last email on this subject

[Maneesh Yadav]

<originially sent to tom..but I really think some people in the group need
to hear this>

Sigh...you guys will realize one day... I made it clear numerous times
that it was not my intent to be offensive... I put out some questions and
all I got were vauge responses that evaluted to "oh yeah I haven't really
thought it out yet, but we'll put that feature in, it'll be easy", I say
this as my final reponse in this thread, you really have no clue how
difficult it is to do what you are doing...I'm looking at your stuff
now...I don't see how you've taken care of simple stuff like tasks sharing
the keyboard, multi-processer safing the kernel, no general application
memory map, I don't see how applications are going to get at systems calls
even.  Have you taken into account what effect multiprocessing will have
on your OS in terms of data structures and protecting them?  And on shared
code such as device drivers? Have you given a sound analysis on your
choice of memory management algorithm or task management?  THis is the
simplest stuff, you really can't code anything until you have that laid 
out.

How is your persistent storage going to tie into VMM? And you're going to
port this to
all those wonderful architectures on your page with all that system
dependant code (the bulk of your code so far is in ASM)?  Distributed
computing too?  All this in some real time span? There is such a thing as
optimism and then there is simple ignorance of the fact.  I garauntee this
won't progress beyond a simple ukernel...I'm not trying to hurt anyone's
feelings, but I can't help but point out something obvious that
apparaently many people don't see. This type of work is not what TUNES
needs, but if you are going to make real new useful OS, you're not doing
it the right way. 

A note to Brian; I am sure it will take me at least till
the end of the summer to be able to give a reasonable opinion on
Arrow...but if you are in SD, maybe we can hook up, sometimes learning
from someone directly helps speed up the process of understanding.  I am
hoping to maybe hook up with Dr. Joseph Gougon and have a talk with him
about his work with CafeObj, hidden algebras etc.  as he is here at
UCSD...I sent him a letter of interest, no response yet, but we will see.