" two days median time" is misleading.
" two days median time" is misleading.
Posted Oct 29, 2025 21:52 UTC (Wed) by ballombe (subscriber, #9523)Parent article: Debian splits ftpmaster team
FTP masters approval is required each time a new binary package name is created.
There are roughly three cases where this happens.
1) when a new source package is introduced.
2) when an preexisting source package generates a new binary package.
3) when an preexisting binary package is renamed.
The case 3) happens any time a shared library soname change, since the soname is part of the package name.
The FTP-masters prioritize (for good reasons) cases 2) and 3) since the source package was already part of Debian.
However this completely skews the statistics.
Saying that 'the median time for packages in the new queue is less than two days' is misleading since this is the combined time for 1)+2)+3) where 2)+3) is much faster than 1).
There are roughly three cases where this happens.
1) when a new source package is introduced.
2) when an preexisting source package generates a new binary package.
3) when an preexisting binary package is renamed.
The case 3) happens any time a shared library soname change, since the soname is part of the package name.
The FTP-masters prioritize (for good reasons) cases 2) and 3) since the source package was already part of Debian.
However this completely skews the statistics.
Saying that 'the median time for packages in the new queue is less than two days' is misleading since this is the combined time for 1)+2)+3) where 2)+3) is much faster than 1).
Note that the full quote from Matthias is
""
* I have learned (thanks @roehling) that the *actual* median time packages spend in NEW is less than two days. In other words, *somebody* must have *some* time available.
""