HARPER@CSC.FI ("Robert Harper Finland", CSC) (02/06/91)
This message was written on Sunday 3rd Feb 1991, and modifed on Monday
and Teusday, after receiving "feedback" from the network. Thanks to
all who were understanding, and to the rest here is my resignation
note.
************************** RESIGNATION NOTE ***********************
On Saturday night 2nd of February, I noticed that there was a loop
starting on the BIONEWS list. The mailer that was responsible for it
was situated at FTP@ukc.ac.uk (UKC FTP Daemon), and the unfortunate user
was from a commercial site in the UK (parsons_a@uk.co.pcr.svax05)
As per the usual proceedure I checked the lists at IRLEARN only to
discover that this user was NOT subscribed to any of the lists, so I
could not REMOVE him and thus stop messages being sent to him, and since
the SERVE OFF command is reserved for people with special privilages at
IRLEARN I could not shut off the rogue mailer in the UK. I felt pretty
helpless. The only sensible thing to do was put ALL of the lists on HOLD.
About midnight on Saturday, I had sent mail messages to Dave
Kristofferson at Genbank, and also to the Bioearn Bboard explaining
what was happening. I also managed to put the slave lists at IRLEARN on
hold, but before I could get to the master lists the link between
Finland and Sweden went down. Sometimes it feels that I am trying to
manage these lists blindfolded with two hands tied behind my back.
Perhaps Helsinki, Finland is not the best place in the world to manage
lists in IRELAND. The job could be better handled locally. I also
informed LSTERN-L that the filters that were supposed to trap error
messages were not working.
On Sunday morning I checked the USENET distribution and it seemed that
BIONEWS, GENBANK, and METHODS were the lists that had been effected.
The loop had increased to about ten entries.
IRLEARN has had a unique role in the BIOSCI network since it allows users
to subscribe to lists without any human intervention, but this has also
caused problems since users continue to send LISTSERV commands to the
BIOSCI lists rather than to LISTSERV itself. The recent bout of CONFIRM
messages is a point of fact. This problem will not go away. It seems that
it is a mistake that the novice will always make, and if BioSci continues
to run in the present setup then it is a fact that users will have to
live with. As someone so aptly pointed out recently... the bug is in the
users not the system.
In private there have been many discussions as how to make the BIOSCI
network better, but perhaps the managers at the BIOSCI nodes have been
too timid in trying to implement other solutions. The main philosophy
being don't interfere with a system that is working. In addition once
you have built up a system, it is not very encouraging to tear it down
and try to build up another one, when you have no assurance that the
new system would work better than the old one. The work that I have
done for BioSci has been on a volunteer basis, but recently I have felt
that management is taking up a considerable part of my working day, and
as the events over the weekend show, it also eats into my free time.
I also receive "private" letters saying why don't you do something
about this LISTSERV business. Now 2 o'clock on a Sunday morning is not
the best time to do your thinking especially if you know that some
lists are looping, and there is nothing you can do about it. But here
is a list of things that it is impossible for me to do.
1) It is impossible for me to keep all the computer networks in the
world up and running: ergo, you will always have bounced mail.
2) It is impossible for me to make subscriber keep their mailboxes
clean: ergo, disk quotas full... bounced mail.
3) It is impossible for me to make users signoff lists when they leave
University: ergo, bounced mail from dead accounts.
I have been a "care-taker" manager at IRLEARN for about a two years,
while the real list owner is in Holland. As a list owner I get:
1) mail messages every time someone joins a BIOSCI list.
2) mail messages every time someone leaves a list.
3) mail messages from dead mailboxes
4) mail messages bad network addresses
5) mail messages from just about everyone under the sun asking about the
workings of LISTSERV.
6) a few mail messages of encouragement.
In actual fact this morning I have 33 mail messages waiting for me from
people who have subscribed to the lists at IRLEARN. Nothing deters some
people:-)
In case you have not realized it, this is an official letter of
resignation. I have asked for my ID to be removed as list owner at all
the BIOSCI lists at IRLEARN. I find it rather ironic that I was
operating as a list owner at IRLEARN, yet reading my BIOSCI news with
NN on USENET. If everyone adopted a news reader solution then none of
these problems would exist, since there would be no bounced mail
messages, and there would be no strain on your personal mailbox.
I have a clear conscience that I have done my best to serve the BIOSCI
network. When I run a status on BIOEARN which is a private conference
set up for the BIOSCI managers I note that there has been an exchange
of over 900 messages just to keep the system running smoothly... this
behind the scenes work is something the BioSci end user knows nothing
about.
The proof if you need it...
Name: BIOEARN (Finland)
Type: Meeting Private Protected Writeable
Organizer: Rob Harper (CSC)
Creation date: 88-05-17 Default expiration time: 365
Number of entries: 958
Everyone wants a simplification. If you can tag the culprit then you
can put a stop to what he is doing. To all the people who want to point
the finger and blame someone for looping messages who do you want to blame?
1) the user in the UK who has a dead mailbox.
2) the friendly mailer that politely returns reject mail.
3) the listserv programme that does not trap error messages.
4) the list owner cos he does not stay awake 24 hours a day
monitoring the lists.
5) the BioSci network for handling the messages in the first place
So what do I hope to achieve by resigning?
1) I will get some sleep at the weekends.
2) I can go on holiday without worry.
3) I might stimulate some open discussion about the restructuring
of the BIOSCI network.
4) I can contribute to the BIOSCI network without worrying about its
management, or being blamed for its "mismanagement".
If I do not resign then I feel that things will just continue as they
have done. Once the crisis is over then everything returns to status
quo... until the next time. I do not want that.
As my trade mark I have always used Rob "insert something here" Harper.
The astute amongst you will have realized that I tell how I am
feeling in between those quotation marks. I usually use lines from songs
or bits of poetry. For this message I will use a line from Robert Burns.
Regards from the Ex-list owner
Rob "aie aft hand yer storie tell when wae a bosom crony
but aie keep somethin' tae yersel' yid scarcely tell tae onie" Harper
P.S. Hmmm... this is 138 lines long... perhaps I should have made a BioBit
out of it:-) kristoff@genbank.bio.net (David Kristofferson) (02/06/91)
I am extremely sorry to see Rob Harper resign from IRLEARN. I am sure that everyone will agree that Rob's contribution to BIOSCI has been invaluable. I understand his motives, but believe that something good *will* come from all of this confusion during the last couple of weeks. BIOSCI has been an international collaboration between four different sites and has worked reasonably well for about 4 years now. However, I am as aware as anyone of all of its warts because I also have devoted a not insignificant amount of time to resolving the problems that Rob mentioned in his message. Because of the variety of systems involved, many compromises were necessary. Overall BIOSCI has been an excellent example of international collaboration in which many groups gave up their parochial interests in the interest of the scientific community as a whole. This enabled BIOSCI to grow rapidly and provide access to users from New Zealand all the way through Asia, the Middle East, Europe, and the Americas. I have even had responses back about BIOSCI messages from research stations in Antartica! However, we can't rest on past accomplishments. Some technological compromises were made in the interest of bringing significant sections of the scientific community on-line as soon as possible. It is obvious that the time has come to review this situation again. I want to assure our users that we are having very frank and direct exchanges between the BIOSCI sites, and I expect that substantial improvements will be forthcoming as a result of this recent disruption. It is unfortunate, but usually it takes incidents like this to add fuel to calls for technological change. I will *once again* encourage our readers to look into getting news software installed at their sites, and also offer a bit of an incentive, i.e., the day may come when we finally decide to simply terminate e-mail subscriptions. I will continue to push in that direction, although I realize that, given our current feedback from readers, e-mail is still necessary. In my mind there is little reason why we could not set a target goal of phasing e-mail out completely within, say, about two more years. It may be necessary to put this kind of stick out in public to motivate people to respond. In the interim there are technological steps that can be taken to reduce and redirect bouncers. We have taken such steps at GenBank, but additional changes are required at some of the other BIOSCI sites. I believe that I will be able to report back to you soon on further progress. HOWEVER, I can not promise that this will be the end of our recent spate of problems. If the system is reconfigured and a mistake is made in the process, the potential for further disruptions still exists. I hope that our readers will bear with us through this transition period because I have full confidence that the final result will be more agreeable to the end user. Sincerely, David Kristofferson, Ph.D. GenBank Manager kristoff@genbank.bio.net
JRAMON@ccuam1.uam.es ("J. R. Valverde") (02/06/91)
I also think that the Rob's resignation is an unvaluable loss. Also
I must recognize here that the work of the maintainers of the lists is a
GIANT's job.
However, I would like to signal a few thougts. Dave and Rob said that
everything would work better using newsreading software. I can't tell since
I don't know but...
Now I'm receiving BIOSCI news via e_mail through a Vax in the Computer
Center of the University. It was my first (and then only) way to the net.
People at the C/C were (and are) VERY busy trying to keep it running, and
to keep (most important) the University with the pace of times, this means
building new local area networks for the faculties, new networks with
consorted research centers, hospitals, new clusters, many of them independent
and all of this in places where no one knew anything of computers. They still
are busy maintaining and growing this from just a single central site.
Obviously, I was able to ask them to install new software (I did),
but they were NOT able to do it at the pace needed, and there are always
more important things.
Now,thanks to them, we have a flamant new computer center at Medicine,
to which my institute is connecting its computer center. I am the current
C/C manager, but Medicine has none, so I must deal with the needs of our
centre, Medicine and the Hospitals. I WOULD like very much to install news
reading soft but...
I must first finish our own LAN, install GCG, attend users AND I
have not enough space nor to install the mailing software so that we must
still use it through a remote computer in the C/C of the University.
Back again to the beginning.
What I want to say is: OK, I understand your "In the end". I
simpathize with them, I will do everything to help (beginning with getting
funds for a hard disk), and I will NEVER be able to pay your devotion. But...
Informatics has just begun to enter in the biological community as
a usual tool (e.g. a microfuge), this means building lots of new things
that weren't planned a year or two ago, and this in turn, convincing many
research fund programs of the need of it. Many of us are still trying to
begin, many even read BIOSCI via a remote site, and this WILL grow, every
day faster, in the next two or three years, with lots of beginners, lots
of C/C managers very busy or even too far from Biology to understand our
needs.
No hope. NONE? Yet there is!
I learnt mainly by trial and error to use the net (nobody here to
help). And more will in the future. I don't find too onerous receiving a
SUB BIOSCI Newcomer Smith msg. from time to time. And I suspect that
if every time such a msg. comes somebody replys a polite and CLEAR answer,
the said Newcomer Smith won't repeat its error. Also, if periodically there
is a post of some simple instructions for readers, I wouldn't mind too
much, and it would help a lot of new users. BioBit was a fantastic solution,
and I have many a time used it to teach new users at our site (THANKS, ROB).
Only that each time it comes with a new theme and going back to past
issues is NOT easy for Newcomer Smith, and no one gives clear notions on
how to get them nor even their existence.
What about a msg. once a month (e.g. the BioSci review) telling that
more information can be get by issuing a message with "GET BIOBIT.XX" to
a given server?
YOU said it, Rob, the fail is not in the net, but in users, and mostly
in new users. So, my gess is that informing better them would hide a lot of
problems. And maybe, teaching system managers how to face some situations
(e.g. a dead account should be unsubscribed, how? -I had to teach this to
some managers no long ago) A msg. ready for these situations to accompany
the letter to inform of the problem could be useful.
In brief, I think that more education -either from us, site managers,
and from the same lists- would be a great relief. And surely, will be the
only solution in next few years.
Sorry for the length. I'm not in a great moment after struggling
the whole day.
Yours sincerely :-) | (-:
J. R. Valverde
Biomedical Research Institute
Madrid - SPAIN
JRamon@EMDCIE51.BitNet
JRamon@ccuam1.uam.es
JRamon@uamed.uam.es
JRamon@cnbvx1.uam.es
Just to give a few and none the definitive at my site.clark@mshri.utoronto.ca (02/06/91)
The news of Rob Harper's resignation from helping to manage the Biosci
newsgroups is sad news indeed. It's too easy for us to sit back and let
others do all the work and then bitch when things go the least bit wrong. I
think this is partly due to the success of the biosci managers at keeping
things going so smoothly. I certainly had no idea of how much work was
involved.
Perhaps the biosci managers could help the situation, and possibly
avoid future similar resignations, by reminding us of what they are up to.
Every month Dave Kristofferson sends out a notice of the available
newsgroups and how to access them. Why not have a second monthly posting
describing some of the more technical aspects of running the groups? I
thought Rob's posting from a couple of days ago describing how the various
nodes interact with each other was interesting and useful. By seeing some
of the complexity behind the system, especially when told of the reasons
for the complexity, one quicky gets the idea that things aren't nearly as
easy as they look on the surface. It may also be useful to let us know
exactly who is managing the newsgroups. How many other unsung heros are
out there?
It's easy to forget that I'm not the only person who can't find
enough time to get everything done that needs doing in a 36 hour day. For
those times when I, and others like me, lose my patience when things aren't
exactly perfect and send off a testy message to Dave or Rob, why not
automatically forward them to a new newsgroup, biosci-flames? If we know
that the whole net will read our silly little nits, we might think harder
before firing them off :-) :-) .
Steve Clark
clark@mshri.utoronto.ca (Internet)
clark@utoroci (Netnorth/Bitnet)kristoff@genbank.bio.net (David Kristofferson) (02/06/91)
You make many good points, and I reiterate, despite my "threats" that we have no intention of dropping e-mail any time soon. I am familiar with the types of problems that you are describing and realize how far down we are on the priority list. However we are investigating ways to reduce the number of problems that crop up. My experience is that for every public bitch about SUBSCRIBE messages there are other responses to me like yours saying that they are tolerable given the other benefits of the system. This is part of the reason that things have sailed along so far. However, if we can better organize the system, that will not only benefit the end users but it would also free up more of our time to go out and look for new uses for BIOSCI, e.g., beef up the journal TOCs, etc. Stay tuned ... I don't think that we have heard the last from Rob yet (he might at a minimum respond and say that we have 8-). Dave Kristofferson
kristoff@genbank.bio.net (David Kristofferson) (02/06/91)
Steve Clark asks for a mention of who is involved in running BIOSCI. These are the people at each of the four nodes. It does require a fair amount of effort, but, to avoid hype, except for occasions like the last week, the total amount of time that I need to keep BIOSCI running, i.e., handle the "mechanics" at our site is typically under two hours per day and often less than an hour. This falls within the scope of my contractual duties at GenBank. Nonetheless we can undoubtedly streamline the system further. IRLEARN Niall O'Reilly, Ireland Vivian Harrington, Ireland Rob Harper, Finland (this split is a long story 8-) U.K. SERC Daresbury Steve Marshall Royd Whittington Chris Dean Univ. Uppsala BMC Mats Sundvall GenBank/IntelliGenetics Eliot Lear Dave Kristofferson This is probably the last posting of this type that I am going to make about the current "crisis" 8-). We need to get our work done behind the scenes now. Dave
Davison@UH.EDU (Dan Davison) (02/06/91)
Dave Kristofferson notes: > and I expect that substantial improvements > will be forthcoming as a result of this recent disruption. It is > unfortunate, but usually it takes incidents like this to add fuel to > calls for technological change. The most important of which is to get IRLEARN out of the loop as soon as possible. The problems caused there (probably because of Behind Its Time NET/LISTSERV protocols) are NOT worth any nebulous benefits of "automatic confusion" -- oh, sorry -- "automatic subscription". Yes, Internet mailers will occasionally go into loops -- this weekend was not bad at all, as far as mailer blowups go -- but as many have noted here over the past couple of years, the "simplifying" via IRLEARN has generally been more consistently problematic, with little *visible* benefit. The monthly "updates" proposed by Steve Clark would have perhaps changed this perception. BIOSCI maintainers, it's NOT the occasional mess-up...it's the consistent ones that cause the perception, rightly or wrongly, of major problems. dan -- dr. dan davison/dept. of biochemical and biophysical sciences/univ. of Houston/4800 Calhoun/Houston,TX 77054-5500/davison@uh.edu/DAVISON@UHOU Disclaimer: As always, I speak only for myself, and, usually, only to myself.
NUM208JN@NRCCAD.NRC.CA (JOHN NASH) (02/06/91)
In a recent message, dan davison writes:
---
......... The problems caused there (probably because of Behind
Its Time NET/LISTSERV protocols) are NOT worth any nebulous benefits
of "automatic confusion" -- oh, sorry -- "automatic subscription".
