Building Angstrom...

Hey Robert,

I need to rebuild my Angstrom development system.

Did you finish reviewing and updating the process?

Thanks!
- dan

which robert are you addressing?

rday

Well, I thought that I picked your email address out of a message but
apparently it just went to 'beaglebord' instead!

Oops...

- dan

ok, enough of that top posting ...

Well, I thought that I picked your email address out of a message
but apparently it just went to 'beaglebord' instead!

  currently (and as i just posted to the OE devel list), i tried a
build of beagleboard-demo-image with the latest git pull of the OE
devel branch and there are two non-building packages:

  * gnome-games, which is apparently a known issue, and
  * e2fsprogs, which no one else has been able to reproduce

this is on fedora 12. so, what exactly were you asking in terms of an
angstrom build? i try the development build on a regular basis and,
when things break, i report them. it's been a few days since i've had
a successful build.

rday

ok, enough of that top posting ...

I second that.

[…]

so, what exactly were you asking in terms of an angstrom build?

I think he refers to your work to update the Wiki page to validate the
BeagleBoard?

Just my guess to add more to the confusion. :wink:

Thanks,

Paul

From: beagleboard@googlegroups.com
[mailto:beagleboard@googlegroups.com] On Behalf Of Robert P. J. Day
Sent: Saturday, December 19, 2009 5:10 AM
To: beagleboard@googlegroups.com
Subject: Re: [beagleboard] Re: Building Angstrom...

  ok, enough of that top posting ...

> Well, I thought that I picked your email address out of a message
> but apparently it just went to 'beaglebord' instead!

  currently (and as i just posted to the OE devel list), i tried a
build of beagleboard-demo-image with the latest git pull of the OE
devel branch and there are two non-building packages:

  * gnome-games, which is apparently a known issue, and
  * e2fsprogs, which no one else has been able to reproduce

this is on fedora 12. so, what exactly were you asking in terms of an
angstrom build? i try the development build on a regular basis and,
when things break, i report them. it's been a few days since i've had
a successful build.

I'm building beagleboard-demo-image on Ubuntu 9.10 x64. I started recording
the Shar1 for each build so that I could revert back to a successful build
when something breaks. Here is an example of my builds.

916ed3c..1e7e232 org.openembedded.dev -> origin/org.openembedded.dev - Nov
15 - Builds
1e7e232..2eb9947 org.openembedded.dev -> origin/org.openembedded.dev - Nov
16 - Builds
2eb9947..0b1f939 org.openembedded.dev -> origin/org.openembedded.dev - Nov
17 - ecore broken
0b1f939..b252fee org.openembedded.dev -> origin/org.openembedded.dev - Nov
18 - Builds
b252fee..9bd7891 org.openembedded.dev -> origin/org.openembedded.dev - Nov
18 - ecore broken
9bd7891..2064416 org.openembedded.dev -> origin/org.openembedded.dev - Nov
19 - Builds
2064416..849cd3a org.openembedded.dev -> origin/org.openembedded.dev - Nov
20 - Builds
849cd3a..238eb99 org.openembedded.dev -> origin/org.openembedded.dev - Nov
20 - Builds
238eb99..18a377d org.openembedded.dev -> origin/org.openembedded.dev - Nov
22 - Builds
18a377d..dfc7ab3 org.openembedded.dev -> origin/org.openembedded.dev - Nov
23 - GCC broken
dfc7ab3..a121173 org.openembedded.dev -> origin/org.openembedded.dev - Nov
24 - Builds
a121173..932a9ec org.openembedded.dev -> origin/org.openembedded.dev - Nov
25 - Builds
932a9ec..e6c9a97 org.openembedded.dev -> origin/org.openembedded.dev - Nov
26 - Builds
e6c9a97..ce34c40 org.openembedded.dev -> origin/org.openembedded.dev - Nov
29 - Builds
Bump to 2009.5 so complete rebuild required
ce34c40..d7b29de org.openembedded.dev -> origin/org.openembedded.dev - Dec
2 - task-beagleboard-demo missing
d7b29de..dd2e82a org.openembedded.dev -> origin/org.openembedded.dev - Dec
3 - Builds
dd2e82a..e8e9c9c org.openembedded.dev -> origin/org.openembedded.dev - Dec
5 - Builds
e8e9c9c..bc563bc org.openembedded.dev -> origin/org.openembedded.dev - Dec
6 - Builds
bc563bc..ff16088 org.openembedded.dev -> origin/org.openembedded.dev - Dec
8 - Builds
ff16088..c561945 org.openembedded.dev -> origin/org.openembedded.dev - Dec
9 - Builds
c561945..5a04169 org.openembedded.dev -> origin/org.openembedded.dev - Dec
9 - Builds
5a04169..9f65815 org.openembedded.dev -> origin/org.openembedded.dev - Dec
13 - mesa-dri broken
9f65815..27e3eff org.openembedded.dev -> origin/org.openembedded.dev - Dec
14 - Builds
27e3eff..47d6cb2 org.openembedded.dev -> origin/org.openembedded.dev - Dec
15 - Builds
47d6cb2..a68775b org.openembedded.dev -> origin/org.openembedded.dev - Dec
16 - Builds
a68775b..9c0a5ca org.openembedded.dev -> origin/org.openembedded.dev - Dec
18 - Builds

I have an i7 975 quad processor machine with 6G or RAM and 5TerraBytes of
disk configured as a RAID 6, which allows me to build beagleboard-demo-image
from scratch in about 1 hour. I have found that sometimes when a build
fails, I delete the complete build folder and build again, then the build
succeeds. If I get the same error, then I wait a day and try to build again.