On Wed, Feb 08, 2006 at 09:36:19AM -0500, [EMAIL PROTECTED] wrote:
> Hi all;
>
> I added this bug tracker address to the CC list of the Savannah bug:
>
> https://savannah.gnu.org/bugs/index.php?func=detailitem&item_id=15182
>
> but for some reason followups did not get sent to the Debian BTS.
>
> Anyway, I wrote:
>
> A complete fix for this problem will have to wait until after 3.81 is
> released. I know how I want to solve it, but I'm trying hard to get 3.81
> out the door and really need to clamp down on changes. However, I'd like
> to fix this issue for gcj. As you've noted, the code in 3.81 has changed
> quite a bit in this area (but not because of the hash_* functions; those
> were used in 3.80 as well--the change is because the code has been
> rearranged to support the secondary expansion feature).
>
> So, let me ask a question: for this specific situation in gcj and GNU
> Classpath, are the dependency lines listed as they are in the makefile
> attached to this bug report? That is, like this:
>
> targ1 targ2 targ3 ... : prereq1 prereq2 prereq3 ...
>
> Or, are they broken out into multiple lines, like this:
>
> targ1: prereq1 prereq2 prereq3 ...
> targ2: prereq1 prereq2 prereq3 ...
> targ3: prereq1 prereq2 prereq3 ...
> ...
>
Neither. It looks like
targ1: prereq11 prereq12 prereq13 ...
targ2: prereq21 prereq22 prereq23 ...
targ3: prereq31 prereq32 prereq33 ...
...
The prereqXX list can be VERY LONG and there may be many duplicates
among prereq1X, prereq2X, prereq3X, ...
H.J.
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
