Package Manager to compile all packages as in Gentoo

Hello,
I love Arch and was looking to get my hands dirty, I there a way where I can compile all the packages that I install like in Gentoo rather than using only the pre-compiled i686 packages? I looked at ABS but was not sure how to keep track of the changes, I believe I simply cant do '-Syu' with ABS. I am just looking for something like Yaourt or Packer AUR helpers but for all the packages from core/extra/community.
Thank You
Gopi

RichAustin wrote:I'm perhaps being a bit dumb here but why not just use Gentoo instead? I am actually quite serious here since I wouldn't have thought there was a major difference between the two except you compile everything in Gentoo. I can't really see why you would want to find a workaround to do what you can already do in Gentoo.
I recently spent many days relearning the Gentoo way, did a stage3 Gentoo install alongside my Arch system, compiled a custom kernel from pf-sources, updated world, spent time reading the forum and web pages.  I won't try to replay "what I found wrong in Gentoo" except for the refrain "Gentoo is broken." That was my take. I found important things broken for long periods with no attention. People rude to each other.   I won't do that scene...
Arch may have a smaller developer team but I've experienced a generally high level of cooperation and discipline here. I've seen complaints about the forum here being moderated -- to me that seems to work for our good.  Thanks to the moderators who keep this place civil and rational!
There seems to be a good knowledge base here and the principle of always keeping up-to-date with the upstream releases forces the developers to constantly parse upstream information to roll a coherent stream of Arch packages. The knowledge here is as good as anywhere.  I just don't expect Arch Linux to be an integrated operating system, like Debian stable or Ubuntu.  In my experience it gets almost as much stability as Debian Testing by virtue of quickly taking in the upstream releases, which then fix the bugs that Arch never has to fix ourselves.  We trust upstream   I didn't see a clear policy in Gentoo -- some upstream releases are held back forever as Gentoo devs argue over it.
I have found it possible to use Arch as the base for a customized, locally compiled system; it's possible to do that with abs/AUR and makepkg.  I've developed my own style.  In place of USE flags I use meld to carry forward my local changes from my local PKGBUILD to the newly released Arch PKGBUILD.  I don't modify many pkgbuilds and mostly just remember which ones I change, for example to exclude gconf from my system. (It's a KDE / Xmonad desktop with a lot of audio tools).
I haven't found it possible or desirable to do bulk builds on Arch. I encounter quite a few pkgbuilds from abs that do not build straight away, sometimes because the package won't build against the library that is in abs at the same time, sometimes because patches are missing or checksums are incorrect:  it becomes clear that the released binary packages are not built from the daily abs update. That would be the job of a build server that takes in abs and spits out a complete set of binaries. That is the missing step that I do myself: take in abs and compile a (daily) coherent set of binaries from it (with local mods). It is not like Gentoo.  I don't think that the Arch team tries to guarantee that the their binaries can be re-built perfectly from abs, so if you try to use abs rather than pacman -S you need to understand that and accept the responsibility. It works well enough for me.

Similar Messages

  • How can I compile all functions, procedures and packages with a script?

    I need to compile all functions, procedures and packages of 5 schemas (users) with a script.
    How can I do it?
    Thanks!

    you can create a script to select all invalid objects in those schemas Since Oracle 8 introduced NDS this approach has struck me as a trifle old fashioned. It's much simpler to loop round the query in PL/SQL and use EXECUTE IMMEDIATE to fire off the DDL statements. No scripts, no muss, no fuss.
    Having said that, the problem with this approach and also with using DBMS_UTILITY.COMPILE_SCHEMA is that they do not compile all the invalid objects in dependency order. This may result in programs being invalidated by the subsequent compilation of dependencies. This is due to the introduction of Java into the database.
    The UTLRP script is much better, because it (usually) avoids cyclic references. But you still may need to run it more than once.
    In general it is better to avoid sledgehammer recompilations (like DBMS_UTILITY.COMPILE_SCHEMA, which starts by invalidating all the objects). If we have twenty invalid objects, nineteen of which are dependencies of the twentieth, we actually only need to recompile the master object, as recompiling it will trigger the recompilation of all the others.
    Cheers, APC

  • Command to compile all installed packages

    hi to all,
    exist a command to compile all packages are installed in your system ?
    thanks

    We know what you want. We also know that instead of searching for it, you want someone to give you "a tip".
    Well, I have a tip for you - don't do it. There is no appreciable benefit gained from recompiling all your packages, so save the time you would have spent on it for something more useful, and we'll leave it at that.
    Alternatively, if you really want to do it - SEARCH THE FORUM.
    Thread closed.

  • How to compile all package body using sqlplus ?

    hi all,
    How to compile all package body using sqlplus ?
    Please help.
    Amy

    dbms_utility.compile_schema will compile all the invalid objects in a schema, including the invalid package bodies. If you only want to compile the invalid package bodies, you could write your own dynamic SQL stored procedure to do that using the user_source view, but that's a lot more effort for a very minimal benefit.
    Justin
    Distributed Database Consulting, Inc.
    http://www.ddbcinc.com/askDDBC

  • Compile all invalid packages together + connect as sysdba only

    Hi
    I've installed a Ora 10.2, and imported a dump. This dump contained some users, a lot of tables and stored procs.
    My first question: There is an imported user with I can connect as sysdba only. I can't connect as normal... it is problem because there is an application upper the db and I can't tell the applicataion that connect as sysdba. (it is only a sandbox for me...) What should I modify on this user to connect as normal user?
    The second question: The schema (upper) contains lot of invalid packages. I have only the free Toad and I can't compile all of them only step by step... (I don't want do this step by step because there are about 300 packages...). How could I complile all of them e.g. from Sqlplus? Any other solution?
    Thanks in advance!
    Alex

    ok, there is a user e.g. "xxxx"
    I can connect with this user as sysdba only. "CONNECT xxxx/yyyy@db as sysdba". Without "as sysdba" I got "ORA-01017: invalid username/password; logon denied" So I can connect only as sysdba... but this is a simple user so I don't want to connect as sysdba...

  • How to compile all objects.....in Oracle10g

    Hi ,
    The Oracle Doc about PL/SQL Packages and Types reffering to DBMS_UTILITY and the COMPILE_SCHEMA procedure....points out :
    "Compiles all procedures, functions, packages, and triggers in the specified schema"
    So how to compile views and possibly other db schema object types ......????
    For example issuing the command :
    select object_name from user_objects where status='INVALID';
    we get all invalid objects, views , packages , e.t.c.
    How to compile all types of objects ... except for EM10g env......
    Many thanks....
    Sim

    You may want to use Solomon Yakobson's utility available here:
    http://apex.oracle.com/pls/otn/f?p=2853:5:4266218307366505

  • Compiling all possible java files

    Hi,
    I have a project in which not all files/packages compiles but a set of packages compiles always.
    Say set of classes which might not compile as 'A' and set of classes which will always compile as 'B'
    On frequent basis i need to get update. And I am only concerned about set 'B' but when i compiles my project using ant
    compilation fails as 'A' compilation fails and because of that it doesn't go further to compile 'B'.
    So how to get eclipse like functionality to compile all possible files?
    what i have is
    ================================
    <javac debug="${debug}" deprecation="${deprication}"
                   optimize="${optimize}" destdir="${build.home}/WEB-INF/classes"
                   listfiles="${listfiles}" target="${jdk.compliance}"
                   verbose="${verbose}" srcdir="${src.home}"
                   memoryMaximumSize="512M" fork="true"
                   source="${src.compliance}">
                   <classpath refid="compile.path"/>
                   <include name="**/*.java"/>
                   <exclude name="HBM/**"/>
                   <compilerarg value="-Xlint:unchecked"/>
    ===================================
    I do not wish to write all the classes specifically that compiles in ant script.
    Thanks in Advance

