/[ddp]/manuals/trunk/developers-reference/pkgs.dbk
ViewVC logotype

Diff of /manuals/trunk/developers-reference/pkgs.dbk

Parent Directory Parent Directory | Revision Log Revision Log | View Patch Patch

revision 5196 by he, Sun Jun 1 14:28:18 2008 UTC revision 7025 by hertzog, Sun Jan 3 23:28:36 2010 UTC
# Line 28  description of the package, the license Line 28  description of the package, the license
28  current URL where it can be downloaded from.  current URL where it can be downloaded from.
29  </para>  </para>
30  <para>  <para>
31  You should set the subject of the bug to ``ITP: <replaceable>foo</replaceable>  You should set the subject of the bug to <literal>ITP:
32  -- <replaceable>short description</replaceable>'', substituting the name of the  <replaceable>foo</replaceable> -- <replaceable>short
33  new package for <replaceable>foo</replaceable>.  The severity of the bug report  description</replaceable></literal>, substituting the name of the new
34  must be set to <literal>wishlist</literal>.  If you feel it's necessary, send  package for <replaceable>foo</replaceable>.
35  a copy to &email-debian-devel; by putting the address in the  The severity of the bug report must be set to <literal>wishlist</literal>.
36  <literal>X-Debbugs-CC:</literal> header of the message (no, don't use  Please send a copy to &email-debian-devel; by using the X-Debbugs-CC
37  <literal>CC:</literal>, because that way the message's subject won't indicate  header (don't use CC:, because that way the message's subject won't
38  the bug number).  indicate the bug number). If you are packaging so many new packages (>10)
39    that notifying the mailing list in seperate messages is too disruptive,
40    do send a summary after filing the bugs to the debian-devel list instead.
41    This will inform the other developers about upcoming packages and will
42    allow a review of your description and package name.
43  </para>  </para>
44  <para>  <para>
45  Please include a <literal>Closes:  Please include a <literal>Closes:
# Line 44  new package in order for the bug report Line 48  new package in order for the bug report
48  package is installed in the archive (see <xref linkend="upload-bugfix"/> ).  package is installed in the archive (see <xref linkend="upload-bugfix"/> ).
49  </para>  </para>
50  <para>  <para>
51    If you think your package needs some explanations for the administrators of the
52    NEW package queue, include them in your changelog, send to ftpmaster@debian.org
53    a reply to the email you receive as a maintainer after your upload, or reply to
54    the rejection email in case you are already re-uploading.
55    </para>
56    <para>
57  When closing security bugs include CVE numbers as well as the Closes: #nnnnn.  When closing security bugs include CVE numbers as well as the Closes: #nnnnn.
58  This is useful for the security team to track vulnerabilities.  If an upload is  This is useful for the security team to track vulnerabilities.  If an upload is
59  made to fix the bug before the advisory ID is known, it is encouraged to modify  made to fix the bug before the advisory ID is known, it is encouraged to modify
# Line 72  already is a volunteer, so efforts may b Line 82  already is a volunteer, so efforts may b
82  <para>  <para>
83  It lets the rest of the maintainers know more about the package than the one  It lets the rest of the maintainers know more about the package than the one
84  line description and the usual changelog entry ``Initial release'' that gets  line description and the usual changelog entry ``Initial release'' that gets
85  posted to <literal>debian-devel-changes</literal>.  posted to &email-debian-devel-changes;.
86  </para>  </para>
87  </listitem>  </listitem>
88  <listitem>  <listitem>
89  <para>  <para>
90  It is helpful to the people who live off unstable (and form our first line of  It is helpful to the people who live off <literal>unstable</literal> (and form
91  testers).  We should encourage these people.  our first line of testers).  We should encourage these people.
92  </para>  </para>
93  </listitem>  </listitem>
94  <listitem>  <listitem>
# Line 269  The package build process extracts this Line 279  The package build process extracts this
279  <literal>Distribution</literal> field of the <literal>.changes</literal> file.  <literal>Distribution</literal> field of the <literal>.changes</literal> file.
280  </para>  </para>
281  <para>  <para>
282  There are several possible values for this field: `stable', `unstable',  There are several possible values for this field: <literal>stable</literal>,
283  `testing-proposed-updates' and `experimental'.  Normally, packages are uploaded  <literal>unstable</literal>, <literal>testing-proposed-updates</literal> and
284  into <literal>unstable</literal>.  <literal>experimental</literal>.  Normally, packages are uploaded into
285    <literal>unstable</literal>.
286  </para>  </para>
287  <para>  <para>
288  Actually, there are two other possible distributions: `stable-security' and  Actually, there are two other possible distributions: <literal>stable-security
289  `testing-security', but read <xref linkend="bug-security"/> for more  </literal> and <literal>testing-security</literal>, but read
290  information on those.  <xref linkend="bug-security"/> for more information on those.
291  </para>  </para>
292  <para>  <para>
293  It is not possible to upload a package into several distributions at the same  It is not possible to upload a package into several distributions at the same
294  time.  time.
295  </para>  </para>
296  <section id="upload-stable">  <section id="upload-stable">
297  <title>Special case: uploads to the <literal>stable</literal> distribution</title>  <title>Special case: uploads to the <literal>stable</literal> and
298    <literal>oldstable</literal> distributions</title>
299  <para>  <para>
300  Uploading to <literal>stable</literal> means that the package will transfered  Uploading to <literal>stable</literal> means that the package will transfered
301  to the <literal>proposed-updates-new</literal>-queue for review by the stable  to the <literal>proposed-updates-new</literal> queue for review by the stable
302  release managers, and if approved will be installed in  release managers, and if approved will be installed in
303  <filename>stable-proposed-updates</filename> directory of the Debian archive.  <filename>stable-proposed-updates</filename> directory of the Debian archive.
304  From there, it will be included in <literal>stable</literal> with the next  From there, it will be included in <literal>stable</literal> with the next
305  point release.  point release.
306  </para>  </para>
307  <para>  <para>
308    To ensure that your upload will be accepted, you should discuss the changes
309    with the stable release team before you upload. For that, send a mail to
310    the &email-debian-release; mailing list, including the patch you want to
311    apply to the package version currently in <literal>stable</literal>. Always
312    be verbose and detailed in your changelog entries for uploads to the
313    <literal>stable</literal> distribution.
314    </para>
315    <para>
316  Extra care should be taken when uploading to <literal>stable</literal>.  Extra care should be taken when uploading to <literal>stable</literal>.
317  Basically, a package should only be uploaded to stable if one of the following  Basically, a package should only be uploaded to <literal>stable</literal> if
318  happens:  one of the following happens:
319  </para>  </para>
320  <itemizedlist>  <itemizedlist>
321  <listitem>  <listitem>
# Line 320  security problems as well.  However, thi Line 340  security problems as well.  However, thi
340  used for Debian security advisories are automatically copied to the appropriate  used for Debian security advisories are automatically copied to the appropriate
341  <filename>proposed-updates</filename> archive when the advisory is released.  <filename>proposed-updates</filename> archive when the advisory is released.
342  See <xref linkend="bug-security"/> for detailed information on handling  See <xref linkend="bug-security"/> for detailed information on handling
343  security problems.  security problems. If the security teams deems the problem to be too
344    benign to be fixed through a <literal>DSA</literal>, the stable release
345    managers are usually willing to include your fix nonetheless in a regular
346    upload to <literal>stable</literal>.
347  </para>  </para>
348  <para>  <para>
349  Changing anything else in the package that isn't important is discouraged,  Changing anything else in the package that isn't important is discouraged,
# Line 331  Packages uploaded to <literal>stable</li Line 354  Packages uploaded to <literal>stable</li
354  running <literal>stable</literal>, so that their dependencies are limited to  running <literal>stable</literal>, so that their dependencies are limited to
355  the libraries (and other packages) available in <literal>stable</literal>;  the libraries (and other packages) available in <literal>stable</literal>;
356  for example, a package uploaded to <literal>stable</literal> that depends on  for example, a package uploaded to <literal>stable</literal> that depends on
357  a library package that only exists in unstable will be rejected.  Making  a library package that only exists in <literal>unstable</literal> will be
358  changes to dependencies of other packages (by messing with  rejected.  Making changes to dependencies of other packages (by messing with
359  <literal>Provides</literal> or shlibs files), possibly making those other  <literal>Provides</literal> or <literal>shlibs</literal> files), possibly
360  packages uninstallable, is strongly discouraged.  making those other packages uninstallable, is strongly discouraged.
361  </para>  </para>
362  <para>  <para>
363  The Release Team (which can be reached at  Uploads to the <literal>oldstable</literal> distributions are possible as
364  &email-debian-release;) will regularly evaluate the uploads to  long as it hasn't been archived. The same rules as for <literal>stable
365  <literal>stable-proposed-updates</literal> and decide if your package can be  </literal> apply.
 included in <literal>stable</literal>.  Please be clear (and verbose, if  
 necessary) in your changelog entries for uploads to  
 <literal>stable</literal>, because otherwise the package won't be considered  
 for inclusion.  
 </para>  
 <para>  
 It's best practice to speak with the stable release manager  
 <emphasis>before</emphasis> uploading to  
 <literal>stable</literal>/<literal>stable-proposed-updates</literal>, so  
 that the uploaded package fits the needs of the next point release.  
366  </para>  </para>
367  </section>  </section>
368    
# Line 394  the Debian package <xref linkend="dcut"/ Line 407  the Debian package <xref linkend="dcut"/
407    
408  <section id="delayed-incoming">  <section id="delayed-incoming">
409  <title>Delayed uploads</title>  <title>Delayed uploads</title>
410    
411  <para>  <para>
412  Delayed uploads are done for the moment via the delayed queue at <literal>gluck  It is sometimes useful to upload a package immediately, but to want this
413  </literal>. The upload-directory is  package to arrive in the archive only a few days later. For example,
414  <literal>gluck:~tfheen/DELAYED/[012345678]-day</literal>. 0-day is uploaded  when preparing a <link linkend="nmu">Non-maintainer Upload</link>,
415  multiple times per day to <literal>&ftp-master-host;</literal>.  you might want to give the maintainer a few days to react.
 </para>  
 <para>  
 With a fairly recent dput, this section  
416  </para>  </para>
417  <screen>  
 [tfheen_delayed]  
 method = scp  
 fqdn = gluck.debian.org  
 incoming = ~tfheen  
 </screen>  
418  <para>  <para>
419  in <filename>~/.dput.cf</filename> should work fine for uploading to the  An upload to the delayed directory keeps the package in
420  <literal>DELAYED</literal> queue.  <ulink url="http://ftp-master.debian.org/deferred.html">
421    the deferred uploads queue"</ulink>.
422    When the specified waiting time is over, the package is moved into
423    the regular incoming directory for processing.
424    This is done through automatic uploading to
425    <literal>&ftp-master-host;</literal> in upload-directory
426    <literal>DELAYED/[012345678]-day</literal>. 0-day is uploaded
427    multiple times per day to <literal>&ftp-master-host;</literal>.
428  </para>  </para>
429  <para>  <para>
430  <emphasis>Note:</emphasis> Since this upload queue goes to  With dput, you can use the <literal>--delayed <replaceable>DELAY</replaceable></literal>
431  <literal>&ftp-master-host;</literal>, the prescription found in <xref  parameter to put the package into one of the queues.
 linkend="upload-ftp-master"/> applies here as well.  
432  </para>  </para>
433  </section>  </section>
434    
# Line 424  linkend="upload-ftp-master"/> applies he Line 436  linkend="upload-ftp-master"/> applies he
436  <title>Security uploads</title>  <title>Security uploads</title>
437  <para>  <para>
438  Do <emphasis role="strong">NOT</emphasis> upload a package to the security  Do <emphasis role="strong">NOT</emphasis> upload a package to the security
439  upload queue (oldstable-security, stable-security, etc.) without prior  upload queue (<literal>oldstable-security</literal>, <literal>stable-security
440  authorization from the security team.  If the package does not exactly meet the  </literal>, etc.) without prior authorization from the security team.  If the
441  team's requirements, it will cause many problems and delays in dealing with the  package does not exactly meet the team's requirements, it will cause many
442  unwanted upload.  For details, please see section <xref  problems and delays in dealing with the unwanted upload.  For details, please
443  linkend="bug-security"/> .  see section <xref linkend="bug-security"/> .
444  </para>  </para>
445  </section>  </section>
446    
447  <section id="s5.6.5">  <section id="s5.6.5">
448  <title>Other upload queues</title>  <title>Other upload queues</title>
449  <para>  <para>
450  The scp queues on <literal>&ftp-master-host;</literal>, and security are mostly  The scp queues on <literal>&ftp-master-host;</literal>, and <literal>
451  unusable due to the login restrictions on those hosts.  security.debian.org</literal> are mostly unusable due to the login restrictions
452    on those hosts.
