From: hunt@hal.gnu.ai.mit.edu () Subject: Re: X11R5 Re: [Bob Pendelton: Re: A question] Date: Thu, 23 Jan 1992 21:10:39 GMT
In article <1992Jan23.031330.2419@henson.cc.wwu.edu> n8141165@henson.cc.wwu.edu (gleason margaret) writes:
>keener@upenn5.hep.upenn.edu (Paul T. Keener) writes:
>
>>In article <32302@mentor.cc.purdue.edu>, chapinsj@mentor.cc.purdue.edu (Steve Chapin) writes:
>>> In article <1992Jan22.231937.3564@ctr.columbia.edu>, ejvette@pace3.cts.mtu.edu (Calvin) writes:
>>> >
>>> > I was led to believe that Linux was still socketless, and that X11 required
>>> > sockets to function at all. This was a good enough reason for the Minix
>>> > people to say that X could not be compiled on that OS. Am I wrong in either
>>> > of these statements?
>>>
>>> Yes and No. X requires a reliable byte stream. How you implement that is up
>>> to you. Of course, to run it "out of the box," it will want either TCP
>>> connections, unix domain sockets, or decnet connections.
>>>
>
>>Before this gets too far, there is a port of the mit X11R4 core to various PC unix
>>platforms and several VGA cards. The one main requirement that the README
>>specifies is that STREAMS must exist. I have included the README below.
>
>>
>> stuff deleted ...
>>
>
>No No No.
>Do not believe anything anyone will tell out about X running without TCP
>It will not. If they say so make them prove it.
>
>Frank Gleason
You obviously don't know what you're talking about. A couple months ago
I ran X11R4 with STREAMS (local) connections only without the use of TCP
on a machine with Interactive Unix.