[chuck-users] chuck 1.2.1.2 still doesn't work on 64-bit linux

[Michael Heuer]

On Mon, Jul 21, 2008 at 1:09 PM, Stephen Sinclair <radarsat1 at gmail.com> wrote:
> On Mon, Jul 21, 2008 at 1:29 PM, Kassen <signal.automatique at gmail.com> wrote:
>> 2008/7/21 Stephen Sinclair <radarsat1 at gmail.com>:
>>>
>>> Sorry, just that this one keeps coming up, and is obviously a bug of
>>> interest to several people.  But a topic like "chuck still doesn't
>>> work on 64-bit linux" sounds like someone wondering why it hasn't been
>>> fixed yet.  And the answer is simply, "because no one's fixed it."
>>>
>>
>> Yes, I see. Still, I didn't read Michael's post that way, I read it more as
>> asking whether anybody has made any progress and implying Michael himself
>> had at least had some results using compatibility drivers (which I think is
>> news?).
>>
>> I think that even asking why something hasn't been fixed yet (though
>> probably not in those exact words) can be a constructive thing and need not
>> be a "complaint". For example; I have been asking from time to time why
>> Philippe's results in getting a ASIO build haven't yet been merged into the
>> main Windows version. To me that's not a complaint as such, there may be
>> good reasons, for example for a long time there was no suitable test system
>> at Princeton, there could be a lack of time.
>>
>> I think that in many cases figuring out where exactly we are stuck can lead
>> to people helping to move things forward. In the case of ASIO there was a
>> set of tests done with a relatively wide variety of ASIO-enabled soundcards,
>> I don't think those tests would've been done as publicly and in that way if
>> Ge hadn't pointed out ASIO was stuck partially because of a lack of testing
>> systems. With Ge's move to Stanford this could have changed, we'll never
>> know if nobody asks and if there are no requests it may be harder to place
>> priorities.
>>
>> Because of this I didn't feel it was benefitial to describe this post as a
>> "complaint"; I think we are all grateful for what *has* been done by those
>> generously donating their time. Michael's post itself didn't strike me as
>> particularly demanding (for lack of a better word) but I can see how the
>> subject title could give that impression.
>
> Yes, I intended to be blunt but not disrespectful.  I agree with you,
> but the 64-bit problem is one that tends to come up quite frequently.
> So, sorry for my negative reply, here's a more positive response:
>
> Has anyone had any success making portions of ChucK 64-bit-ready?
> I think taking things one small chunk at a time might be the way to
> go, though that could be easier said than done.

I gave it a try with the last source release, by adjusting the size of
the pointers in the main header file and fixing the compile errors.
That worked on a few more examples than the binary that is built by
default, but it still blew whenever certian library calls were made.
Unfortunately I am not much of a C programmer, and threw out whatever
I had accomplished before.

If anyone who knows what they are doing wants to jump in, I would be
very willing to help test it.

As far as the 32-bit compat libs, I thought someone on Ubuntu-64 had
gotten chuck to work?  I don't have any of those on my machine (gentoo
amd64/2008.0/no-multilib profile).

   michael