Wednesday, March 1, 1995 6:41:04 PM
GenWeb Item
From: Chuck Tucker,ctucker@coyote.csusm.edu,Internet
Subject: WWW Home Page
To: GenWeb
Can anyone recommend a source of instructions on how to make up a home page
on the web (or is it more proper to say make up an html document)? I'm just
a little lost in all of this hyper text type stuff.
Thanks,
Chuck
Thursday, March 2, 1995 7:44:12 AM
GenWeb Item
From: Jim Isaak-
isaak@csac.enet.dec.com,isaak@csac.ljo.dec.com,Internet
Subject: Sources of info on HTML structure and tools
To: GenWeb
Chuck,
here are some web pointers to information about HTML
and tools available to create web pages. I just installed
the Microsoft package on my PC w/Word 6, and it works fairly
well (although is not as well documented as I would like, but
that is the norm for many of these packages available over
the net).
What I'm interested in is the "style" that we should use
to create genealogical pages. I just joined the distribution list,
so this may be an old topic (can someone point me to an archive, or
a web page that expands on the concepts) ... Questions I see in this
reguard:
Family Group pages vs individual/marriage pages (this is
"open" from what I can tell)
norm for "title" on a page (name, bdate??) [wais can search this]
sequence of information on a page
norm for references - (end of page, hot links...)
should pages have internal "anchors" {#labels} and what?
for example ... akin to GEDCOM
or ... for spouse, kids, etc.
Jim Isaak, N.H.
For information on the structure of HTML
========================================
http://info.cern.ch/hypertext/WWW/MarkUp/MarkUp.html
points to both the "formal" documents describing the
language, and also the tutorials & primers online.
Tools for creating HTML (Web) documents:
========================================
see:
http://info.cern.ch/hypertext/WWW/Tools/Overview.htm
[Microsoft word package, see:]
http://www.microsoft.com/pages/deskapps/word/ia/default.htm
Thursday, March 2, 1995 7:44:12 AM
GenWeb Item
From: Jim Isaak-
isaak@csac.enet.dec.com,isaak@csac.ljo.dec.com,Internet
Subject: Sources of info on HTML structure and tools
To: GenWeb
Chuck,
here are some web pointers to information about HTML
and tools available to create web pages. I just installed
the Microsoft package on my PC w/Word 6, and it works fairly
well (although is not as well documented as I would like, but
that is the norm for many of these packages available over
the net).
What I'm interested in is the "style" that we should use
to create genealogical pages. I just joined the distribution list,
so this may be an old topic (can someone point me to an archive, or
a web page that expands on the concepts) ... Questions I see in this
reguard:
Family Group pages vs individual/marriage pages (this is
"open" from what I can tell)
norm for "title" on a page (name, bdate??) [wais can search this]
sequence of information on a page
norm for references - (end of page, hot links...)
should pages have internal "anchors" {#labels} and what?
for example ... akin to GEDCOM
or ... for spouse, kids, etc.
Jim Isaak, N.H.
For information on the structure of HTML
========================================
http://info.cern.ch/hypertext/WWW/MarkUp/MarkUp.html
points to both the "formal" documents describing the
language, and also the tutorials & primers online.
Tools for creating HTML (Web) documents:
========================================
see:
http://info.cern.ch/hypertext/WWW/Tools/Overview.htm
[Microsoft word package, see:]
http://www.microsoft.com/pages/deskapps/word/ia/default.htm
Thursday, March 2, 1995 3:48:19 PM
GenWeb Item
From: R.Mclee,R.McLee@sheffield.ac.uk,Internet
Subject: THE BIG R....contact address.......
To: GenWeb
#####################################
# #
# #### ### #### #### #
# # # # # # # #
# # # # # # # #
# #### # # # # #
# # # # # ### #### #
# # # # # # # # #
# #### ### #### # # #
# #
#####################################
Those of you with roots in the uk but have very few contacts with
those interested in the same family name can get in touch with each
other via the BIGR; The British Isles Genealogical Register.
It is a compiled database of subscribers who have included their
interests in all of Surname, Town, Area/County, Period. The option
to say anywhere at anytime was also given.....stretching the field of
coverage even more. I played a small part in the running of this huge
project, which (I am told) will be running again soon. My personal
belief is that this information is vital for anyone who has roots in
Britain, linking two parties with a common interest.
INTERESTED?.......
The full Register is available on microfiche ONLY (it is to big for
paper that's for sure)....As a whole it covers 23 fiche costing 25
pounds STERLING.
You may wish to purchase the whole thing but I believe it is
available from societies in County sections.
CONTACT:
FFHS Publications
2-4 Killing Street
Ramsbottom
Lancashire
BL0 9B7
England
PLEASE NOTE....STIRLING MONEY ORDERS ARE PREFERED!!!!!!!
If I can help anyone else with their research in the UK...though I
must stress I am NO GENEALOGIST!!!! I'm a student in Electronic
engineering, Feel free to e-mail me.
Happy Hunting!
Yours Ross McLee.
Friday, March 3, 1995 10:36:10 AM
GenWeb Item
From: Timothy B. Doyle,tdoyle@netcom.com,Internet
Subject: Questions...
To: GenWeb
Hello all:
Please excuse this post if it repeats questions asked often. If there is a
FAQ available, please point the way.
Several months ago I investigated the possibility of placing my data on the
net and looked at several different approaches. I quickly realized that
because my service provider (Netcom) did not have http access (they only
have ftp), cgi scripts were out of the question and thus I put my effort on
hold.
I have since joined forces with two co-workers and we now have three
ethernet-linked computers running Linux and connected to the net via a 28.8
dedicated line. The 28.8 line will probably be upgraded in the semi-near
future. Our http server is up and I now have the ability to utilize cgi
scripts.
The intent of this post is to request some guidance. As I see it, there are
several approaches to placing genealogical data on the net. They include:
1. Generate one html for the entire database. This isn't practical for
anything but a tiny database as the resulting html file would be too large
and would take way too long to send to the user.
2. Generate html documents for each person (or family). These files would be
small enough to quickly send to the user, but there would potentially be a
huge number of files.
3. Use cgi scripts to access the GEDCOM and generate html documents on the
fly. This seems like the best method (for me at least).
I have been able to locate and install Lifelines, and I have the ll and
standardperson scripts, which are needed to use approach number 3. My next
task, I believe, is to find a program that will index my GEDCOM file. I
remember working with such a program when I was investigating this whole
thing earlier, but I cannot remember where I found it. Basically, when run,
it would generate an index for every 500 people in the database and create
an 'index of indexes'. Does anyone know if this utility is still available?
Are there any other approaches that you might recommend?
Tim Doyle | "I do not feel obligated to believe that the same
tdoyle@netcom.com | God who has endowed us with sense, reason, and
| intellect has intended us to forego their use."
| --Galileo Galilei
WWW homepage: ftp://ftp.netcom.com/pub/td/tdoyle/welcome.html
ftp directory: ftp.netcom.com pub/td/tdoyle
Monday, March 6, 1995 3:41:15 PM
GenWeb Item
From: Timothy B. Doyle,tdoyle@netcom.com,Internet
Subject: Doyle Database online - GenWeb Index online
To: GenWeb
I have succeeded in getting my 7,000+ person database on the web. My
database has strong collections for the
Chardron/Cherdron/Shedron/Shetron/Shetrone and Treffeisen families (scant
Doyle info). The method that I am currently using is the one that Vic Abell
devised. This method keeps the number of files to a minimum, but is
relatively quick. There are still a few bugs in this system (some parent
links are missing, only one line of notes print, etc.), so I'm not sure if
this is the method that I will stay with, but it is working for now.
I have also created an Index of GenWeb Databases containing all 27 that I
was able to locate. This index is located at
http://www.doit.com/tdoyle/genweb/genweb.html. The index is currently
categorized by the method being used to create the html files, but the
organization may change in the future. If you know of other databases
online, please let me know. Also, if I have any information incorrect on
this index, I would appreciate hearing from you.
My homepage, which contains other genealogical links is located at
http://www.doit.com/tdoyle/
Thanks to everyone that sent replies to my plea for assistance.
Tim Doyle
tdoyle@doit.com / tdoyle@netcom.com
WWW homepage: ftp://www.doit.com/tdoyle/
ftp directory: doit.com pub/tdoyle
Tuesday, March 7, 1995 7:13:01 AM
GenWeb Item
From: Scott McGee,smcgee@microware.com,Internet
Subject: mail failed to Tim Doyle
To: GenWeb
Tim,
I tried to send you a comment from your genweb page, but my mailer says
that doit.com does not exist.
What I wanted to say is that my genweb stuff could be put under the
"Personal HTML Method" heading. They were generated with the LifeLines
ll2html report, and then hand modified. I have a much better genweb
site online (over 4000 names) that is handled by LifeLines and CGI
but not quite the way you describe the LifeLines method. It doesn't
generate a gedcom for the individual and then convert that to HTML, but
directly generates an HTML page from the database. Out server currently
does not allow offsite access so I can't give you a working URL. When I
can, I'll replace my former stuff with a link to it and anounce it here.
I like your genweb index! Good information!
Scott
PS Anyone have a GenWeb logo? I want to build a GenWeb page with links
into my own stuff. It would make a better front end to my genweb stuff.
When in danger, | If it has my name on it, it must be MY opinion!
or in doubt, |______________________________________________________
run in circles, | Email: smcgee@microware.com (Scott McGee)
scream and shout! | Web: http://www.cc.utah.edu/~sam8644/homepage.html
Tuesday, March 7, 1995 8:30:12 AM
GenWeb Item
From: David J. Ward,David_Ward@sirius.UVic.CA,Internet
Subject: Re: Doyle Database online - GenWeb Index online
To: GenWeb
> The fact that I am new to this group will quickly be obvious, but
> the genealogy info that I have set up has narrative history plus
> the various dates for family members of my many-great grandparents.
