSD.org Sat Apr 11 09:20:39 2009 From: pav at FreeBSD.org (Pav Lucistnik) Date: Sat Apr 11 09:21:11 2009 Subject: */rubygem-* => MAKE_JOBS_SAFE=yes In-Reply-To: <49E0AAE2.9090803@p6m7g8.com> References: <49DFD535.2000107@p6m7g8.com> <1239429944.70926.1.camel@hood.oook.cz> <49E0AAE2.9090803@p6m7g8.com> Message-ID: <49E0C34A.2090107@FreeBSD.org> Philip M. Gollucci wrote: > Pav Lucistnik wrote: >> Philip M. Gollucci p??e v p? 10. 04. 2009 v 19:24 -0400: >>> Is it worth setting this ? >>> >>> They all have NO_BUILD= yes from Mk/bsd.ruby.mk >> >> So it would be no-op? > Thats correct, but I forgot if we were adding it to all ports for Tracking. Not at this point. But adding a note about it to Wiki page would be nice. Pav From stas at FreeBSD.org Sat Apr 11 09:31:59 2009 From: stas at FreeBSD.org (Stanislav Sedov) Date: Sat Apr 11 09:32:05 2009 Subject: [RFC]: RUBYGEM_AUTOPLIST implied by USE_RUBYGEMS=yes In-Reply-To: <49E0AB5D.60403@p6m7g8.com> References: <49DFC680.9000104@ridecharge.com> <49DFCAEB.60505@p6m7g8.com> <20090411113216.24170d6f.stas@FreeBSD.org> <49E0AB5D.60403@p6m7g8.com> Message-ID: <20090411203159.25fd4c48.stas@FreeBSD.org> -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On Sat, 11 Apr 2009 10:38:21 -0400 "Philip M. Gollucci" mentioned: > > I don't think it is a good idead as historically we prefer static pkg-plists > > over dynamic ones. For ports where names of the files are perfectly known > > AUTOPLIST feature is not required. When I introduced this option my intention > > was to work around rdoc names generation problem and to not eliminate rubygem > > pkg-plists entirely. > *sigh*. > > Wouldn't every rubygem port have to use it to work around that issue > (the zfs one?)? Nope, as not all gem ports contain auto-generated rdoc files. > > That means every rubygem should fail in QAT then b/c of the NOPORTDOCS > .... he must be excluding them. > Well, as I said before I like the idea of automatic handle of NOPORTDOCS key. Often, docs included are just static files with known names, and in this case AUTOPLIST feature is not needed. Otherwise, the port can use AUTOPLIST feature. - -- Stanislav Sedov ST4096-RIPE -----BEGIN PGP SIGNATURE----- iEYEARECAAYFAkngxgMACgkQK/VZk+smlYHlgQCffRvPSYjYDFU5LRr3LiYunC2E sHIAn23DUOgquFzhQpTRlUhLGT9rBhvM =8scD -----END PGP SIGNATURE----- !DSPAM:49e0c5fc967001346518677!