453  </para>  </para>
454  <para>  <para>
455  The anonymous queues on ftp.uni-erlangen.de and ftp.uk.debian.org are currently  The anonymous queues on ftp.uni-erlangen.de and ftp.uk.debian.org are currently
# Line 445  down.  Work is underway to resurrect the Line 458  down.  Work is underway to resurrect the
458  <para>  <para>
459  The queues on master.debian.org, samosa.debian.org, master.debian.or.jp, and  The queues on master.debian.org, samosa.debian.org, master.debian.or.jp, and
460  ftp.chiark.greenend.org.uk are down permanently, and will not be resurrected.  ftp.chiark.greenend.org.uk are down permanently, and will not be resurrected.
 The queue in Japan will be replaced with a new queue on hp.debian.or.jp some  
 day.  
 </para>  
 <para>  
 For the time being, the anonymous ftp queue on auric.debian.org (the former  
 ftp-master) works, but it is deprecated and will be removed at some point in  
 the future.  
461  </para>  </para>
462  </section>  </section>
463    
# Line 461  the future. Line 467  the future.
467  The Debian archive maintainers are responsible for handling package uploads.  The Debian archive maintainers are responsible for handling package uploads.
468  For the most part, uploads are automatically handled on a daily basis by the  For the most part, uploads are automatically handled on a daily basis by the
469  archive maintenance tools, <command>katie</command>.  Specifically, updates to  archive maintenance tools, <command>katie</command>.  Specifically, updates to
470  existing packages to the `unstable' distribution are handled automatically.  In  existing packages to the <literal>unstable</literal> distribution are handled
471  other cases, notably new packages, placing the uploaded package into the  automatically.  In other cases, notably new packages, placing the uploaded
472  distribution is handled manually.  When uploads are handled manually, the  package into the distribution is handled manually.  When uploads are handled
473  change to the archive may take up to a month to occur.  Please be patient.  manually, the change to the archive may take up to a month to occur.  Please
474    be patient.
475  </para>  </para>
476  <para>  <para>
477  In any case, you will receive an email notification indicating that the package  In any case, you will receive an email notification indicating that the package
# Line 508  file</literal>. Line 515  file</literal>.
515  <para>  <para>
516  To alter the actual section that a package is put in, you need to first make  To alter the actual section that a package is put in, you need to first make
517  sure that the <filename>debian/control</filename> file in your package is  sure that the <filename>debian/control</filename> file in your package is
518  accurate.  Next, send an email &email-override; or submit a  accurate.  Next, submit a
519  bug against <systemitem role="package">ftp.debian.org</systemitem> requesting  bug against <systemitem role="package">ftp.debian.org</systemitem> requesting
520  that the section or priority for your package be changed from the old section  that the section or priority for your package be changed from the old section
521  or priority to the new one.  Be sure to explain your reasoning.  or priority to the new one. Use a Subject like
522    <literal>override: PACKAGE1:section/priority, [...],
523      PACKAGEX:section/priority</literal>, and include the justification for the
524    change in the body of the bug report.
525  </para>  </para>
526  <para>  <para>
527  For more information about <literal>override files</literal>, see  For more information about <literal>override files</literal>, see
# Line 662  procedure</ulink>. Line 672  procedure</ulink>.
672  If the bug is real but it's caused by another package, just reassign the bug to  If the bug is real but it's caused by another package, just reassign the bug to
673  the right package.  If you don't know which package it should be reassigned to,  the right package.  If you don't know which package it should be reassigned to,
674  you should ask for help on <link linkend="irc-channels">IRC</link> or  you should ask for help on <link linkend="irc-channels">IRC</link> or
675  on &email-debian-devel;.  Please make sure that the  on &email-debian-devel;.  Please inform the maintainer(s) of the package
676  maintainer(s) of the package the bug is reassigned to know why you reassigned  you reassign the bug to, for example by Cc:ing the message that does the
677  it.  reassign to <email>packagename@packages.debian.org</email> and explaining
678    your reasons in that mail. Please note that a simple reassignment is
679    <emphasis>not</emphasis> e-mailed to the maintainers of the package
680    being reassigned to, so they won't know about it until they look at
681    a bug overview for their packages.
682  </para>  </para>
683  <para>  <para>
684    If the bug affects the operation of your package, please consider
685    cloning the bug and reassigning the clone to the package that really
686    causes the behavior. Otherwise, the bug will not be shown in your
687    package's bug list, possibly causing users to report the same bug over
688    and over again. You should block "your" bug with the reassigned, cloned
689    bug to document the relationship.
690    </para>
691    </listitem>
692    <listitem>
693    <para>
694  Sometimes you also have to adjust the severity of the bug so that it matches  Sometimes you also have to adjust the severity of the bug so that it matches
695  our definition of the severity.  That's because people tend to inflate the  our definition of the severity.  That's because people tend to inflate the
696  severity of bugs to make sure their bugs are fixed quickly.  Some bugs may even  severity of bugs to make sure their bugs are fixed quickly.  Some bugs may even
# Line 721  several developers working on the same p Line 745  several developers working on the same p
745  </listitem>  </listitem>
746  <listitem>  <listitem>
747  <para>  <para>
748  Once a corrected package is available in the <literal>unstable</literal>  Once a corrected package is available in the archive, the bug should be
749  distribution, you can close the bug.  This can be done automatically, read  closed indicating the version in which it was fixed. This can be done
750  <xref linkend="upload-bugfix"/> .  automatically, read <xref linkend="upload-bugfix"/>.
751  </para>  </para>
752  </listitem>  </listitem>
753  </orderedlist>  </orderedlist>
# Line 808  Due to their sensitive nature, security- Line 832  Due to their sensitive nature, security-
832  The Debian Security Team exists to coordinate this activity, keeping track of  The Debian Security Team exists to coordinate this activity, keeping track of
833  outstanding security problems, helping maintainers with security problems or  outstanding security problems, helping maintainers with security problems or
834  fixing them themselves, sending security advisories, and maintaining  fixing them themselves, sending security advisories, and maintaining
835  security.debian.org.  <literal>security.debian.org</literal>.
836  </para>  </para>
 <!-- information about the security database goes here once it's ready -->  
 <!-- (mdz) -->  
837  <para>  <para>
838  When you become aware of a security-related bug in a Debian package, whether or  When you become aware of a security-related bug in a Debian package, whether or
839  not you are the maintainer, collect pertinent information about the problem,  not you are the maintainer, collect pertinent information about the problem,
840  and promptly contact the security team at  and promptly contact the security team at
841  &email-security-team; as soon as possible.  <emphasis  &email-security-team; as soon as possible.  <emphasis
842  role="strong">DO NOT UPLOAD</emphasis> any packages for stable; the security  role="strong">DO NOT UPLOAD</emphasis> any packages for <literal>stable</literal>
843  team will do that.  Useful information includes, for example:  without contacting the team.  Useful information includes, for example:
844  </para>  </para>
845  <itemizedlist>  <itemizedlist>
846  <listitem>  <listitem>
847  <para>  <para>
848  Which versions of the package are known to be affected by the bug.  Check each  Which versions of the package are known to be affected by the bug.  Check each
849  version that is present in a supported Debian release, as well as testing and  version that is present in a supported Debian release, as well as
850  unstable.  <literal>testing</literal> and <literal>unstable</literal>.
851  </para>  </para>
852  </listitem>  </listitem>
853  <listitem>  <listitem>
# Line 853  linkend="bug-security-advisories"/> ) Line 875  linkend="bug-security-advisories"/> )
875  </para>  </para>
876  </listitem>  </listitem>
877  </itemizedlist>  </itemizedlist>
878    <para>As the maintainer of the package, you have the responsibility to
879    maintain it, even in the stable release. You are in the best position
880    to evaluate patches and test updated packages, so please see the sections
881    below on how to prepare packages for the Security Team to handle.</para>
882    
883    <section id="bug-security-tracker">
884    <title>The Security Tracker</title>
885    <para>
886    The security team maintains a central database, the
887    <ulink url="http://security-tracker.debian.net/">Debian Security Tracker</ulink>.
888    This contains all public information that is known about security issues:
889    which packages and versions are affected or fixed, and thus whether stable,
890    testing and/or unstable are vulnerable. Information that is still confidential
891    is not added to the tracker.
892    </para>
893    <para>
894    You can search it for a specific issue, but also on package name. Look
895    for your package to see which issues are still open. If you can, please provide
896    more information about those issues, or help to address them in your package.
897    Instructions are on the tracker web pages.
898    </para>
899    </section>
900    
901  <section id="bug-security-confidentiality">  <section id="bug-security-confidentiality">
902  <title>Confidentiality</title>  <title>Confidentiality</title>
903  <para>  <para>
# Line 911  release of Debian.  When sending confide Line 956  release of Debian.  When sending confide
956  be sure to mention this fact.  be sure to mention this fact.
957  </para>  </para>
958  <para>  <para>
959  Please note that if secrecy is needed you may not upload a fix to unstable (or  Please note that if secrecy is needed you may not upload a fix to
960    <literal>unstable</literal> (or
961  anywhere else, such as a public CVS repository).  It is not sufficient to  anywhere else, such as a public CVS repository).  It is not sufficient to
962  obfuscate the details of the change, as the code itself is public, and can (and  obfuscate the details of the change, as the code itself is public, and can (and
963  will) be examined by the general public.  will) be examined by the general public.
# Line 921  There are two reasons for releasing info Line 967  There are two reasons for releasing info
967  requested: the problem has been known for a while, or the problem or exploit  requested: the problem has been known for a while, or the problem or exploit
968  has become public.  has become public.
969  </para>  </para>
970    <para>
971    The Security Team has a PGP-key to enable encrypted communication about
972    sensitive issues. See the <ulink url="http://www.debian.org/security/faq.en.html#contact">Security Team FAQ</ulink> for details.
973    </para>
974  </section>  </section>
975    
976  <section id="bug-security-advisories">  <section id="bug-security-advisories">
977  <title>Security Advisories</title>  <title>Security Advisories</title>
978  <para>  <para>
979  Security advisories are only issued for the current, released stable  Security advisories are only issued for the current, released stable
980  distribution, and <emphasis>not</emphasis> for testing or unstable.  When  distribution, and <emphasis>not</emphasis> for <literal>testing</literal>
981  released, advisories are sent to the  or <literal>unstable</literal>.  When released, advisories are sent to the
982  &email-debian-security-announce; mailing list and posted on  &email-debian-security-announce; mailing list and posted on
983  <ulink url="&url-debian-security-advisories;">the security web  <ulink url="&url-debian-security-advisories;">the security web
984  page</ulink>.  Security advisories are written and posted by the security team.  page</ulink>.  Security advisories are written and posted by the security team.
# Line 1057  Be sure to verify the following items: Line 1107  Be sure to verify the following items:
1107  <itemizedlist>  <itemizedlist>
1108  <listitem>  <listitem>
1109  <para>  <para>
1110  Target the right distribution in your <filename>debian/changelog</filename>.  <emphasis role="strong">Target the right distribution</emphasis>
1111  For stable this is <literal>stable-security</literal> and for testing this is  in your <filename>debian/changelog</filename>.
1112  <literal>testing-security</literal>, and for the previous stable release, this  For <literal>stable</literal> this is <literal>stable-security</literal> and
1113  is <literal>oldstable-security</literal>.  Do not target  for testing this is <literal>testing-security</literal>, and for the previous
1114    stable release, this is <literal>oldstable-security</literal>.  Do not target
1115  <replaceable>distribution</replaceable><literal>-proposed-updates</literal> or  <replaceable>distribution</replaceable><literal>-proposed-updates</literal> or
1116  <literal>stable</literal>!  <literal>stable</literal>!
1117  </para>  </para>
1118  </listitem>  </listitem>
1119  <listitem>  <listitem>
1120  <para>  <para>
1121  The upload should have urgency=high.  The upload should have <emphasis role="strong">urgency=high</emphasis>.
1122  </para>  </para>
1123  </listitem>  </listitem>
1124  <listitem>  <listitem>
1125  <para>  <para>
1126  Make descriptive, meaningful changelog entries.  Others will rely on them to  Make descriptive, meaningful changelog entries.  Others will rely on them to
1127  determine whether a particular bug was fixed.  Always include an external  determine whether a particular bug was fixed.  Add <literal>closes:</literal>
1128  reference, preferably a CVE identifier, so that it can be cross-referenced.  statements for any <emphasis role="strong">Debian bugs</emphasis> filed.
1129  Include the same information in the changelog for unstable, so that it is clear  Always include an external reference, preferably a <emphasis role="strong">CVE
1130  that the same bug was fixed, as this is very helpful when verifying that the  identifier</emphasis>, so that it can be cross-referenced. However, if a CVE
1131  bug is fixed in the next stable release.  If a CVE identifier has not yet been  identifier has not yet been assigned, do not wait for it but continue the
1132  assigned, the security team will request one so that it can be included in the  process. The identifier can be cross-referenced later.
 package and in the advisory.  