> I see Timothy's site and most of the rest also use index format.
> Is this because it takes up too much space to give the history
> or because indexing is enough for Internet researching?
Nancy,
In my opinion the narrative history is just as good (or better)
than the indexing method.
The reason that I am currently using the index method
(http://www-engr.uvic.ca/~dward/gen/index.html) is that I am
just playing with getting my information on-line and am not
yet sure how I want it to look. I have already tried a couple
of methods, and will probably change it again. Another item that
is important for me is that I want the method I use in generating
the html documents to be fairly automatic. I use lifelines under
unix for my database, and am starting to change scripts that other
individuals have written so that I can get the information that I
want into the html documents.
I think that the indexing methods is a good COMPLEMENT to the narative
approach as it does aid in searching, but it depends upon the purpose
of the information as to how you want it normally displayed.
David
-------------------------------------------------------------------------------
| David J. Ward | Email: David_Ward@CSc.UVic.CA |
| CSCU Vice President | ub057@freenet.Victoria.BC.CA |
| University of Victoria | Phone: (604) 727-0929 |
| Victoria, B.C., Canada | Radio: VE7DWJ (usually 146.84 MHz in Victoria) |
-------------------------------------------------------------------------------
Tuesday, March 7, 1995 9:39:59 AM
GenWeb Item
From: Nancy Sween,nsween@falcon.cc.ukans.edu,Internet
Subject: Re: Doyle Database online - GenWeb Index online
To: GenWeb
On Tue, 7 Mar 1995, David J. Ward wrote:
> (http://www-engr.uvic.ca/~dward/gen/index.html)
> ... I want the method I use in generating
> the html documents to be fairly automatic. I use lifelines under
> unix for my database
To those of you with more experience in genealogy html doc sites,
is it impractical to set up files with a header and then just use
and
to turn family history files into html docs? That
way the only non-automatic thing left to do is link names in one
document to relatives in other documents. Doesn't it take awhile
to convert info into lifelines? I don't think I have that kind of
time, unless your experience shows that it's worth the time.
Thanks.
Nancy Sween "The Interactive Medical Student Lounge"
MD Student Records URL: http://falcon.cc.ukans.edu/~nsween
Tuesday, March 7, 1995 9:50:43 AM
GenWeb Item
From: Martin Ott,p581mao@sun91.mpifr-bonn.mpg.de,Internet
Subject: Re: Doyle Database online - GenWeb Index online
To: GenWeb
> was able to locate. This index is located at
> http://www.doit.com/tdoyle/genweb/genweb.html. The index is currently
403 Forbidden
> categorized by the method being used to create the html files, but the
> organization may change in the future. If you know of other databases
> online, please let me know. Also, if I have any information incorrect on
> this index, I would appreciate hearing from you.
>
> My homepage, which contains other genealogical links is located at
> http://www.doit.com/tdoyle/
"Read 4926 of 4744 bytes of data."
... and hung.
Martin
Wednesday, March 8, 1995 6:30:38 AM
GenWeb Item
From: Nancy Sween,nsween@falcon.cc.ukans.edu,Internet
Subject: Re: Doyle Database online - GenWeb Index online
To: GenWeb
The fact that I am new to this group will quickly be obvious, but
the genealogy info that I have set up has narrative history plus
the various dates for family members of my many-great grandparents.
I see Timothy's site and most of the rest also use index format.
Is this because it takes up too much space to give the history
or because indexing is enough for Internet researching?
Tim, since my new site is so different maybe it's not ready for
prime time yet. But if variety is really the spice of life,
would you consider linking to it from your home page?
GEORGE & HANDLEY families
http://falcon.cc.ukans.edu/~nsween/genealogy.html
/george.html or /handley.html for faster direct access
Nancy Sween nsween@falcon.cc.ukans.edu
Wednesday, March 8, 1995 11:34:03 AM
GenWeb Item
From: MR RICHARD E VAN METER,LDMS30A@prodigy.com,Internet
Subject: The Van Meter / Van Metre Family
To: GenWeb
Grettings,
My name is Richard E. (Rick) Van Meter.
I'm attempting to gather as much information as possible
on the Van Meter / Van Metre Family. Any assistance would
be very much appreciated.
Thank you
Rick
Thursday, March 9, 1995 2:52:07 PM
GenWeb Item
From: ANNELISE ANDERSON,ANDRSN@HOOVER.STANFORD.EDU,Internet
Subject: GEDCOM to HTML
To: GenWeb
I downloaded a file from the home page of someone whose last name is Stark
that is a try at GEDCOM to HTML but the file has a .gz extension and I'm
not sure what to do with it now. Would appreciate any suggestions.
I too am unable to access T. Doyle's home page.
Annelise Anderson
andrsn@hoover.stanford.edu
Thursday, March 9, 1995 6:03:20 PM
GenWeb Item
From: Phlete Teachout,fteachou@eagle1.eaglenet.com,Internet
Subject: Re: GEDCOM to HTML
To: GenWeb
On Thu, 9 Mar 1995, ANNELISE ANDERSON wrote:
>
> I too am unable to access T. Doyle's home page.
>
I had no trouble getting to GENWEB site this evening (9 Mar, 2100 EST) at
the following URL:
http://www.doit.com/tdoyle/genweb/genweb.html
- fleet -
<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
< P.R. "Fleet" Teachout - Net Surfer >
< fteachou@eagle1.eaglenet.com >
< >
< "Knowledge is of two kinds. We know a subject ourselves, or >
< we know where we can find information upon it." >
< Samuel Johnson 1709-1784 >
<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
Friday, March 10, 1995 6:18:25 AM
GenWeb Item
From: trevor_western@ptltd.com,Internet
Subject: Re: GEDCOM to HTML
To: GenWeb
I used the same URL a few minutes ago (9:00AM EST) and it is correct and
"up and running".
-- Trevor
-------------
Original Text
From Phlete Teachout , on 3/9/95 4:01 PM:
On Thu, 9 Mar 1995, ANNELISE ANDERSON wrote:
>
> I too am unable to access T. Doyle's home page.
>
I had no trouble getting to GENWEB site this evening (9 Mar, 2100 EST) at
the following URL:
http://www.doit.com/tdoyle/genweb/genweb.html
- fleet -
<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
< P.R. "Fleet" Teachout - Net Surfer >
< fteachou@eagle1.eaglenet.com >
< >
< "Knowledge is of two kinds. We know a subject ourselves, or >
< we know where we can find information upon it." >
< Samuel Johnson 1709-1784 >
<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
Monday, March 13, 1995 1:31:01 PM
GenWeb Item
From: Gary Hoffman,ghoffman@ucsd.edu,Internet
Subject: genweb.org established
To: GenWeb
After some delays, we have finally established the genweb.org domain with
ucsd.edu acting as the primary name server. The first task was to put my
GenWeb
proposal and supporting files into the genweb domain in anticipation of a
server
change in my department. I now request everyone maintaining links to URL
http://irpsbbs.ucsd.edu/gene/genedemo.html to please change them to URL
http://demo.genweb.org/gene/genedemo.html.
Here are the CNAMEs or host aliases I have set up so far:
demo.genweb.org -> irpsbbs.ucsd.edu
austin.genweb.org -> www.rahul.net
abell.genweb.org -> www-rcd.cc.purdue.edu
charles.genweb.org-> faui80.informatik.uni-erlangen.de
Others interested in joining the genweb.org domain are welcome to e-mail
me.
Please include the host name and the full path to your GenWeb entry page.
The above CNAMEs only replace the host name in the URL for a particular
GenWeb
site. Please note that the internal path to the html pages in question must
still
be maintained. However, I have a suggestion that will simplify that
process, too.
Please put an index page at the root level accessible to the http process.
The
index page should be named "gendex.html". This will allow the following
address
yield a good page: http://xxx.genweb.org/gendex.html. In some cases, this
may
involve a special request to your system administrator in order to put this
page
in place. The gendex.html document should include links to all GenWeb
resources
on that server and may also include other genealogy related materials.
Cheers,
Gary
***************************************************************************
*Gary B. Hoffman, Computer/Language Lab Director e-mail: ghoffman@ucsd.edu*
*Graduate School of International Relations and Pacific Studies (IR/PS)*
*University of California, San Diego (UCSD) voice: (619) 534-7733*
*9500 Gilman Dr., La Jolla, CA 92093-0519 USA fax: (619) 534-3939*
***************************************************************************
Monday, March 13, 1995 3:54:47 PM
GenWeb Item
From: Anders Andersson,andersa@Mizar.DoCS.UU.SE,Internet
Subject: Re: genweb.org established
To: GenWeb
Gary Hoffman writes:
>demo.genweb.org -> irpsbbs.ucsd.edu
>austin.genweb.org -> www.rahul.net
>abell.genweb.org -> www-rcd.cc.purdue.edu
>charles.genweb.org-> faui80.informatik.uni-erlangen.de
As suggested by both me and Mike Patton earlier, *please* put an
intermediary domain label between "GENWEB.ORG" and the individual
database names, such as "DEMO.DB1.GENWEB.ORG". Chances are that
the method of implementation will change as we get along, and you
may want to start all over with another set of the same database
names, though with different kinds of records.
Specifically, you may (sometime in the future) want to add a TXT
record providing additional information about each GENWEB database.
As "DEMO.GENWEB.ORG" is a CNAME, you can't use a TXT record with
the same name, and you would have to create a new namespace anyway
(such as "DEMO.DB2.GENWEB.ORG").
A separate database domain would also clarify the logical structure.
