[email protected] (John Gilmore) writes:
> The elephant in the room is being studiously ignored.
>
> The crucial objection to C's nul-delimited strings of 'conceptually
> unlimited' length has so far gone unmentioned here.  They have been
> the all but exclusive foci of security breaches, thousands of them, of
> systems written in C/C++ and their dialects.


I had exceeded my quota on null-delimited string rants for the month.

recent posts about IBM research review of Multics security ... written
in PLI and not having any of the string related issues epidemic
in c-language implementation
http://www.garlic.com/~lynn/2013h.html#35 Some Things Never Die
http://www.garlic.com/~lynn/2013i.html#11 EBCDIC and the P-Bit

also, original mainframe tcp/ip stack was implemented in vs/pascal and
had none of the string related issues epidemic in c-language
implementations
http://www.garlic.com/~lynn/2013h.html#34 CLECs, Barbara, and the Phone Geek
http://www.garlic.com/~lynn/2013h.html#80 Minicomputer Pricing
as well as
http://www.garlic.com/~lynn/2013i.html#11 EBCDIC and the P-Bit

lots of past posts on the subject
http://www.garlic.com/~lynn/subintegrity.html#buffer

for various reasons the original vs/pascal tcp/ip implementation had
some performance issues ... only getting about 44kbytes/sec throughput
using full 3090 processor. I did the rfc1044 changes and in some tuning
tests at cray research between 4341 and cray ... got sustained channel
speed using only modest amount of 4341 processor (possibly 500 times
improvement in bytes moved per instruction executed) ... some past posts
http://www.garlic.com/~lynn/subnetwork.html#1044

but not as bad as the subsequent implementation that the communication
group contracted for a tcp/ip stack support in vtam. Initial demo had
tcp much faster than lu6.2. The communication group told the contractor
that everybody knows that for a valid tcp/ip implementation, lu6.2 is
much faster than tcp ... and they would only be paying for a *valid*
implementation.

-- 
virtualization experience starting Jan1968, online at home since Mar1970

----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN

Reply via email to