 </para>  
 </listitem>  
 <listitem>  
 <para>  
 Make sure the version number is proper.  It must be greater than the current  
 package, but less than package versions in later distributions.  If in doubt,  
 test it with <literal>dpkg --compare-versions</literal>.  Be careful not to  
 re-use a version number that you have already used for a previous upload.  For  
 <literal>testing</literal>, there must be a higher version in  
 <literal>unstable</literal>.  If there is none yet (for example, if  
 <literal>testing</literal> and <literal>unstable</literal> have the same  
 version) you must upload a new version to unstable first.  
1133  </para>  </para>
1134  </listitem>  </listitem>
1135  <listitem>  <listitem>
1136  <para>  <para>
1137  Do not make source-only uploads if your package has any binary-all packages (do  Make sure the <emphasis role="strong">version number</emphasis> is proper.
1138  not use the <literal>-S</literal> option to  It must be greater than the current package, but less than package versions in
1139  <command>dpkg-buildpackage</command>).  The <command>buildd</command>  later distributions.  If in doubt, test it with <literal>dpkg
1140  infrastructure will not build those.  This point applies to normal package  --compare-versions</literal>.  Be careful not to re-use a version number that
1141  uploads as well.  you have already used for a previous upload, or one that conflicts with a
1142    binNMU. The convention is to append
1143    <literal>+</literal><replaceable>codename</replaceable><literal>1</literal>, e.g.
1144    <literal>1:2.4.3-4+etch1</literal>, of course increasing 1 for any subsequent
1145    uploads.
1146  </para>  </para>
1147  </listitem>  </listitem>
1148  <listitem>  <listitem>
1149  <para>  <para>
1150  Unless the upstream source has been uploaded to security.debian.org before (by  Unless the upstream source has been uploaded to <literal>security.debian.org
1151  a previous security update), build the upload with full upstream source  </literal> before (by a previous security update), build the upload <emphasis
1152  (<literal>dpkg-buildpackage -sa</literal>).  If there has been a previous  role="strong">with full upstream source</emphasis> (<literal>dpkg-buildpackage
1153  upload to security.debian.org with the same upstream version, you may upload  -sa</literal>).  If there has been a previous upload to
1154  without upstream source (<literal>dpkg-buildpackage -sd</literal>).  <literal>security.debian.org</literal> with the same upstream version, you may
1155    upload without upstream source (<literal> dpkg-buildpackage -sd</literal>).
1156  </para>  </para>
1157  </listitem>  </listitem>
1158  <listitem>  <listitem>
1159  <para>  <para>
1160  Be sure to use the exact same <filename>*.orig.tar.gz</filename> as used in the  Be sure to use the <emphasis role="strong">exact same
1161    <filename>*.orig.tar.gz</filename></emphasis> as used in the
1162  normal archive, otherwise it is not possible to move the security fix into the  normal archive, otherwise it is not possible to move the security fix into the
1163  main archives later.  main archives later.
1164  </para>  </para>
1165  </listitem>  </listitem>
1166  <listitem>  <listitem>
1167  <para>  <para>
1168  Build the package on a clean system which only has packages installed from the  Build the package on a <emphasis role="strong">clean system</emphasis> which only
1169  distribution you are building for.  If you do not have such a system yourself,  has packages installed from the distribution you are building for. If you do not
1170  you can use a debian.org machine (see <xref linkend="server-machines"/> ) or  have such a system yourself, you can use a debian.org machine (see
1171  setup a chroot (see <xref linkend="pbuilder"/> and <xref  <xref linkend="server-machines"/> ) or setup a chroot (see
1172  linkend="debootstrap"/> ).  <xref linkend="pbuilder"/> and <xref linkend="debootstrap"/> ).
1173  </para>  </para>
1174  </listitem>  </listitem>
1175  </itemizedlist>  </itemizedlist>
# Line 1135  linkend="debootstrap"/> ). Line 1179  linkend="debootstrap"/> ).
1179  <title>Uploading the fixed package</title>  <title>Uploading the fixed package</title>
1180  <para>  <para>
1181  Do <emphasis role="strong">NOT</emphasis> upload a package to the security  Do <emphasis role="strong">NOT</emphasis> upload a package to the security
1182  upload queue (oldstable-security, stable-security, etc.) without prior  upload queue (<literal>oldstable-security</literal>, <literal>stable-security
1183  authorization from the security team.  If the package does not exactly meet the  </literal>, etc.) without prior authorization from the security team.  If the
1184  team's requirements, it will cause many problems and delays in dealing with the  package does not exactly meet the team's requirements, it will cause many
1185  unwanted upload.  problems and delays in dealing with the unwanted upload.
1186  </para>  </para>
1187  <para>  <para>
1188  Do <emphasis role="strong">NOT</emphasis> upload your fix to proposed-updates  Do <emphasis role="strong">NOT</emphasis> upload your fix to <literal>
1189  without coordinating with the security team.  Packages from security.debian.org  proposed-updates</literal> without coordinating with the security team.
1190  will be copied into the proposed-updates directory automatically.  If a package  Packages from <literal>security.debian.org</literal> will be copied into
1191    the <literal>proposed-updates</literal> directory automatically.  If a package
1192  with the same or a higher version number is already installed into the archive,  with the same or a higher version number is already installed into the archive,
1193  the security update will be rejected by the archive system.  That way, the  the security update will be rejected by the archive system.  That way, the
1194  stable distribution will end up without a security update for this package  stable distribution will end up without a security update for this package
# Line 1157  archives.  For security uploads, the pla Line 1202  archives.  For security uploads, the pla
1202  </para>  </para>
1203  <para>  <para>
1204  Once an upload to the security queue has been accepted, the package will  Once an upload to the security queue has been accepted, the package will
1205  automatically be rebuilt for all architectures and stored for verification by  automatically be built for all architectures and stored for verification by
1206  the security team.  the security team.
1207  </para>  </para>
1208  <para>  <para>
# Line 1167  problems that cannot be disclosed yet. Line 1212  problems that cannot be disclosed yet.
1212  </para>  </para>
1213  <para>  <para>
1214  If a member of the security team accepts a package, it will be installed on  If a member of the security team accepts a package, it will be installed on
1215  security.debian.org as well as proposed for the proper  <literal>security.debian.org</literal> as well as proposed for the proper
1216  <replaceable>distribution</replaceable><literal>-proposed-updates</literal>  <replaceable>distribution</replaceable><literal>-proposed-updates</literal>
1217  on <literal>&ftp-master-host;</literal>.  on <literal>&ftp-master-host;</literal>.
1218  </para>  </para>
# Line 1220  described in <xref linkend="override-fil Line 1265  described in <xref linkend="override-fil
1265  If for some reason you want to completely remove a package (say, if it is an  If for some reason you want to completely remove a package (say, if it is an
1266  old compatibility library which is no longer required), you need to file a bug  old compatibility library which is no longer required), you need to file a bug
1267  against <literal>ftp.debian.org</literal> asking that the package be removed;  against <literal>ftp.debian.org</literal> asking that the package be removed;
1268  as all bugs, this bug should normally have normal severity.  Make sure you  as all bugs, this bug should normally have normal severity.
1269  indicate which distribution the package should be removed from.  Normally, you  The bug title should be in the form <literal>RM: <replaceable>package
1270  can only have packages removed from <literal>unstable</literal> and  </replaceable> <replaceable>[architecture list]</replaceable> --
1271  <literal>experimental</literal>.  Packages are not removed from  <replaceable>reason</replaceable></literal>, where <replaceable>package</replaceable>
1272    is the package to be removed and <replaceable>reason</replaceable> is a
1273    short summary of the reason for the removal request.
1274    <replaceable>[architecture list]</replaceable> is optional and only needed
1275    if the removal request only applies to some architectures, not all. Note
1276    that the <command>reportbug</command> will create a title conforming
1277    to these rules when you use it to report a bug against the <literal>
1278    ftp.debian.org</literal> pseudo-package.
1279    </para>
1280    
1281    <para>
1282    If you want to remove a package you maintain, you should note this in
1283    the bug title by prepending <literal>ROM</literal> (Request Of Maintainer).
1284    There are several other standard acronyms used in the reasoning for a package
1285    removal, see <ulink url="http://&ftp-master-host;/removals.html"></ulink>
1286    for a complete list. That page also provides a convenient overview of
1287    pending removal requests.
1288    </para>
1289    
1290    <para>
1291    Note that removals can only be done for the <literal>unstable
1292    </literal>, <literal>experimental</literal> and <literal>stable
1293    </literal> distribution.  Packages are not removed from
1294  <literal>testing</literal> directly.  Rather, they will be removed  <literal>testing</literal> directly.  Rather, they will be removed
1295  automatically after the package has been removed from  automatically after the package has been removed from
1296  <literal>unstable</literal> and no package in <literal>testing</literal>  <literal>unstable</literal> and no package in <literal>testing
1297  depends on it.  </literal> depends on it.
1298  </para>  </para>
1299  <para>  <para>
1300  There is one exception when an explicit removal request is not necessary: If a  There is one exception when an explicit removal request is not necessary: If a
# Line 1247  supersedes the one to be removed. Line 1314  supersedes the one to be removed.
1314  <para>  <para>
1315  Usually you only ask for the removal of a package maintained by yourself.  If  Usually you only ask for the removal of a package maintained by yourself.  If
1316  you want to remove another package, you have to get the approval of its  you want to remove another package, you have to get the approval of its
1317  maintainer.  maintainer. Should the package be orphaned and thus have no maintainer,
1318    you should first discuss the removal request on &email-debian-qa;. If
1319    there is a consensus that the package should be removed, you should
1320    reassign and retitle the <literal>O:</literal> bug filed against the
1321    <literal>wnpp</literal> package instead of filing a new bug as
1322    removal request.
1323  </para>  </para>
1324  <para>  <para>
1325  Further information relating to these and other package removal related topics  Further information relating to these and other package removal related topics
# Line 1262  role="package">apt</systemitem> package. Line 1334  role="package">apt</systemitem> package.
1334  showpkg <replaceable>package</replaceable></literal>, the program will show  showpkg <replaceable>package</replaceable></literal>, the program will show
1335  details for <replaceable>package</replaceable>, including reverse depends.  details for <replaceable>package</replaceable>, including reverse depends.
1336  Other useful programs include <literal>apt-cache rdepends</literal>,  Other useful programs include <literal>apt-cache rdepends</literal>,
1337  <command>apt-rdepends</command> and <command>grep-dctrl</command>.  Removal of  <command>apt-rdepends</command>, <command>build-rdeps</command> (in the
1338    <systemitem role="package">devscripts</systemitem> package) and
1339    <command>grep-dctrl</command>.  Removal of
1340  orphaned packages is discussed on &email-debian-qa;.  orphaned packages is discussed on &email-debian-qa;.
1341  </para>  </para>
1342  <para>  <para>
# Line 1271  should either be reassigned to another p Line 1345  should either be reassigned to another p
1345  code has evolved into another package (e.g.  <literal>libfoo12</literal> was  code has evolved into another package (e.g.  <literal>libfoo12</literal> was
1346  removed because <literal>libfoo13</literal> supersedes it) or closed if the  removed because <literal>libfoo13</literal> supersedes it) or closed if the
1347  software is simply no longer part of Debian.  software is simply no longer part of Debian.
1348    When closing the bugs,
1349    to avoid marking the bugs as fixed in versions of the packages
1350    in previous Debian releases, they should be marked as fixed
1351    in the version <literal>&lt;most-recent-version-ever-in-Debian&gt;+rm</literal>.
1352  </para>  </para>
1353  <section id="s5.9.2.1">  <section id="s5.9.2.1">
1354  <title>Removing packages from <filename>Incoming</filename></title>  <title>Removing packages from <filename>Incoming</filename></title>
# Line 1292  occur too often anyway. Line 1370  occur too often anyway.
1370  <section id="s5.9.3">  <section id="s5.9.3">
1371  <title>Replacing or renaming packages</title>  <title>Replacing or renaming packages</title>
1372  <para>  <para>
1373  When you make a mistake naming your package, you should follow a two-step  When the upstream maintainers for one of your packages chose to
1374  process to rename it.  First, set your <filename>debian/control</filename> file  rename their software (or you made a mistake naming your package),
1375  to replace and conflict with the obsolete name of the package (see the <ulink  you should follow a two-step process to rename it. In the first
1376  url="&url-debian-policy;">Debian Policy Manual</ulink> for  step, change the <filename>debian/control</filename> file to
1377  details).  Once you've uploaded the package and the package has moved into the  reflect the new name and to replace, provide and conflict with the
1378  archive, file a bug against <literal>ftp.debian.org</literal> asking to remove  obsolete package name (see the <ulink url="&url-debian-policy;">
1379  the package with the obsolete name.  Do not forget to properly reassign the  Debian Policy Manual</ulink> for details).  Please note that you
1380  package's bugs at the same time.  should only add a <literal>Provides</literal> relation if all
1381    packages depending on the obsolete package name continue to work
1382    after the renaming. Once you've uploaded the package and the package
1383    has moved into the archive, file a bug against <literal>
1384    ftp.debian.org</literal> asking to remove the package with the
1385    obsolete name (see <xref linkend="removing-pkgs"/>).  Do not forget
1386    to properly reassign the package's bugs at the same time.
