glibc bugs affecting Ethiopian

2007-10-08 Thread Bernardo Innocenti
John,

to enable the am_ET locale, we need a few glibc bugfixes
from F8.  Ulrich Drepper, the glibc maintainer, says a
backport would be quite invasive and highly recommends
updating instead.  Additionally, until now glibc wasn't
forked in OLPC-2.

I installed 2.6.90-17 and everything appears to work, but
of course it would need more testing.

So I propose we update glibc post Trial-3.  If that's not
acceptable, the alternatives I see are:

1) attempt a backport and fork glibc.  That would give us a
   good excuse to enable the Geode optimizations too.

2) Leave Ethiopian as an unsopported language in FD1.0 and
   tell people to follow the procedure on the wiki to add it
   manually.

-- 
 \___/
 |___|   Bernardo Innocenti - http://www.codewiz.org/
  \___\  One Laptop Per Child - http://www.laptop.org/
___
Devel mailing list
Devel@lists.laptop.org
http://lists.laptop.org/listinfo/devel


Re: glibc bugs affecting Ethiopian

2007-10-08 Thread John (J5) Palmieri
On Mon, 2007-10-08 at 14:46 -0400, Bernardo Innocenti wrote:
 John,
 
 to enable the am_ET locale, we need a few glibc bugfixes
 from F8.  Ulrich Drepper, the glibc maintainer, says a
 backport would be quite invasive and highly recommends
 updating instead.  Additionally, until now glibc wasn't
 forked in OLPC-2.
 
 I installed 2.6.90-17 and everything appears to work, but
 of course it would need more testing.
 
 So I propose we update glibc post Trial-3.  If that's not
 acceptable, the alternatives I see are:
 
 1) attempt a backport and fork glibc.  That would give us a
good excuse to enable the Geode optimizations too.
 
 2) Leave Ethiopian as an unsopported language in FD1.0 and
tell people to follow the procedure on the wiki to add it
manuall


This is a business decision.  How likely is a Ethiopian sale in the
short term?  Does it pay right now to get distracted by Ethiopian
support (other than planning for the future) or concentrate more on the
places we are shipping?  Talk to Uli and find out if he knows of any
interactions with other libraries.  If the risk is low and the benefits
high then go for it.  However for FRS 1.0 I suspect it would be best to
not make this change.

-- 
John (J5) Palmieri [EMAIL PROTECTED]

___
Devel mailing list
Devel@lists.laptop.org
http://lists.laptop.org/listinfo/devel


Re: glibc bugs affecting Ethiopian

2007-10-08 Thread John (J5) Palmieri
On Mon, 2007-10-08 at 14:58 -0400, Walter Bender wrote:
   How likely is a Ethiopian sale in the short term? 
 
 Ethiopia is going to be one of the launch countries, unless
 negotiations between the Prodi government and Ethiopia somehow go
 awry.
 
 -walter

In which case the decision is made and you should build and heavily test
using the F8 glibc if there is no other way to fix the issues with
Ethiopian support.

 
 On 10/8/07, John (J5) Palmieri [EMAIL PROTECTED] wrote:
 On Mon, 2007-10-08 at 14:46 -0400, Bernardo Innocenti wrote:
  John,
 
  to enable the am_ET locale, we need a few glibc bugfixes
  from F8.  Ulrich Drepper, the glibc maintainer, says a
  backport would be quite invasive and highly recommends 
  updating instead.  Additionally, until now glibc wasn't
  forked in OLPC-2.
 
  I installed 2.6.90-17 and everything appears to work, but
  of course it would need more testing.
  
  So I propose we update glibc post Trial-3.  If that's not
  acceptable, the alternatives I see are:
 
  1) attempt a backport and fork glibc.  That would give us a
 good excuse to enable the Geode optimizations too. 
 
  2) Leave Ethiopian as an unsopported language in FD1.0 and
 tell people to follow the procedure on the wiki to add it
 manuall
 
 
 This is a business decision.  How likely is a Ethiopian sale
 in the 
 short term?  Does it pay right now to get distracted by
 Ethiopian
 support (other than planning for the future) or concentrate
 more on the
 places we are shipping?  Talk to Uli and find out if he knows
 of any
 interactions with other libraries.  If the risk is low and the
 benefits 
 high then go for it.  However for FRS 1.0 I suspect it would
 be best to
 not make this change.
 
 --
 John (J5) Palmieri [EMAIL PROTECTED]
 
 ___ 
 Devel mailing list
 Devel@lists.laptop.org
 http://lists.laptop.org/listinfo/devel
 
 
 
 -- 
 Walter Bender
 One Laptop per Child
 http://laptop.org
-- 
John (J5) Palmieri [EMAIL PROTECTED]

___
Devel mailing list
Devel@lists.laptop.org
http://lists.laptop.org/listinfo/devel