> > I wouldn't say that nobody cares about your precious coordination, but that
> > people (including me) often fails in packing, documenting and releasing the
> > work they have done. A bit more effort in this would help a lot.
OK, well. As Milo has already been mentioned - what patches for Milo yet
unreleased? From my side it's only minor stuff. Milo 0.25 still works
fine with my current kernel.
> Perfectly right. It doesn't help that someone fixes his specific
> problems. The solution must be made available, and in such a way
> that it does not produce new problems.
> > See above - Since i'm kinda stalled with the video board i'll release the
> > crossdev that i built, unless there are newer versions on theyr (close)
> Sorry, I don't remember which version you run. I'm running gcc-2.7.0-2
> or so, but I'm never sure if it's really the *right* version. It would
> be way better if Ralf would release binaries since he's the one that
> knows what can go wrong. Imagine H.J.Lu would have thrown gcc sources
> after early Linux users. Do you think Linux would have had the same success?
> It is not the question wether we're developers and able to build our
> own compilers or not. It's necessary to have a reference to make sure
> that certain problems do not depend on "personal" compiler bugs!
You're fine with gcc-2.7.0-2, gcc-2.7.1-1. The old gcc-2.6.3-whatever should
also still be working though I'd not bet on that.
> > You're right. I would categorize the editor into this kind of applications,
> > anyway :).
> Right. And we're lightyears away from an editor!
Not at all. I've got a port of the GNU libc-951116 more or less working.
Various kind of applications do compile. I just need to do some minor fixes
and this thing could be more or less working. If only the damn streamer'd
read that tape I could even upload the stuff ...
Why I hesitated to do uploads in the last time? Well, I was heavily modifying
kernel interfaces to be something conforming to some standard. I'm pretty
shure you people wouldn't like to update libc and the kernel daily :-) The
tricky thing is that designing these basic system interfaces is a once in a
lifetime chance. If you bug it, ... The truly evil thing is that I have more
or less no documentation about all the stuff. Guess what fun it is to
disassemble the kernels & libs of a certain company (I won't call names :-)
which is well known for it's graphic computers ... A view things like
complete ABI (not the cutdown version of www.mipsabi.org) or even deeper
documentation could speed up the whole mess a *lot*.
Apropos speed - credits go to Wayne Hodgen for borrowing me his whistles and
bells 586 to me. Without it I'd not be where I am now.
> > Gee, if this crazyness will take me the right way you will call me Paolo
> > Bevilacqua!
> Take a look at the subscriber list: There is already a Drew :-)
> Ask him. Perhaps he can help you!
I guess Drew should take that as a compliment :-)