1387  </para>  </para>
1388  <para>  <para>
1389  At other times, you may make a mistake in constructing your package and wish to  At other times, you may make a mistake in constructing your package and wish to
# Line 1397  you are not a porter, you should read mo Line 1481  you are not a porter, you should read mo
1481  Porting is the act of building Debian packages for architectures that are  Porting is the act of building Debian packages for architectures that are
1482  different from the original architecture of the package maintainer's binary  different from the original architecture of the package maintainer's binary
1483  package.  It is a unique and essential activity.  In fact, porters do most of  package.  It is a unique and essential activity.  In fact, porters do most of
1484  the actual compiling of Debian packages.  For instance, for a single  the actual compiling of Debian packages.  For instance, when a maintainer
1485  <literal>i386</literal> binary package, there must be a recompile for each  uploads a (portable) source packages with binaries for the <literal>i386
1486  architecture, which amounts to &number-of-arches; more builds.  </literal> architecture, it will be built for each of the other architectures,
1487    amounting to &number-of-arches; more builds.
1488  </para>  </para>
1489  <section id="kind-to-porters">  <section id="kind-to-porters">
1490  <title>Being kind to porters</title>  <title>Being kind to porters</title>
# Line 1430  Make sure that your <literal>Build-Depen Line 1515  Make sure that your <literal>Build-Depen
1515  <literal>Build-Depends-Indep</literal> settings in  <literal>Build-Depends-Indep</literal> settings in
1516  <filename>debian/control</filename> are set properly.  The best way to validate  <filename>debian/control</filename> are set properly.  The best way to validate
1517  this is to use the <systemitem role="package">debootstrap</systemitem> package  this is to use the <systemitem role="package">debootstrap</systemitem> package
1518  to create an unstable chroot environment (see <xref linkend="debootstrap"/> ).  to create an <literal>unstable</literal> chroot environment (see <xref
1519    linkend="debootstrap"/> ).
1520  Within that chrooted environment, install the <systemitem  Within that chrooted environment, install the <systemitem
1521  role="package">build-essential</systemitem> package and any package  role="package">build-essential</systemitem> package and any package
1522  dependencies mentioned in <literal>Build-Depends</literal> and/or  dependencies mentioned in <literal>Build-Depends</literal> and/or
# Line 1450  Manual</ulink> for instructions on setti Line 1536  Manual</ulink> for instructions on setti
1536  </listitem>  </listitem>
1537  <listitem>  <listitem>
1538  <para>  <para>
1539  Don't set architecture to a value other than ``all'' or ``any'' unless you  Don't set architecture to a value other than <literal>all</literal> or
1540  really mean it.  In too many cases, maintainers don't follow the instructions  <literal>any</literal> unless you really mean it.  In too many cases,
1541  in the <ulink url="&url-debian-policy;">Debian Policy  maintainers don't follow the instructions in the <ulink
1542  Manual</ulink>.  Setting your architecture to ``i386'' is usually incorrect.  url="&url-debian-policy;">Debian Policy Manual</ulink>.  Setting your
1543    architecture to only one architecture (such as <literal>i386</literal>
1544    or <literal>amd64</literal>) is usually incorrect.
1545  </para>  </para>
1546  </listitem>  </listitem>
1547  <listitem>  <listitem>
# Line 1468  scratch with <command>dpkg-buildpackage< Line 1556  scratch with <command>dpkg-buildpackage<
1556  <para>  <para>
1557  Make sure you don't ship your source package with the  Make sure you don't ship your source package with the
1558  <filename>debian/files</filename> or <filename>debian/substvars</filename>  <filename>debian/files</filename> or <filename>debian/substvars</filename>
1559  files.  They should be removed by the `clean' target of  files.  They should be removed by the <literal>clean</literal> target of
1560  <filename>debian/rules</filename>.  <filename>debian/rules</filename>.
1561  </para>  </para>
1562  </listitem>  </listitem>
# Line 1484  even if it's the same architecture. Line 1572  even if it's the same architecture.
1572  <listitem>  <listitem>
1573  <para>  <para>
1574  Don't depend on the package you're building being installed already (a sub-case  Don't depend on the package you're building being installed already (a sub-case
1575  of the above issue).  of the above issue). There are, of course, exceptions to this rule, but be
1576    aware that any case like this needs manual bootstrapping and cannot be done
1577    by automated package builders.
1578  </para>  </para>
1579  </listitem>  </listitem>
1580  <listitem>  <listitem>
# Line 1497  standardize on different compilers. Line 1587  standardize on different compilers.
1587  </listitem>  </listitem>
1588  <listitem>  <listitem>
1589  <para>  <para>
1590  Make sure your debian/rules contains separate ``binary-arch'' and  Make sure your debian/rules contains separate <literal>binary-arch</literal>
1591  ``binary-indep'' targets, as the Debian Policy Manual requires.  Make sure that  and <literal>binary-indep</literal> targets, as the Debian Policy Manual
1592  both targets work independently, that is, that you can call the target without  requires.  Make sure that both targets work independently, that is, that you
1593  having called the other before.  To test this, try to run  can call the target without having called the other before.  To test this,
1594  <literal>dpkg-buildpackage -B</literal>.  try to run <command>dpkg-buildpackage -B</command>.
1595  </para>  </para>
1596  </listitem>  </listitem>
1597  </orderedlist>  </orderedlist>
# Line 1528  The way to invoke <command>dpkg-buildpac Line 1618  The way to invoke <command>dpkg-buildpac
1618  -m<replaceable>porter-email</replaceable></literal>.  Of course, set  -m<replaceable>porter-email</replaceable></literal>.  Of course, set
1619  <replaceable>porter-email</replaceable> to your email address.  This will do a  <replaceable>porter-email</replaceable> to your email address.  This will do a
1620  binary-only build of only the architecture-dependent portions of the package,  binary-only build of only the architecture-dependent portions of the package,
1621  using the `binary-arch' target in <filename>debian/rules</filename>.  using the <literal>binary-arch</literal> target in <filename>debian/rules
1622    </filename>.
1623  </para>  </para>
1624  <para>  <para>
1625  If you are working on a Debian machine for your porting efforts and you need to  If you are working on a Debian machine for your porting efforts and you need to
# Line 1545  which the package was built was not good Line 1636  which the package was built was not good
1636  bad compiler, ...).  Then you may just need to recompile it in an updated  bad compiler, ...).  Then you may just need to recompile it in an updated
1637  environment.  However, you have to bump the version number in this case, so  environment.  However, you have to bump the version number in this case, so
1638  that the old bad package can be replaced in the Debian archive  that the old bad package can be replaced in the Debian archive
1639  (<command>katie</command> refuses to install new packages if they don't have a  (<command>dak</command> refuses to install new packages if they don't have a
1640  version number greater than the currently available one).  version number greater than the currently available one).
1641  </para>  </para>
1642  <para>  <para>
1643  You have to make sure that your binary-only NMU doesn't render the package  You have to make sure that your binary-only NMU doesn't render the package
1644  uninstallable.  This could happen when a source package generates  uninstallable.  This could happen when a source package generates
1645  arch-dependent and arch-independent packages that depend on each other via  arch-dependent and arch-independent packages that have inter-dependencies
1646  $(Source-Version).  generated using dpkg's substitution variable <literal>$(Source-Version)
1647    </literal>.
1648  </para>  </para>
1649  <para>  <para>
1650  Despite the required modification of the changelog, these are called  Despite the required modification of the changelog, these are called
# Line 1568  source code). Line 1660  source code).
1660  </para>  </para>
1661  <para>  <para>
1662  The ``magic'' for a recompilation-only NMU is triggered by using a suffix  The ``magic'' for a recompilation-only NMU is triggered by using a suffix
1663  appended to the package version number, following the form b&lt;number&gt;.  appended to the package version number, following the form <literal>
1664    b<replaceable>number</replaceable></literal>.
1665  For instance, if the latest version you are recompiling against was version  For instance, if the latest version you are recompiling against was version
1666  ``2.9-3'', your NMU should carry a version of ``2.9-3+b1''.  If the latest  <literal>2.9-3</literal>, your binary-only NMU should carry a version of
1667  version was ``3.4+b1'' (i.e, a native package with a previous recompilation  <literal>2.9-3+b1</literal>.  If the latest version was <literal>3.4+b1
1668  NMU), your NMU should have a version number of ``3.4+b2''.  <footnote><para> In  </literal> (i.e, a native package with a previous recompilation NMU), your
1669  the past, such NMUs used the third-level number on the Debian part of the  binary-only NMU should have a version number of <literal>3.4+b2</literal>.
1670  revision to denote their recompilation-only status; however, this syntax was  <footnote><para> In the past, such NMUs used the third-level number on the
1671  ambiguous with native packages and did not allow proper ordering of  Debian part of the revision to denote their recompilation-only status;
1672  recompile-only NMUs, source NMUs, and security NMUs on the same package, and  however, this syntax was ambiguous with native packages and did not allow
1673  has therefore been abandoned in favor of this new syntax.  </para> </footnote>  proper ordering of recompile-only NMUs, source NMUs, and security NMUs on
1674    the same package, and has therefore been abandoned in favor of this new syntax.
1675    </para> </footnote>
1676  </para>  </para>
1677  <para>  <para>
1678  Similar to initial porter uploads, the correct way of invoking  Similar to initial porter uploads, the correct way of invoking
# Line 1598  the architecture is a candidate for incl Line 1693  the architecture is a candidate for incl
1693  release managers decide and announce which architectures are candidates.  release managers decide and announce which architectures are candidates.
1694  </para>  </para>
1695  <para>  <para>
1696  If you are a porter doing an NMU for `unstable', the above guidelines for  If you are a porter doing an NMU for <literal>unstable</literal>, the above
1697  porting should be followed, with two variations.  Firstly, the acceptable  guidelines for porting should be followed, with two variations.  Firstly, the
1698  waiting period — the time between when the bug is submitted to the BTS and  acceptable waiting period — the time between when the bug is submitted to
1699  when it is OK to do an NMU — is seven days for porters working on the  the BTS and when it is OK to do an NMU — is seven days for porters working
1700  unstable distribution.  This period can be shortened if the problem is critical  on the <literal>unstable</literal> distribution.  This period can be shortened
1701  and imposes hardship on the porting effort, at the discretion of the porter  if the problem is critical and imposes hardship on the porting effort, at the
1702  group.  (Remember, none of this is Policy, just mutually agreed upon  discretion of the porter group.  (Remember, none of this is Policy, just
1703  guidelines.) For uploads to stable or testing, please coordinate with the  mutually agreed upon guidelines.) For uploads to <literal>stable</literal> or
1704  appropriate release team first.  <literal>testing </literal>, please coordinate with the appropriate release
1705    team first.
1706  </para>  </para>
1707  <para>  <para>
1708  Secondly, porters doing source NMUs should make sure that the bug they submit  Secondly, porters doing source NMUs should make sure that the bug they submit
1709  to the BTS should be of severity `serious' or greater.  This ensures that a  to the BTS should be of severity <literal>serious</literal> or greater.  This
1710  single source package can be used to compile every supported Debian  ensures that a single source package can be used to compile every supported
1711  architecture by release time.  It is very important that we have one version of  Debian architecture by release time.  It is very important that we have one
1712  the binary and source package for all architecture in order to comply with many  version of the binary and source package for all architectures in order to
1713  licenses.  comply with many licenses.
1714  </para>  </para>
1715  <para>  <para>
1716  Porters should try to avoid patches which simply kludge around bugs in the  Porters should try to avoid patches which simply kludge around bugs in the
# Line 1663  linkend="tools-porting"/> . Line 1759  linkend="tools-porting"/> .
1759  </para>  </para>
1760  </section>  </section>
1761    
1762  <section id="buildd">  <section id="wanna-build">
1763  <title><systemitem role="package">buildd</systemitem></title>  <title><systemitem role="package">wanna-build</systemitem></title>
1764  <para>  <para>
1765  The <systemitem role="package">buildd</systemitem> system is used as a  The <systemitem role="package">wanna-build</systemitem> system is used as a
1766  distributed, client-server build distribution system.  It is usually used in  distributed, client-server build distribution system.  It is usually used in
1767  conjunction with <literal>build daemons</literal>, which are ``slave'' hosts  conjunction with build daemons running the <systemitem role="package">buildd
1768  which simply check out and attempt to auto-build packages which need to be  </systemitem> program. <literal>Build daemons</literal> are ``slave'' hosts
1769  ported.  There is also an email interface to the system, which allows porters  which contact the central <systemitem role="package"> wanna-build</systemitem>
1770  to ``check out'' a source package (usually one which cannot yet be auto-built)  system to receive a list of packages that need to be built.
