2008 Face-to-Face Meeting
ACM History Committee February 2008 Face-to-Face Call
February 11, 2008, 8am to 4pm EST
Present: Bill Aspray, Mary Hall, Mike Mahoney, Pat Ryan (late afternoon; at SGB meeting the rest of the time), Len Shusek, Rick Snodgrass, and David S. Wise
- Bill, Mike, Pat, Len, Rick, and David had dinner the night before with Donna Cappo (ACM SIG Director), Wayne Graves (Director of Information Systems, Andrew McGettrick (chair of ACM Education Board), and John White (ACM CEO).
- Meeting with Eric Altman, chair of SIGMicro
We discussed possibilities. Eric had lots of questions, as did we (!).
- What is an oral history? How does it differ from an interview? What
is its cost? Who does it?
- Is SIGMicro interested in discovering the history of the SIG or the
history of the discipline?
- What was HOPL like? Does it make sense for SIGMicro to do something
- The IEEE History Center has an interest in history of the discipline,
as does the CHM.
It seems that a FAQ of how to go about thinking about what a SIG might do would be helpful. The HC will prepare such a FAQ.
We then went over to the SGB for mingling and an introduction at the start of their meeting. David introduced us and Mary gave a short summary of history activities, including a several-page handout.
We then came back to our meeting room.
We reviewed oral histories under consideration.
Several options were discussed, including the following.
- Expanding our charge to the history of the computing discipline
We felt that the topic was way too large and our resources way too
small to take this on.
- More focus on SIG institutional history
This is squarely in our court. While ACM values this activity, no one
else in ACM is positioned to do this task.
There exist historians who study organizations.
This relates directly to our mission of teasing out how ACM has impacted
the discipline of CS.
- The HOPL experience was very good.
Note that the HOPLs said nothing about how ACM or SIGPLAN has impacted
the discipline of CS. They were about the history of programming
languages irrespective of the involvement of either group. In some
cases, however, they had a lot to say about the role of other institutions.
If SIGs do want to move out to the history of their field, a HOx
delivers more bang for the buck than do interviews. First and foremost,
just choosing the topics and the speakers means thinking very hard
about what contributions shaped the field. If the group wants to
issue a call for papers, then again it has to specify the parameters
by which it will judge whether the subject is of historical importance
to the field. The critical care with which that is done will then
shape the critiques and revisions of the submissions and the direction
of the discussions at the sessions.
If SIGs continue to be interested in the history of their discipline,
perhaps they might want to take the HOx plunge. Possibilities
include SIGARCH, SIGART, SIGCHI, SIGDA, and SIGGRAPH. SIGOPS seems less
- The conversation should start with "what kind of question do you hope
to answer" from which appropriate approaches would then follow. (See FAQ,
Our conclusion was that we need to expand from the first part of our mission ("history of ACM") to do more on the second part (ACM's "role in the development of computing").
An organization may be thought of as an ecology, encompassing a tension between areas. A "discipline" is a shared agenda, a shared understanding og the questions and important problems. Disciplines can arise, split, and merge, as these shared agendas evolve over time.
We can focus on such events in part by studying closely how SIGs arose and evolved.
Meeting with Stu Feldman, ACM President
Stu shared his viewpoint.
- Let's not lose it all (history). We need to catch what is available,
- It is important to understand the implications and the impact that
ACM's activities have have on society.
- We can use history to impact the image of the field. It is hard to
underestimate the impact of compelling heroes (Stu's term) of the field.
- ACM was formative of the field. It is important to capture ACM's role
as shaper of the field, including turning-point issues.
- HOPL was very successful; other SIGs should produce their version.
- A history of the formation of SIGs would indicate coalescence of the
field (for example, the dot-com phenomenon in SIGecomm).
- There are very roughly 5-10M IT workers, whereas professional
societies total less than 200K members. The history of IT or IS is much larger
than the history of CS.
- ACM has been formative in the systematization and formalization of
industry innovations; it is important to capture this activity.
- The discipline needs to celebrate its heroes, and tell its stories.
It is better if the stories are historically accurate.
Meeting with Diana Marculescu of SIGDA
Diana presented an organized and ambitious list of possible initiatives for discussion and feedback from the HC.
- SIGDA was started in 1965, so 2010 is its 45th anniversary. It would
be good to celebrate at DAC (Design Automation Conference) and ICCAD
(International Conference on CAD).
- Educational activities, to feature anniversaries, such as travel
grants to students (10-year anniversary) and University booths (20-year
- Feature history on the DAC exhibit floor.
- Have speakers at the 2009 ICCAD member meeting.
- It doesn't seem useful to involve history in the Asian and European
- Perhaps a History of Design Automation (HDA) (or perhaps
HoDA or DAH) workshop/affiliated conference co-located with DAC.
- Interviews focusing on breakthroughs of the field.
- Wants to "do it right".
- Possibility of CHM loaning artifacts for exhibits.
We'll provide a SIGChairs History FAQ (see below) and will be available to help with any other history activity.
Meeting with John White, ACM CEO
- HC is increasingly visible and getting traction with the SIGs.
- Interviews with past presidents are important.
- We need more technical meetings on the history of the discipline (ala
(Mentioned History of Personal Workstations, 1986 that John arranged,
History of Scientific Computing and History of Medical Informatics).
Excited by possibility of HDA conference.
- HC needs to reach out to other parts of ACM, to leverage funding.
- Aiding collection of history of individual SIGs is an important
activity, as is history of individual boards
- Mentioned that while ACM is 80% pubs and conferences, most of his time
as CEO (and before as President) involves other issues, such as ethics,
public policy, gender diversity. Can get at these issues through
interviews of ACM presidents.
- It is important to attend to history in a more systematic way. Council
can benefit from knowing its history.
