SYSKEY, TNG freeze, 2.0.x->TNG merge and other thoughts

[Jean Francois Micouleau]

On Thu, 10 Feb 2000, Luke Kenneth Casson Leighton wrote:

> On Wed, 9 Feb 100 jeremy at varesearch.com wrote:

> > code in 2.0.x is much more reliable than the TNG code (not
> > the actual server/client functions though).
> 
> just because it's been running in production for two yeasrs doesn't mimply
> that uit's good code or does the correct job.

We're talking about different things. With Jeremy we're talking about the
prs_struct itself and the prs_init(), prs_grow(), prs_whatever()
functions. Jeremy rewrote them a year ago, between 2.0.0 and 2.0.2

And these are much better than in TNG. Or at least they don't LEAK MEMORY
like in TNG. Comprendo ?

> because i certainly don't trust any of the [very limited, small, alpha,
> first-version, development-quality] 2_0 rpc code -- server, marshalling
> _or_ client.

I can't stand that any more. You're totally over-exagerating. The diff is
not so big. It's a fact, I MADE MYSELF THE DIFF. 

OK, as I'm honest, you're right on the client code.

	J.F.