In case you add a CNAME record for, say, "WWW.GENWEB.ORG", you have
no straightforward way of telling that this particular CNAME isn't
like "DEMO.GENWEB.ORG" and the other databases (other than mentally
recognizing that particular acronym "WWW", but don't expect humans
to be around interpreting these domain names forever). You will
be cluttering up your namespace with lots of different names for
different purposes, and you will make certain experimental
configurations of the GENWEB.ORG zone more difficult to set up
(I can offer examples with lots of technical detail, but that's
pretty much beyond the scope of the GENWEB mailing list right now),
unless you use an intermediary domain label.
It doesn't really matter what particular domain label you choose to
indicate this first set of databases. It could be "DB", "DB1", "FOO"
or whatever. Just avoid those which are likely to be mistaken for
entirely different services (such as "FTP" or "MAIL"). You certainly
won't lose anything by adding a level at this early stage, but you
may gain some flexibility as the GENWEB grows.
--
Anders Andersson, Dept. of Computer Systems, Uppsala University
Paper Mail: Box 325, S-751 05 UPPSALA, Sweden
Phone: +46 18 183170 EMail: andersa@DoCS.UU.SE
Tuesday, March 14, 1995 2:45:50 AM
GenWeb Item
From: Gene Stark,starkhome!gene@sbstark.cs.sunysb.edu,Internet
Subject: Re: genweb.org established
To: GenWeb
It would be nice if the server for genweb.org permitted queries to
list all the domain information. Right now I get:
% nslookup
> set debug
> ls -d genweb.org
;; res_mkquery(0, genweb.org, 1, 252)
[localhost.cs.sunysb.edu]
*** Can't list domain genweb.org: Query refused
>
I would think that the normal security reasons for prohibiting this kind of
query would not apply to the use of DNS for genweb.
- Gene Stark
Tuesday, March 14, 1995 4:39:07 AM
GenWeb Item
From: Anders Andersson,andersa@Mizar.DoCS.UU.SE,Internet
Subject: Re: genweb.org established
To: GenWeb
Gene Stark writes:
> > ls -d genweb.org
> ;; res_mkquery(0, genweb.org, 1, 252)
> [localhost.cs.sunysb.edu]
> *** Can't list domain genweb.org: Query refused
>I would think that the normal security reasons for prohibiting this kind of
>query would not apply to the use of DNS for genweb.
Which server were you querying? From the above, it seems that you
queried LOCALHOST.CS.SUNYSB.EDU (which is your local 127.0.0.1), not
UCSD.EDU or NETWORK.UCSD.EDU (the official name servers for GENWEB.ORG).
I have no problem obtaining a copy of the entire zone from UCSD.EDU:
>;
>; Last update Tue Mar 14 13:09:59 1995
>;
>$ORIGIN ORG.
>GENWEB IN SOA genweb.org. ghoffman.ucsd.edu. (
> 95031201 36000 1800 360000 86400 )
> IN NS ucsd.edu.
> IN NS network.ucsd.edu.
>$ORIGIN GENWEB.ORG.
>austin IN CNAME www.rahul.net.
>demo IN CNAME irpsbbs.ucsd.edu.
>abell IN CNAME www-rcd.cc.purdue.edu.
>charles IN CNAME faui80.informatik.uni-erlangen.de.
>mailhost IN CNAME localhost.GENWEB.ORG.
>localhost IN A 127.0.0.1
>loghost IN CNAME localhost.GENWEB.ORG.
>hoffman IN CNAME irpsbbs.ucsd.edu.
However, I don't understand why this domain already has a "localhost",
a "mailhost" and a "loghost" when there are still no physical hosts in
GENWEB.ORG. Is some host somewhere else set up to use GENWEB.ORG as
its default domain, or what?
I don't see any use for those three, but if they are going to stay,
it further stresses the point of having a separate subdomain for the
database alias names, just so that we can avoid having to check for
silly name conflicts for every new database.
--
Anders Andersson, Dept. of Computer Systems, Uppsala University
Paper Mail: Box 325, S-751 05 UPPSALA, Sweden
Phone: +46 18 183170 EMail: andersa@DoCS.UU.SE
Tuesday, March 14, 1995 1:48:16 PM
GenWeb Item
From: Gene Stark,starkhome!gene@sbstark.cs.sunysb.edu,Internet
Subject: genweb.org established
To: GenWeb
>Which server were you querying? From the above, it seems that you
>queried LOCALHOST.CS.SUNYSB.EDU (which is your local 127.0.0.1), not
>UCSD.EDU or NETWORK.UCSD.EDU (the official name servers for GENWEB.ORG).
Yes, that is what I did. I was under the impression that normal use
of DNS involved querying a local server, which then delegates queries
for domains for which it is not a primary or secondary.
The delegation worked just fine, as I was able to resolve the individual
demo entries. I just couldn't list the entire domain. I suppose I can
configure my server as a secondary, but I don't really want to.
- Gene
Tuesday, March 14, 1995 2:08:42 PM
GenWeb Item
From: Gary Hoffman,ghoffman@ucsd.edu,Internet
Subject: subdomains to genweb.org domain
To: GenWeb
To the GenWeb mailing list:
Anders Andersson wrote:
As suggested by both me and Mike Patton earlier, *please* put an
intermediary domain label between "GENWEB.ORG" and the individual
database names, such as "DEMO.DB1.GENWEB.ORG". Chances are that
the method of implementation will change as we get along, and you
may want to start all over with another set of the same database
names, though with different kinds of records.
Gary replies:
I appreciate your desire to prepare for any eventuality with regard to how
GenWeb might develop. However, I still do not appreciate the advantage that
an additional layer of domain name gives us. As I see it, if we change our
method of implementation in the future, we should probably create a new
domain to describe it (son of genweb, for example). After all, these are
not primary names, just aliases and we can change them at will.
I think the bigger problem is unifying the pathway within each host in the
domain so that we can easily reach the genealogy databases on that host. In
fact, if someone does change their method of implementing a genealogy
database, they will probably change the directly pathway to their new
implementation anyway.
I am presently in favor of keeping things simple in the beginning and
creating layers of complexity only when it is useful. I have seen brillant
minds bog down trying to anticipate every possible outcome of an idea such
that they never commit themselves to a firm expression of that idea, and no
one every hears about it. On the other hand, I will strike out in some
(any) direction to get something started, even if it means that later I
have to backtrack and start all over again in a different direction. That
is what I have done with genweb.org.
Cheers,
Gary
***************************************************************************
*Gary B. Hoffman, Computer/Language Lab Director e-mail: ghoffman@ucsd.edu*
*Graduate School of International Relations and Pacific Studies (IR/PS)*
*University of California, San Diego (UCSD) voice: (619) 534-7733*
*9500 Gilman Dr., La Jolla, CA 92093-0519 USA fax: (619) 534-3939*
***************************************************************************
Tuesday, March 14, 1995 2:50:51 PM
GenWeb Item
From: Chris Garrigues,cwg@DeepEddy.Com,Internet
Subject: Re: genweb.org established
To: GenWeb
At 2:10 PM 3/14/95, Gene Stark wrote:
> >Which server were you querying? From the above, it seems that you
> >queried LOCALHOST.CS.SUNYSB.EDU (which is your local 127.0.0.1), not
> >UCSD.EDU or NETWORK.UCSD.EDU (the official name servers for GENWEB.ORG).
>
> Yes, that is what I did. I was under the impression that normal use
> of DNS involved querying a local server, which then delegates queries
> for domains for which it is not a primary or secondary.
>
> The delegation worked just fine, as I was able to resolve the individual
> demo entries. I just couldn't list the entire domain. I suppose I can
> configure my server as a secondary, but I don't really want to.
If you ask for the entire zone, you have to go to one of the servers for
the zone.
Chris
Chris Garrigues cwg@DeepEddy.Com
My pgp public key is on my homepage: http://www.DeepEddy.Com/~cwg/
Tuesday, March 14, 1995 4:03:41 PM
GenWeb Item
From: Gene Stark,starkhome!gene@sbstark.cs.sunysb.edu,Internet
Subject: subdomains to genweb.org domain
To: GenWeb
I also agree with Anders Andersson and Mike Patton regarding the
advisability of adding an intermediary domain between the database
names and GENWEB.ORG.
With them, I strongly urge you to do this, and believe it will
turn out to be a serious mistake if you do not. The point is to not
have to throw away everything and start over. The method they propose
will make it possible to experiment with multiple naming schemes for
the same databases simultaneously, as well as to evolve seamlessly
to new versions of GenWeb.
- Gene Stark
Wednesday, March 15, 1995 9:15:01 AM
GenWeb Item
From: Everton,jayhall@xmission.com,Internet
Subject: Adress change for Boyle's Civil War pages
To: GenWeb
The 1 March issue of Everton's Genealogical Helper: online edition
(http://www.xmission.com/~jayhall/ghonline.html) contains a review of
Boyle's excellent pages of U.S. Civil War information. The address for
Boyle's pages has been changed to:
http://www.access.digex.net/~bdboyle/cw.html
If you are seeking information on the U.S. Civil War, or just want some
background "flavor" on this time in history, I heartily recommend this
site.
------------------------------------------------------------------------------
Jay Hall, Everton Publishers | E-Mail = jayhall@xmission.com
165 South West Temple, Suite 200 | WWW = http://www.xmission.com/~jayhall/
Salt Lake City, UT 84101 | ftp = ftp.xmission.com/pub/users/j/jayhall/
Wednesday, March 15, 1995 12:45:30 PM
GenWeb Item
From: Gary Hoffman,ghoffman@ucsd.edu,Internet
Subject: Intermediate Domain Name
To: GenWeb
To GenWebbers:
I admit I am still very much a novice when it comes to Domain Name Service,
so I am asking these questions as inquiries, not as challenges. Any light
you can lend to the subject will be appreciated.
