April 2011 - macports-tickets - lists.macosforge.org

Re: [MacPorts] #10879: RFE: apache2 and other ports won't install as non-root users
by MacPorts 02 Apr '11

02 Apr '11

#10879: RFE: apache2 and other ports won't install as non-root users ------------------------------------+--------------------------------------- Reporter: plambert@… | Owner: macports-tickets@… Type: enhancement | Status: closed Priority: Normal | Milestone: MacPorts 2.0.0 Component: base | Version: Resolution: fixed | Keywords: Port: | ------------------------------------+--------------------------------------- Changes (by jmr@…): * status: assigned => closed * resolution: => fixed * milestone: MacPorts Future => MacPorts 2.0.0 Comment: r66700, r75143 -- Ticket URL: <https://trac.macports.org/ticket/10879#comment:12> MacPorts <http://www.macports.org/> Ports system for Mac OS

1 0

Re: [MacPorts] #11563: RFE: adduser should be done the same way as startupitems
by MacPorts 02 Apr '11

02 Apr '11

#11563: RFE: adduser should be done the same way as startupitems -------------------------------+-------------------------------------------- Reporter: yves@… | Owner: macports-tickets@… Type: enhancement | Status: new Priority: Normal | Milestone: MacPorts 2.0.0 Component: base | Version: Keywords: | Port: -------------------------------+-------------------------------------------- Changes (by jmr@…): * priority: Low => Normal * milestone: MacPorts Future => MacPorts 2.0.0 Comment: This is going to be important to allow installing from binaries to work correctly. -- Ticket URL: <https://trac.macports.org/ticket/11563#comment:4> MacPorts <http://www.macports.org/> Ports system for Mac OS

1 0

[MacPorts] #28847: mfiler3
by MacPorts 02 Apr '11

02 Apr '11

#28847: mfiler3 ------------------------------+--------------------------------------------- Reporter: ab25cq@… | Owner: macports-tickets@… Type: submission | Status: new Priority: Normal | Milestone: Component: ports | Version: 1.9.2 Keywords: | Port: ------------------------------+--------------------------------------------- a console file manger with saphire shell which is a shell script language. mfiler3 requires saphire shell. -- Ticket URL: <https://trac.macports.org/ticket/28847> MacPorts <http://www.macports.org/> Ports system for Mac OS

1 7

[MacPorts] #28846: saphire shell
by MacPorts 02 Apr '11

02 Apr '11

#28846: saphire shell ------------------------------+--------------------------------------------- Reporter: ab25cq@… | Owner: macports-tickets@… Type: submission | Status: new Priority: Normal | Milestone: Component: ports | Version: 1.9.2 Keywords: | Port: ------------------------------+--------------------------------------------- -- Ticket URL: <https://trac.macports.org/ticket/28846> MacPorts <http://www.macports.org/> Ports system for Mac OS

1 7

[MacPorts] #27506: Incorrect SQL in reg_entry_dependents() from cregistry/entry.c
by MacPorts 02 Apr '11

02 Apr '11

#27506: Incorrect SQL in reg_entry_dependents() from cregistry/entry.c -------------------------------------+-------------------------------------- Reporter: mmpestorich@… | Owner: macports-tickets@… Type: defect | Status: new Priority: Normal | Milestone: Component: base | Version: 1.9.2 Keywords: sqlite entry dependents | Port: -------------------------------------+-------------------------------------- I recently noticed a problem when running {{{port dependents ...some port...}}} in that it would return only the name of the port queired instead of a list of its dependents. (i.e. {{{port dependents openssl}}} would show {{{openssl depends on openssl}}} instead of the expected {{{openssh depends on openssl}}} ...). The actual dependencies however were still being calculated correctly. {{{port uninstall ...some port...}}} would still complain about dependent ports being installed, but as in the case above would display lines of dependent ports consisting only the current port's name and not it's dependents. I had recently upgraded my SQLite version to 3.7.3 and upgraded from macports flat file registry to the newer sqlite registry. This made me think that there might the issue with how information is being pulled from the database. I found this; [http://www.sqlite.org/lang_select.html#fromclause SQLite Documentation] states: > ''For each pair of columns identified by a USING clause, the column from the right-hand dataset is omitted from the joined dataset. This is the only difference between a USING clause and its equivalent ON constraint.'' Currently, the SQL statement in reg_entry_dependents is: {{{ #!sql SELECT dependent.id FROM ports port INNER JOIN dependencies USING(name) INNER JOIN ports dependent USING(id) WHERE port.id=%lld ORDER BY dependent.name, dependent.epoch, dependent.version, dependent.revision, dependent.variants }}} When I ran the above statement in the sqlite3 command line utility I did infact get back the same incorrect result that the port command was giving me. When I replaced the USING clauses with ON clauses, I received the correct expected result. I then read the snippet of docmentation above and tried the modified query below. It worked as well. Based on that SQLite documentation and my testing, I believe ([source:trunk/base/src/cregistry/entry.c#L1232 cregistry/entry.c Line1232]) should instead be: {{{ #!sql SELECT dependencies.id FROM ports port }}} Side Note: I have a couple of computers and this was only an issue on one of them. The only difference that I could tell between the two installations was the version of SQLite. The computer with SQLite version 3.6.23.1 worked properly while the computer that had version 3.7.3 did not. MacPorts 1.9.2 is installed on both of them. I have attached my patch that fixed this issue for me. -- Ticket URL: <https://trac.macports.org/ticket/27506> MacPorts <http://www.macports.org/> Ports system for Mac OS

1 5

Re: [MacPorts] #13143: Hard-links used for image-mode don't really fit Time Machine
by MacPorts 02 Apr '11

02 Apr '11

#13143: Hard-links used for image-mode don't really fit Time Machine -------------------------------------------+-------------------------------- Reporter: spiralyoda-macports@… | Owner: macports-tickets@… Type: defect | Status: closed Priority: Low | Milestone: MacPorts 2.0.0 Component: base | Version: 1.5.0 Resolution: fixed | Keywords: Port: | -------------------------------------------+-------------------------------- Changes (by jmr@…): * status: new => closed * resolution: => fixed * milestone: MacPorts Future => MacPorts 2.0.0 Comment: No longer using hardlinks as of r77511. -- Ticket URL: <https://trac.macports.org/ticket/13143#comment:8> MacPorts <http://www.macports.org/> Ports system for Mac OS

1 0

[MacPorts] #19458: Combine image and archive modes
by MacPorts 02 Apr '11

02 Apr '11

#19458: Combine image and archive modes ------------------------------+--------------------------------------------- Reporter: blb@… | Owner: blb@… Type: enhancement | Status: new Priority: Normal | Milestone: MacPorts Future Component: base | Version: 1.8.0 Keywords: archive image | Port: ------------------------------+--------------------------------------------- Combining the image and archive modes into a new image-through-archives and always enabling it buys several advantages: - no more need to worry about links for image mode as all inactive ports would simply exist in the archive instead of being fully installed in ${prefix}/var/macports/software - configuration settings for these (portinstalltype, portarchivemode, portarchivepath, and portarchivetype) can be removed simplifying code and macports.conf - always building/using archives should bring MacPorts closer to supporting install-from-binaries instead of just from source - removal of archive-specific targets/actions (since they'd be merged into the activate area) simplifies MacPorts base code I'll create a [browser:branches/images-and-archives branch for development] to keep this off of trunk until it is ready for more widespread testing. -- Ticket URL: <http://trac.macports.org/ticket/19458> MacPorts <http://www.macports.org/> Ports system for Mac OS

1 4

[MacPorts] #28425: A port shouldn't depend on itself
by MacPorts 01 Apr '11

01 Apr '11

#28425: A port shouldn't depend on itself ---------------------------------+------------------------------------------ Reporter: vinc17@… | Owner: macports-tickets@… Type: defect | Status: new Priority: Normal | Milestone: Component: base | Version: 1.9.2 Keywords: | Port: ---------------------------------+------------------------------------------ I have the following problem: {{{ $ sudo port -v uninstall atlas @3.8.3_4+gcc44 ---> Unable to uninstall/deactivate atlas @3.8.3_4+gcc44, the following ports depend on it: ---> atlas @3.8.3_4+gcc44 ---> atlas @3.8.3_4+gcc44 Error: Target org.macports.uninstall returned: Please uninstall the ports that depend on atlas first. Warning: the following items did not execute (for atlas): org.macports.uninstall Log for atlas is at: /opt/local/var/macports/logs/_opt_local_var_macports_registry_portfiles_atlas_3.8.3_4+gcc44/main.log Warning: Failed to execute portfile from registry for atlas @3.8.3_4+gcc44 ---> Unable to uninstall/deactivate atlas @3.8.3_4+gcc44, the following ports depend on it: ---> atlas @3.8.3_4+gcc44 ---> atlas @3.8.3_4+gcc44 Error: port uninstall failed: Please uninstall the ports that depend on atlas first. }}} i.e. atlas depend on itself! The Portfile seems correct, so I assume that it's a bug in base. -- Ticket URL: <https://trac.macports.org/ticket/28425> MacPorts <http://www.macports.org/> Ports system for Mac OS

1 7

[MacPorts] #28698: llvm 2.8 buildroot leaky
by MacPorts 01 Apr '11

01 Apr '11

#28698: llvm 2.8 buildroot leaky --------------------------------------+------------------------------------- Reporter: howarth@… | Owner: macports-tickets@… Type: defect | Status: new Priority: Normal | Milestone: Component: ports | Version: 1.9.2 Keywords: languages | Port: llvm --------------------------------------+------------------------------------- The llvm build root in v2.8 is leaky because of a flaw in Makefile.rules. The attached patch (which has been committed to llvm svn for 2.9) fixes this by making sure that LD.Flags precedes LDFLAGS. Otherwise, if -L/opt/local/lib is passed to LDFLAGS, the current llvm 2.8 Makefile.rules allows that path to override those set for the newly built libraries in the buildroot. -- Ticket URL: <https://trac.macports.org/ticket/28698> MacPorts <http://www.macports.org/> Ports system for Mac OS

1 10

[MacPorts] #28999: Upgrade: p5-freezethaw to 0.5001
by MacPorts 01 Apr '11

01 Apr '11

#28999: Upgrade: p5-freezethaw to 0.5001 --------------------------------+------------------------------------------- Reporter: ricci@… | Owner: dluke@… Type: update | Status: new Priority: Normal | Milestone: Component: ports | Version: 1.9.2 Keywords: | Port: p5-freezethaw --------------------------------+------------------------------------------- Upgrade p5-freezethaw to 0.5001 -- Ticket URL: <https://trac.macports.org/ticket/28999> MacPorts <http://www.macports.org/> Ports system for Mac OS

1 1