1771  and work on it.  </para>
1772  </para>  <para>
1773  <para>  <systemitem role="package">wanna-build</systemitem> is not yet available as a
1774  <systemitem role="package">buildd</systemitem> is not yet available as a  package; however, all Debian porting efforts are using it for automated
1775  package; however, most porting efforts are either using it currently or  package building.  The tool used to do the actual package builds, <systemitem
1776  planning to use it in the near future.  The actual automated builder is  role="package">sbuild</systemitem> is available as a package, see its
1777  packaged as <systemitem role="package">sbuild</systemitem>, see its description  description in <xref linkend="sbuild"/> .  Please note that the packaged
1778  in <xref linkend="sbuild"/> .  The complete <systemitem  version is not the same as the one used on build daemons, but it is close
1779  role="package">buildd</systemitem> system also collects a number of as yet  enough to reproduce problems.
1780  unpackaged components which are currently very useful and in use continually,  </para>
1781  such as <command>andrea</command> and <command>wanna-build</command>.  <para>
1782  </para>  Most of the data produced by <systemitem role="package">wanna-build
1783  <para>  </systemitem> which is generally useful to porters is available on the
1784  Some of the data produced by <systemitem role="package">buildd</systemitem>  web at <ulink url="&url-buildd;"></ulink>.  This data includes nightly
1785  which is generally useful to porters is available on the web at <ulink  updated statistics, queueing information and logs for build attempts.
 url="&url-buildd;"></ulink>.  This data includes nightly updated  
 information from <command>andrea</command> (source dependencies) and  
 <systemitem role="package">quinn-diff</systemitem> (packages needing  
 recompilation).  
1786  </para>  </para>
1787  <para>  <para>
1788  We are quite proud of this system, since it has so many possible uses.  We are quite proud of this system, since it has so many possible uses.
# Line 1700  flavor of Debian built with <command>gcc Line 1792  flavor of Debian built with <command>gcc
1792  also enable Debian to recompile entire distributions quickly.  also enable Debian to recompile entire distributions quickly.
1793  </para>  </para>
1794  <para>  <para>
1795  The buildds admins of each arch can be contacted at the mail address  The wanna-build team, in charge of the buildds,
1796  $arch@buildd.debian.org.  can be reached at <literal>debian-wb-team@lists.debian.org</literal>.
1797    To determine who (wanna-build team, release team) and how (mail, BTS)
1798    to contact, refer to <ulink url="&url-wb-team;"></ulink>.
1799    </para>
1800    
1801    <para>
1802    When requesting binNMUs or give-backs (retries after a failed build),
1803    please use the format described at <ulink url="&url-release-wb;"/>.
1804  </para>  </para>
1805    
1806  </section>  </section>
1807    
1808  </section>  </section>
# Line 1713  $arch@buildd.debian.org. Line 1813  $arch@buildd.debian.org.
1813  Some packages still have issues with building and/or working on some of the  Some packages still have issues with building and/or working on some of the
1814  architectures supported by Debian, and cannot be ported at all, or not within a  architectures supported by Debian, and cannot be ported at all, or not within a
1815  reasonable amount of time.  An example is a package that is SVGA-specific (only  reasonable amount of time.  An example is a package that is SVGA-specific (only
1816  i386), or uses other hardware-specific features not supported on all  available for <literal>i386</literal> and <literal>amd64</literal>), or uses
1817  architectures.  other hardware-specific features not supported on all architectures.
1818  </para>  </para>
1819  <para>  <para>
1820  In order to prevent broken packages from being uploaded to the archive, and  In order to prevent broken packages from being uploaded to the archive, and
# Line 1732  allow the package to build as soon as th Line 1832  allow the package to build as soon as th
1832  </para>  </para>
1833  <para>  <para>
1834  Additionally, if you believe the list of supported architectures is pretty  Additionally, if you believe the list of supported architectures is pretty
1835  constant, you should change 'any' to a list of supported architectures in  constant, you should change <literal>any</literal> to a list of supported
1836  debian/control.  This way, the build will fail also, and indicate this to a  architectures in <filename>debian/control</filename>.  This way, the build will
1837  human reader without actually trying.  fail also, and indicate this to a human reader without actually trying.
1838  </para>  </para>
1839  </listitem>  </listitem>
1840  <listitem>  <listitem>
# Line 1764  role="package">ftp.debian.org</systemite Line 1864  role="package">ftp.debian.org</systemite
1864  <section id="nmu">  <section id="nmu">
1865  <title>Non-Maintainer Uploads (NMUs)</title>  <title>Non-Maintainer Uploads (NMUs)</title>
1866  <para>  <para>
1867  Under certain circumstances it is necessary for someone other than the official  Every package has one or more maintainers. Normally, these are the people who
1868  package maintainer to make a release of a package.  This is called a  work on and upload new versions of the package. In some situations, it is
1869  non-maintainer upload, or NMU.  useful that other developers can upload a new version as well, for example if
1870  </para>  they want to fix a bug in a package they don't maintain, when the maintainer
1871  <para>  needs help to respond to issues.  Such uploads are called
1872  This section handles only source NMUs, i.e.  NMUs which upload a new version of  <emphasis>Non-Maintainer Uploads (NMU)</emphasis>.
 the package.  For binary-only NMUs by porters or QA members, please see <xref  
 linkend="binary-only-nmu"/> .  If a buildd builds and uploads a package, that  
 too is strictly speaking a binary NMU.  See <xref linkend="buildd"/> for some  
 more information.  
 </para>  
 <para>  
 The main reason why NMUs are done is when a developer needs to fix another  
 developer's package in order to address serious problems or crippling bugs or  
 when the package maintainer is unable to release a fix in a timely fashion.  
 </para>  
 <para>  
 First and foremost, it is critical that NMU patches to source should be as  
 non-disruptive as possible.  Do not do housekeeping tasks, do not change the  
 name of modules or files, do not move directories; in general, do not fix  
 things which are not broken.  Keep the patch as small as possible.  If things  
 bother you aesthetically, talk to the Debian maintainer, talk to the upstream  
 maintainer, or submit a bug.  However, aesthetic changes must  
 <emphasis>not</emphasis> be made in a non-maintainer upload.  
 </para>  
 <para>  
 And please remember the Hippocratic Oath: Above all, do no harm.  It is better  
 to leave a package with an open grave bug than applying a non-functional patch,  
 or one that hides the bug instead of resolving it.  
1873  </para>  </para>
1874    
1875  <section id="nmu-guidelines">  <section id="nmu-guidelines">
1876  <title>How to do a NMU</title>  <title>When and how to do an NMU</title>
1877  <para>  
 NMUs which fix important, serious or higher severity bugs are encouraged and  
 accepted.  You should endeavor to reach the current maintainer of the package;  
 they might be just about to upload a fix for the problem, or have a better  
 solution.  
 </para>  
 <para>  
 NMUs should be made to assist a package's maintainer in resolving bugs.  
 Maintainers should be thankful for that help, and NMUers should respect the  
 decisions of maintainers, and try to personally help the maintainer by their  
 work.  
 </para>  
1878  <para>  <para>
1879  A NMU should follow all conventions, written down in this section.  For an  Before doing an NMU, consider the following questions:
 upload to testing or unstable, this order of steps is recommended:  
1880  </para>  </para>
1881  <itemizedlist>  <itemizedlist>
1882  <listitem>  <listitem>
1883  <para>  <para>
1884  Make sure that the package's bugs that the NMU is meant to address are all  Does your NMU really fix bugs? Fixing cosmetic issues or changing the
1885  filed in the Debian Bug Tracking System (BTS).  If they are not, submit them  packaging style in NMUs is discouraged.
 immediately.  
1886  </para>  </para>
1887  </listitem>  </listitem>
1888  <listitem>  <listitem>
1889  <para>  <para>
1890  Wait a few days for the response from the maintainer.  If you don't get any  Did you give enough time to the maintainer? When was the bug reported to the
1891  response, you may want to help them by sending the patch that fixes the bug.  BTS? Being busy for a week or two isn't unusual.  Is the bug so severe that it
1892  Don't forget to tag the bug with the patch keyword.  needs to be fixed right now, or can it wait a few more days?
1893  </para>  </para>
1894  </listitem>  </listitem>
1895  <listitem>  <listitem>
1896  <para>  <para>
1897  Wait a few more days.  If you still haven't got an answer from the maintainer,  How confident are you about your changes? Please remember the Hippocratic Oath:
1898  send them a mail announcing your intent to NMU the package.  Prepare an NMU as  "Above all, do no harm." It is better to leave a package with an open grave bug
1899  described in this section, and test it carefully on your machine (cf.  <xref  than applying a non-functional patch, or one that hides the bug instead of
1900  linkend="sanitycheck"/> ).  Double check that your patch doesn't have any  resolving it. If you are not 100% sure of what you did, it might be a good idea
1901  unexpected side effects.  Make sure your patch is as small and as  to seek advice from others. Remember that if you break something in your NMU,
1902  non-disruptive as it can be.  many people will be very unhappy about it.
1903  </para>  </para>
1904  </listitem>  </listitem>
1905  <listitem>  <listitem>
1906  <para>  <para>
1907  Upload your package to incoming in <filename>DELAYED/7-day</filename> (cf.  Have you clearly expressed your intention to NMU, at least in the BTS?
1908  <xref linkend="delayed-incoming"/> ), send the final patch to the maintainer  It is also a good idea to try to contact the
1909  via the BTS, and explain to them that they have 7 days to react if they want to  maintainer by other means (private email, IRC).
 cancel the NMU.  
1910  </para>  </para>
1911  </listitem>  </listitem>
1912  <listitem>  <listitem>
1913  <para>  <para>
1914  Follow what happens, you're responsible for any bug that you introduced with  If the maintainer is usually active and responsive, have you tried to contact
1915  your NMU.  You should probably use <xref linkend="pkg-tracking-system"/> (PTS)  him? In general it should be considered preferable that a maintainer takes care
1916  to stay informed of the state of the package after your NMU.  of an issue himself and that he is given the chance to review and correct your
1917    patch, because he can be expected to be more aware of potential issues which an
1918    NMUer might miss. It is often a better use of everyone's time if the maintainer
1919    is given an opportunity to upload a fix on their own.
1920  </para>  </para>
1921  </listitem>  </listitem>
1922  </itemizedlist>  </itemizedlist>
1923  <para>  <para>
1924  At times, the release manager or an organized group of developers can announce  When doing an NMU, you must first make sure that your intention to NMU is
1925  a certain period of time in which the NMU rules are relaxed.  This usually  clear.  Then, you must send a patch with the differences between the
1926  involves shortening the period during which one is to wait before uploading the  current package and your proposed NMU to the BTS. The
1927  fixes, and shortening the DELAYED period.  It is important to notice that even  <literal>nmudiff</literal> script in the <literal>devscripts</literal> package
1928  in these so-called bug squashing party times, the NMU'er has to file bugs and  might be helpful.
1929  contact the developer first, and act later.  Please see <xref  </para>
1930  linkend="qa-bsp"/> for details.  <para>
1931  </para>  While preparing the patch, you should better be aware of any package-specific
1932  <para>  practices that the maintainer might be using. Taking them into account reduces
1933  For the testing distribution, the rules may be changed by the release managers.  the burden of getting your changes integrated back in the normal package
1934  Please take additional care, and acknowledge that the usual way for a package  workflow and thus increases the possibilities that that will happen. A good
1935  to enter testing is through unstable.  place where to look for for possible package-specific practices is
1936  </para>  <ulink url="&url-debian-policy;ch-source.html#s-readmesource"><literal>debian/README.source</literal></ulink>.
1937  <para>  </para>
1938  For the stable distribution, please take extra care.  Of course, the release  <para>
1939  managers may also change the rules here.  Please verify before you upload that  Unless you have an excellent reason not to do so, you must then give some time
1940  all your changes are OK for inclusion into the next stable release by the  to the maintainer to react (for example, by uploading to the
1941  release manager.  <literal>DELAYED</literal> queue).  Here are some recommended values to use for delays:
1942  </para>  </para>
1943    <itemizedlist>
1944    <listitem>
1945  <para>  <para>
1946  When a security bug is detected, the security team may do an NMU, using their  Upload fixing only release-critical bugs older than 7 days: 2 days
 own rules.  Please refer to <xref linkend="bug-security"/> for more  
 information.  
