A warm and non-denominational Season's Greetings to all!
Alan has already sent out our "Bunyip Xmas Card", but I'd
like to add a few words, as well. In particular, I thought
I'd share some news about where we are and where we're
going in the next little while with you who have help so
much in supporting our little enterprise. In all the
excitement of getting a company off the ground it's
sometimes easy to loose sight of where we're heading and
so I'd like to spend a few minutes checkpointing our
progress with those of you who have done so much to get us
where we are today.
New Additions to the Gang...
-----------------------------
Bunyip has definitely taken on a life of its own, with not
one, but two new employees joining the family full time in
the past two months.
Wanda Pierce began working for us full time in October and
will concentrate her initial efforts on product
development for a range of Internet user clients. Alan and
I have worked both back at the McGill School of Computer
Science for nearly five years and during our stay at the
McGill Computing Centre over the past 18 months, so she
_should_ know what she's getting into! :-)
We have now embarked upon a rather ambitious program to
develop client tools to help ensure better access to
Internet information and Wanda will play a strategic role
in making sure are products in this area do what real
people want them to do. She has a interesting mixed
background in both Computer Science and Fine Arts and has
been tasked with keeping the company's techno-weenies in
line whenever she detects that they are forgetting the
needs of the "real" people using our programs.
We don't want to say too much more about this effort at
this point (basically, because we don't want to promise
vapourware) but hope to have more concrete announcements
about our information clients later in the coming year.
Bill Heelan has also joined the gang full time as of this
week to work on systems programming for a variety of
projects. Again, we have all worked together while at the
School of Computer Science and so he, too has been given
fair warning of what he's in for...
Those who have been following the archie story closely
(and like to read the credits at the end while everyone
else is filing out of the cinema) will recognize Bill as
one of the authors of the original archie implementation.
When we first decided to open Alan's nifty little database
to the world Bill was responsible for the first telnet
archie client program and as his first task for Bunyip has
taken over the final push to get the long-promised new
version of this component out the door.
This particular task has been in my lap for the past
several months but has languished "almost finished" for
some time as I've been swamped with company business (such
as organizing our orderly departure from the university,
negotiating final details with McGill for network access
after we move on in January, locating our new premises,
shopping for furniture, bookkeeping, etc). I'm very
pleased to have Bill on line for "real work" since I'm
obviously going to devote more and more of my time to such
paper shuffling. The sad part is I can't get any sympathy
out of the rest of the gang for all the dross I help save
them from!
We estimate that we need only a few days of work to get
the last pieces of the client (output formatting and
paging) working and we're now hoping to get an initial
release of this badly needed piece out to the archie sites
at the start of January. This will give archie operators a
telnet client that speaks Prospero to the server (as do
all the freeware archie clients), reducing demand on
resources and giving better control over the load an
archie will put on your system.
Other archie news..
--------------------
So far, all but one of the original operators of archie
have taken on the supported version (and we're still
hoping to make that 100 percent in January!). In addition,
of course, there are a number of new additions to the
family who should be going operational in the new year.
We're obviously far from done developing archie and we'd
especially like to extend a big "Thank You" for this vote
of confidence to our customers in what is obviously still
an embryonic product line.
The current schedule calls for a number of improvements to
archie in the next month, including the long-awaited new
release of the telnet client. This will feature, among
other improvements, the new Prospero interface to reduce
the demand upon system resources plus the changes needed
to provide multiple language support. The Help system has
been completely redone (making it much easier to extend or
modify the Help text) and we've also added a ".archierc"
file to allow sites to better customize their installation.
Of course, as Alan has written earlier, there are also a
number of new additions at the server end, including
support for domain specific searches, plus pathname limits
on searches (allowing the user to specific only files that
match _and_ contain specific strings in the directory
path). This code is implemented in the current release but
is awaiting the final freeze on the Prospero protocol to
allow the clients to use it. Long promised, we're hoping
for good news on this from the Prospero group shortly.
We have also received a commission from one of the local
universities to prepare and operate a French-language
archie here in Montreal for an initial period of five
months. As part of this deal we will be able to license
the French translation to those needing it for an
additional fee.