Anders Andersson wrote:
As suggested by both me and Mike Patton earlier, *please* put an
intermediary domain label between "GENWEB.ORG" and the individual
database names, such as "DEMO.DB1.GENWEB.ORG". Chances are that
the method of implementation will change as we get along, and you
may want to start all over with another set of the same database
names, though with different kinds of records.
And Gene Stark wrote:
I also agree with Anders Andersson and Mike Patton regarding the
advisability of adding an intermediary domain between the database
names and GENWEB.ORG.
With them, I strongly urge you to do this, and believe it will
turn out to be a serious mistake if you do not. The point is to not
have to throw away everything and start over. The method they propose
will make it possible to experiment with multiple naming schemes for
the same databases simultaneously, as well as to evolve seamlessly
to new versions of GenWeb.
So I wonder:
I assume you recommend a format like "demo.db1.genweb.org" so that it can
migrate to "demo.db2.genweb.org" some day. Perhaps I don't appreciate the
subtleties involved here. Why won't the following work just as well:
"demo1.genweb.org" migrating to "demo2.genweb.org"? Since these are merely
CNAMEs anyway, they have to point to a real canonical name which may or may
not change as needed. Is there some management issue that I'm not
perceiving?
Thanks for your assistance.
Gary
***************************************************************************
*Gary B. Hoffman, Computer/Language Lab Director e-mail: ghoffman@ucsd.edu*
*Graduate School of International Relations and Pacific Studies (IR/PS)*
*University of California, San Diego (UCSD) voice: (619) 534-7733*
*9500 Gilman Dr., La Jolla, CA 92093-0519 USA fax: (619) 534-3939*
***************************************************************************
Wednesday, March 15, 1995 1:46:33 PM
GenWeb Item
From: Nancy Sween,nsween@falcon.cc.ukans.edu,Internet
Subject: Civil War site
To: GenWeb
Permission has not yet been given to outside readers for Boyle's
Civil War site, it looks like.
Nancy Sween
Wednesday, March 15, 1995 5:38:22 PM
GenWeb Item
From: Anders Andersson,andersa@Mizar.DoCS.UU.SE,Internet
Subject: Re: Intermediate Domain Name
To: GenWeb
Gary Hoffman writes:
>I assume you recommend a format like "demo.db1.genweb.org" so that it can
>migrate to "demo.db2.genweb.org" some day. Perhaps I don't appreciate the
>subtleties involved here. Why won't the following work just as well:
>"demo1.genweb.org" migrating to "demo2.genweb.org"? Since these are merely
>CNAMEs anyway, they have to point to a real canonical name which may or may
>not change as needed. Is there some management issue that I'm not
>perceiving?
Thank you for the opportunity. I'll take it to offer a longer
response, hoping that I can clarify my position. Those who aren't
concerned with the Domain Name System or naming issues in general
can safely skip the rest of the message. You have been warned...
1. Why is changing alias names a problem?
Whether the names are CNAMEs or something else isn't really a big
matter. Yes, you can easily change what *they* point to, but the
issue here is to decide what those names will be before you suggest
*referring* to them in the hundreds (thousands?) from several
genealogical databases around the Internet.
Changing DEMO.GENWEB.ORG from pointing (via the CNAME record) at
"irpsbbs.ucsd.edu" (which is does right now) to point at, say,
"irpsbbs2.ucsd.edu" (because you moved your database to a second
machine) is a small matter, which you can handle without even
telling the GenWeb community. It doesn't matter at all how many
GenWeb users there are, or how far references to the domain names
have spread. This is (partly) why we have CNAME records.
Changing the supposedly well-used alias "DEMO.GENWEB.ORG" itself
to instead be "DEMO.DB1.GENWEB.ORG" (or anything else) is quite a
different thing, assuming that you ultimately want to remove the
original name from the DNS (if the alias change implies keeping the
old name around forever, then I think the entire point of changing
it must have been lost somewhere).
Assuming a few dozen people have started to cross-reference individual
records in each other's databases by means of GENWEB.ORG alias names,
they all have to update their databases before you can remove the old
names from the DNS. What about occurrences of these cross-references
in copies of GEDCOM records or entire files which have migrated to
off-line media, such as the disks of non-networked genealogists, or
even anyone's long-term backup storage? They may want to use those
references once they get on-line again. Depending on their skill,
enthusiasm, available tools and other circumstances, they may or may
not be able to manually or semi-automatically convert their old data,
but given that we today have experience from multiple generations of
various software applications having required "conversion" of old
files (if it has been possible at all), it seems wise to try to avoid
such conversions whenever they can be foreseen and easily avoided.
2. Can we anticipate future GenWeb development in any way?
I think (and I hope) that every GenWeb list member has at least some
idea of what GenWeb may look like in the future, and that only a small
fraction of those ideas have so far been discussed publicly. Here I
can of course only provide my own ideas.
There seems to be little dispute against the opinion that we *will*
try more than one solution for mapping cross-reference identifiers to
actual records and, indirectly, their physical locations. The CNAME
convention being employed now is just a crude beginning, and only a
very partial solution (in that we haven't resolved how to identify
the actual database or the records it contains, once we have found out
which server it sits on). Gary mentioned the "gendex.html" index in
a recent message, and we need to discuss that further.
However, another mechanism need not, for instance, depend on the use
of a particular access protocol such as HTTP. I anticipate that for
really long-lived cross-references in genealogical records, we will
need something like a : pair, with every
reference to specific access methods such as "http", "genweb.org" or
"gendex.html" removed from the records, and rather put in software
implementing the mapping mechanism (meaning it can be more easily
updated). Given suitably flexible software, the mechanism may even
be defined in the DNS, meaning that updating can be performed by a
single person for all of the GenWeb!
Regardless of the technical advantages a future mechanism may have over
its predecessors, it may essentially fail if a lot of users are forced
to abandon a less powerful, but working, mechanism in the process. We
should not require them to update their software simply for them to be
able to keep the existing functionality. Therefore, we should continue
to serve the old mechanisms as long as they are useful to some users.
The smaller an effort which is needed to provide the old services, the
longer they can remain in function for a smaller community, in the hope
that also the last person to upgrade will gladly do so for the purpose
of gaining new features, rather than to merely maintain existing ones.
The conclusion is of course that the services we establish now should be
as easy as possible to maintain also long after they have been replaced
by better services in our (supposedly) published recommendations. That
may involve setting up specialized server software using a single GenWeb
index to provide data files for specific portions of the GENWEB.ORG name
space, and potentially for other distributed indexing services.
3. If the domains DB{2,3,4,...}.GENWEB.ORG are reserved for future
experiments like those outlined above, why can't GENWEB.ORG simply
be seen as synonymous to the suggested DB1.GENWEB.ORG?
- What's in a name? That, which we call a rose,
by any other name would smell as sweet.
/Romeo and Juliet
Shakespeare has a good point, although the `name conflict' which he
has in mind isn't quite as trivial as the one I'm addressing here.
Whether the domain chosen to contain our initial database server
name space is called DB1.GENWEB.ORG, GENWEB.ORG, LEISURE.IBM.COM,
or GHOFFMAN.UCSD.EDU doesn't really matter in principle, although
we can probably rule out the last two examples immediately on the
grounds that we cannot expect to get or to maintain any control over
a name space which belongs to a different organization.
My point #1 is that whatever domain we choose, it should be easy for
us to reconfigure and relocate it to whatever DNS servers we find
appropriate, for the purpose of providing this `old' service also in
the future. We know from experience that the NIC staff, who must be
involved every time we change official servers for GENWEB.ORG, have
more than enough to do dealing with organizations asking for their
own domains on the Internet. We shouldn't have to bother them with
server change requests resulting from experiments being performed
essentially within the GenWeb community.
My point #2 is that whatever domain we choose, it should be reserved in
its entirety for this particular service, i.e. mapping a set of GenWeb
database names to Internet server hosts by means of CNAME records.
The GENWEB.ORG domain already contains a number of subdomains which
don't serve this purpose, and it is likely to contain more of that kind
(such as the hypothetical "DB2" subdomain) in the future. There is at
least a theoretical risk for name conflicts here.
Even though I don't expect a real GenWeb database to be called "DB2",
"mailhost", "WWW", "postmaster", "NTP1", or "ghoffman", it's generally
not a good idea to formalize a specification such that "you may give
your GenWeb database any name that isn't already taken by some other
database, or that isn't one of mailhost, www, postmaster, or ghoffman,
or that consists of either of the abbreviations DB or NTP followed by a
decimal number". Maintaining such historically motivated restrictions
in the future, when the majority of the GenWeb users need to deal only
with names in the DB4.GENWEB.ORG domain, will seem plain silly, even if
it's technically needed for backward-compatibility.
There may also be a psychological side of the issue. For those users
concerned enough to actually look at the domain names, it would be good
if their syntax somehow suggest how they are used. It's easy to
document that "every subdomain of DB1.GENWEB.ORG represents a database"
but not as easy to do for the relevant subset of the GENWEB.ORG names,
if they seem mixed up with other functions.
In the case of having specialized server software for the name space,
it's easier to have it serve a particular zone with limited contents,
than to hook it onto the UCSD.EDU master server and have it serve the
entire GENWEB.ORG domain with all its existing peculiarities.
4. Does it all really matter? If GenWeb is a failure, why waste time
on these details?
Of course, we can never guarantee anything about the future. We may
lose Internet support of the GENWEB.ORG domain, or the Internet may
break down entirely. Who knows? However, planning for general
failure isn't going to get us anywhere. We participate in this work
because we believe it may lead forward, and while we are at it, we
may just as well try to learn from history and avoid repeating some
well-known mistakes of the past, especially if the cost is as small
as four extra octets per database, and even though the advantages
can still only be seen by a small clan of secretive magicians with
crystal balls... ;-)
To summarize, I believe that defining a specific subdomain for a
specific technical purpose within the GenWeb initiative will have
technical as well as educational benefits, rather than add any
considerable complexity. Note that the majority of GenWeb users
shouldn't have to worry about DNS issues; if they are simply told
a domain name to be used in a particular way, Gary and I (and maybe
a few other daring souls) can get as involved in the technicalities
of DNS zone management, NIC interaction, CNAME, MX, and TXT records
as we like (this is not to keep anyone out against their will, but
just to provide some relief to those who think there is too much
DNS and too little genealogy on this mailing list).
The length of the above isn't representative for my list of Things to
be Implemented in GenWeb in the Future, but rather for my fear of not
being understood on this single issue. I have used a lot of words,
but content-wise there is actually not that much to it. Sorry...
--
Anders Andersson, Dept. of Computer Systems, Uppsala University
Paper Mail: Box 325, S-751 05 UPPSALA, Sweden
Phone: +46 18 183170 EMail: andersa@DoCS.UU.SE
Wednesday, March 15, 1995 7:50:47 PM
GenWeb Item
From: Gene Stark,starkhome!gene@sbstark.cs.sunysb.edu,Internet
Subject: Intermediate Domain Name
To: GenWeb
>I assume you recommend a format like "demo.db1.genweb.org" so that it can
>migrate to "demo.db2.genweb.org" some day. Perhaps I don't appreciate the
>subtleties involved here. Why won't the following work just as well:
>"demo1.genweb.org" migrating to "demo2.genweb.org"? Since these are merely
>CNAMEs anyway, they have to point to a real canonical name which may or may
>not change as needed. Is there some management issue that I'm not
>perceiving?
Here's the idea as I see it. First let's consider the scheme with no
intermediate domain name. We will start with the "Version 1" GenWeb,
and will have databases: demo1.genweb.org, charles1.genweb.org, etc.
These databases will create links to each other using these names.
Eventually we will find that we want to change some sort of organizational
principle underlying GenWeb, and we will want to change everything to
"Version 2" GenWeb. If we change over to "demo2.genweb.org",
"charles2.genweb.org", etc., then the owner of, say, the "charles1"
database has to go through and edit every one of his links now to refer.
to "charles2.genweb.org". Not only that, everyone else that has any
links to data in this database will have to change their links as well.
Any host that wants to have the same data be part of both versions at the
same time will have to have two copies of that data.
On the other hand, lets suppose we start out with "demo.v1.genweb.org"
and "charles.v2.genweb.org". Now a host serving a "Version 1" GenWeb
database can exist in subdomain "v1.genweb.org". The links in the database
files can perhaps refer to domains "demo", "charles", etc., without giving
a full path to the root domain. If host "demo.v1.genweb.org" has a link
to "http://charles/foo/bar", it will refer to host "charles.v1.genweb.org".
When the changeover is made to "Version 2" GenWeb, then (unless changes
are made to the format of the URL's themselves) simply by changing the default
domain of the host "demo.v1.genweb.org" to "demo.v2.genweb.org" we now
automatically interpret links "http://charles/foo/bar" as links to the
Version 2 database on "charles.v2.genweb.org", rather than the Version 1
database on "charles.v1.genweb.org". No editing of the data itself is
required. What is more, "Version 1" and "Version 2" can exist simultaneously,
thereby accomodating sites that are slow to change over to the new version,
or supporting experimentation with multiple versions at the same time.
To see the advantage of multiple versions, consider the following.
Initially, suppose there is only one site that is serving surname "Smith",
and this site receives the domain "smith.genweb.org". But soon ten other
sites come along also with data on the "Smith" surname. Under the scheme
without the intermediate domain name, we decide that these latecomers are
to be assigned subdomains "smith1", "smith2", etc. However, a better
alternative might be to use the domain "smith.genweb.org" as an intermediate
domain within which to locate all other hosts serving surname "Smith".
Under your naming scheme, at this point we have to ask everyone with links
to "smith.genweb.org" to change their data to the new naming scheme.
Anybody who doesn't change is going to have broken links. On the other
hand, with the "extra subdomain" scheme, we can simply announce a transition
from "Version X" GenWeb to "Version Y" GenWeb. Version X GenWeb has
the domain "smith.vX.genweb.org" with no subdomains. In Version Y GenWeb,
the domain "smith.vY.genweb.org" will have subdomains for each of the
Smith databases. Anyone who wants to upgrade completely from Version X
to Version Y will of course have to edit their links to refer to the proper
subdomain of "smith.vY.genweb.org" instead of "smith.vX.genweb.org". However,
anybody who hasn't gotten around to doing it yet will still be able to
link to "smith.vX.genweb.org" and their links will work OK, though they
might refer to old data. The original Smith site could exist simultaneously
in "smith.vX.genweb.org" and in "smith0.smith.vY.genweb.org", so as to
continue to handle links from Version X GenWeb sites at the same time as it
has outgoing links to both Version X and Version Y GenWeb sites.
I hope these arguments show that the additional flexibility over the long
run of the additional naming level will be well worth any minor added
complexity.
- Gene Stark
Thursday, March 16, 1995 8:07:36 AM
GenWeb Item
From: olsen@cs.byu.edu,Internet
Subject: Re: Intermediate Domain Name
To: GenWeb
>Date: Thu, 16 Mar 95 02:34:20 +0100
>From: Anders Andersson
>To: genweb@UCSD.EDU, ghoffman@UCSD.EDU
>Subject: Re: Intermediate Domain Name
>However, another mechanism need not, for instance, depend on the use
>of a particular access protocol such as HTTP. I anticipate that for
>really long-lived cross-references in genealogical records, we will
>need something like a : pair, with every
>reference to specific access methods such as "http", "genweb.org" or
>"gendex.html" removed from the records, and rather put in software
>implementing the mapping mechanism (meaning it can be more easily
>updated). Given suitably flexible software, the mechanism may even
>be defined in the DNS, meaning that updating can be performed by a
>single person for all of the GenWeb!
>
This is the heart of what we really need to make networked genealogy work.
There are essentially three components to the kind of electronic
referencing that is needed in family history. I want to point out that
linking to individuals isn't the only need in building good online
genealogy. We also need a clean mechanism to reference the electronic or
physical source material from which the family relationships were built.
The three parts are:
This is the electronic or physical repository where one wants to look
for a resource. In WWW terms this would be a URL. It is a location
physical or network where such materials might be stored. This
information, as discussed above, should not be included in the actual
electronic materials
This uniquely identifies the book, GEDCOM file, database or other
organized unit of information. Just as an ISBN identifies a book
uniquely, regardless of what library or bookseller has the book, this
should identify a particular electronic file or database, regardless
of how many sites it is stored on, or where it is. This corresponds
to the unit described above.
This is the individual, family, birth record entry, paragraph or other
fragment of the electronic book which is being referenced.
The problem with the current domain name discussion is that domain names
are ultimately location names. Mixing location with content identification
just will not work in the long term.
The problems to be solved then are:
This is simple. The current URL structure of the WWW provide a fine
mechanism for identifying a library in which to look.
This is problematic. We need some scheme for assigning identifiers
which are unique but don't create a bottleneck. Perhaps a format of
- would work where only publishers need to be
globally unique, with each publisher being responsible for the
uniqueness of the electronic resources they create. This would require
one or more publishers for information created by individuals so that
anyone could easily get a unique indentifier.
This would vary widely with the type of "book" being accessed. A
GEDCOM identifier would be different from one in a census index or
into a chapter of some person's biography. Perhaps here we need
something of the form - where the identifies
the algorithm required to map to a record in the specified
electronic book.
If we do this well, one special case of such a scheme would also support
bibliographic references into physical sources such as books or microfilm.
____________________
Dan R. Olsen Jr. Computer Science Department
Brigham Young University Provo, UT 84602
(801) 378-2225 FAX 801-378-7775
Thursday, March 16, 1995 12:49:42 PM
GenWeb Item
From: Genealogical Data Communications
Specs,GEDCOM-L@VM1.NoDak.EDU,Internet
Subject: Linked databases
To: GEDCOM
Anders Andersson writes:
>However, another mechanism need not, for instance, depend on the use
>of a particular access protocol such as HTTP. I anticipate that for
>really long-lived cross-references in genealogical records, we will
>need something like a : pair, with every
>reference to specific access methods such as "http", "genweb.org" or
>"gendex.html" removed from the records, and rather put in software
>implementing the mapping mechanism...
Dan Olsen writes:
>This is the heart of what we really need to make networked genealogy work.
>There are essentially three components to the kind of electronic
>referencing that is needed in family history. I want to point out that
>linking to individuals isn't the only need in building good online
>genealogy. We also need a clean mechanism to reference the electronic or
>physical source material from which the family relationships were built.
>
>The three parts are:
>
>
[physical address]
>
[logical address]
>
[entry number]
>
>The problem with the current domain name discussion is that domain names
>are ultimately location names. Mixing location with content identification
>just will not work in the long term.
>
>The problems to be solved then are:
I haven't been following this business about domain names very closely
but the above perked up my interest... Y'all mind if I jump in here for
a bit?
I've been a staunch (and solitary) advocate of linked genealogy databases
for many years. So much so, in fact, that I wrote my own code rather than
make do with something else. (Being a software engineer had nothing to do
with it...:-).
Anyway, here's how I solved this problem.
Database SMITH Database JONES
A:102 A:17
H:1:17 H:2:102
A:103
H:1:19
Entry 102 in the database SMITH is married to entry 17 in JONES.
Files SMITH.GDB and JONES.GDB can be anywhere - different directories,
drives, machines, countries, etc.
When the processing software runs, it first reads a file containing
the location of all known database files - or at least all that you
wish to deal with at any one time.
As each record is read from the file, it is parsed into a book
structure:
#define BOOKLOG_NODE_LEN_128
#define BOOKLOG_NAME_LEN 30
#define BOOKLOG_DIR_COUNT 10
struct booklog_struct {
struct booklog_struct *flink;
struct booklog_struct *blink;
int booknumber;
char node[BOOKLOG_NODE_LEN];
char name[BOOKLOG_NAME_LEN];
char *dirlist[BOOKLOG_DIR_COUNT];
char device[30];
char extension[6];
char composit[128];
int open;
int accessed;
int new;
int old;
int index;
int tested;
FILE *fp;
};
typedef struct booklog_struct BOOKLOG;
By convention, the first record in the list is the last book the
software used. The user is asked which book name is desired
(without reference to it's location) and the default is the 1st
book name. Let us assume the user types in the name of another
book.
The list is reordered so that the user's selection is now 1st.
The order of the others is unimportant. The booknumber element
int the struct above is updated throughout the list, 0...N.
Back to the examples above - there are references like "19" and
"17." The first, without the <...> construct, refers to
an entry in the current database. If prefixed with the <...>,
the entry is in the named book.
The key to making all this work is the linked list of book structs.
(...which also enable this stuff to run on multiple operating systems
since the machine|device|directory|file|extension syntax oddities
are stripped out.)
At the beginning of any program activity, the book list is ordered
with the current book at position 0. All other books are 1...N.
As something is read from the database, the <...> string constructs
are converted to the 1...N integer values. No <...>?, use 0. All
further reference to entries is by struct:
struct BKENT_PAIR {
int bk,ent;
};
During the course of events, if reference is made to a new book,
it's added to the end of the book list and the 'new' flag is set. If it's
ever needed, an attempt is made to find it - which may break down to
asking to user to enter it's location. At the end of operations,
the entire book list is saved back out to the file - which accounts
for how the last used book is the 1st record. Note that new books are
learned...
Access to the database information is made by issuing a request to
a caching mechanism with the bkent_pair argument. The cache keeps
some number of entries from any number of databases on hand as caches
are inclined to do :-). The active list may look like 0/2, 2/2, 2/3, 17/9,
22/1, 0/55,0/56.... etc.
Having taken up more of your time than you were probably prepared to
give up, I'll close. If you want to go into this more, I've got all
sorts of examples and stuff. At one point when I was learning NT,
I rewrote the bit that reads the list of files to be a client to a server.
The server actively finds all the database files on the system and
reports them to all interested parties, including other servers, via
named pipes.... I removed all that stuff - I was beginning to loose
touch with what the program is supposed to do. Fun stuff, anyway.
Mickey.
Friday, March 17, 1995 6:05:39 AM
GenWeb Item
From: Anders Andersson,andersa@Mizar.DoCS.UU.SE,Internet
Subject: Re: Intermediate Domain Name
To: GenWeb
Dan R. Olsen writes:
>The problem with the current domain name discussion is that domain names
>are ultimately location names. Mixing location with content identification
>just will not work in the long term.
You are right that we need to identify the separate issues here
(persistent resource identifiers vs. physical document locators).
The current discussion is about a particular use of CNAME records
to identify (supposedly physical) servers providing (supposedly
physical) files to the users. This should be viewed as a fairly
low-level component in a more complex hierarchy, and chances are
that it will be replaced by a more refined mechanism pretty soon.
This is part of my reason for not wanting it to dominate the
GENWEB.ORG domain.
Domain names as such, however, are not limited to physical locations.
For instance, the GENWEB.ORG domain has no particular "location" on
the net, as far as I can see. This is what makes it useful also for
implementing Internet-wide resource indexing mechanisms that don't
have anything to do with locations (whether DNS is the optimal vehicle
for such mechanisms is another matter, but that's something we may
want to experiment with).
--
Anders Andersson, Dept. of Computer Systems, Uppsala University
Paper Mail: Box 325, S-751 05 UPPSALA, Sweden
Phone: +46 18 183170 EMail: andersa@DoCS.UU.SE
Friday, March 17, 1995 7:52:22 PM
GenWeb Item
From: Jim Dial,Jim.Dial@UC.EDU,Internet
Subject: Dial database finally online!
To: GenWeb
I am pleased to announce that I have finally finished the laborious task of
getting my database
online. I haven't checked ALL of the links yet, but it's there. The
address is:
http://ucunix.san.uc.edu/~dialjm/jimdial.html
This goes back approx. 9 generations down the Dial side and back approx. 5
generations
on all of the others, please tell me what you think. This was all done by
hand. Is there
a GEDCOM ---> HTML convertor available for DOS/Windows machines? What a time-
saver that would've been!
Anyway, if you visit, please let me know what you think of it! Thanx
_______________________________________________________________________________
Jim Dial
LAN Administrator, WGUC Radio, Cincinnati, OH
Internet: Jim.Dial@uc.edu WWW: http://ucunix.san.uc.edu/~dialjm/
_______________________________________________________________________________
Friday, March 17, 1995 7:52:22 PM
GenWeb Item
From: Jim Dial,Jim.Dial@UC.EDU,Internet
Subject: Dial database finally online!
To: GenWeb
I am pleased to announce that I have finally finished the laborious task of
getting my database
online. I haven't checked ALL of the links yet, but it's there. The
address is:
http://ucunix.san.uc.edu/~dialjm/jimdial.html
This goes back approx. 9 generations down the Dial side and back approx. 5
generations
on all of the others, please tell me what you think. This was all done by
hand. Is there
a GEDCOM ---> HTML convertor available for DOS/Windows machines? What a time-
saver that would've been!
Anyway, if you visit, please let me know what you think of it! Thanx
_______________________________________________________________________________
Jim Dial
LAN Administrator, WGUC Radio, Cincinnati, OH
Internet: Jim.Dial@uc.edu WWW: http://ucunix.san.uc.edu/~dialjm/
_______________________________________________________________________________
Sunday, March 19, 1995 6:12:13 AM
GenWeb Item
From: Phlete Teachout,fteachou@eagle1.eaglenet.com,Internet
Subject: Re: My tree online (fixed)
To: GenWeb
On Sat, 18 Mar 1995, Jim Dial wrote:
> Well, I think I've got most of the bugs fixed, it was late last night and I
> wanted to get it out there.
> Hopefully everything is working properly now. Let me know what you think.
> Sorry for the inconvenience!
Just checked in! Everything seems to be working swell! I have only one
suggestion and that would be to add an index in surname order.
Best of luck in your future research!
- fleet -
Monday, March 20, 1995 7:25:07 AM
GenWeb Item
From: Hal Bodin,hwbodin@smtpgate.read.tasc.com,Internet
Subject: Information on the Bodin Family
To: GenWeb
Hello,
I'm new to all the net stuff... and would like to see if anyone has
information about the Bodin family.
Tanks a bunch!!!!!
Hal Bodin
e-mail
hwbodin@tasc.com
Monday, March 20, 1995 10:58:58 AM
GenWeb Item
From: Jan Alden Cornish,jac@cadence.com,Internet
Subject: Searching Genweb
To: GenWeb
> From list-relay@UCSD.EDU Mon Mar 20 10:14:25 1995
> Received-Date: Mon, 20 Mar 1995 10:14:24 -0800
> X-Mailer: WordPerfect Office 4.0
> Date: Fri, 17 Mar 1995 14:02:30 -0500
> From: Hal Bodin
> To: genweb@UCSD.EDU
> Subject: Information on the Bodin Family
> Content-Length: 167
>
How does one search for particular a family ?
> Hello,
> I'm new to all the net stuff... and would like to see if anyone has
> information about the Bodin family.
>
> Tanks a bunch!!!!!
> Hal Bodin
> e-mail
> hwbodin@tasc.com
>
>
Monday, March 20, 1995 1:43:05 PM
GenWeb Item
From: Nancy Sween,nsween@falcon.cc.ukans.edu,Internet
Subject: Re: Searching Genweb
To: GenWeb
On Mon, 20 Mar 1995, Jan Alden Cornish wrote:
> How does one search for particular a family ?
If you have access to the Web (with Mosaic, Netscape, or lynx), try
http://ftp.cac.psu.edu/~saw/genealogy.html
and look for the direct search to the Roots surname list.
Another site that lets you search is
http://uxl.cso.uiuc.edu/~al-helm/genealogy.html
I'll bet there are others, but that's all I have notes on just now.
Nancy Sween
Genealogy URL: http://falcon.cc.ukans.edu/~nsween/genealogy.html
Monday, March 20, 1995 2:14:08 PM
GenWeb Item
From: Chuck Wheeler,chuckw@cnw.com,Internet
Subject: newbeee
To: GenWeb
I just read a note from H Bodin re his family on the genweb list. I too am=
=20
new to this, how does one determine which files are in the list? Is there a=
=20
way to peruse the list or see whether your particular files made it to the=
=20
list? Why all the secrecy or where does my ignorance begin?
Chuck Wheeler
chuckw@cnw.com
Monday, March 20, 1995 4:46:24 PM
GenWeb Item
From: Async User,gandalf@epix.net,Internet
Subject: RE: Revolutionary War
To: GenWeb
I am searching for information about specific revolutionary battles.
Several of my ancesters (The Mechlings from the Palitinade) were involved
in the war and I wish to have some background information. Is there
anyone who can recommend a detailed book about this war or an availible
E-TEXT.
Thank you in advance,
Gandalf the Geneologist
Monday, March 20, 1995 4:49:14 PM
GenWeb Item
From: Michael A. PattonĄ GenWeb
discussion,MAP=GenWeb@BBN.COM,Internet
Subject: Re: Information on the Bodin Family
To: GenWeb
Date: Fri, 17 Mar 1995 14:02:30 -0500
From: Hal Bodin
I'm new to all the net stuff... and would like to see if anyone has
information about the Bodin family.
The GenWeb list is NOT a good place for this sort of question. This
list is for discussing the details of how genealogy info might be
organized for the Web. If you're actually interested in researching
actual names, you are better off with soc.genealogy.surnames (and the
associated search engine and Roots Surname List).
-MAP
Monday, March 20, 1995 4:50:55 PM
GenWeb Item
From: Michael A. PattonĄ GenWeb
discussion,MAP=GenWeb@BBN.COM,Internet
Subject: Re: Searching Genweb
To: GenWeb
Date: Mon, 20 Mar 1995 10:52:23 -0800
From: Jan Alden Cornish
How does one search for particular a family ?
The GenWeb demos set up so far are nowhere near the stage of having
comprehensive searching. It's certainly one of the goals, but we've
really only just started... For searching, I recommend the Roots
Surname List maintained by Karen Isaacson or asking on the Newsgroup
soc.genealogy.surnames or searching its archives.
-MAP
Tuesday, March 21, 1995 5:07:57 AM
GenWeb Item
From: Birger A. Wathne,Birger.Wathne@vest.sdata.no,Internet
Subject: Re: genweb.org established
To: GenWeb
I would like the genweb domain to have one additional level, as well.
It could have a level called db.genweb.org containing pointers to the
actual databases, and other sublevels with links to information,
homepages, etc.
There should perhaps be a separate search.genweb.org with pointers to
servers with some (future) standardized search interface (making
distributed searching possible).
Perhaps we could add some info into the records in the future, with
hints on what the rest of the URL should look like, so you can get
directly to the requested database (as there will be more than one
database on each host). I guess we could put this into the HINFO
field?
How will we assign the database names? Should we try to keep names like
stoyan.genweb.org, or should we let each database owner get some unique
prefix, and let them add a unique suffix? E.g. let someone have the
prefix A0000 and let him/her have the 3 last letters to identify
individual databases?
In this case, some more info on the individual databases should also
be added to the name server info. I know these issues will force us
to move away from CNAME records, but it should be possible to have a
mail gateway and/or WWW form to let the users register their databases
automatically.
Lots of thoughts. To summarize: genweb.org may well contain a lot of
different info in the future. So let's build it right from the start.
Birger
Tuesday, March 21, 1995 8:55:33 AM
GenWeb Item
From: andersa@Mizar.DoCS.UU.SE,Internet
Subject: Re: genweb.org established
To: GenWeb
Birger Wathne writes:
>There should perhaps be a separate search.genweb.org with pointers to
>servers with some (future) standardized search interface (making
>distributed searching possible).
I believe we agree here, though I'd just like to point out that we
should avoid actually implementing those domain names until we have
reached agreement on how they should be used. I have seen a few
times how people who meet each other in "real time" seem to agree
on a design solution in theory, only to run off in totally different
directions when they start implementing the stuff. Individual
experiments are fine, as long as we don't pretend that they will
remain mutually compatible and constitute parts of a coherent system.
>Perhaps we could add some info into the records in the future, with
>hints on what the rest of the URL should look like, so you can get
>directly to the requested database (as there will be more than one
>database on each host). I guess we could put this into the HINFO
>field?
In the standard, HINFO has been reserved for the CPU hardware and
software identification strings. Since there is no enforcement of
some restricted syntax or set of valid identification strings for
these fields, people have come up with all sorts of innovative uses
for them, such as physical location or legal owner of a machine.
As long as humans are the only ones trying to interpret these fields,
there is little risk of confusion, but here we are talking about an
automated mechanism trying to do the same.
I'd prefer employing the generic text (TXT) record, as its use isn't
prescribed in a similar way. There is even a proposed (?) standard
(RFC 1464) on how to store arbitrary string attributes in the DNS
using the TXT records, and I believe that GenWeb data would fit
right in here very well.
--
Anders Andersson, Dept. of Computer Systems, Uppsala University
Paper Mail: Box 325, S-751 05 UPPSALA, Sweden
Phone: +46 18 183170 EMail: andersa@DoCS.UU.SE
Tuesday, March 21, 1995 1:13:11 PM
GenWeb Item
From: Everton,jayhall@xmission.com,Internet
Subject: Everton's Genealogical Helper: online edition
To: GenWeb
Volume 1, Number 2 (21 March 1995) of "Everton's Genealogical Helper:
online edition" is now available on the World Wide Web. The magazine is
accessed through the URL:
http://www.xmission.com/~jayhall/ghonline.html
"Everton's Genealogical Helper: online edition" is a free e-zine focusing
on genealogy in the online world, and published by Everton Publishers.
------------------------------------------------------------------------------
Jay Hall, Everton Publishers | E-Mail = jayhall@xmission.com
165 South West Temple, Suite 200 | WWW = http://www.xmission.com/~jayhall/
Salt Lake City, UT 84101 | ftp = ftp.xmission.com/pub/users/j/jayhall/
Tuesday, March 21, 1995 2:07:00 PM
GenWeb Item
From: fred@compu.com,Internet
Subject: Re: genweb.org established
To: GenWeb
also sagte Birger A. Wathne:
>
> Lots of thoughts. To summarize: genweb.org may well contain a lot of
> different info in the future. So let's build it right from the start.
>
I don't have much to add about CNAME records and stuff like that because I
have no idea what you are talking about anyway. :-)
I only see problems ahead if we don't organize the whole genealogical web
scene. Every day someone else tells us about his homepage and there is really
no place to find all of these things.
Should not GENWEB's first and foremost goal be to serve as a central
information point? Sort of like anybody who is publishing a homepage should
first register this at genweb.org where this entry is then included in a
master lookup page?
For example, we are slowly building data resources for German genealogy at
www.genealogy.com (qrz.com). Here we hope to gather other pages which carry
appropriate German related data from various other places.
The idea in my mind would be to look at the genealogy master at genweb.org
which then has other major categories which the user may follow down the line.
Is this what we are trying to do conceptually? Punch me if I'm wrong.
Fred
--
W. Fred Rump or CompuData, Inc. | 'When we were young, we did not ask
26 Warren St | 10501 Drummond Rd | questions; now that we're old, there's
Beverly NJ 08010 | Phila. PA 19154| no one to answer them.'
609-386-6846 | 215-824-3000 | fred@compu.com, 73512.2234@compuserve.com
Wednesday, March 22, 1995 10:48:12 AM
GenWeb Item
From: ANNELISE ANDERSON,ANDRSN@HOOVER.STANFORD.EDU,Internet
Subject: GEDCOM>HTML
To: GenWeb
I would like to clarify whether or not there is a converter for
GEDCOM to HTML. I suppose what I would like are family pages
plus an index of names, with links from the index to the pages
and links on each family page for children who also have family
pages. Alternatively, whatever is the standard.
I was thinking of adding my genealogical information, which is
about 225 names, to my home page, which already exists.
HTML is not a mystery to me and I could do the links myself.
However I'm not sure I can easily create a text file of family
pages. I have PAF and FTW. I don't think PAF will print to
a file. FTW has a "print to file" option but not for batch
printing of family pages. Second, printing in FTW to a file
results in an unreadable non-text file (which cannot be
reimported into FTW!). I'm not sure how I'd set up a "report"
in FTW to print, in effect, family pages, and then save the
report to the clipboard for latter processing. I don't think
the clipboard would take such a large file.
Also I don't know if it's possible in Windows, which is doing
the FTW printing, to print to a file instead of a printer, i.e.,
to redirect the output from LPT1:>filename.
Sorry to bother you with these issues; perhaps most of this
should be in soc.g.computing, if one could see through the
smoke created by the flames.
But presumably those who run this group are interested (at least
eventually) in encouraging people to make databases available
on the Web.
Let me make a suggestion. The conversion program that may be
needed should possibly be written as a program or macro in a
word processor. Of existing word processors WordPerfect is the
most common. (I use XyWrite, uncommon, and keep WordPerfect for
"compatibility" services.) Second and becoming more standard
is Microsoft Word for Windows.
Please advise. Thanks.
Annelise Anderson
Wednesday, March 22, 1995 12:10:50 PM
GenWeb Item
From: T.T.Wetmore,ttw@beltway.att.com,Internet
Subject: Re: GEDCOM>HTML
To: GenWeb
Annelise,
There are a number of GEDCOM to HTML converters, both as stand alone
programs, and as features of genealogical database programs.
I have seen reference to stand alone programs that run on MS-DOS systems,
but since I don't use such systems, I don't pay attention to them. Sorry.
The method I am most aware of is the use of a LifeLines report program to
convert a GEDCOM record or set of records to an HTML index file or person
or family group HTML file. I know this because I am the author of
LifeLines and the author of a couple such GEDCOM to HTML converter report
programs. This is the technique used by at least of couple WWW GenWeb
sites to provide their database browsing services. That is, LifeLines
provides the genealogical engine, and browsers' actions are converted to
requests to generate HTML files from GEDCOM records stored in the LifeLines
database in an "on demand" basis. You must use UNIX however to use this
approach.
Good luck finding more.
Tom Wetmore
Wednesday, March 22, 1995 12:18:09 PM
GenWeb Item
From: Terrance E Sterkel,tsterkel@mtgbcs.mt.att.com,Internet
Subject: Re: GEDCOM>HTML
To: GenWeb
I think that your conversion program will get off
the ground much quicker, given the availability of
the MSWord 2. => HTML program that is available.
