[ih] Dan Lynch has passed away

[Wayne Hathaway]

When I worked for Auspex Systems, I added code to our devices to automatically take and email me memory dumps on system failure.  So often I would know of a problem, debug it, and download a fix long before the clients even reported the reboot.  But as Dave noted, often times it was some other device that was causing the issues with a faulty implementation (usually NFS rather than TCP/IP), so occasionally I would start my email to the customer’s IT people with something like “When did you install the machine at 192.36.72.14?  it is doing so-and-so in violation of the NFS protocol spec and it is causing problems for others on your network.”  The responses were often hilarious.  😊


Wayne Hathaway
wayne at Ames-67 in the good old days

On Mon, 1 Apr 2024 09:38:35 -0700, Dave Crocker via Internet-history <internet-history at elists.isoc.org> wrote:

>> On 4/1/2024 9:23 AM, vinton cerf via Internet-history wrote:
>> > We also did TCP interop tests - whoever broke the
>> > most TCPs got a bottle of champagne from yours truly.
>>
>>
>> An anecdote about interop testing outside of a planned event:
>>
>> Roughly 1987. At Ungerman-Bass we'd put TCP onto the the company's
>> 'intelligent' Ethernet card.  The card had a 80186 processor, whicht was
>> more powerful than the chip on many PCs. Our Telnet and FTP ran on the PC.
>>
>> We'd been deploying for some time, with no interoperability problems. 
>> Then one day I get a call from Boeing which was testing our product. 
>> (Boeing was big on OSI in those days.  This group was off in a corner of
>> the company.)  They reported that our telnet was not working against
>> another company's telnet server.
>>
>> We had a primitive packet-capture tool and I had them use it to record
>> the session.  On reviewing this I found that the other company's product
>> was doing the telnet protocol only to accept the connection and it then
>> just handed the session over to the regular terminal handler, with no
>> protocol engine mediating.
>>
>> Unfortunately, our telnet client was aggressive and initiated some
>> options.  But the requests were just treated as user data ans were
>> getting echoed back to our telnet client.
>>
>> I called the customer up and went through the details with them. They
>> had been working with us for awhile and were technically quite
>> competent.  So I was thrown off-balance when their response to my
>> showing them what was happening was to asked what /we/ were going to do
>> to fix it.
>>
>> I repeated the summary of the problem, noting that the other company's
>> product was in completely violation of the telnet spec and Boeing should
>> get them to change it.
>>
>> Their response was that they understood this and would certainly talk
>> with the other company.  But ours was the product causing the problem --
>> the other company's product worked fine with everyone else --  AND... we
>> were more responsive...
>>
>> I'd never thought of interoperability testing as a having marketing
>> potential until that point.
>>
>> Dan, on the other hand, saw this from the start.
>>
>>
>> d/
>>
>> --
>> Dave Crocker
>> Brandenburg InternetWorking
>> bbiw.net
>> mast:@dcrocker at mastodon.social
>> --
>> Internet-history mailing list
>> Internet-history at elists.isoc.org
>> https://elists.isoc.org/mailman/listinfo/internet-history