Although we don't have plans to prepare additional
translations ourselves at this point, the telnet client is
being reworked specifically to make it easier to prepare
and operate foreign-language version for those customers
who wish to implement their own help text. Details will be
made available to those licensees who might be interested
in preparing their own text when the telnet client ships,
and of course we would be happy to investigate the
possibilities for translations of our text, if you'd
prefer.
Future plans also call for multiple database support,
greater browsing capability and a few more gadgets, but
these additional features are being held for the next
release, which will follow once Bill has it ready to roll.
We realize how important it is to get the Prospero client
out the door and don't want to hold up the release for
features which are less in demand at this point. Still,
the hooks are now in place and it is not expected to take
too long to fold them in and get a follow-up release out,
as well. As usual, your suggestions for further
improvements and additions are welcome.
New Services...
---------------
Also planned for January are at least one and hopefully
several new freely-available archie databases built using
the additional data-gathering and client capabilities of
the new release. Pending final approval after the holiday
period we appear to have secured a site here in Montreal
to host this new service and with Bill on board we finally
have the time needed to get this off the ground.
Once we have things up and running we plan to make the
additional data-gathering scripts available to archie
licensees wishing to extend their offerings. As with the
anonFTP database, we will be offering coordination and
communications services as part of our archie system
support. We'll post details of the new databases to the
archie-maint list in (hopefully) late January and those of
you with licenses can then decide if they want to extend
your range of wares. More on this below...
Other Projects...
------------------
As they say "Bunyip does not discuss unannounced projects"
but we have a number of other activities we are involved
in that we'd like to let you know about.
IETF:
Our largest timesink, outside of our efforts on archie of
course, is the volunteer work we do through the IETF. Both
Alan and I have been heavily involved with such efforts in
the past year and plan to continue this work in 1993.
Alan and I co-chaired the "Internet Anonymous FTP
Archives" working group to develop standardized templates
for entering descriptive information into anonFTP
archives. This working group came to completion at the
Washington IETF in October (or was it November? It's all
becoming a blur!) and we will now be coordinating efforts
to get such templates filled out and distributed on the
Internet throughout 1993.
Several volunteer sites have agreed to act as repositories
for such templates so smaller sites need not fill in
templates for the most common or popular packages (and to
help avoid wasteful duplication or misclassification
during the cataloging phase). A mailing list is also being
set up to coordinate donations and otherwise help the work
along.
Alan and I are also working closely with others in the
field of Internet resource discovery and access on a
number of basic research issues. Perhaps the most urgent
of these involve attempts to define and standardize
mechanisms for naming and referencing objects in an
internet environment.
Working with Joyce Reynolds and Russ Hobby (IETF Area
Directors for User Services and Applications,
respectively) Alan and I inaugurated a new working group
at the last session to develop what is currently being
referred to as "Uniform Resource Identifiers", which in
our lexicon include both "Locators" (pointers to
resources) and "Names" (a more nebulous concept, but
conceptually a location-independent identifier of the
information content of an object). Alan will co-chair the
working group with Jim Fullton of CNIDR, while I plan to
take an activist role as an advocate for a particular
mechanism for defining and creating URIs. For example, I
am currently working on a paper with Chris Weider of Merit
which will be our proposal for URNs.
This work is of more than academic interest to us here at
Bunyip since we plan to migrate all our Internet-based
projects towards the use of URIs as soon as they are
standardized. This would allow client programs, for
example, to perform searches in archie and directly pass
off the identifiers returned to compliant systems without
further processing or conversion. We have high hopes for
URIs and their component parts in the next few years as an
integrating force for future Internet tool development and
plan to keep Bunyip at the forefront of such development
work.
The WHOIS++ project:
Another major project of mine at the moment (and hopefully
a future product line for Bunyip) is my participation in
the WHOIS++ effort. This project came about as a result of
a BOF at the Boston IETF in July, in which a number of
people participated in an examination into the reasons for
the lack a practical directory service for the Internet.
X.500, which was supposed to fill this role, has never
taken off and there are still a number of major
impediments to its success.