1947  </para>  </para>
1948    </listitem>
1949    <listitem>
1950  <para>  <para>
1951  For the differences for Porters NMUs, please see <xref  Upload fixing only release-critical and important bugs: 5 days
 linkend="source-nmu-when-porter"/> .  
1952  </para>  </para>
1953    </listitem>
1954    <listitem>
1955  <para>  <para>
1956  Of course, it is always possible to agree on special rules with a maintainer  Other NMUs: 10 days
 (like the maintainer asking please upload this fix directly for me, and no diff  
 required).  
1957  </para>  </para>
1958  </section>  </listitem>
1959    </itemizedlist>
1960    
 <section id="nmu-version">  
 <title>NMU version numbering</title>  
1961  <para>  <para>
1962  Whenever you have made a change to a package, no matter how trivial, the  Those delays are only examples. In some cases, such as uploads fixing security
1963  version number needs to change.  This enables our packing system to function.  issues, or fixes for trivial bugs that blocking a transition, it is desirable
1964    that the fixed package reaches <literal>unstable</literal> sooner.
1965  </para>  </para>
1966    
1967  <para>  <para>
1968  If you are doing a non-maintainer upload (NMU), you should add a new minor  Sometimes, release managers decide to allow NMUs with shorter delays for a
1969  version number to the <replaceable>debian-revision</replaceable> part of the  subset of bugs (e.g release-critical bugs older than 7 days). Also, some
1970  version number (the portion after the last hyphen).  This extra minor number  maintainers list themselves in the <ulink url="&url-low-threshold-nmu;">Low
1971  will start at `1'.  For example, consider the package `foo', which is at  Threshold NMU list</ulink>, and accept that NMUs are uploaded without delay. But
1972  version 1.1-3.  In the archive, the source package control file would be  even in those cases, it's still a good idea to give the maintainer a few days
1973  <filename>foo_1.1-3.dsc</filename>.  The upstream version is `1.1' and the  to react before you upload, especially if the patch wasn't available in the BTS
1974  Debian revision is `3'.  The next NMU would add a new minor number `.1' to the  before, or if you know that the maintainer is generally active.
 Debian revision; the new source control file would be  
 <filename>foo_1.1-3.1.dsc</filename>.  
 </para>  
 <para>  
 The Debian revision minor number is needed to avoid stealing one of the package  
 maintainer's version numbers, which might disrupt their work.  It also has the  
 benefit of making it visually clear that a package in the archive was not made  
 by the official maintainer.  
1975  </para>  </para>
1976    
1977  <para>  <para>
1978  If there is no <replaceable>debian-revision</replaceable> component in the  After you upload an NMU, you are responsible for the possible problems that you
1979  version number then one should be created, starting at `0.1' (but in case of a  might have introduced. You must keep an eye on the package (subscribing to the
1980  debian native package still upload it as native package).  If it is absolutely  package on the PTS is a good way to achieve this).
 necessary for someone other than the usual maintainer to make a release based  
 on a new upstream version then the person making the release should start with  
 the <replaceable>debian-revision</replaceable> value `0.1'.  The usual  
 maintainer of a package should start their  
 <replaceable>debian-revision</replaceable> numbering at `1'.  
1981  </para>  </para>
1982    
1983  <para>  <para>
1984  If you upload a package to testing or stable, sometimes, you need to fork the  This is not a license to perform NMUs thoughtlessly.  If you NMU when it is
1985  version number tree.  For this, version numbers like 1.1-3sarge0.1 could be  clear that the maintainers are active and would have acknowledged a patch in a
1986  used.  timely manner, or if you ignore the recommendations of this document, your
1987    upload might be a cause of conflict with the maintainer.
1988    You should always be prepared to
1989    defend the wisdom of any NMU you perform on its own merits.
1990  </para>  </para>
1991  </section>  </section>
1992    
1993  <section id="nmu-changelog">  <section id="nmu-changelog">
1994  <title>Source NMUs must have a new changelog entry</title>  <title>NMUs and debian/changelog</title>
 <para>  
 Anyone who is doing a source NMU must create a changelog entry, describing  
 which bugs are fixed by the NMU, and generally why the NMU was required and  
 what it fixed.  The changelog entry will have the email address of the person  
 who uploaded it in the log entry and the NMU version number in it.  
 </para>  
1995  <para>  <para>
1996  By convention, source NMU changelog entries start with the line  Just like any other (source) upload, NMUs must add an entry to
1997    <literal>debian/changelog</literal>, telling what has changed with this
1998    upload.  The first line of this entry must explicitely mention that this upload is an NMU, e.g.:
1999  </para>  </para>
2000  <screen>  <screen>
2001    * Non-maintainer upload    * Non-maintainer upload.
2002  </screen>  </screen>
 </section>  
2003    
 <section id="nmu-patch">  
 <title>Source NMUs and the Bug Tracking System</title>  
2004  <para>  <para>
2005  Maintainers other than the official package maintainer should make as few  The way to version NMUs differs for native and non-native packages.
2006  changes to the package as possible, and they should always send a patch as a  </para>
2007  unified context diff (<literal>diff -u</literal>) detailing their changes to  <para>
2008  the Bug Tracking System.  If the package is a native package (without a debian revision in the version number),
2009    the version must be the version of the last maintainer upload, plus
2010    <literal>+nmu<replaceable>X</replaceable></literal>, where
2011    <replaceable>X</replaceable> is a counter starting at <literal>1</literal>.
2012    If
2013    the last upload was also an NMU, the counter should be increased.  For example,
2014    if the current version is <literal>1.5</literal>, then an NMU would get
2015    version <literal>1.5+nmu1</literal>.
2016    </para>
2017    <para>
2018    If the package is a not a native package, you should add a minor version number
2019    to the debian revision part of the version number (the portion after the last
2020    hyphen). This extra number must start at 1.  For example,
2021    if the current version is <literal>1.5-2</literal>, then an NMU would get
2022    version <literal>1.5-2.1</literal>. If a new upstream version
2023    is packaged in the NMU, the debian revision is set to <literal>0</literal>, for
2024    example <literal>1.6-0.1</literal>.
2025    </para>
2026    <para>
2027    In both cases, if the last upload was also an NMU, the counter should
2028    be increased. For example, if the current version is
2029    <literal>1.5+nmu3</literal> (a native package which has already been
2030    NMUed), the NMU would get version <literal>1.5+nmu4</literal>.  .
2031    </para>
2032    <para>
2033    A special versioning scheme is needed to avoid disrupting the maintainer's
2034    work, since using an integer for the Debian revision will potentially
2035    conflict with a maintainer upload already in preparation at the time of an
2036    NMU, or even one sitting in the ftp NEW queue.
2037    It also has the
2038    benefit of making it visually clear that a package in the archive was not made
2039    by the official maintainer.
2040  </para>  </para>
2041    
2042  <para>  <para>
2043  What if you are simply recompiling the package?  If you just need to recompile  If you upload a package to testing or stable, you sometimes need to "fork" the
2044  it for a single architecture, then you may do a binary-only NMU as described in  version number tree. This is the case for security uploads, for example.  For
2045  <xref linkend="binary-only-nmu"/> which doesn't require any patch to be sent.  this, a version of the form
2046  If you want the package to be recompiled for all architectures, then you do a  <literal>+deb<replaceable>XY</replaceable>u<replaceable>Z</replaceable></literal>
2047  source NMU as usual and you will have to send a patch.  should be used, where <replaceable>X</replaceable> and
2048    <replaceable>Y</replaceable> are the major and minor release numbers, and
2049    <replaceable>Z</replaceable> is a counter starting at <literal>1</literal>.
2050    When the release number is not yet known (often the case for
2051    <literal>testing</literal>, at the beginning of release cycles), the lowest
2052    release number higher than the last stable release number must be used.  For
2053    example, while Etch (Debian 4.0) is stable, a security NMU to stable for a
2054    package at version <literal>1.5-3</literal> would have version
2055    <literal>1.5-3+deb40u1</literal>, whereas a security NMU to Lenny would get
2056    version <literal>1.5-3+deb50u1</literal>. After the release of Lenny, security
2057    uploads to the <literal>testing</literal> distribution will be versioned
2058    <literal>+deb51uZ</literal>, until it is known whether that release will be
2059    Debian 5.1 or Debian 6.0 (if that becomes the case, uploads will be versioned
2060    as <literal>+deb60uZ</literal>.
2061  </para>  </para>
2062    </section>
2063    
2064    <section id="nmu-delayed">
2065    <title>Using the <literal>DELAYED/</literal> queue</title>
2066    
2067  <para>  <para>
2068  Bugs fixed by source NMUs used to be tagged fixed instead of closed, but since  Having to wait for a response after you request permission to NMU is
2069  version tracking is in place, such bugs are now also closed with the NMU  inefficient, because it costs the NMUer a context switch to come back to the
2070  version.  issue.
2071    The <literal>DELAYED</literal> queue (see <xref linkend="delayed-incoming"/>)
2072    allows the developer doing the NMU to perform all the necessary tasks at the
2073    same time. For instance, instead of telling the maintainer that you will
2074    upload the updated
2075    package in 7 days, you should upload the package to
2076    <literal>DELAYED/7</literal> and tell the maintainer that he has 7 days to
2077    react.  During this time, the maintainer can ask you to delay the upload some
2078    more, or cancel your upload.
2079  </para>  </para>
2080    
2081  <para>  <para>
2082  Also, after doing an NMU, you have to send the information to the existing bugs  The <literal>DELAYED</literal> queue should not be used to put additional
2083  that are fixed by your NMU, including the unified diff.  Historically, it was  pressure on the maintainer. In particular, it's important that you are
2084  custom to open a new bug and include a patch showing all the changes you have  available to cancel or delay the upload before the delay expires since the
2085  made.  The normal maintainer will either apply the patch or employ an alternate  maintainer cannot cancel the upload himself.
 method of fixing the problem.  Sometimes bugs are fixed independently upstream,  
 which is another good reason to back out an NMU's patch.  If the maintainer  
 decides not to apply the NMU's patch but to release a new version, the  
 maintainer needs to ensure that the new upstream version really fixes each  
 problem that was fixed in the non-maintainer release.  
2086  </para>  </para>
2087    
2088  <para>  <para>
2089  In addition, the normal maintainer should <emphasis>always</emphasis> retain  If you make an NMU to <literal>DELAYED</literal> and the maintainer updates
2090  the entry in the changelog file documenting the non-maintainer upload -- and of  his package before the delay expires, your upload will be rejected because a
2091  course, also keep the changes.  If you revert some of the changes, please  newer version is already available in the archive.
2092  reopen the relevant bug reports.  Ideally, the maintainer will take care to include your proposed changes (or
2093    at least a solution for the problems they address) in that upload.
2094  </para>  </para>
2095    
2096  </section>  </section>
2097    
2098  <section id="nmu-build">  <section id="nmu-maintainer">
2099  <title>Building source NMUs</title>  <title>NMUs from the maintainer's point of view</title>
2100    
2101  <para>  <para>
2102  Source NMU packages are built normally.  Pick a distribution using the same  When someone NMUs your package, this means they want to help you to keep it in
2103  rules as found in <xref linkend="distribution"/> , follow the other  good shape.  This gives users fixed packages faster.  You
2104  instructions in <xref linkend="upload"/> .  can consider asking the NMUer to become a co-maintainer of the package.
2105    Receiving an NMU on a package is not a bad
2106    thing; it just means that the package is interesting enough for other people to
2107    work on it.
2108  </para>  </para>
2109    
2110  <para>  <para>
2111  Make sure you do <emphasis>not</emphasis> change the value of the maintainer in  To acknowledge an NMU, include its changes and changelog entry in your next
2112  the <filename>debian/control</filename> file.  Your name as given in the NMU  maintainer upload.  If you do not acknowledge the NMU by including the
2113  entry of the <filename>debian/changelog</filename> file will be used for  NMU changelog entry in your changelog, the bugs will remain closed in the
2114  signing the changes file.  BTS but will be listed as affecting your maintainer version of the package.
2115  </para>  </para>
2116    
2117  </section>  </section>
2118    
2119  <section id="ack-nmu">  <section id="nmu-binnmu">
2120  <title>Acknowledging an NMU</title>  <title>Source NMUs vs Binary-only NMUs (binNMUs)</title>
2121    
2122  <para>  <para>
2123  If one of your packages has been NMU'ed, you have to incorporate the changes in  The full name of an NMU is <emphasis>source NMU</emphasis>.  There is also
2124  your copy of the sources.  This is easy, you just have to apply the patch that  another type, namely the <emphasis>binary-only NMU</emphasis>, or
2125  has been sent to you.  Once this is done, you have to close the bugs that have  <emphasis>binNMU</emphasis>.  A binNMU is also a package upload by someone
2126  been tagged fixed by the NMU.  The easiest way is to use the  other than the package's maintainer.  However, it is a binary-only upload.