    Try with failonerror option in your javac.
            <javac debug="${debug}" deprecation="${deprication}" optimize="${optimize}" destdir="${build.home}/WEB-INF/classes" listfiles="${listfiles}" target="${jdk.compliance}" verbose="${verbose}" srcdir="${src.home}" memoryMaximumSize="512M" fork="true" source="${src.compliance}" {color:#0000ff}*failonerror="false"*{color}>
                <classpath refid="compile.path" />
                <include name="*/.java" />
                <exclude name="HBM/**" />
                <compilerarg value="-Xlint:unchecked" />Regards,
    Prashanth.

  • To compile all invalid objects

    how we can compile all invalid objects? is there any sample plsql code?

    dbms_utility.compile_schema('shema_name');Please note that from Oracle 9i onwards this does not work. It compiles objects in alphabetical order rather than reverse dependency order, so you may end up with more objects uncompiled than you started with. This change occurred due to the introduction of Java (boo! hiss!) which allows circular dependencies.
    Since then, the only safe way to compile objects is to use the %ORACLE_HOME%\rdbms\admin\utlirp.sql script (which requires sysdba access).
    This has been packaged up in 10gR2 as UTL_RECOMP.
    Thereby proving that searching is easy but finding the right answer is difficult.
    Cheers, APC
    Blog : http://radiofreetooting.blogspot.com/

  • Making Arch packages out off Gentoo Binary packages

    emerge in portage can build binary packages of what it compiles.
    emerge -b konqueror konq-plugins nsplugins
    and we get splik KDE pkg of konqueror with all it dependencies
    arch pkg vs gentoo pkg
    gentoo pkg doesn't have 2 files: .FILELIST and .PKGINFO + it makes tar.bz2 not tar.gz packages
    how to convert a gentoo package
    - unpack
    - add .FILELIST (list all files >> .FILELIST)
    - create .PKGINFO data from package Ebuild can be used
    - tar.gz it and done
    Someone good in bash/sh scripts could write something so it would convert those packages more automaticly. To make gentoo packages you need gentoo but you don't have to install it like a normal distro:
    1. get i686 stage from here 85MB
    - untar it to /home/user/gentoo
    tar -xvjpf file.tar.bz2
    - cp /etc/resolv.conf to gentoo/etc/
    - modify /etc/make.conf (in gentoo)  add:
    ACCEPT_KEYWORDS="~x86"
    - add USE flags
    and you can set CFLAGS and CXXFLAGS... jackass (unofficial stages) uses someting like:
    CFLAGS="-march=i686"
    CFLAGS="${CFLAGS} -pipe"
    CFLAGS="${CFLAGS} -O2"
    CFLAGS="${CFLAGS} -fweb"
    CFLAGS="${CFLAGS} -frename-registers"
    CFLAGS="${CFLAGS} -fforce-addr"
    CFLAGS="${CFLAGS} -momit-leaf-frame-pointer"
    CFLAGS="${CFLAGS} -fomit-frame-pointer"
    CXXFLAGS="${CFLAGS} -fvisibility-inlines-hidden"
    as root in console chroot to gentoo:
    - go to folder with gentoo
    chroot ./ bin/bash
    env-update
    source /etc/profile
    then if net works:
    emerge sync
    after updating portage tree you can emerge packages

    Separate KDE packages are bit hard to make via pkgbuild because you have to:
    - make spited source packages
    - set dependencies for them
    And it isn't easy Gentoo has splitted KDE sources, but ebuilds aren't to friendly for pkgbuilds But running gentoo binaries on arch will help find all required dependencies for those splitted source packages and based on that – make pkgbuilds
    Currently I'm using Gentoo with splitted KDE packages and they do work faster – less dependencies etc. + they are easy to update (1MB K*something instead whole kdebase*or*something). My vacations starts next week and I'll try to make some gentoo -> arch work in the next months ;]
    (and I switched to Gentoo to get all the soft I wanted "on-demand" – I have only modem at home and now I have some exams so I can't play with arch to get them.)