- The history of ACM chapters is an important story.
- It would be good to get the story on controversial issues, such
as human rights and the fellows program.
- It might be useful to gather an "official history," defined as
"authorized by ACM," with total access accorded, warts and all. It is
not clear that the cost/benefit ratio is favorable. [We also mentioned
the hidden costs, e.g., of access to staff and volunteers.] Companies
commission such official histories to provide a compelling story to
employees, customers, and vendors.
Meeting with Wayne Graves, ACM Information Systems Director
We had a detailed discussion of how the digital archive would work and formulated the following policy.
ACM Digital Archive Policy
Approved February 25, 2008
- The name of the digital archive is ftp://digitalarchive.acm.org
- The digital archive is separate from the record of an active unit,
which is stored online at http://acm.org, e.g., history.acm.org
In particular, access to the digital archive will be password-protected
and will not be served via http (except for a web page for manual deposit,
see below). The digital archive is designed as a storage area for material,
some of which will later be organized by an archivist and some of that
will later be served up on web pages, again, separate from the
digital archive. (As an example, the plan is to serve up conference
- The structure will be Type/Unit/Subunit/MajorItem. Initial units
- Individual Committees, e.g., Committees/HistoryCommittee/
- Individual SIGs, e.g., SIGs/SIGMOD/
- Boards, e.g., Boards/EducationBoard/
- Council and EC: Council/ and ExecutiveCommittee/
- HQ, e.g., HQ/CEO/, HQ/COO/, HQ/Pubs/
- Individual publications, e.g.,
Publications/Transactions/DatabaseSystems/ for TODS
- Conferences, e.g., Conferences/SIGMOD
Within Conferences will be years, and within years will be web site,
- Workshops, e.g., Workshops/AdvancesinModelBasedTesting/
- Periodic scrapings of the entire ACM web site, e.g.,
The scraping will be done by a program that will generate
HTML from server-side scripts such as PHP. Admittedly this
will not capture all the information in the "dark web" of
databases behind some ACM web pages.
- Conferences and workshops are managed separately from SIGs,
because some conferences are sponsored by multiple SIGs, and some workshops
are associated with multiple conferences, or not with a conference at all.
- Each directory should include a README file that describes the content
of that directory.
- Each directory will be "owned" by an "information director"
designated by the unit's chair or responsible officer. For example,
the information director for the Conferences/SIGMOD directory will be
designated by the SIGMOD chair, the Council information director by
the ACM President, and the HQ/CEO by the CEO.
- There will be guidelines provided to the SIG chairs, as
copyright verbiage for use on conference web sites, as verbiage
in the conference/workshop authorization forms, and as helpful advice,
e.g., on acquiring permissions.
SIG chairs and by delegation conference general chairs and webmasters
should ensure that ACM has full rights to publicly host all of the
conference/workshop website except for easily
identifiable content, such as papers, powerpoint, and documents authored
by other than the conference organizers or on subjects other than
the conference organization, such as tutorials or presentations.
SIG chairs will also be advised to manage the evolution of the conference
web site so that it is *cumulative*. For example, deadlines should be
retained even after they are passed.
The conference sponsorship and incooperation and proceedings workflows
will be augmented to (a) inform the SIG chairs of requirements on
conference web pages during the initial phase and (b) capture the
conference website by copying it into the digital archive.
- In some cases, e.g., conference and workshop web sites, material will
be automatically deposited in the digital archive as a component of
workflows managed by ACM staff. But there will also be a web form for
manually depositing material into the digital archive. This web form
will capture minimal provenance, including who is depositing the
material, a description of where the material is coming from, which
directory the deposit is going into, what form of permission is being
given, whether there are any restrictions on
distribution, and a description of the format and content of the deposit.
The date of deposit will be automatically captured.
- Later, a copy of the conference and workshop web sites will be served
by the ACM DL and linked from the associated papers in the DL. These
web sites will also be indexed by the DL.
If it turns out that there is material that shouldn't be served,
a take-down request will be honored. (In doing so, ACM is thus
free of liability.)
- The HC will coordinate with the SIGs to capture past conference
web sites, via the manual deposit web form.
- Wayne will help develop an index of the ACM Digital Archive for use
by information directors and will set up a separate mailing list
of Digital Archive Information Directors.
The HC then individually met with SIG chairs during lunch.
Meeting with Moshe Vardi, EiC of CACM
Moshe provided an overview of the new CACM. The model is Science and Nature, with sections on news, viewpoints, practice, reviewed articles, contributed articles, and research highlights.
Moshe felt that history could appear in viewpoints (one of the columnists is Martin Kelly-Campbell, a noted historian) and contributed scholarly articles (one of the Associate Editors is Bill Aspray). Edited interviews are a possibility for the latter (6 pages at 800 words per page). Future interviews should be of people that would enjoy broad interest to the readership.
The last page is tentatively entitled Last Byte, and could be a question-and-answer discussion of an historically-significant person.
CACM thus presents an opportunity to reach out to the community. Moshe predicts interviews on a quarterly basis, perhaps retrospective articles, and prose articles that review history, again, if of broad interest.
Meeting with Doug Terry, chair of SIGOPS
We asked about HOPS. Doug responded that he wanted to start small, and later build up. Some ideas we discussed include
- A session on Multics at SOSP'10
- Build HOPS around best papers
After discussions with several SIG chairs, we noticed that similar questions kept coming up.
We decided that a useful future emphasis for the HC would be to help SIGs identify their founding chair, to generate a history of how that SIG was founded, as a way of characterizing the formation of a subdiscipline as a shared agenda. Specifically, how did that shared agenda come about?
From what other disciplines did it arise? How did it differentiate itself from these other disciplines? What other factors (political, social, personalities, economic) were involved?