 <literal>-v</literal> option of <command>dpkg-buildpackage</command>, as this  
 allows you to include just all changes since your last maintainer upload.  
 Alternatively, you can close them manually by sending the required mails to the  
 BTS or by adding the required <literal>closes: #nnnn</literal> in the changelog  
 entry of your next upload.  
2127  </para>  </para>
2128    
2129  <para>  <para>
2130  In any case, you should not be upset by the NMU.  An NMU is not a personal  When a library (or other dependency) is updated, the packages using it may need
2131  attack against the maintainer.  It is a proof that someone cares enough about  to be rebuilt.  Since no changes to the source are needed, the same source
2132  the package that they were willing to help you in your work, so you should be  package is used.
 thankful.  You may also want to ask them if they would be interested in helping  
 you on a more frequent basis as co-maintainer or backup maintainer (see <xref  
 linkend="collaborative-maint"/> ).  
2133  </para>  </para>
 </section>  
2134    
 <section id="nmu-vs-qa">  
 <title>NMU vs QA uploads</title>  
2135  <para>  <para>
2136  Unless you know the maintainer is still active, it is wise to check the package  BinNMUs are usually triggered on the buildds by wanna-build.
2137  to see if it has been orphaned.  The current list of orphaned packages which  An entry is added to debian/changelog,
2138  haven't had their maintainer set correctly is available at <ulink  explaining why the upload was needed and increasing the version number as
2139  url="&url-debian-qa-orphaned;"></ulink>.  If you perform an NMU on an  described in <xref linkend="binary-only-nmu"/>.
2140  improperly orphaned package, please set the maintainer to <literal>Debian QA Group  This entry should not be included in the next upload.
 &lt;packages@qa.debian.org&gt;</literal>.  
2141  </para>  </para>
 </section>  
2142    
 <section id="nmu-who">  
 <title>Who can do an NMU</title>  
2143  <para>  <para>
2144  Only official, registered Debian Developers can do binary or source NMUs.  A  Buildds upload packages for their architecture to the archive as binary-only
2145  Debian Developer is someone who has their key in the Debian key ring.  uploads.  Strictly speaking, these are binNMUs.  However, they are not normally
2146  Non-developers, however, are encouraged to download the source package and  called NMU, and they don't add an entry to debian/changelog.
 start hacking on it to fix problems; however, rather than doing an NMU, they  
 should just submit worthwhile patches to the Bug Tracking System.  Maintainers  
 almost always appreciate quality patches and bug reports.  
2147  </para>  </para>
2148    
2149  </section>  </section>
2150    
2151  <section id="nmu-terms">  <section id="nmu-qa-upload">
2152  <title>Terminology</title>  <title>NMUs vs QA uploads</title>
2153    
2154  <para>  <para>
2155  There are two new terms used throughout this section: ``binary-only NMU'' and  NMUs are uploads of packages by somebody else than their assigned maintainer.
2156  ``source NMU''.  These terms are used with specific technical meaning  There is
2157  throughout this document.  Both binary-only and source NMUs are similar, since  another type of upload where the uploaded package is not yours: QA uploads. QA
2158  they involve an upload of a package by a developer who is not the official  uploads are uploads of orphaned packages.
 maintainer of that package.  That is why it's a  
 <literal>non-maintainer</literal> upload.  
2159  </para>  </para>
2160    
2161  <para>  <para>
2162  A source NMU is an upload of a package by a developer who is not the official  QA uploads are very much like normal maintainer uploads: they may fix anything,
2163  maintainer, for the purposes of fixing a bug in the package.  Source NMUs  even minor issues; the version numbering is normal, and there is no need to use
2164  always involves changes to the source (even if it is just a change to  a delayed upload.  The difference is that you are not listed as the Maintainer
2165  <filename>debian/changelog</filename>).  This can be either a change to the  or Uploader for the package.  Also, the changelog entry of a QA upload has a
2166  upstream source, or a change to the Debian bits of the source.  Note, however,  special first line:
 that source NMUs may also include architecture-dependent packages, as well as  
 an updated Debian diff.  
2167  </para>  </para>
2168    
2169    <screen>
2170     * QA upload.
2171    </screen>
2172    
2173  <para>  <para>
2174  A binary-only NMU is a recompilation and upload of a binary package for a given  If you want to do an NMU, and it seems that the maintainer is not active, it is
2175  architecture.  As such, it is usually part of a porting effort.  A binary-only  wise to check if the package is orphaned
2176  NMU is a non-maintainer uploaded binary version of a package, with no source  (this information is displayed on the package's Package Tracking System page).
2177  changes required.  There are many cases where porters must fix problems in the  When doing the first QA upload to an
2178  source in order to get them to compile for their target architecture; that  orphaned package, the maintainer should be set to <literal>Debian QA Group
2179  would be considered a source NMU rather than a binary-only NMU.  As you can  &lt;packages@qa.debian.org&gt;</literal>.  Orphaned packages which did
2180  see, we don't distinguish in terminology between porter NMUs and non-porter  not yet have a QA upload still have their old maintainer set.  There is a list
2181  NMUs.  of them at <ulink url="&url-orphaned-not-qa;"/>.
2182  </para>  </para>
2183    
2184  <para>  <para>
2185  Both classes of NMUs, source and binary-only, can be lumped under the term  Instead of doing a QA upload, you can also consider adopting the package by
2186  ``NMU''.  However, this often leads to confusion, since most people think  making yourself the maintainer.  You don't need permission from anybody to
2187  ``source NMU'' when they think ``NMU''.  So it's best to be careful: always use  adopt an orphaned package, you can just set yourself as maintainer and upload
2188  ``binary NMU'' or ``binNMU'' for binary-only NMUs.  the new version (see <xref linkend="adopting"/>).
2189  </para>  </para>
2190    
2191  </section>  </section>
2192    
2193  </section>  </section>
# Line 2170  a false sense of good maintenance. Line 2277  a false sense of good maintenance.
2277  <section id="testing-basics">  <section id="testing-basics">
2278  <title>Basics</title>  <title>Basics</title>
2279  <para>  <para>
2280  Packages are usually installed into the `testing' distribution after they have  Packages are usually installed into the <literal>testing</literal> distribution
2281  undergone some degree of testing in unstable.  after they have undergone some degree of <literal>testing</literal> in
2282    <literal>unstable</literal>.
2283  </para>  </para>
2284  <para>  <para>
2285  They must be in sync on all architectures and mustn't have dependencies that  They must be in sync on all architectures and mustn't have dependencies that
2286  make them uninstallable; they also have to have generally no known  make them uninstallable; they also have to have generally no known
2287  release-critical bugs at the time they're installed into testing.  This way,  release-critical bugs at the time they're installed into <literal>testing
2288  `testing' should always be close to being a release candidate.  Please see  </literal>.  This way, <literal>testing</literal> should always be close to
2289  below for details.  being a release candidate.  Please see below for details.
2290  </para>  </para>
2291  </section>  </section>
2292    
# Line 2186  below for details. Line 2294  below for details.
2294  <title>Updates from unstable</title>  <title>Updates from unstable</title>
2295  <para>  <para>
2296  The scripts that update the <literal>testing</literal> distribution are run  The scripts that update the <literal>testing</literal> distribution are run
2297  each day after the installation of the updated packages; these scripts are  twice each day, right after the installation of the updated packages; these
2298  called <literal>britney</literal>.  They generate the  scripts are called <literal>britney</literal>.  They generate the
2299  <filename>Packages</filename> files for the <literal>testing</literal>  <filename>Packages</filename> files for the <literal>testing</literal>
2300  distribution, but they do so in an intelligent manner; they try to avoid any  distribution, but they do so in an intelligent manner; they try to avoid any
2301  inconsistency and to use only non-buggy packages.  inconsistency and to use only non-buggy packages.
# Line 2202  the following: Line 2310  the following:
2310  The package must have been available in <literal>unstable</literal> for 2, 5  The package must have been available in <literal>unstable</literal> for 2, 5
2311  or 10 days, depending on the urgency (high, medium or low).  Please note that  or 10 days, depending on the urgency (high, medium or low).  Please note that
2312  the urgency is sticky, meaning that the highest urgency uploaded since the  the urgency is sticky, meaning that the highest urgency uploaded since the
2313  previous testing transition is taken into account.  Those delays may be doubled  previous <literal>testing</literal> transition is taken into account.  Those
2314  during a freeze, or testing transitions may be switched off altogether;  delays may be doubled during a freeze, or <literal>testing</literal>
2315    transitions may be switched off altogether;
2316  </para>  </para>
2317  </listitem>  </listitem>
2318  <listitem>  <listitem>
# Line 2216  available in <literal>unstable</literal> Line 2325  available in <literal>unstable</literal>
2325  <listitem>  <listitem>
2326  <para>  <para>
2327  It must be available on all architectures on which it has previously been built  It must be available on all architectures on which it has previously been built
2328  in unstable.  <xref linkend="madison"/> may be of interest to check that  in <literal>unstable</literal>.  <xref linkend="dak-ls"/> may be of interest
2329  information;  to check that information;
2330  </para>  </para>
2331  </listitem>  </listitem>
2332  <listitem>  <listitem>
# Line 2236  all the necessary criteria); Line 2345  all the necessary criteria);
2345  </listitem>  </listitem>
2346  </itemizedlist>  </itemizedlist>
2347  <para>  <para>
2348  To find out whether a package is progressing into testing or not, see the  To find out whether a package is progressing into <literal>testing</literal>
2349  testing script output on the <ulink  or not, see the <literal>testing</literal> script output on the <ulink
2350  url="&url-testing-maint;">web page of the testing  url="&url-testing-maint;">web page of the testing
2351  distribution</ulink>, or use the program <command>grep-excuses</command> which  distribution</ulink>, or use the program <command>grep-excuses</command> which
2352  is in the <systemitem role="package">devscripts</systemitem> package.  This  is in the <systemitem role="package">devscripts</systemitem> package.  This
# Line 2260  scripts.  See below for details. Line 2369  scripts.  See below for details.
2369  </para>  </para>
2370  <para>  <para>
2371  Some further dependency analysis is shown on <ulink  Some further dependency analysis is shown on <ulink
2372  url="http://bjorn.haxx.se/debian/"></ulink> — but be warned, this page also  url="http://release.debian.org/migration/"></ulink> — but be warned, this page also
2373  shows build dependencies which are not considered by britney.  shows build dependencies which are not considered by britney.
2374  </para>  </para>
2375  <section id="outdated">  <section id="outdated">
2376  <title>out-of-date</title>  <title>out-of-date</title>
2377  <para>  <para>
2378  <!-- FIXME: better rename this file than document rampant professionalism? -->  <!-- FIXME: better rename this file than document rampant professionalism? -->
2379  For the testing migration script, outdated means: There are different versions  For the <literal>testing</literal> migration script, outdated means: There are
2380  in unstable for the release architectures (except for the architectures in  different versions in <literal>unstable</literal> for the release architectures
2381  fuckedarches; fuckedarches is a list of architectures that don't keep up (in  (except for the architectures in fuckedarches; fuckedarches is a list of
2382  update_out.py), but currently, it's empty).  outdated has nothing whatsoever to  architectures that don't keep up (in <filename>update_out.py</filename>), but
2383  do with the architectures this package has in testing.  currently, it's empty).  outdated has nothing whatsoever to do with the
2384    architectures this package has in <literal>testing</literal>.
2385  </para>  </para>
2386  <para>  <para>
2387  Consider this example:  Consider this example:
# Line 2300  Consider this example: Line 2410  Consider this example:
2410  </tgroup>  </tgroup>
2411  </informaltable>  </informaltable>
2412  <para>  <para>
2413  The package is out of date on alpha in unstable, and will not go to testing.  The package is out of date on alpha in <literal>unstable</literal>, and will
2414  And removing foo from testing would not help at all, the package is still out  not go to <literal>testing</literal>. Removing the package would not help at all, the
2415  of date on alpha, and will not propagate to testing.  package is still out of date on <literal>alpha</literal>, and will not
2416    propagate to testing.
2417  </para>  </para>
2418  <para>  <para>
2419  However, if ftp-master removes a package in unstable (here on arm):  However, if ftp-master removes a package in <literal>unstable</literal> (here
2420    on <literal>arm</literal>):
2421  </para>  </para>
2422  <informaltable pgwide="1">  <informaltable pgwide="1">
2423  <tgroup cols="4">  <tgroup cols="4">
# Line 2335  However, if ftp-master removes a package Line 2447  However, if ftp-master removes a package
2447  </informaltable>  </informaltable>
2448  <para>  <para>
2449  In this case, the package is up to date on all release architectures in  In this case, the package is up to date on all release architectures in
2450  unstable (and the extra hurd-i386 doesn't matter, as it's not a release  <literal>unstable</literal> (and the extra <literal>hurd-i386</literal>
2451  architecture).  doesn't matter, as it's not a release architecture).
