Unstable to testing delay
Description
Before being included in the stable distributions, packages first go into
unstable, then into testing (see
Debian releases). From 2005-11-02 to 2006-01-17,
we have measured the time taken for a package from getting from unstable
to testing (as packages seldom move from testing to stable).
Results
| Minimum delay (days) | Average | Maximum |
|---|
| -28 | 6.0696 | 75 |
It is surprising to note that packages appear in testing BEFORE they
appear in unstable. This needs further investigation ; is that a bug in our software ?
| Delay <= | Number of packages | Percentage |
|---|
| 0 | 13146 | 62% |
| 7 days | 14804 | 70% |
| 2 weeks | 18443 | 87% |
| 3 weeks | 19243 | 91% |
| 6 weeks | 20651 | 98% |
| 3 months | 21080 | 100% |
Top 30 slowest packages:
| Name | Version | Delay |
|---|
| libapache-miniwiki-perl | 0.92-1 | 75 |
| libproc-process-perl | 0.39-1 | 75 |
| libspoon-perl | 0.20-1 | 75 |
| r5rs-doc | 20010328-6 | 75 |
| gmt-doc-pdf | 4.0-2 | 70 |
| mysqltcl | 3.01-1 | 70 |
| gmt-examples | 4.0-2 | 70 |
| gmt-tutorial-pdf | 4.0-2 | 70 |
| gmt-doc | 4.0-2 | 70 |
| lvm-common | 1.5.20 | 70 |
| gmt | 4.0-2 | 70 |
| omniidl4 | 4.0.6-2 | 68 |
| pppstatus | 0.4.2-8 | 67 |
| libcos4-dev | 4.0.6-2 | 66 |
| omniorb4-nameserver | 4.0.6-2 | 66 |
| libomniorb4c2 | 4.0.6-2 | 66 |
| libcos4c2 | 4.0.6-2 | 66 |
| python2.4-omniorb2-omg | 2.6-2 | 66 |
| omniorb4-doc | 4.0.6-2 | 66 |
| omniorb4-idl | 4.0.6-2 | 66 |
| omniidl4-python | 2.6-2 | 66 |
| libomnithread3-dev | 4.0.6-2 | 66 |
| python2.3-omniorb2-omg | 2.6-2 | 66 |
| python2.4-omniorb2 | 2.6-2 | 66 |
| python-omniorb2 | 2.6-2 | 66 |
| python2.3-omniorb2 | 2.6-2 | 66 |
| libomnithread3c2 | 4.0.6-2 | 66 |
| python2.2-omniorb2 | 2.6-2 | 66 |
| omniorb4 | 4.0.6-2 | 66 |
| python2.1-omniorb2 | 2.6-2 | 66 |
Packages with negative delays:
| Name | Version | Delay |
|---|
| openvpn | 2.0.5-1 | -4 |
| liblog-log4perl-perl | 1.01-1 | -10 |
| libgnomedb2-4 | 1.2.1-4 | -13 |
| ifenslave-2.4 | 0.07+2.5.15-7 | -13 |
| xtoolwait | 1.3-5 | -13 |
| xrootconsole | 1:0.5-1 | -13 |
| gkrellmitime | 1.0.1-2 | -13 |
| libmdsp-dev | 0.11-7 | -13 |
| sg-utils | 1.02-1 | -13 |
| unicon-im | 3.0.4-9 | -13 |
| gimp-dcraw | 1.21-1 | -13 |
| libgsm-tools | 1.0.10-13 | -13 |
| freewnn-common | 1.1.0+1.1.1-a020-1 | -13 |
| airstrike-common | 0.99+1.0pre6a-3 | -13 |
| cdrdao | 1:1.1.9-3 | -13 |
| libguilegtk-1.2-dev | 0.31-3 | -13 |
| tea | 10.1-1 | -13 |
| libsyck0-dev | 0.55-3 | -13 |
| octave-forge | 2005.06.13-2 | -13 |
| gstreamer0.8-misc | 0.8.11-2 | -13 |
| i2c-2.4.27-2-386 | 1:2.9.2-1 | -13 |
| zope-cmfplone | 2.1-2 | -13 |
| spew | 1.0.5-1 | -13 |
| xfce4-mcs-manager-dev | 4.2.2-1 | -13 |
| libgeda-doc | 20050820-1 | -13 |
| mono-classlib-2.0 | 1.1.9.2-1 | -13 |
| leaktracer | 2.4-5 | -13 |
| libibtk0 | 0.0.14-10 | -13 |
| gstreamer0.8-oss | 0.8.11-2 | -13 |
| libxalan18c2 | 1.8-5 | -13 |
| libwxbase2.4-dev | 2.4.4.1.1 | -13 |
| xfont-nexus | 0.0.1-3 | -28 |
SQL queries
insert into movement select t.pid,t.testing,t.unstable from (select distinct
l1.package_id as pid, min(l1.day) as testing, min(l2.day) as unstable from lifetimes as l1 left join lifetimes as l2 on
l1.package_id=l2.package_id and l1.archive_id = 11 and l2.archive_id = 4 where l1.archive_id is not null and
l2.archive_id is not null group by l1.package_id) as t;
create table delay (name varchar(64), version varchar(64), delay int);
insert into delay select u.name,v.version,datediff(m.unstable,m.testing) as delay from movement as m, units as u, packages as p, versions as v where p.id = m.package_id and v.id = p.version_id and u.id = p.unit_id order by delay desc;
select min(delay),avg(delay),max(delay) from delay;
Comments: 0