After this BOF illuminated the problems (which seems as
much political as technical at this point), a small group
of us, including Joan Gargano of UCDavis, Jim Fullton of
CNIDR, Chris Weider of Merit, Simon Spero of UNC-Chapel
Hill and myself set out to architect and deploy a
light-weight and practical directory service that could
address the problems we see holding back x.500.
We believe that so far we've made excellent progress, with
no less than three papers now in draft form that explain
our solution starting down the path through the IETF
towards eventual RFC status. A new working group (the
WHOIS and Network Information Lookup Service Working
Group) has been set up to coordinate this work and had its
first meeting in Washington.
The heart of the new service (currently called "WHOIS++"
in honour of our original modest ambition to develop a
follow-on to the existing NIC WHOIS service) is a flexible
template oriented data model and an architecture that
allows the operator to use any database system desired for
holding the information to be served. The new system has
now moved far beyond its original modest goals, allowing
operators to define and serve _any_ template-like
information they desire.
We obviously hope and expect at least some sites to pick
up, index and serve collections of the newly created IAFA
templates. This will allow serving collections of
information about users, services, software packages and
more to the entire Internet. In addition to more extensive
and useful databases of information about anonFTP, we
foresee a range of other useful collections and a demo
facility for this service is an important part of our
plans for the
A pilot implementation of WHOIS++ is now deployed
at several sites for testing and our plan is to release the
public domain implementation of our work in January or
February.
Here at Bunyip we intend to offer a combination of archie
and a supported version of a WHOIS++ server as a new
addition to our product line, with preferential pricing
for those who already have archie to enable you to offer
distributed directory support at your site if you want it.
Details on this should be available in February or March.
The Buddy Project:
Alan is now working with Carl Malamud and others to set up
a "buddy project", in which individuals and institutions
will be able to act as sponsors or coordinators for
donations and other assistance aimed at newly connected
sites, especially in the developing world. Alan has
already sent a first collection of donated books and other
materials to Estonia and as one of their first enlisted
volunteers I am currently gathering material to forward to
the University of Swaziland. In the new year.
It is hoped that this work will become a sponsored project
of the Internet Society, with a number of individual
coordinating projects feeding badly needed material where
it will do the most good. In the long run, Carl and Alan
have ambitious plans for organizing donations of
educational material and even hardware directly from
manufacturers, but to me the greatest potential in their
plans is for making available the expertise of individuals
to those who need it most. I expect we'll be hearing much
more about this project in 1993.
The Wrap-Up...
Well, that's a quick survey of the activities of Bunyip
over the past year, along with a few pointers towards
where we hope to see the company go in the new year. Now
we have Wanda and Bill on staff we expect to be able to
offer a much better level of user support to our existing
client base, while continuing the all-important research
and develop into new tools and services that we must
perform to keep alive (and keep up with the demand from
our users for a more useful Internet). I'd be happy to
pass on further information on any of the development
projects we're participating in through the IETF, and
would encourage anyone who is interested to join the
appropriate mailing lists and get involved. Even if you
can't make it to the meetings, you can have a real impact
on development with your comments and suggestions.
Peter's Soapbox
----------------
As for us, we promise to try to extract every possible
drop out of the funding we're receiving through archie
license fees to push out both an improved archie and our
related follow-on products. As I mentioned at the
beginning of this note, we're very grateful for the early
(and tangible!) support our licensees have show us and
will do our best to deliver value for money. Although we
are playing in perhaps one of the most interesting and
complex technological sandboxes available today, we
definitely hope to be perceived as a "people" company,
offering quality products and responding to our customer's
needs.
As a small start-up it will obviously be hard to always
fulfill expectations but we're going to give it our best
shot. If we're _NOT_ performing up to scratch, please tell
us. And if we are, as they say "Tell your friends"!
Enough drivel - may you all enjoy a happy holiday season
and a prosperous and successful new year.
Peter (speaking on behalf of Alan,
Wanda and Bill, plus Alan's fish
which are doing better last time we
checked, now that the filter is
working properly...)
--