On Fri, 2008-04-25 at 10:47 +0200, Martin Hollmichel wrote:
> How does
>
> http://www.openoffice.org/issues/show_bug.cgi?id=88700
> to-be-BEA300_m1: extensions and arch in Identifier
> fit into this context ?
To cut to the chase of what's causing a problem for us, we simply want
to be able to continue to do (for new extensions) what we can do now,
i.e. simply package an extension and in the package removal trigger have
e.g.
unopkg remove org.openoffice.legacy.writer2latex.uno.pkg
and that works regardless of the arch that the extension has been built
for. The linked-to blog says "Because the extension identifier
represents the 'logical extension' they must be the same for both
extensions." So that would seem to indicate that the basic extension
identifier as used by e.g. unopkg should be the same regardless of the
architecture, yes ?
i.e. we have an arch-independent identifier for the extension without
having to do something of the nature of...
uname -m
switch `uname -m`
case sparc:
unoarchname = "somethingA";
case i386:
unoarchname = "somethingB";
case x86_64:
unoarchname = "somethingC";
case ia64:
unoarchname = "somethingD";
case s390:
unoarchname = "somethingE";
case s390x:
unoarchname = "somethingF";
case arm-gnueabi:
unoarchname = "somethingG";
case arm:
unoarchname = "somethingH";
case ppc64:
unoarchname = "somethingI";
case ppc:
unoarchname = "somethingJ";
esac
unopkg remove org.openoffice.legacy.writer2latex-$unoarchname.uno.pkg
And that all the rest of the support for a "combined" identifier to
identify the platforms that the built extension supports is carried
around in additional fields in the description.xml so the extension
manager can find suitable extensions in the online repo to upgrade to.
C.
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
