[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: volatile.debian.org: Does Debian still support it?



You made a subtle mistake. You checked the source package for tzdata
and the changes are really small. Then you checked the differences for
libdatetime-timezone-perl package, but this source package is already
compiled. You should check the difference between binary packages for
tzdata.

The real source data for DateTime::TimeZone is the same as for tzdata
source package. Then, the data is precompiled into Perl source. It
makes a Debian source package.

Did you really read this patch?

It contains the series of such differences:

diff -Nru DateTime-TimeZone-0.42/lib/DateTime/TimeZone/Africa/Addis_Ababa.pm
DateTime-TimeZone-0.42
--- DateTime-TimeZone-0.42/lib/DateTime/TimeZone/Africa/Addis_Ababa.pm
 2006-02-20 16:45:58.0000000
+++ DateTime-TimeZone-0.42.2007g/lib/DateTime/TimeZone/Africa/Addis_Ababa.pm
   2007-09-05 14:29:11
@@ -3,7 +3,7 @@
 # DateTime::TimeZone module distribution in the tools/ directory

 #
-# Generated from /tmp/3PactztUqR/africa.  Olson data version 1
+# Generated from tzdata/africa.  Olson data version 2007g
 #
 # Do not edit this file directly.
 #
@@ -50,7 +50,7 @@

 sub has_dst_changes { 0 }

-sub _max_year { 2016 }
+sub _max_year { 2017 }

 sub _new_instance
 {

If you multiply this chunk by 400 timezones, you can get 40K of patch.
Another patch changes test units so "make test" can be passed with new
data. Also I provided a shell script which precompiles tzdata source
into Perl source so updating is quite simplier.

As I sad previously, the timezones data are precompiled into Perl
source because this library runs on non-Linux platforms, too, and in
this case it has a meaning.

You just compared an apples with oranges.

I understand, it's not comfortable situation with updates, but this
library already exists, and is used by many Perl projects. I see you
proposed to replace this library with other, which doesn't provide own
timezone data, but it is too late and not possible just now.

2007/9/25, Martin Zobel-Helas <zobel@ftbfs.de>:
> So where are 38k lines here, which changed?
>
> In general i have no problem to accept a package into volatile, if one can give
> me GOOD reasons, why all these changes are necessary. What i do not understand
> is, why tzdata has about 300 lines changed, while your package needs 38k lines
> changed to achive the same effect.

-- 
 .''`.    Piotr Roszatycki
: :' :    mailto:Piotr.Roszatycki@gmail.com
`. `'     mailto:dexter@debian.org
  `-



Reply to: