|
|
From semonch@gibbs.oit.unc.eduMon Nov 10 15:41:55 1997
Date: Mon, 10 Nov 1997 15:15:26 -0500
From: Barbara Semonche
To: DBQC@listserv.oit.unc.edu
Subject: FW: Indy session on Cx (fwd)
To all:
Here is the first of some messages that Jackie Chamberlain, Bruce
Oakley, and I exchanged in the early phase of planning for the Indy
conference. Most of this is FYI, but your comments and suggestions are
solicited. Consider this the first part of getting ourselves acquainted
with what we want to accomplish with our program. BTW, I am updating the
DBQC web page. The URL again is:
http://sunsite.unc.edu/journalism/databasequality.html
[Bene Note: please send me (email is preferred) your bio briefs so that I
can load them on the web site. It will save on introductions at the
program if we have this stuff ready early. And if you can also send me a
photo of yourself (no complaining, now) I will scan it and load it on the
DBQC web page so we can all get a look at ourselves *before* we start.
However, if you have already have this info on your own web site, please
send me the URL. I'll simply link to it.
Barbara P. Semonche, Library Director (Listowner: NewsLib, NCSLA)
UNC-CH School of Journalism and Mass Communication
Howell Hall, CB# 3365 Phone: 919/962-1204, Fax: 962-0620
Chapel Hill, NC 27599-3365 E-mail: semonch@gibbs.oit.unc.edu
Semonche's URL http://sunsite.unc.edu/journalism/
---------- Forwarded message ----------
Date: Wed, 15 Oct 1997 15:32:05 -0700
From: Jackie Chamberlain
To: "'semonch@gibbs.oit.unc.edu'"
Subject: FW: Indy session on Cx
Barbara,
Here's what I sent to Carol last week -- as mentioned in my message of
moments ago.
Jackie Chamberlain
----------
From: Jackie Chamberlain[SMTP:jchamberlain@pe.com]
Sent: Thursday, October 09, 1997 3:54 PM
To: 'carolc@oklahoman.net'
Subject: Indy session on Cx
Okay Carol, here it is:
--------------------
To err is human...Cleaning it up is in our control:
Corrections and quality in news databases.
Are the standards for handling corrections in full-text
news databases changing? Are Web editions, where
there are no rules, affecting the way we look at
quality control in general, or were the cracks in the
QC foundation already there? Where do we draw the
line? How do we keep it clean, credible and legal?
Moderator: Jackie Chamberlain, The Press-Enterprise
Panel: TBA
-----------------------
I see this as a group of short presentations from various perspectives,
followed by a round-table discussion (thus the description in the form
of questions to suggest "debate") in which the participants, and I hope the
"attendees" will discuss the issues. I hope this will segue nicely into the
general roundtable, and you may want to reformat this in a way that suggests
that (?). I'm also looking at the presenters each taking on one of the
issues rather than simply do a show and tell. This is all still perculating.
I'm hoping to get a spectrum of participants. I've so far asked (e-mailed) but
not talked to): Mike Jesse (on vacation until Monday)
Cathy Tierney (out until today, will call Monday)
Tim Rozgonyi (who can't -- no SLA next year, has to go to
techie things -- but I've asked him to make
a written contribution)
Bruce Oakley (Online Editor in Little Rock who researched
the subject as a KR Fellow at Chapel Hill --
Me (as "moderator" and presenter)
Barbara Semonche is going to help with a Website and the technicalities of
graphic presentations, as well as annotated bibliography.
I'm looking at the possibility of a survey, the results of which would go
on the Web site and in handouts. Bruce is shopping around his research
paper to publishers and may have one -- we may be able to get reprints.
Jackie Chamberlain jchamberlain@pe.com
From semonch@gibbs.oit.unc.eduMon Nov 10 15:42:08 1997
Date: Mon, 10 Nov 1997 15:20:46 -0500
From: Barbara Semonche
To: DBQC@listserv.oit.unc.edu
Subject: Indy session equipment (fwd)
To all:
Second installment on the Indy Conference.
Barbara P. Semonche, Library Director (Listowner: NewsLib, NCSLA)
UNC-CH School of Journalism and Mass Communication
Howell Hall, CB# 3365 Phone: 919/962-1204, Fax: 962-0620
Chapel Hill, NC 27599-3365 E-mail: semonch@gibbs.oit.unc.edu
Semonche's URL http://sunsite.unc.edu/journalism/
---------- Forwarded message ----------
Date: Fri, 24 Oct 1997 10:33:36 -0700
From: Jackie Chamberlain
To: "'semonch@gibbs.oit.unc.edu'"
Subject: Indy session equipment
Barbara,
Wow! I had no idea that stuff was so expensive! As a seriously AV challnged
individual I'm astounded. Carol told me this morning she was floored too.
After pondering your incredible wealth of ideas, and thinking about how
this should work, I talked to Pete Johnson at the LA Times yesterday. He
volunteered to coordinated the equipment "stuff" for Carol, bless his
heart. After we brainstormed the ideas you had given me and I thought
about what would work best, I have a pretty good idea of what will work
best for us. I called Carol this morning to get a "feel" for the room
setup to make sure what we'd need. She also said we now have until
mid-Novemberish to firm things up (Pete said for Seattle they had until
March -- go figure).
She says it will be quite informal -- the round tables for 6 or 8 kind of
setup -- since there will be breakfast and the idea is informality anyway.
She said we could either work from one of the tables or have a long table
up front. Hmmmm. The idea is to get others involved, discuss and keep show
and tell and pontificating to a minimum. (I'm not sure I could pontificate
at 7:30 a.m. anyway!)
Both were delighted that you would be involved in helping, that you'd
bring your own laptop (as Pete said, "Don't work with a laptop you don't
know). And of course it saves $$. He also suggested the more fancy you get
the more that can go wrong -- thus verifying my KISS principle (Keep it
simple....)
So here's what I'm thinking:
Mikes -- clip-ons for the panelists and a hand-held one for me on a long
cord (for those soft-spoken in the audience (?). Pete said that regular
mikes are $25, clip-ons $35 -- but he agrees that it would be worth the
extra cost to avoid interruptions in the flow of the discussion. And of
course amplification.
Obviously the data video projector and screen -- tho I guess in a $ pinch
we could use an overhead, but hey. And cables.
My big hangup is whether to go "live" which is fraught with potential
glitches. *How* much time consuming preparation (and what kind) needs to
be done to do this "off-line'? I don't want too much of this to fall on
your shoulders. If each of us provided you with samples to load, what
would you need? Again, I don't see this as being so much "show-and-tell"
and "Ok here's what we do and why -- now let's talk about it. Would it be
do-able to project the points of discussion as they come up -- as if on a
blackboard but on screen? Or is that overkill? Hmmm, seems a little much.
As for A-V staffers, I assume that the plan is to have someone there
anyway since this is part of a larger session, but I'll ask Carol. Getting
set up that early will be a challenge in itself.
Your mentioning trying to get vendors to help with cost reminds me of your
earlier reference to getting vendors involved. Absolutely, if they're
willing. I've already had some interest shown by DataTimes to be there, a
conviction that not only should providers adhere to certain standards but
that vendors need to take standards seriously themselves. So I'm pretty
sure I can get DataTimes there (even tho they won't be vendors by then,
they still feel strongly about standardization), I'll have to figure out
how to convince the others. I don't know that DT would be willing to
absorb costs all by themselves. Will think on this too. Carol said that if
I can get the vendors to commit she will get a standard agreement from
them about what they've agreed to.
As to sending things to you for mounting on the Web site, when would you
need it and in what form?
If it's feasible I'd like to (audio) tape the session -- and I can
certainly come up with something here that will suit, I think
(certainly...I think? does the ambivalence show?) If we could do that,
maybe a transcript could be put up on the Web site (?)
Except for Tim Rozgonyi who can't attend, all have agreed to participate
-- Mike Jesse, Bruce, Cathy Tierney. I've asked Mike for the name of
another "loyal oppositionist" to balance things out. Haven't heard, so
will call. Tim suggested Jim Hunter as someone who always has an opinion,
but I'll have to find out what his opinion is in this case!
Anyway, things are moving along. Let me know what you think, when you get
a chance.
More later.
Jackie Chamberlain jchamberlain@pe.com
The Press-Enterprise Phone - (909) 782-7579
Riverside CA Fax - (909) 782-7572
From semonch@gibbs.oit.unc.eduMon Nov 10 15:42:16 1997
Date: Mon, 10 Nov 1997 15:26:21 -0500
From: Barbara Semonche
To: DBQC@listserv.oit.unc.edu
Subject: One more chapter to the Nexis/DataTimes Saga -- Let's hope
that's it (fwd)
To all:
Here is the fourth installment.
Barbara P. Semonche, Library Director (Listowner: NewsLib, NCSLA)
UNC-CH School of Journalism and Mass Communication
Howell Hall, CB# 3365 Phone: 919/962-1204, Fax: 962-0620
Chapel Hill, NC 27599-3365 E-mail: semonch@gibbs.oit.unc.edu
Semonche's URL http://sunsite.unc.edu/journalism/
---------- Forwarded message ----------
Date: Tue, 28 Oct 1997 14:54:26 -0800
From: Jackie Chamberlain
To: "'sclairemont@link.freedom.com'" ,
"'sharon.reeves@uniontrib.com'" ,
"'sherry.adams@chron.com'" ,
"'tglib@telegram.infi.net'" ,
"'beltonj@washpost.com'"
Cc: "'semonch@gibbs.oit.unc.edu'" ,
"'bruce_oakley@adg.ardemgaz.com'"
Subject: One more chapter to the Nexis/DataTimes Saga -- Let's hope that's it
Sharon, Sharon, Sherry, George and Jennifer:
Because I talked to each of you personally about the problems Basis papers
have with corrected long docs sent to Nexis, and the solution to it, I'm
letting you all know what I hope is the final chapter in this saga as well
as posting much the same thing on NewsLib.
You'll recall that last summer I discovered that documents with overflow
records (OK Sherry, you can go ahead and gloat now that you have PLS!) --
when corrected in any way, lost their bottoms -- as well as all
corrections, notes, captions -- at Nexis. After months of wrestling with
Nexis, we came up with the solution: when correcting a doc, "put" not only
the home record, but the overflow record(s) as well and they'd get
stitched back together at Nexis.
We had been talking to DataTimes about this all along, and before we took
this step we checked with 2 people (one of whom I considered
unimpeachable) at DataTimes to be sure there would be no negative result
there. We were assured that there would be no problem. (Neither of those
people is at DT any longer.)
Recently I had reason to go back to re-correct some long docs that were
older than 2 months (simply a matter of style change requested by my ME).
I discovered that not only were my re-corrections not there, but neither
were the original correction, nor -- nor -- the originals. The stories
were simply gone. That meant all corrected long docs sent since
mid-December were missing. Granted that's not a large number -- we don't
have *that* many long docs and only some of them are corrected. But they
do tend to be more important stories.
I called DataTimes. They were appalled, apologetic and quick to get on the
problem of why it had happened and how to fix it. Unlike Nexis, they were
fully cooperative and took the responsibility to work it out.
Here's what they discovered: Nexis insisted that we extract data for
transmission to them differently than we extract data for DataTimes, i.e.
Nexis insisted we use HTPrint. Unfortunatley, this bypasses a little
program called GenData (sp?) which is where (note my technical expertise
here) the parts of a long doc hook up, so to speak. Now, the "Nexis fix"
worked by artificially hooking up the parts. But the result at DataTimes
was a fatal flaw: if you "put the home record, and also "put" the overflow
record(s) you create a "double delete" effect: The corrected home record
says "delete the original or resident doc at DT because here comes a
corrected version; but right behind it comes its partner *also* with a
delete command saying "get out of my way, I'm taking over." (Don't you
love this techie talk?) The result: The original or resident doc deletes--
and so does the replacement: No doc.
The solution, is elegantly simple: If you are extracting data for Nexis
using HTPrint and bypassing GenData, don't. Extract data for Nexis in the
same way you do for DataTimes, passing thru GenData. If you do that, you
don't need to "put" overflow records. Just do corrections the old
fashioned way.
So what happens with all the stuff you double-put? My systems guy Jim
ZIkratch assures me he can create a file of potentially affected docs. He
may, in fact, be able to find the specifically affected docs. They can
then be re-sent and all will be well. (If you've been following my
now-obviously-flawed advice, please let me know. Jim and I will be glad to
help in any way we can.)
We have tested this and it works -- both at DT and Nexis. In fact I sent a
*twice*-corrected long doc in the test and it worked fine.
I must give tremendous credit to DataTimes for solving this as soon as it
came to light. They were fast, responsive and cooperative -- and they did
all the work and took responsbility. But I felt responsible for letting
you 5 and any others who may have been affected know personally. Mike
Harris at DT says they will more than willingly help anyone who has
questions or needs help: (405) 749-6351.
Jackie Chamberlain jchamberlain@pe.com
The Press-Enterprise Phone - (909) 782-7579
Riverside CA Fax - (909) 782-7572
From semonch@gibbs.oit.unc.eduMon Nov 10 15:42:23 1997
Date: Mon, 10 Nov 1997 15:24:09 -0500
From: Barbara Semonche
To: DBQC@listserv.oit.unc.edu
Subject: RE: Indy session equipment (fwd)
To all:
This is the third installment. We're coming to the end of this
"thread." Maybe one or two more.
Barbara P. Semonche, Library Director (Listowner: NewsLib, NCSLA)
UNC-CH School of Journalism and Mass Communication
Howell Hall, CB# 3365 Phone: 919/962-1204, Fax: 962-0620
Chapel Hill, NC 27599-3365 E-mail: semonch@gibbs.oit.unc.edu
Semonche's URL http://sunsite.unc.edu/journalism/
---------- Forwarded message ----------
Date: Fri, 24 Oct 1997 13:55:23 -0700
From: Jackie Chamberlain
To: 'Barbara Semonche'
Subject: RE: Indy session equipment
Barbara,
Great minds! You've captured *exactly* what I envisioned. Thanks for
the perspective.
Isn't it funny how quickly we get caught up in technology that we forget the
simple stuff. Of *course* what I really want is something like a simple
paper flip chart!
I'll work up a schedule and let you know. As for the DBQC list, that's all
I have for now, tho I may have one or two more a little later.
This is great, and I appreciate your help so much.
Will be in touch soon.
Jackie Chamberlain
----------
From: Barbara Semonche[SMTP:semonch@gibbs.oit.unc.edu]
Sent: Friday, October 24, 1997 1:08 PM
To: Jackie Chamberlain
Subject: Re: Indy session equipment
Jackie:
Good on you for making such progress. Let's indeed try to keep it
simple.
First, if the approach is informal, perhaps we should do it
something like "Wall Street Week" or "News of the Week in Review" or
whatever. A roundtable, a moderator who introduces the panel, quickly
sets the scene, then get the panelists to make their opening remarks,
briefly. Back to the moderator who may summarize, and then tosses out the
first of "challenging" questions to a specific panelist. Responses brief,
pointed, and perhaps controversial. If other panelists are not joining in
at this time, then moderator solicits responses. EVERYONE comes prepared.
To do this, panelists must have their opening remarks (probably no more
than one page, that is about 250 words requiring about 3-4 minutes each,
ready at least a month ahead of time. These remarks are sent via email to
all the panelists simultaneously. I suggest that I launch a "mini" list
so that we can post messages to the group as a whole. This a simple thing
to do. Then I can capture the panelists' prepared remarks and mount them
on the web site.
You as moderator, open the discussion to the group as quickly as
possible in order to get reactions from the audience. Ask for
experiences, ideas, comments, suggestions, etc. If you wish, I'll remain
in the back of the room, looking for people who might want to speak up
and signal you. At the end of the session, remember to do a 4 or 5 point
speed wrap-up. And then direct them for the next step, whatever that is.
When the web site is complete ready, I'll make a master printout
of it and give it to someone on our committee (you, you lucky girl!) to
make 60-70 copies to distribute. Doubtless there will be other packets of
info that will need to be incorporated into the total session handout
package, but we'll need them early in order to get them together.
Next, I'll do a "Web Whacker" capture of the web site so that it
can be available on my laptop at the meeting. If anyone wants to see it
there, he/she (small groups) can check it out after the session.
Advantage? We won't have to have any electronic equipment (other than the
microphones, of course). I'll bring my own cable and look for the nearest
power outlet. If desired, I may bring some extra discs with the web site
downloaded on it if people want the material that way. In short, we've
kept it short, sweet, and simple.
Perhaps we could ask for a flip chart, just in case you want to
do a written summary of suggestions for the audience (and for me; I'll be
keeping notes for addition to the web site.) A chart is not necessary,
just a possibility if you think it might be useful. It may be just
another thing to keep track of in an already overpacked program.
To summarize:
1. I'll prepare the web site (including my bibliography on database
quality control, perhaps Bruce's research paper, the NewsLib thread on
"online news database errors and quality control," and perhaps your
research results.)
2. I'll also get the "DBQC" (database quality control) list
started. Let's see, that would be:
Jackie Chamberlain jchamberlain@pe.com
Bruce Oakley bruce_oakley@ardemgaz.com
Cathy Tierney ctierney@mail.bright.net
Mike Jesse Michael_Jesse@dni.com
Barbara Semonche semonch@gibbs.oit.unc.edu
and
Tim Rozgonyi rozgonyi@post-gazette.com
Anyone else?
3. You'll need to prepare an "agenda" of the meeting as well as a
timeline (a sort of "I" Day minus 7 months, etc.) for when things are due
to me and you so that everyone will know what is in the works, who will be
doing what when.
4. And, of course, you will be coordinating all the rest! AND doing
a presentation as well. You are entitled, after all, to have a prominent
voice in this program. You are the one who sounded the clarion call to
action!
Cheers!
Barbara P. Semonche, Library Director (Listowner: NewsLib, NCSLA)
UNC-CH School of Journalism and Mass Communication
Howell Hall, CB# 3365 Phone: 919/962-1204, Fax: 962-0620
Chapel Hill, NC 27599-3365 E-mail: semonch@gibbs.oit.unc.edu
Semonche's URL http://sunsite.unc.edu/journalism/
On Fri, 24 Oct 1997, Jackie Chamberlain wrote:
> Date: Fri, 24 Oct 1997 10:33:36 -0700
> From: Jackie Chamberlain
> To: "'semonch@gibbs.oit.unc.edu'"
> Subject: Indy session equipment
>
> Barbara,
>
> Wow! I had no idea that stuff was so expensive! As a seriously AV challnged
> individual I'm astounded. Carol told me this morning she was floored too.
> > After pondering your incredible wealth of ideas, and thinking about
how this should work, I talked to Pete Johnson at the LA Times yesterday.
He volunteered to coordinated the equipment "stuff" for Carol, bless his
heart.
> After we brainstormed the ideas you had given me and I thought
about what would work best, I have a pretty good idea of what will work
best for us. I called Carol this morning to get a "feel" for the room
setup to make sure what we'd need. She also said we now have until
mid-Novemberish to firm things up (Pete said for Seattle they had until
March -- go figure).
>
> She says it will be quite informal -- the round tables for 6 or 8 kind of
setup --
> since there will be breakfast and the idea is informality anyway. She said
> we could either work from one of the tables or have a long table up front.
> Hmmmm. The idea is to get others involved, discuss and keep show and
> tell and pontificating to a minimum. (I'm not sure I could pontificate
at 7:30 a.m. anyway!)
>
> Both were delighted that you would be involved in helping, that you'd bring
your own laptop (as Pete said, "Don't work with a laptop you don't know).
And of course it saves $$. He also suggested the more fancy you get the more
that can go wrong -- thus verifying my KISS principle (Keep it simple....)
>
> So here's what I'm thinking:
>
> Mikes -- clip-ons for the panelists and a hand-held one for me on a long cord
> (for those soft-spoken in the audience (?). Pete said that regular mikes
> are $25, clip-ons $35 -- but he agrees that it would be worth the extra cost
to avoid interruptions in the flow of the discussion. And of course
amplification.
>
> Obviously the data video projector and screen -- tho I guess in a $
pinch we could use an overhead, but hey. And cables.
> My big hangup is whether to go "live" which is fraught with potential
glitches. *How* much time consuming preparation (and what kind) needs to
be done
> to do this "off-line'? I don't want too much of this to fall on your
shoulders.If each of us provided you with samples to load, what would you need?
> Again, I don't see this as being so much "show-and-tell" and "Ok here's what
> we do and why -- now let's talk about it. Would it be do-able to project the
> points of discussion as they come up -- as if on a blackboard but on screen?
> Or is that overkill? Hmmm, seems a little much.
>
> As for A-V staffers, I assume that the plan is to have someone there anyway
> since this is part of a larger session, but I'll ask Carol. Getting set
up that early will be a challenge in itself.
>
> Your mentioning trying to get vendors to help with cost reminds me of your
> earlier reference to getting vendors involved. Absolutely, if they're willing.
> I've already had some interest shown by DataTimes to be there, a conviction
> that not only should providers adhere to certain standards but
that vendors need to take standards seriously themselves. So I'm pretty
sure I can get DataTimes there (even tho they won't be vendors by then,
they still
> feel strongly about standardization), I'll have to figure out how to
convince the others. I don't know that DT would be willing to absorb costs
all by themselves. Will think on this too. Carol said that if I can get
the vendors to commit she will get a standard agreement from them about
what they've agreed to.
>
> As to sending things to you for mounting on the Web site, when would
you need it and in what form?
>
> If it's feasible I'd like to (audio) tape the session -- and I can
certainly come up with something here that will suit, I think
(certainly...I think? does the
> ambivalence show?) If we could do that, maybe a transcript could be put up on
> the Web site (?)
>
> Except for Tim Rozgonyi who can't attend, all have agreed to participate --
> Mike Jesse, Bruce, Cathy Tierney. I've asked Mike for the name of
another "loyal oppositionist" to balance things out. Haven't heard, so
will call. Tim suggested
> Jim Hunter as someone who always has an opinion, but I'll have to find
> out what his opinion is in this case!
>
> Anyway, things are moving along. Let me know what you think, when you
get a chance.
> More later.
>
> Jackie Chamberlain jchamberlain@pe.com
> The Press-Enterprise (909) 782-7579
> Riverside CA Fax - (909) 782-7572
>
>
>
>
From semonch@gibbs.oit.unc.eduMon Nov 10 15:42:30 1997
Date: Mon, 10 Nov 1997 15:27:12 -0500
From: Barbara Semonche
To: DBQC@listserv.oit.unc.edu
Subject: RE: One more chapter to the Nexis/DataTimes Saga -- Let's hope
that's it (fwd)
To all:
And here is the fifth installment. Not the last one, however.
Barbara P. Semonche, Library Director (Listowner: NewsLib, NCSLA)
UNC-CH School of Journalism and Mass Communication
Howell Hall, CB# 3365 Phone: 919/962-1204, Fax: 962-0620
Chapel Hill, NC 27599-3365 E-mail: semonch@gibbs.oit.unc.edu
Semonche's URL http://sunsite.unc.edu/journalism/
---------- Forwarded message ----------
Date: 28 Oct 1997 17:06:15 U
From: Bruce Oakley
To: jchamberlain@pe.com
Cc: Barbara Semonche
Subject: RE: One more chapter to the Nexis/DataTimes Saga -- Let's
hope that's it
Sigh. Good grief! Glad you found folks with good answers.
We're still wrestling with cleaning up our archive for our own use, much less
getting it out right to Nexis. It has gotten better -- librarians are now on
the same floor as the newsroom (out of the basement). And they have terminals
hooked up to the newsroom database. And we are working on a software package
that will make the library seamless with the newsroom database -- and
eventually allow us to open it read-only to the public through our web site.
In the meantime, we are still having your kind of software collisions with our
various newsroom packages, but the photo editor is now our director of
archives, too, and he has the voice of authority to get the newsroom to follow
procedures so that the archival feed is clean. Alfred Thomas, who was stuck in
the basement, never found that kind of voice.
Stay in touch -- let me know as plans take shape for June.
_________________________________________________________
Bruce W. Oakley, Arkansas Online editor, www.ardemgaz.com
Arkansas Democrat-Gazette, Little Rock
email: bruce_oakley@adg.ardemgaz.com
Fax: (501) 372-3908 (501) 399-3602
_________________________________________________________
From: jchamberlain@pe.com on Tue, Oct 28, 1997 4:52 PM
>Sharon, Sharon, Sherry, George and Jennifer:
>
>Because I talked to each of you personally about the problems Basis papers
>have with corrected long docs sent to Nexis...
>
>Jackie Chamberlain jchamberlain@pe.com
>The Press-Enterprise (909) 782-7579
>Riverside CA Fax: (909) 782-7572
From semonch@gibbs.oit.unc.eduMon Nov 10 15:42:40 1997
Date: Mon, 10 Nov 1997 15:33:13 -0500
From: Barbara Semonche
To: DBQC@listserv.oit.unc.edu
Subject: One more, final, chapter to the Nexis/DataTimes Saga (fwd)
To all:
Here is the sixth installment for the Indy conference "threads."
Barbara P. Semonche, Library Director (Listowner: NewsLib, NCSLA)
UNC-CH School of Journalism and Mass Communication
Howell Hall, CB# 3365 Phone: 919/962-1204, Fax: 962-0620
Chapel Hill, NC 27599-3365 E-mail: semonch@gibbs.oit.unc.edu
Semonche's URL http://sunsite.unc.edu/journalism/
---------- Forwarded message ----------
Date: Tue, 28 Oct 1997 16:16:03 -0800
From: Jackie Chamberlain
To: "'newslib@listserv.oit.unc.edu'"
Cc: "'jzikratch@pe.com'" ,
"'nkapper@pe.com'"
Subject: One more, final, chapter to the Nexis/DataTimes Saga
NewsLibbers,
I've written a version of this message to those with whom I've talked about
this (Jennifer Belton your message came back -- what's your address?),
but I'm concerned some Newsies who use(d) Basis and saw my earlier
chapters, several months ago, will want to hear this (maybe "want" isn't
the right word, "need") , especially if they acted on that information.
OK, you're all saying, here comes another endless message. Please
read on.
If you don't uses Basis, you're welcome to read this for whatwever value it
has as a lesson in dealing with vendors.
A brief review: about a year ago I discovered that documents with overflow
records, when corrected in any way, lost their bottoms at Nexis -- as well
as all the corrections, notes, captions etc that Nexis puts at the bottom.
After months of wrestling with Nexis in general, but a lot of cooperation from
Mike Kent, we came up with a solution: when correcting a doc, "put" not only
the home record, but the overflow record(s) as well and they'd get stitched
back togethr at Nexis.
We did not do this blindly. We had been talking to DataTimes about the
problems all along, and before we took this step ("putting" all the parts
of a story) we checked with 2 people, independently, at DataTimes to be sure
that unusual step would not have a negative result at DT. We were assured
there would be no problem. WIth that assurance we went ahead with the
"Nexis fix" in mid-December and I posted several detailed alerts on
NewsLib.
Recently I went back to re-correct some long docs older than 2 months old
(which has meaning only for Basis/DT people). (The changes were a matter
of style which could look like "editorial" comment. No on point right now, but
I know you're saying "why-ever for?") The important thing:
I discovered that not only were my re-corrections not there (in
DT -- they were oK in Nexis), but neither were the original
corrections, nor - nor -- nor people, were the original stories.
*The stories were simply gone.*
That meant all corrected long docs sent since mid-December were missing --
not a large number, but they are stories that tend to be more important.
DataTimes was appalled, apologetic and quick to get on the problem of why
it happened and how to fix it. Unlike Nexis (at the time), they were fully
cooperative and took the responsibility to work it out.
Here's what they discovered: Nexis insisted that we extract data for
transmission to them using HTPrint. Unfortunately that bypasses a little
program called GenData (sp?) which is where (this is technical talk here)
the parts of a long doc hook up. No GenData process, no hookup. Now, the
"Nexis fix" -- re-sending all the parts of a story, works for Nexis, the
result at DataTimes is -- hold on -- a fatal flaw: if you "put" the home
record and also "put" the overflow record(s), you create a "double delete
effect: The corrected home record says "delete the original or resident
doc at DT because here comes a corrected version. This is supposed to
happen. But if you have also re-sent the overflow record, along comes
another "delete" command, right behind the home record, which deletes its
own top and since it's a stump bombs out. (Don't you love this techie
talk?) The result: The original or resident doc deletes -- and so does its
replacement. No doc.
The solution is elegantly simple: If you are extracting data for Nexis
using HTPrint and bypassing GenData...Don't. Extract data in exactly the
same way you do for DataTimes, passing thru GenData. If you do that, you
don't need to "put" overflow records. Just do corrections the
old-fashioned way.
We have tested this and it works -- both at DataTimes and Nexis.
So what about the stuff you may have already double-put"???? My systems
person, Jim Zikratch assures me he can create a file of potentially
affected docs. He may, in fact, be able to find the specifically affected
docs. They can then be re-sent and all will be well. If you have any
questions please let me know. I may not be able to explain the technical
stuff, but Jim certainly can. Mike Harris at DataTimes (405) 749-6351 has
told me to pass along word that they will gladly work with anyone with any
questions or problems related to this, too
I must give a lot of credit to DataTimes for solving this as soon as it came
to light. They were fast, responsive and cooperative. They did all the
work in short order and, perhaps most importantly in this whole saga,
took responsibility.
If any lesson is to be learned from this, perhaps it's that we, as Information
Providers have to take in active role in what happens to what is, after all,
*our* body of work.
(Ok, who's going to write our fight song???)
Jackie Chamberlain jchamberlain@pe.com
The Press-Enterprise Phone - (909) 782-7579
Riverside CA Fax - (909) 782-7572
From semonch@gibbs.oit.unc.eduMon Nov 10 15:43:07 1997
Date: Mon, 10 Nov 1997 15:35:36 -0500
From: Barbara Semonche
To: DBQC@listserv.oit.unc.edu
Subject: Cx in news databases - Equipment for Indy (fwd)
To all:
Yet another, the seventh (?) in the Indy series.
Barbara P. Semonche, Library Director (Listowner: NewsLib, NCSLA)
UNC-CH School of Journalism and Mass Communication
Howell Hall, CB# 3365 Phone: 919/962-1204, Fax: 962-0620
Chapel Hill, NC 27599-3365 E-mail: semonch@gibbs.oit.unc.edu
Semonche's URL http://sunsite.unc.edu/journalism/
---------- Forwarded message ----------
Date: Fri, 31 Oct 1997 10:43:45 -0800
From: Jackie Chamberlain
To: "'carolc@connectok.com'"
Cc: "'Peter.Johnson@latimes.com'" ,
"'semonch@gibbs.oit.unc.edu'"
Subject: Cx in news databases - Equipment for Indy
Carol,
OK, here it is.
The session on corrections will be informal and (I hope) lively -- truly
a roundtable much like, as Barbara has characterized it, "Wall Street
Week" et al.
Table: It would be nice to have the panel actually sit at a round table, as
everyone else will be. Only problem: half would have their backs to the
rest of the room. We could scrunch together on one side and get cozy,
but I think the better thing would be to have a rectangular table up front
so we'd all face the "audience" and have room for our stuff (and breakfast!)
However, no podium, I think. That's too formal.
MIkes: There may be a total of 6 panelists including me. To keep the
discussion flowing, it would be preferable for the panel to have clip-on
mikes so one doesn't have to be passed around. However, it might be
desirable to also have one hand-holdable mike (Pete?? Barbara??) for me if
I'm moving around -- *unless* the clip-on would give me enough mobility
(again, Pete?? Barbara??). (All those years I taught I relied on my
ability to project, never had to use a mike!) So 6 clip-ons or 5 clip-ons
and one hand-holdable. I think I need some help here.
Laptop: Barbara has volunteered to use her laptop and sufficient cable to
have the Web site she'll create for the session available for viewing. She
has said she may also bring some discs with the web site downloaded for
those interested. (Barbara, maybe I can work something out to help on this
-- i.e. getting the discs.)
A flip chart easle (I'll bring the marker! -- might even be able to bring
sheets of newsprint -- *if* I have to).
That's it! No fancy electronic graphics, no projection screen to detract
from the discussion. Just talk. It'll be my job to keep it moving but I
have a feeling that won't take much (except lots of coffee).
Each of the panelists, in a short presentation, can verbally describe what
he/she does re corrections and include samples in the handouts and on the
web site.
So far, Carol, I have commitments from Mike Jesse, Cathy Tierney and Bruce
Oakley (Little Rock), and of course Barbara's involvement. I'm nudging
Mike to come up with someone else for his "side" of the debate. And I may
get one more panelist. Still toying with that.
On a related note:
Vendors: I have a verbal commitment from Nexis for a representative to be
there and enough interest expressed by DataTimes I assume they'll agree to
be there. I'm tracking down contacts for Dow-Jones and Dialogue. What
about UMI? WIll DT be sufficient for them (this is DT as a processor
rather than a direct vendor). I'm pitching this as something in which they
have an obvious vested interest, which they recognize. I'm saying they
don't have to come prepared to say anything. They can answer questions if
they choose, but most importantly they would be there to listen -- and, we
hope, heed.
Pete, Barbara, if there's anything I missed or misstated, please let me
know.
Carol, is this sufficient to your needs?
Hey, it's Friday. That's all we know and all we need to know.
Jackie Chamberlain jchamberlain@pe.com
The Press-Enterprise Phone - (909) 782-7579
Riverside CA Fax - (909) 782-7572
From semonch@gibbs.oit.unc.eduMon Nov 10 15:43:19 1997
Date: Mon, 10 Nov 1997 15:38:13 -0500
From: Barbara Semonche
To: DBQC@listserv.oit.unc.edu
Subject: RE: One more, final, chapter to the Nexis/DataTimes Saga (fwd)
To all:
Ok, Ok, so I don't know how to count. I guess this is # 8.
Barbara P. Semonche, Library Director (Listowner: NewsLib, NCSLA)
UNC-CH School of Journalism and Mass Communication
Howell Hall, CB# 3365 Phone: 919/962-1204, Fax: 962-0620
Chapel Hill, NC 27599-3365 E-mail: semonch@gibbs.oit.unc.edu
Semonche's URL http://sunsite.unc.edu/journalism/
---------- Forwarded message ----------
Date: Fri, 31 Oct 1997 12:40:05 -0800
From: Jackie Chamberlain
To: "'newslib@listserv.oit.unc.edu'"
Subject: RE: One more, final, chapter to the Nexis/DataTimes Saga
Kathy,
Thanks for the kind words. The sad part is all this could have been
avoided. Actually this may not be *quite* the final chapter. My eagle-eyed
systems guy has some question, now, about whether re-sending affected docs
thru the GenData process (as opposed to new corrections) may cause the
same problem. I'm posting a separate addendum to that effect, right now,
on list.
Jackie Chamberlain
----------
From: Kathy Foley[SMTP:kfoley@express-news.net]
Sent: Friday, October 31, 1997 12:24 PM
To: newslib@listserv.oit.unc.edu
Subject: Re: One more, final, chapter to the Nexis/DataTimes Saga
Jackie - Thanks for taking the time to explain this very techie, but
very critical horror story. Even though I know longer feel responsible
for the integrity of a Basis database - I remember the terms you are
using and shiver to think of the problems others might have also been
having without realizing. This is not to speak ill of Basis - just the
opposite. I have never felt as secure about my newspaper archives as I
did when I did with PostHaste at the Washington Post. I'm also pleased
that you reported such good things about DataTimes and Mike Harris. He
runs a great team!
We all bear a great deal of responsibility for our newspaper's precious
history and content. I'm not sure we have received the recognition we
deserve for how well we have handled this responsibility. But our
children and our grandchildren will thank us!
From semonch@gibbs.oit.unc.eduMon Nov 10 15:48:29 1997
Date: Mon, 10 Nov 1997 15:41:02 -0500
From: Barbara Semonche
To: DBQC@listserv.oit.unc.edu
Subject: Addendum to Nexis/DataTimes Saga....... (fwd)
To all:
Gee Whiz! This is more reading than I thought. I hereby confer on
all of you a M.A. P. [That stands for "master's in patience"]. Here is # 9.
BTW, I'm caching this thread for the web page. Updates will be
added. More anon.
Barbara P. Semonche, Library Director (Listowner: NewsLib, NCSLA)
UNC-CH School of Journalism and Mass Communication
Howell Hall, CB# 3365 Phone: 919/962-1204, Fax: 962-0620
Chapel Hill, NC 27599-3365 E-mail: semonch@gibbs.oit.unc.edu
Semonche's URL http://sunsite.unc.edu/journalism/
---------- Forwarded message ----------
Date: Fri, 31 Oct 1997 13:03:45 -0800
From: Jackie Chamberlain
To: "'newslib@listserv.oit.unc.edu'"
Cc: "'sclairemont@link.freedom.com'" ,
"'sharon.rreves@uniontrib.com'" ,
"'sherry.adams@chron.com'" ,
"'tglib@telegram.infi.net'"
Subject: Addendum to Nexis/DataTimes Saga.......
There's always something.
Tuesday I wrote, especially for Basis users, what I thought was the
"final" chapter in the saga: The "Nexis Fix" for corrected long docs caused
them to bomb at DataTimes, but when discovered was promptly solved by
DataTimes. (If you didn't get the Tuesday message and may be affected,
let me know.)
One possible hitch -- and this has only to do with "retro" stories, stories
you'd have to send again to DT to fix the problem if you were using the
Nexis "solution" and extracting using HTPrint for Nexis:
The new DT fix, which guarantees that long corrected documents make it to
both DataTimes & Nexis OK may -- or may *not* -- work for re-sending the
docs affected by "putting' all parts of a story. My eagle/eyed sysyems guy
Jim Zikratch, is a little skeptical about send the affected docs (anything
"double-putted") back through GenData because they may double delete
again. He's researching that and will be talking to DataTimes next week.
Again, if your technical people have any questions, Mike Harris at DT is
more than willing to help, and once we have a solution I'll let everyone
know. Meanwhile my advice is go ahead and send newly corrected docs the
old-fashioned way, but hold off on doing anything about stuff you "double-
putted" earlier.
Does this make any sense?
Jackie Chamberlain jchamberlain@pe.com
The Press-Enterprise Phone - (909) 782-7579
Riverside CA Fax - (909) 782-7572
|