Subject: Re: softfloat - IEEE ready yet?
To: Ross Harvey <>
From: Thomas T. Thai <>
List: port-alpha
Date: 01/25/2001 00:38:57
On Wed, 24 Jan 2001, Ross Harvey wrote:

> Date: Wed, 24 Jan 2001 22:11:28 -0800 (PST)
> From: Ross Harvey <>
> To:
> Cc:
> Subject: Re: softfloat - IEEE ready yet?
> > From: "Thomas T. Thai" <>
> >:::
> > your kind and considerate words are appreciated...
> I suppose I could have said "it's not done yet", but you already knew that,
> before you posted the same question twice. I suppose I could have said "I
> don't know when it will be done". I guess, after this, that I won't be
> posting any more estimates of feature completions.

ross, you just dropped my jaw to the floor with your initial response. i
was stunned to say the least.

> The only other possible response would consist of excuses, of which I have
> several, including a distant funeral, an unanticpated deadline at work, HW
> failures in my two fastest alphas, both of which were required for the testing
> of a platform-specific feature like this -- but excuses originally seemed
> rather dishonerable and I wanted to avoid making them.

i hope things work out for you. priority is priority and i wouldn't
say IEEE is a priority. if i knew i wouldn't have even asked.

> I don't believe these arguments justify my nuking you over your question,
> but please drop the innocent-and-offended act. You might want to think
> about why, *exactly*, you *really* posted those two tweaky notes.

i've repeatedly stated exactly why. there isn't a stable RDMS for our
NetBSD/Alpha. i tested MySQL for a length of time, and even tried to work
with the developers from MySQL to get it to run stable on NetBSD/Alpha as
it does with Linux and FreeBSD. however, they repeatedly stated that i
should compile with the latest release of GCC. unfortunately, that version
in our package collection fails to compile on NetBSD/Alpha.

next i tested postgresql. the one from package collection failed many
regression tests. upon working with the postgresql developers, we got rid
of all but two problems which seemed to be due to floating point. at that
point i posted the initial message about IEEE.

innocent-and-offended act? no, i was just simply stunned.

to conclude this thread, i thank you again, for your wonderful
contributions for NetBSD -- of which i and many others have benefitted.