2452  </para>  </para>
2453  <para>  <para>
2454  Sometimes, the question is raised if it is possible to allow packages in that  Sometimes, the question is raised if it is possible to allow packages in that
# Line 2355  with the new version of <literal>b</lite Line 2467  with the new version of <literal>b</lite
2467  be removed to allow <literal>b</literal> in.  be removed to allow <literal>b</literal> in.
2468  </para>  </para>
2469  <para>  <para>
2470  Of course, there is another reason to remove a package from testing: It's just  Of course, there is another reason to remove a package from <literal>testing
2471  too buggy (and having a single RC-bug is enough to be in this state).  </literal>: It's just too buggy (and having a single RC-bug is enough to be
2472    in this state).
2473  </para>  </para>
2474  <para>  <para>
2475  Furthermore, if a package has been removed from unstable, and no package in  Furthermore, if a package has been removed from <literal>unstable</literal>,
2476  testing depends on it any more, then it will automatically be removed.  and no package in <literal>testing</literal> depends on it any more, then it
2477    will automatically be removed.
2478  </para>  </para>
2479  </section>  </section>
2480    
# Line 2411  happens to one of your packages. Line 2525  happens to one of your packages.
2525  <section id="s5.13.2.4">  <section id="s5.13.2.4">
2526  <title>influence of package in testing</title>  <title>influence of package in testing</title>
2527  <para>  <para>
2528  Generally, there is nothing that the status of a package in testing means for  Generally, there is nothing that the status of a package in <literal>testing
2529  transition of the next version from unstable to testing, with two exceptions:  </literal> means for transition of the next version from <literal>unstable
2530    </literal> to <literal>testing</literal>, with two exceptions:
2531  If the RC-bugginess of the package goes down, it may go in even if it is still  If the RC-bugginess of the package goes down, it may go in even if it is still
2532  RC-buggy.  The second exception is if the version of the package in testing is  RC-buggy.  The second exception is if the version of the package in <literal>
2533  out of sync on the different arches: Then any arch might just upgrade to the  testing</literal> is out of sync on the different arches: Then any arch might
2534  version of the source package; however, this can happen only if the package was  just upgrade to the version of the source package; however, this can happen
2535  previously forced through, the arch is in fuckedarches, or there was no binary  only if the package was previously forced through, the arch is in fuckedarches,
2536  package of that arch present in unstable at all during the testing migration.  or there was no binary package of that arch present in <literal>unstable
2537    </literal> at all during the <literal>testing</literal> migration.
2538  </para>  </para>
2539  <para>  <para>
2540  In summary this means: The only influence that a package being in testing has  In summary this means: The only influence that a package being in <literal>
2541  on a new version of the same package is that the new version might go in  testing</literal> has on a new version of the same package is that the new
2542  easier.  version might go in easier.
2543  </para>  </para>
2544  </section>  </section>
2545    
# Line 2442  part of britney.) (There is a similar th Line 2558  part of britney.) (There is a similar th
2558  is not described here.  If you're interested in that, please peruse the code.)  is not described here.  If you're interested in that, please peruse the code.)
2559  </para>  </para>
2560  <para>  <para>
2561  Now, the more complex part happens: Britney tries to update testing with the  Now, the more complex part happens: Britney tries to update <literal>testing
2562  valid candidates; first, each package alone, and then larger and even larger  </literal> with the valid candidates. For that, britney tries to add each
2563  sets of packages together.  Each try is accepted if testing is not more  valid candidate to the testing distribution. If the number of uninstallable
2564  uninstallable after the update than before.  (Before and after this part, some  packages in <literal>testing</literal> doesn't increase, the package is
2565  hints are processed; but as only release masters can hint, this is probably not  accepted. From that point on, the accepted package is considered to be part
2566  so important for you.)  of <literal>testing</literal>, such that all subsequent installability
2567    tests include this package.  Hints from the release team are processed
2568    before or after this main run, depending on the exact type.
2569  </para>  </para>
2570  <para>  <para>
2571  If you want to see more details, you can look it up on  If you want to see more details, you can look it up on
2572  merkel:/org/&ftp-debian-org;/testing/update_out/ (or there in  <filename>merkel:/org/&ftp-debian-org;/testing/update_out/</filename> (or
2573  ~aba/testing/update_out to see a setup with a smaller packages file).  Via web,  in <filename>merkel:~aba/testing/update_out</filename> to see a setup with
2574  it's at <ulink  a smaller packages file).  Via web, it's at <ulink
2575  url="http://&ftp-master-host;/testing/update_out_code/"></ulink>  url="http://&ftp-master-host;/testing/update_out_code/"></ulink>
2576  </para>  </para>
2577  <para>  <para>
# Line 2467  url="http://&ftp-master-host;/testing/hi Line 2585  url="http://&ftp-master-host;/testing/hi
2585  <section id="t-p-u">  <section id="t-p-u">
2586  <title>Direct updates to testing</title>  <title>Direct updates to testing</title>
2587  <para>  <para>
2588  The testing distribution is fed with packages from unstable according to the  The <literal>testing</literal> distribution is fed with packages from
2589  rules explained above.  However, in some cases, it is necessary to upload  <literal>unstable</literal> according to the rules explained above.  However,
2590  packages built only for testing.  For that, you may want to upload to  in some cases, it is necessary to upload packages built only for <literal>
2591  <literal>testing-proposed-updates</literal>.  testing</literal>.  For that, you may want to upload to <literal>
2592    testing-proposed-updates</literal>.
2593  </para>  </para>
2594  <para>  <para>
2595  Keep in mind that packages uploaded there are not automatically processed, they  Keep in mind that packages uploaded there are not automatically processed, they
# Line 2482  give on &email-debian-devel-announce;. Line 2601  give on &email-debian-devel-announce;.
2601  <para>  <para>
2602  You should not upload to <literal>testing-proposed-updates</literal> when you  You should not upload to <literal>testing-proposed-updates</literal> when you
2603  can update your packages through <literal>unstable</literal>.  If you can't  can update your packages through <literal>unstable</literal>.  If you can't
2604  (for example because you have a newer development version in unstable), you may  (for example because you have a newer development version in <literal>unstable
2605  use this facility, but it is recommended that you ask for authorization from  </literal>), you may use this facility, but it is recommended that you ask for
2606  the release manager first.  Even if a package is frozen, updates through  authorization from the release manager first.  Even if a package is frozen,
2607  unstable are possible, if the upload via unstable does not pull in any new  updates through <literal>unstable</literal> are possible, if the upload via
2608  dependencies.  <literal>unstable</literal> does not pull in any new dependencies.
2609  </para>  </para>
2610  <para>  <para>
2611  Version numbers are usually selected by adding the codename of the testing  Version numbers are usually selected by adding the codename of the
2612  distribution and a running number, like 1.2sarge1 for the first upload through  <literal>testing</literal> distribution and a running number, like
2613  <literal>testing-proposed-updates</literal> of package version 1.2.  <literal>1.2sarge1</literal> for the first upload through
2614    <literal>testing-proposed-updates</literal> of package version
2615    <literal>1.2</literal>.
2616  </para>  </para>
2617  <para>  <para>
2618  Please make sure you didn't miss any of these items in your upload:  Please make sure you didn't miss any of these items in your upload:
# Line 2500  Please make sure you didn't miss any of Line 2621  Please make sure you didn't miss any of
2621  <listitem>  <listitem>
2622  <para>  <para>
2623  Make sure that your package really needs to go through  Make sure that your package really needs to go through
2624  <literal>testing-proposed-updates</literal>, and can't go through unstable;  <literal>testing-proposed-updates</literal>, and can't go through <literal>
2625    unstable</literal>;
2626  </para>  </para>
2627  </listitem>  </listitem>
2628  <listitem>  <listitem>
# Line 2547  at &email-debian-release; and ask them t Line 2669  at &email-debian-release; and ask them t
2669  <title>What are release-critical bugs, and how do they get counted?</title>  <title>What are release-critical bugs, and how do they get counted?</title>
2670  <para>  <para>
2671  All bugs of some higher severities are by default considered release-critical;  All bugs of some higher severities are by default considered release-critical;
2672  currently, these are critical, grave, and serious bugs.  currently, these are <literal>critical</literal>, <literal>grave</literal> and
2673    <literal>serious</literal> bugs.
2674  </para>  </para>
2675  <para>  <para>
2676  Such bugs are presumed to have an impact on the chances that the package will  Such bugs are presumed to have an impact on the chances that the package will
2677  be released with the stable release of Debian: in general, if a package has  be released with the <literal>stable</literal> release of Debian: in general,
2678  open release-critical bugs filed on it, it won't get into testing, and  if a package has open release-critical bugs filed on it, it won't get into
2679  consequently won't be released in stable.  <literal>testing</literal>, and consequently won't be released in <literal>
2680  </para>  stable</literal>.
 <para>  
 The unstable bug count are all release-critical bugs without either any  
 release-tag (such as potato, woody) or with release-tag sid; also, only if they  
 are neither fixed nor set to sarge-ignore.  The testing bug count for a package  
 is considered to be roughly the bug count of unstable count at the last point  
 when the testing version equalled the unstable version.  
2681  </para>  </para>
2682  <para>  <para>
2683  This will change post-sarge, as soon as we have versions in the bug tracking  The <literal>unstable</literal> bug count are all release-critical bugs which
2684  system.  are marked to apply to <replaceable>package</replaceable>/<replaceable>version
2685    </replaceable> combinations that are available in unstable for a release
2686    architecture. The <literal>testing</literal> bug count is defined analogously.
2687  </para>  </para>
2688  </section>  </section>
2689    
2690  <section id="s5.13.4.2">  <section id="s5.13.4.2">
2691  <title>How could installing a package into testing possibly break other packages?</title>  <title>How could installing a package into <literal>testing</literal> possibly
2692    break other packages?</title>
2693  <para>  <para>
2694  The structure of the distribution archives is such that they can only contain  The structure of the distribution archives is such that they can only contain
2695  one version of a package; a package is defined by its name.  So when the source  one version of a package; a package is defined by its name.  So when the source
2696  package acmefoo is installed into testing, along with its binary packages  package <literal>acmefoo</literal> is installed into <literal>testing</literal>,
2697  acme-foo-bin, acme-bar-bin, libacme-foo1 and libacme-foo-dev, the old version  along with its binary packages <literal>acme-foo-bin</literal>, <literal>
2698  is removed.  acme-bar-bin</literal>, <literal>libacme-foo1</literal> and <literal>
2699    libacme-foo-dev</literal>, the old version is removed.
2700  </para>  </para>
2701  <para>  <para>
2702  However, the old version may have provided a binary package with an old soname  However, the old version may have provided a binary package with an old soname
2703  of a library, such as libacme-foo0.  Removing the old acmefoo will remove  of a library, such as <literal>libacme-foo0</literal>.  Removing the old
2704  libacme-foo0, which will break any packages which depend on it.  <literal>acmefoo</literal> will remove <literal>libacme-foo0</literal>, which
2705    will break any packages which depend on it.
2706  </para>  </para>
2707  <para>  <para>
2708  Evidently, this mainly affects packages which provide changing sets of binary  Evidently, this mainly affects packages which provide changing sets of binary
# Line 2592  the ==, &lt;=, or &lt;&lt; varieties. Line 2714  the ==, &lt;=, or &lt;&lt; varieties.
2714  When the set of binary packages provided by a source package change in this  When the set of binary packages provided by a source package change in this
2715  way, all the packages that depended on the old binaries will have to be updated  way, all the packages that depended on the old binaries will have to be updated
2716  to depend on the new binaries instead.  Because installing such a source  to depend on the new binaries instead.  Because installing such a source
2717  package into testing breaks all the packages that depended on it in testing,  package into <literal>testing</literal> breaks all the packages that depended on
2718    it in <literal>testing</literal>,
2719  some care has to be taken now: all the depending packages must be updated and  some care has to be taken now: all the depending packages must be updated and
2720  ready to be installed themselves so that they won't be broken, and, once  ready to be installed themselves so that they won't be broken, and, once
2721  everything is ready, manual intervention by the release manager or an assistant  everything is ready, manual intervention by the release manager or an assistant
# Line 2600  is normally required. Line 2723  is normally required.
2723  </para>  </para>
2724  <para>  <para>
2725  If you are having problems with complicated groups of packages like this,  If you are having problems with complicated groups of packages like this,
2726  contact debian-devel or debian-release for help.  contact &email-debian-devel; or &email-debian-release; for help.
2727  </para>  </para>
2728  </section>  </section>
2729    

Legend:
Removed from v.5196  
changed lines
  Added in v.7025

  ViewVC Help
Powered by ViewVC 1.1.5