  • How do I make 'javac' to compile all subdirectories?

    Is there any option of 'javac' command, so javac compiles all .java files in all subdirectories?

    Hi ,
    No recursive way of compiling. You can use packages which will
    be searched to be recomplied. Or if you do not want that, but the
    java files have different names then you can use a file composed of the java files and the @file name.
    ie
    foo1/TestJava1.java
    foo1/foo2/TestJava2.java
    put these two java files into a file say srcfile
    srcfile
    foo1/TestJava1.java
    foo1/foo2/TestJava2.java
    javac @srcfile
    Dave Nicholson Sun Developer Tech Support

  • Compile File vs. Compile All

    Hi All,
    This is probably an easy one. What is the difference between File>Administration>Compile File (ctrl t) and Program>Compile>All (shift+ctrl+k)?
    And does one need to do both before running a form?
    Bradley

    Control+Shift+k recompiles all PL/SQL inside the module, but does not generate any external file.
    Control+t generates a new executable (*.fmx) and in the process will only re-compile PL/SQL in the module that it thinks are dirty.
    As a matter of course I tend to do a Control+Shift+K before doing a Control+t. This is most sensible when you have a lot of changing external dependances in Database packages and libraries.

  • Please give me idea now I compile all branch data in head office only new r

    Hi master
    Sir I have 5 office In different city and one head office in Karachi all branch have same oracle system my question is how I get new data and compile for accumulative reporting
    I want only new record and modify record not old record
    Old record already I import
    When I use oracle import and export tool that no give me right result
    If xxx table exists in database then import tool not replace and no insert or no replace with previous data with new modify data
    Such as
    First time
    V_no=897 have debit amount 3998 is Islamabad branch and i export form Islamabad and import in Karachi office
    Next time
    Islamabad office change v_no=897 debit amount with 76555 and add many new record
    I export form Islamabad and import in Karachi office but system no change and not add new record in Karachi office
    Please give me idea now I compile all branch data in head office only new record and modify record
    Thanks
    aamir

    Here a very simple example with table EMP, assuming source table has a primary key. Firstly you create a materialized view log on source table (necessary for fast refresh) :
    SYS@db102 SQL> conn test/test
    Connected.
    TEST@db102 SQL> create materialized view log on emp including new values;
    Materialized view log created.
    TEST@db102 SQL> then at destination DB/user :
    TEST@db102 SQL> conn scott/tiger@test10
    Connected.
    SCOTT@test10 SQL> create database link test
      2  connect to test identified by test
      3  using 'db102';
    Database link created.
    SCOTT@test10 SQL> create materialized view emp_mv
      2  refresh fast
      3  as select * from emp@test;
    Materialized view created.
    SCOTT@test10 SQL>Now emp_mv is the exact copy of emp. To refresh the MV, to reflect changes :
    SCOTT@test10 SQL> exec dbms_mview.refresh('EMP_MV','F');
    PL/SQL procedure successfully completed.
    SCOTT@test10 SQL>                                                                  ...but I strongly recommend you to read the documentation...

  • My husband recently tried to sync my ipod to itunes and in the process he managed to loose all the photos I had stored on the ipod.  When I go to my pictures folder on my (Dell) laptop there is now a folder called ipod photo cache.  How do I view?

    My husband recently tried to sync my ipod to itunes and in the process he managed to loose all the photos I had stored on the ipod.  When I go to my pictures folder on my (Dell) laptop there is now a folder called ipod photo cache.  How do I view these files? They are now show as ITHMB files!

