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

Re: State of Debian blends, notes and questions on debian-ezgo

At bottom :-

On 14/02/2019, Franklin Weng <franklin@goodhorse.idv.tw> wrote:
> Andreas Tille <tille@debian.org> 於 2019年2月15日 週五 00:07寫道:
>> Hi Franklin,
>> > Rewriting the PhET package hasn't started yet.  I'm considering to ask
>> > if
>> > University of Colorado would like to have such packages or not.  At
>> > least
>> > I'll get newest offline version as they have completed many new HTML5
>> > simulations recently.
>> Whatever you do about PhET - it should definitely not be part of the
>> ezgo source package but rather a separate package.
> Originally ezgo-phet is a empty package to download a zip file in the
> postinst script.  But the download ftp site was not stable.
> So I'm considering to make another deb package for different languages
> offline version and see if it can be published from USC _OR_ from my own
> ezgo repository, which is mirroring to some other places everyday.  Then
> the ezgo-phet in blends is still a meta package.
> Franklin

Hi all,

Please CC me as I'm not subscribed to the list.

@Franklin - While I'm not sure whether you know the reason debian
dislikes large tarballs or everything in one repo. kinda scenario. As
a debian user I can share from just a user POV why it makes sense to
have anything in small sections.

A similar example I could give is 0ad [1] . If you look at the 0ad
package today in debian you would see it is split into various
sub-parts as can be seen below -

$ aptitude show 0ad
Package: 0ad
State: installed
Automatically installed: no
Priority: optional
Section: games
Maintainer: Debian Games Team <pkg-games-devel@lists.alioth.debian.org>
Architecture: amd64
Uncompressed Size: 20.3 M
Depends: 0ad-data (>=, 0ad-data (<=,
0ad-data-common (>=, 0ad-data-common (<=,
libboost-filesystem1.67.0, libboost-system1.67.0, libc6 (>= 2.27),
libcurl3-gnutls (>= 7.16.2), libenet7, libgcc1 (>= 1:3.4), libgl1,
libgloox17, libicu63 (>= 63.1-1~), libminiupnpc17 (>= 1.9.20140610),
libnspr4 (>= 2:4.9.2), libnvtt2, libopenal1 (>= 1.14), libpng16-16 (>=
1.6.2-1), libsdl2-2.0-0 (>= 2.0.8), libsodium23 (>= 1.0.14),
libstdc++6 (>= 5.2), libvorbisfile3 (>= 1.1.2), libwxbase3.0-0v5 (>=
3.0.4+dfsg), libwxgtk3.0-0v5 (>= 3.0.4+dfsg), libx11-6, libxcursor1 (>
1.1.2), libxml2 (>= 2.9.0), zlib1g (>= 1:1.2.0)
PreDepends: dpkg (>= 1.15.6~)
Description: Real-time strategy game of ancient warfare
 0 A.D. (pronounced "zero ey-dee") is a free, open-source,
cross-platform real-time strategy (RTS) game of ancient warfare. In
short, it is a historically-based war/economy game that allows players
to relive or rewrite the history of Western civilizations, focusing on
the years between 500 B.C. and 500 A.D. The project is highly
ambitious, involving state-of-the-art 3D graphics, detailed artwork,
sound, and a flexible and powerful custom-built game engine.
Homepage: http://play0ad.com/

a. So the first thing is many people in today's environments use
metered connections. Me and some of the people to whom I provide the
service have some sort of metered connections. In such a scenario, it
is much more easier to install few packages and over time update and
install the whole package without dedicated network bandwidth.

b. One of the probably most under-appreciated but probably one of the
more widely used tool is debdelta. I have written about it in 2011 [2]
. If you have packages which are split it would be easier to use
debdelta on them rather than on the whole.

c. Re-using libraries -  This is best explained in Raphael's book
chapter 5 [3] . It makes sense to have shared libraries and split
packages so its easier to maintain. Bug-reporting, bug-triaging and
maintainance in both short, medium and long-term would bd easier. I
hope I was able to shed some light. It is possible that you may
already have known this, if so please excuse for the noise.

@Andreas - thank you for sharing the link about the debconf 13 talk.
It was good because it gave me some insight but not enough as to what
came first i.e. the idea of debian pure blends or Debian med ?

I am/was looking at the Debian Pure blend page history [4] and the
time-stamp of that says 2008 while Debian Med seems to have much more
longer history [5] circa 2004. I see you joined the team or at least
started contributing via wiki since 2007 [6] . If there is a
possibility of giving some insight of what started before it would
make it easier for me to share the same in a chronological order. I do
have some questions for debian-med but would probably ask them in the
debian-med mailing list as it would be more pertinent there.

Looking forward to info.

1. https://tracker.debian.org/pkg/0ad
2. https://flossexperiences.wordpress.com/2011/05/30/debian-and-deb-delta/
3. https://debian-handbook.info/browse/stable/sect.source-package-structure.html
4. https://wiki.debian.org/DebianPureBlends?action=recall&rev=1
5. https://wiki.debian.org/DebianMed?action=recall&rev=1
6. https://wiki.debian.org/DebianMed?action=recall&rev=42

          Shirish Agarwal  शिरीष अग्रवाल
  My quotes in this email licensed under CC 3.0
EB80 462B 08E1 A0DE A73A  2C2F 9F3D C7A4 E1C4 D2D8

Reply to: