While we’re still waiting for OpenOffice.org to compile in the last run, linimon compiled the next set of patches for testing on pointyhat. This will also include the first results of the Summer of Code project on the ports infrastructure. The following patches are scheduled for testing:
89398: [PATCH] bsd.wx.mk and patches
92445: current-us [patch] change all bogus uses of BROKEN to IGNORE
93601: new feature: make missing
98105: new feature: IA32_BINARY_PORT macro
98206: bsd.port.mk – whitespace and CHECKSUM_ALGO fix
98565: Add PKGVERSION to bsd.port.mk
98731: add bsd.scons.mk to Mk/
98891: bsd.port.mk: Fix DESKTOP_ENTRIES processing on 4.x
99370: USE_LDCONFIG fails installation for non-default prefix
Monthly Archives: June 2006
New bsd.port.mk patch set
59. Belgian ale
Belgian ale
4100g pale malt
112g cara-munich
112g brown malt
Mash @70°C for 60 min.
Add 500g dark candi sugar at the start of boil
28g Styrian goldings for 60 min.
10g Hallertau for 60 min.
10g Saaz for 15 min.
White Labs WLP530 Abbey Ale
21L
OG: 12.8 brix (1.052)
FG: 7.7 brix (1.011)
Racked: 2/7
Bottled: 22/7
with 150g glucose
New ports committer: Rong-En Fan
Symposium at Træmuseet (wood museum)
Every second year, the museum for woodwork (Træmuseet – Museum for træbearbejdning) aranges a symposium with international sculptors who create a sculpture on the spot within one week. I went to have a look on how it was going yesterday, and as always when I go to visit Fins parents, they have some work for me to do. This time, they would like to see some pictures from the symposium on the site of the museum (danish – click “billeder” on the front page), so Fin took some pictures and small videos which I put on the site today. Go and have a look.
New ports committer: Boris Samorodov
Ports infrastructure imported to perforce
gtetlow has been so kind to add the ports infrastructure files to the import from the FreeBSD cvs repository to a vendor branch perforce. The files under ports/ include CHANGES, KNOBS, LEGAL, MOVED, Makefile, Mk/, Templates/, Tools/ and UPDATING. This gives people wanting to do development on the ports infrastructure and scripts a real RCS before things are added to cvs. In general, before patches are committed to the ports infrastructure they are thoroughly tested by a build of all ports on the pointyhat cluster. Perforce give people the possiblity to keep patches in a RCS before they are ready for a pointyhat test run. I already started my own branch.
Google works in mysterious ways
SSLUUG/DKUUG had arranged to get Ole Tange, a dane working for Google, to do a presentation about Linux and Google. Christian Laursen was quick to his mail and managed to set up the same presentation for AaUUG and even found a big enough room for it on very short notice.
Unfortunately, shortly after the presentation was announced, it had to be cancelled because the legal department at Google hadn’t sanctioned it (yet). Google works in mysterious ways… Of course, I really like what they are doing with the Summer of Code, but with all the work Yahoo! has done for FreeBSD, I’m more and more thinking that Simon is right. Of course we all know that Google is slowly but certainly working behind the scenes on building a do-no-evil empire with all the new features they add just about every second day. Hey, what are those black helicopters doing outsiNOCARRIER
Certified
Ports infrastructure ideas and USE_* cleanup
After several months on my TODO-list, I finally pulled myself together and rewrote some of the ideas for the ports infrastructure in a bit more readable english, instead “only I need to read and understand this”, and SGMLify it for the webpage with some help from joel@. See the result here.
Nothing better than to take some of my own medicine, so I started working on one of the items on the list immediately. The use of USE_* and WITH(OUT)_* is quite a mess throughout the tree and could do with a bit of a cleanup. USE_* macros can only be defined internally to the ports infrastructure, not by ports themselves where WITH_* should be used. Looking through the tree, I found several cases of misspelling, which were easy to fix (USE_GNAKE, PREFIY, etc.) There were also a number of cases where macros were called that never existed (USE_PERL=USE_PERL5(_BUILD|_RUN), USE_GZIP) and some experimental macros that never made it into the tree (USE_GPG). The list of suspicious entries already is a lot lower than when I started, and next in line will be fixing the abuse of WITH_* in the ports themselves.