    Those photos taken or bumped photos should bi in the iPod's Camera Roll album
    - If they are in an iPod backup then restore from that backup. Note however, that the the backup that had the photos may have been overwritten by a subsequent backup that did hot have the photos on the iPod.See the restore topic of:
    iOS: How to back up
    - If you used PhotoStream then try getting them from your PhotoStream. See that topic of:
    iOS: Importing personal photos and videos from iOS devices to your computer
    - Have you looked on the computer? Maybe they were imported to the computer and then deleted from the iPod. Since the photos were not synced to the iPod they are not in the iPod Photo Cache folder .

  • How to compile all forms in a directoy at time?

    Hi friends, thank for your replies for the previous questions.
    Now
    How to compile all form modules located in a directory
    at a time? usually what I know is opening form one by one
    and compile it. But I want to compile all forms and create
    executable files at a time.
    Is project Builder helpfule to achive this??
    thanks
    madhu

    Yes, his file is useful -- it should run on Windows XP. You need to copy the above text into a file using a standard text editor, something like Notepad. To run it, you open a Cmd.exe DOS window on your pc, change directory (CD) to the folder where your forms are, and enter the command script file name.
    Here are several more links to other topics in this forum asking how to do the same thing:
    Re: How To Compile 370 forms ?
    Re: Creating Oracle 10g Forms and Reports Builds

  • I have a new mac and want to sync my phone to the iTunes on it, but it won't let me manually manage without deleting all music on the phone already from another iTunes. is there a way around this?

    I have a new mac and want to sync my phone to the iTunes on it, but it won't let me manually manage without deleting all music on the phone already from another iTunes. is there a way around this?

    The best solution:
    iTunes: How to move [or copy] your music to a new computer [or another drive] - http://support.apple.com/kb/HT4527
    Quick answer if you let iTunes manage your music:  Copy the entire iTunes folder (and in doing so all its subfolders and files) intact to the other drive.  Start iTunes with the option (shift on Windows) key held down and guide it to the new location of the library.
    An alternative which is really not what you should have to do, or the way it is intended to be done, if you properly manage your computer and backups:
    iTunes Store: Transferring purchases from your iOS device or iPod to a computer - http://support.apple.com/kb/HT1848 - only media purchased from iTunes Store
    Sync transfer is one way, computer to device, matching the device content to the content on the computer (except purchased content as mentioned above).  For transferring other items from an i-device to a computer you will probably have to use third party commercial software unless you have an older model iPod. Examples (check the web for others; this is not an exhaustive listing, nor do I have any idea if they are any good):
    - Senuti - http://www.fadingred.com/senuti/
    - Phoneview - http://www.ecamm.com/mac/phoneview/
    - MusicRescue - http://www.kennettnet.co.uk/products/musicrescue/ - Mac & Windows
    - Sharepod (free) - http://download.cnet.com/SharePod/3000-2141_4-10794489.html?tag=mncol;2 - Windows
    - Snowfox/iMedia - http://www.mac-videoconverter.com/imedia-transfer-mac.html - Mac & PC
    - Yamipod (free) - http://www.yamipod.com/main/modules/downloads/ - PC, Linux, Mac [Still updated for use on newer devices? No edits to site since 2010.]
    - Post by Zevoneer: iPod media recovery options - https://discussions.apple.com/message/11624224 - this is an older post and many of the links are also for old posts, so bear this in mind when reading them.
    Syncing to a "New" Computer or replacing a "crashed" Hard Drive - https://discussions.apple.com/docs/DOC-3141 - dates from 2008 and some outdated information now.
    Copying Content from your iPod to your Computer - The Definitive Guide - http://www.ilounge.com/index.php/articles/comments/copying-music-from-ipod-to-co mputer/ - Information about use in disk mode pertains only to older model iPods.
    Additional information here https://discussions.apple.com/message/18324797
    Do not regard your i-device as a reliable unique storage for your media. It is not a backup device and everything is structured around you maintaining your media on a computer which is itself backed up.

Maybe you are looking for