All of lore.kernel.org
 help / color / mirror / Atom feed
From: Martyn Welch <martyn.welch@ge.com>
To: openembedded-devel@lists.openembedded.org
Subject: Re: [RFC][PATCH v2] Alternative to busybox, using "proper tools"
Date: Tue, 23 Mar 2010 10:32:08 +0000	[thread overview]
Message-ID: <4BA898A8.4040701@ge.com> (raw)
In-Reply-To: <201003221733.58036.marcin@juszkiewicz.com.pl>

Marcin Juszkiewicz wrote:
> Dnia poniedziałek, 22 marca 2010 o 17:08:20 Martyn Welch napisał(a):
>
>   
>> This patch provides a mechanism to replace busybox by defining a busybox
>> package which actually just requires a set of other packages.
>>
>>     To get it to be used rather than the standard busybox, define:
>>
>>     PREFERRED_PROVIDER_busybox = "busybox-alt"
>> ---
>>
>>  Marcin: Something like this you mean?
>>     
>
> Yes, few notes:
>  
>   

No prob.

>> diff --git a/recipes/busybox/busybox-alt_1.bb
>> b/recipes/busybox/busybox-alt_1.bb new file mode 100644
>>     
>
> busybox-alt.bb is fine too - PV will be 1.0 then
>
>   
>> +++ b/recipes/busybox/busybox-alt_1.bb
>>     
>
>   
>> +ALLOW_EMPTY = "1"
>> +PACKAGES = "busybox"
>> +PACKAGE_ARCH = "all"
>> +
>> +PROVIDES = "busybox"
>> +DEFAULT_PREFERENCE = "1"
>>     
>
> "-1" is more proper - we do not want people to get busybox-alt instead of 
> busybox by mistake.
>
>   

Agreed - definitely left over from prototyping.

> RDEPENDS = "task-proper-tools"
>
> One line looks better. DEPENDS are not needed.
>
>   

Ok.

I understand that RDEPENDS are runtime dependancies. Am I right to
surmise that DEPENDS are configure/compile/packaging time dependences?

>> +++ b/recipes/tasks/task-proper-tools.bb
>> @@ -4,33 +4,107 @@ PR = "r11"
>>     
>
> Bump PR.
>
>   
>>  RDEPENDS_${PN} = "\
>> +		bash \
>> +		bc \
>> +		bind-utils \
>>     
>
>   
>> +		binutils \
>>     
>
> Not needed - 'binutils-symlinks' depends on it.
>
>   

Ok.

>> +		bzip2 \
>> +		console-tools \
>>     
>
>   
>> +		dhcp-client \
>>     
>
> I added 'pump' - but thats minor.
>
>   

As long as it works, I'm not fussed either way.

>> +		e2fsprogs-fsck \
>>     
>
> In .dev this is in util-linux-ng now.
>   

So I assume I can remove e2fsprogs from this list?

>   
>> +		e2fsprogs-libs \
>>     
>
> What for this is?
>
>   

Shared libraries for e2fsprogs. I'll remove this, I assume this is
either automatically added or not needed.

>> +		ifupdown-ubuntu \
>>     
>
> Can we rename it to 'ifupdown'?
>
>   

We added a recipe to our overlay which used the Ubuntu version of
ifupdown rather than using the Debian version already in OE. I'm afraid
I didn't create the original recipe for our overlay (one of my
colleagues did), my understanding is as follows:

The ifupdown tool does not have an active upstream maintainer, as a
result (at least Debian and Ubuntu) have differently patched copies of
version 0.6.8. The functionality provided by the Ubuntu version provided
a better drop in replacement for the required functionality than the
Debian version and hence we have used this one. As the version numbers
were the same, I created a "ifupdown-ubuntu" recipe for this package. To
ensure it was used here instead of the Debian version, I put
"ifupdown-ubuntu" in the RDEPENDS. Whilst we were prototyping this we
had a "PREFERRED_PROVIDER_ifupdown" entry in our local.conf to force
it's use.

If you think this is unwarranted, we can try using the Debian version
instead.

>> +		iputils-ping \
>> +		iputils-ping6 \
>> +		iputils-arping \
>>     
>
>
>   
>> +		minicom \
>>     
>
> Maybe 'picocom' would be nice too.
>
>   

Never heard of it - up to you.

>> +		mktemp \
>>     
>
>   
>> +		ncurses \
>>     
>
> I added "ncurses-tools" - gives few utils more.
>
>   

Ok.

>> +		openrdate \
>> +		realpath \
>>     
>
>
>   
>> +		start-stop-daemon \
>>     
>
> Nice - I added 'dpkg' instead but it was overkill.
>
>   
>> +		syslog-ng \
>> +		sysvinit \
>> +		tftp-hpa \
>> +		time \
>>     
>
> I added 'shadow' to the list.
>
>
>   
>> +DEPENDS_${PN} = "\
>>     
>
> DEPENDS are not needed.
>   

Martyn


-- 
Martyn Welch (Principal Software Engineer)   |   Registered in England and
GE Intelligent Platforms                     |   Wales (3828642) at 100
T +44(0)127322748                            |   Barbirolli Square, Manchester,
E martyn.welch@ge.com                        |   M2 3AB  VAT:GB 927559189




  reply	other threads:[~2010-03-23 10:32 UTC|newest]

Thread overview: 6+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2010-03-22 16:08 [RFC][PATCH v2] Alternative to busybox, using "proper tools" Martyn Welch
2010-03-22 16:33 ` Marcin Juszkiewicz
2010-03-23 10:32   ` Martyn Welch [this message]
2010-11-06 14:00     ` Stefan Schmidt
2011-04-16 15:13 ` [oe,RFC,v2] " Ben Gamari
2011-04-18 12:31   ` Martyn Welch

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=4BA898A8.4040701@ge.com \
    --to=martyn.welch@ge.com \
    --cc=openembedded-devel@lists.openembedded.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.