Provenance of Binder?

[Shmuel Metz]

Was the Binder part of the original Multics design? If not, when was
it introduced. Is there an online reference containing those data?
This query was prompted by Message-ID: <kb0bq2$51u$1...@dont-email.me> in
alt.folklore.computers.

--
Shmuel (Seymour J.) Metz, SysProg and JOAT <http://patriot.net/~shmuel>

Unsolicited bulk E-mail subject to legal action. I reserve the
right to publicly post or ridicule any abusive E-mail. Reply to
domain Patriot dot net user shmuel+news to contact me. Do not
reply to spam...@library.lspace.org

[Daiyu Hurst]

On Dec 22, 5:46 pm, Shmuel (Seymour J.) Metz

<spamt...@library.lspace.org.invalid> wrote:
> Was the Binder part of the original Multics design? If not, when was
> it introduced. Is there an online reference containing those data?

It's described in "Introduction and Overview of the Multics System" by

F. J. Corbató
Massachusetts Institute of Technology
Cambridge, Massachusetts
and
V. A. Vyssotsky
Bell Telephone Laboratories, Inc.
Murray Hill, New Jersey

as presented at the 1965 Fall Joint Computer Conference.

You can see that paper here:

http://www.multicians.org/fjcc1.html

The source for bind_.pl1 includes the comments:

Designed and initially coded by Michael J. Spier, August 19, 1970,
Completely revised by Michael J. Spier, December 23, 1971 */

and was revised as recently as 1986 (I'm looking at the sources from
MR11
and MR12.3).

-dai

[John W Gintell]

On 12/22/12 9:52 PM, Daiyu Hurst wrote:
> On Dec 22, 5:46 pm, Shmuel (Seymour J.) Metz
> <spamt...@library.lspace.org.invalid> wrote:
>> Was the Binder part of the original Multics design? If not, when was
>> it introduced. Is there an online reference containing those data?
>
> It's described in "Introduction and Overview of the Multics System" by
>

> F. J. Corbat�

> Massachusetts Institute of Technology
> Cambridge, Massachusetts
> and
> V. A. Vyssotsky
> Bell Telephone Laboratories, Inc.
> Murray Hill, New Jersey
>
> as presented at the 1965 Fall Joint Computer Conference.
>
> You can see that paper here:
>
> http://www.multicians.org/fjcc1.html
>
> The source for bind_.pl1 includes the comments:
>
> Designed and initially coded by Michael J. Spier, August 19, 1970,
> Completely revised by Michael J. Spier, December 23, 1971 */
>
> and was revised as recently as 1986 (I'm looking at the sources from
> MR11
> and MR12.3).
>
> -dai
>

In the early days of Multics: 1967 - 1969 there was no binder and that was a
huge cause of a slow system. Many more linkage faults were needed of course
during initialization of the supervisor and every time a new command was used.
Also many more page faults, particularly since many procedures were much smaller
than one page so binding a few things together reduced memory usage and page
faults. Also for each procedure segment there were two segments since the
linkage data had to be in a different segment than the procedure since it wasn't
pure. Another big change was the "combined linkage segment" so that all the
linkage (and static storage) for a process was combined into one segment. This
essentially divided the number of segments needed by 2. This was done before the
binder was created.

[Daiyu Hurst]

On Dec 23, 12:58 am, John W Gintell <j...@gintell.org> wrote:
> On 12/22/12 9:52 PM, Daiyu Hurst wrote:
>
>
>
>
>
>
>
>
>
> > On Dec 22, 5:46 pm, Shmuel (Seymour J.) Metz
> > <spamt...@library.lspace.org.invalid> wrote:
> >> Was the Binder part of the original Multics design? If not, when was
> >> it introduced. Is there an online reference containing those data?
>
> > It's described in "Introduction and Overview of the Multics System" by
>

> > F. J. Corbató

So the term bind(ing) as used in the Overview paper refers instead to
dynamic linking?

Or is it looking ahead?

-dai

[John W Gintell]

On 12/23/12 9:33 AM, Daiyu Hurst wrote:
> On Dec 23, 12:58 am, John W Gintell <j...@gintell.org> wrote:
>> On 12/22/12 9:52 PM, Daiyu Hurst wrote:
>>> On Dec 22, 5:46 pm, Shmuel (Seymour J.) Metz
>>> <spamt...@library.lspace.org.invalid> wrote:
>>>> Was the Binder part of the original Multics design? If not, when was
>>>> it introduced. Is there an online reference containing those data?
>>
>>> It's described in "Introduction and Overview of the Multics System" by
>>

>>> F. J. Corbat�

I just sped read the Overview paper and a couple of the others and do think that
they meant dynamic linking. Not yo say that the idea of a static binder hadn't
been thought about in the conceptual stages of the project, but I doubt it. (I
started working on the project in February of 1967; nothing worked yet at that
time).