Missing t2k.lib, critical bug in Windows binary plugs

[Phil Race]

Dan Fabulich wrote:
> Phil Race wrote:
>
>> replacing T2k will take months of hard effort but that is mostly 
>> behind us so its more like a few weeks
>> of effort .. and some inevitable build cycle time lag.
>
> Naturally, I have no idea how far we are along with replacing T2K with 
> unencumbered code.  I assume that Sun developers are working on that 
> in the Sun proprietary source control system, since the public OpenJDK 
> svn is read-only, even to Sun developers, right?

Actually the 2D openjdk component page documents the status of this work.
>
> If that's so, am I right in thinking that developers from the 
> community (such as myself) won't be able to build a Windows OpenJDK 
> until June, at the earliest?
With the current approach, that's a fair estimate.
>> And its more legally complicated too .. suppose we can't reship the 
>> binaries except as part of Java .. not sure of the answer to that but 
>> putting out the plug directory separately means more work for legal 
>> and that is a bigger bottleneck than might be apparent.
>
> It's surprising to me that t2k.dll, t2k.pdb, and t2k.map can all ship 
> in the binary plugs, but adding t2k.lib would be a big legal hassle.  

That it was mentioned as a possibility but we don't ship any .lib files 
there so I think
it would be an oddity. I meant shipping a 'zip' of the directory 
'closed' built by the closed-binaries-dir
as another download bundle. In other words the code to make a buildable 
windows JDK was
done, its just that the bits aren't built and shipped.

So this isn't an engineering/bug type issue .. its more a matter of as 
you say documenting it, or
getting it resolved that another bundle can be shipped or cleared.

-phil