Yes, Internet mailers will occasionally go into loops -- this weekend
was not bad at all, as far as mailer blowups go -- but as many have
noted here over the past couple of years, the "simplifying" via
IRLEARN has generally been more consistently problematic, with little
*visible* benefit. The monthly "updates" proposed by Steve Clark
would have perhaps changed this perception.
BIOSCI maintainers, it's NOT the occasional mess-up...it's the
consistent ones that cause the perception, rightly or wrongly, of
major problems.
---
Well, at least we didn't get dozens of complaints landing back on our laps as
bounced messages. At the moment the C programmers' list (C-L@UIUCVMD) has
gone completely nuts, and even the complaints are bouncing (38 this morning and
still counting!).
I suppose the moral of this post is to suggest that if you do get bounced mail,
please don't complain about it to the originating LIST, as it adds to the
confusion. Just patiently do a "dir" and block delete the messages.
cheers,
John Nash. <bitnet: NUM208JN@NRCCAD>BROE@AARDVARK.UCS.UOKNOR.EDU (Bruce Roe) (02/06/91)
Gentle Readers,
As we progress deeper and deeper into the unknown,
pushing back the frontiers of knowledge, learning more about
ourselves, our environment and how the physical, chemical
and biological world works, we push ourselves to the limits
of our capabilities and beyond. We succeed in some efforts
and fail in others, but we keep on pushing forward until we
reach our breaking point, and then we quit when the light at
the end of the tunnel is no longer visible.
Such has just happenned with the resignation of
Rob "pushing back the frontiers" Harper in his position with
IRLEARN. Rob's contributions, along with Dave and others,
have allowed many of us who use BIOSCI as a tool for our
research to get about our business, while they help us
communicate with others and teach us how to broaden our
horrizons and look towards the future where we all will
communicate easily via computers. We sometimes forget that
their work is research too, as they also are pushing back
frontiers of knowledge and doing experiments to improve
our abilities to communicate, discuss, educate and grow.
Some experiments succeed while some almost succeed
and yet others fail miserably. Such is life and life goes
on. When our experiments fail or even when they succeed,
we re-adjust our priorities and we push forward in a new
direction. We make our contribution in one area and then
move on to new areas. Some of my collegues have left the
bench for administration, full-time classroom teaching, or
selling insurance. All continue to make a contribution,
some more visible than others, but a contribution none the
less.
To Rob, good luck, hug your kids for me, continue to
be yourself, thanks for your efforts in helping us communicate
and keep pushing back those frontiers.
To Dave and others who are "working behind the scenes",
you have our support and go-for-it.
To those of us who are users of the network, let us
remember the 11th commandment:
"Thou Shalt Not Take Thyself To Seriously"
Life is too short to get all bent out of shape when
someone doesn't use the network correctly or equipment or
human made programs don't work properly. The novice user
eventually will learn the correct proceedure, the engineers
will fix the equipment and the programmers will eliminate the
bug. People are killing each other over oil, children are
starving to death in under developed countries, the AIDs
crisis is upon us, drugs are taking over the lives of other
wise productive people, and some of us are upset because of
bounced mail messages. As my teenage daughter would say.
"Get real and get a life".
I'm not sorry for this long message and just hope it doesn't
fall on blind eyes.
Best to one and all,
Bruce A. Roe
Professor of Chemistry and Biochemistry
INTERNET: BROE@aardvark.ucs.uoknor.edu
BITNET: BROE@uokucsvx
AT&TNET: 405-325-4912 or 405-325-7610
SnailNet: Department of Chemistry and Biochemistry
University of Oklahoma
620 Parrington Oval, Rm 208
Norman, Oklahoma 73019
FAXnet: 405-325-6111
ICBMnet: 35 deg 14 min North, 97 deg 27 min WestHARPER@CSC.FI ("Robert Harper Finland", CSC) (02/07/91)
Just a short note about the support I have been receiving on Bionews and also from many people on the net who I did not know even existed. I would just like to say to all who have written to me on a positive note thank you very much for your letters of encouragement. BioSci has always been rather professional in its outlook, and Dave has always kept a tight grip on nettiquette, so to save him the trouble of getting Bionews back on the scientific track again, I would ask you please be about your ordinary business, and leave and the accolades until I pick up my nobel peace prize:-) Rob "this is to embarrassing for words" Harper