Also, programs such as The Master Genealogist (TMG)
outputs directly into MS Word.
best wishes,
terry
Wednesday, March 22, 1995 12:23:10 PM
GenWeb Item
From: shadow@mindspring.com,Internet
Subject: Re: GEDCOM>HTML
To: GenWeb
Gene Stark has created a Dos/Windows GED>HTML conversion program that
works well. I have a copy of it that you may get from my homepage at:
http://www.mindspring.com/~shadow/GEN/gen.html
My pages were created using this program, so you can sample them to
see what kind of output to expect.
Hope this helps,
Bill Spurlock
Wednesday, March 22, 1995 2:26:51 PM
GenWeb Item
From: Nancy Sween,nsween@falcon.cc.ukans.edu,Internet
Subject: Re: GEDCOM>HTML
To: GenWeb
On Wed, 22 Mar 1995, ANNELISE ANDERSON wrote:
> HTML is not a mystery to me and I could do the links myself.
> However I'm not sure I can easily create a text file of family
> pages.
When I did my page, I used before and
after to
keep the spacing intact, then recopied the original complete
file into new files by deleting children from one file and
deleting parents from the new file, etc, then linking parent
file to child file. Clear as mud?
> reimported into FTW!). I'm not sure how I'd set up a "report"
> in FTW to print, in effect, family pages, and then save the
> report to the clipboard for latter processing. I don't think
> the clipboard would take such a large file.
Hope someone has the answer to this too.
>
Nancy Sween |"The Interactive Medical Student Lounge"
nsween@falcon.cc.ukans.edu| URL: http://falcon.cc.ukans.edu/~nsween
===================================================================
Genealogy site: http://falcon.cc.ukans.edu/~nsween/genealogy.html
Europa Kansas site: http://falcon.cc.ukans.edu/~nsween/europa.html
Wednesday, March 22, 1995 2:26:51 PM
GenWeb Item
From: Nancy Sween,nsween@falcon.cc.ukans.edu,Internet
Subject: Re: GEDCOM>HTML
To: GenWeb
On Wed, 22 Mar 1995, ANNELISE ANDERSON wrote:
> HTML is not a mystery to me and I could do the links myself.
> However I'm not sure I can easily create a text file of family
> pages.
When I did my page, I used before and
after to
keep the spacing intact, then recopied the original complete
file into new files by deleting children from one file and
deleting parents from the new file, etc, then linking parent
file to child file. Clear as mud?
> reimported into FTW!). I'm not sure how I'd set up a "report"
> in FTW to print, in effect, family pages, and then save the
> report to the clipboard for latter processing. I don't think
> the clipboard would take such a large file.
Hope someone has the answer to this too.
>
Nancy Sween |"The Interactive Medical Student Lounge"
nsween@falcon.cc.ukans.edu| URL: http://falcon.cc.ukans.edu/~nsween
===================================================================
Genealogy site: http://falcon.cc.ukans.edu/~nsween/genealogy.html
Europa Kansas site: http://falcon.cc.ukans.edu/~nsween/europa.html
Thursday, March 23, 1995 6:39:31 AM
GenWeb Item
From: Mickey Lane,MLANE@csi.compuserve.com,Internet
Subject: GEDCOM>HTML
To: GenWeb
ANNELISE ANDERSON writes:
>I was thinking of adding my genealogical information, which is
>about 225 names, to my home page, which already exists.
>HTML is not a mystery to me and I could do the links myself.
>However I'm not sure I can easily create a text file of family
If you can mail a gedcom file to me, I'll run it through ROOTSBOOK
and mail the people pages + index back to you. It won't have HTML tags
in it but it will be in a good format to support them.
I have other motives here :-). I'm in the process of adding HTML
support to ROOTSBOOK and I'd like to see how others would tag the
output text.
Mickey.
Sunday, March 26, 1995 9:17:18 PM
GenWeb Item
From: Cleve Lowry,cleve@nucleus.com,Internet
Subject: genealogy
To: GenWeb
Hi. I am interested in reaching others re: my work in the family history
of these three names. LOWRY, DEGRAFF, RIEKEN. Please advise as to the
route to take etc. All help will be greatly appreciated. cleve@nucleus.com
Monday, March 27, 1995 4:59:17 PM
GenWeb Item
From: Kip Sperry,SPERRYK@rel1.byu.edu,Internet
Subject: Computer Lectures and Workshops, San Diego, May 1995
To: GenWeb
The 1995 National Genealogical Society (NGS) Conference will be held
at the Town & Country Hotel and Convention Center 3-6 May and will
include a number of computer lectures and workshops. Sample topics:
GenWeb--A Distributed World Wide Genealogy Database on the Internet
Surfing the Internet
Family Tree Maker for Windows
Roots IV
Personal Ancestral File (PAF)
Focus and Design with PAF
GEDCOM
Family Origins for DOS and Windows
Comparison of Genealogy Software
Utilities for Enhancing PAF
FamilySearch
The Master Genealogist
Graphic Techniques
Scanning
BBS's
Multimedia and Genealogy Software
Windows Applications
High-Teching Your Ancestors
CD-Rom Selection
Utilities to Make FamilySearch More Effective
and others, including computer demonstrations
Other classes will cover research methodology, state and regional
sources, foreign topics, National Archives sources, family history
topics, historical sources, library topics, etc.
For a free registration brochure, contact the conference program
chair:
Cindy Conser
9074 Meadowrun Way
San Diego, CA 92129-3330
PHONE: 619-484-3700
FAX: 619-484-5100
Tuesday, March 28, 1995 10:11:35 AM
GenWeb Item
From: Scott McGee,smcgee@microware.com,Internet
Subject: genweb index pages
To: GenWeb
In my web surfing, I know there is at least one, and I am thinking two or
three different pages that have an index to other genweb sites. I know
my little micro example off of my homepage was listed in one.
I am working on a GenWeb page that is primarily intended as an interface
to my own genweb , but I would like to point to some other pages that
contain general genweb information, and other genweb sites. If anyone
maintains either type of page (or knows of one) please let me know.
Thanks
Scott
When in danger, | If it has my name on it, it must be MY opinion!
or in doubt, |______________________________________________________
run in circles, | Email: smcgee@microware.com (Scott McGee)
scream and shout! | Web: http://www.cc.utah.edu/~sam8644/homepage.html
Tuesday, March 28, 1995 12:49:26 PM
GenWeb Item
From: Timothy B. Doyle,tdoyle@netcom.com,Internet
Subject: Re: genweb index pages
To: GenWeb
> If anyone maintains either type of page (or knows of one) please let me know.
Scott:
I have a page that lists all known (to me) GenWeb databases. I have
categorized the databases by the method used to create them and include a
short description of each of the methods. When available, I also provide
information on how others can implement the methods and provide links to any
related utilities.
This page is located at:
http://www.doit.com/tdoyle/genweb/genweb.html
If anyone has any additions or corrections to any of the information on my
page, I would appreciate hearing from you.
Tim Doyle
tdoyle@doit.com / tdoyle@netcom.com
WWW homepage: ftp://www.doit.com/tdoyle/
ftp directory: doit.com pub/tdoyle
Wednesday, March 29, 1995 8:41:50 AM
GenWeb Item
From: Herbert
Stoyan,Herbert.Stoyan@informatik.uni-erlangen.de,Internet
Subject: family serach
To: GenWeb
In my applications, use
http://www8.informatik.uni-erlangen.de/cgi-bin/genweb-index/F=/N=familyname
Wednesday, March 29, 1995 10:19:32 AM
GenWeb Item
From: Gary Hoffman,ghoffman@ucsd.edu,Internet
Subject: Fwd: Re: Trial Hypertext for West Family
To: GenWeb
mficklin@sierra.net wrote:
Thanks for replying to my message re standards for an html doc for genweb.
I know how to
write html. I have documentation on the proper form for source citations.
Let's see if I can restate my question. -- I'm looking for a 'standard'
for submitting to
genweb. --I would like to see a genweb html source file. I really don't
know how to state it
any other way.
The following is just background. Skip it if you understand the above.
From what I read on
the listserver there are programs that convert a gedcom file to some html
genweb
standard. What I have decided to do is look at the output of one or more
of those
conversion programs to see how the html source file is layed out. Then I
will set up my
html source files to be structured the same way. I don't plan to start
with a gedcom
file--although I could. It seems to me in the future GEDCOM will be
obsolete. Everyone
will just input to html docs to begin with because everything will be on
the web.
I just want my html doc to look like your standard genweb submission.
Standards I would
want to know would be things like: What is the acceptable size of a graphic
image? How
many individuals will be allowed to be within a single file (URL)? That
sort of thing.
Standards specific to html on genweb. If this doesn't explain it, I'll
just continue to do it my
own way and hope I don't have to rewrite too much in the future when I
actually submit
them to genweb.
My reply:
Welcome to the frontier of genealogy. GenWeb is an emerging standard which
is another way of saying we don't know what the standard is yet. We are
still experimenting. However, one of the interesting things about HTML
development is that you can easily capture someone else's HTML code to see
how they put it together and then put yours together the same way. One
danger of setting standards too early is that we may inhibit development.
Remember, this stuff is only a couple of months old.
You are welcome to look at my pages and see how I built them. Then look at
some of the others out there. At some point, after experimenting with
various formats, we will get together and decide how we want to standardize
things. Photos are a good example. You asked, what is the acceptable size
of a graphic image? I answer, acceptable to whom? To your server? To the
poor schmuck downloading at the end of a 2400 baud SLIP connection? or
acceptable to me at the university with high speed connectivity?
Likewise with how you want the data to display. Bill Minnick likes to show
the individual only on one HTML document with a separate doc for marriage
and kids. My pages show the kids on each parents' doc. We may standardize
on one of these or yet another display format.
The GenWeb Foundation is being formed to work out these standards. You will
hear talk about standards more and more in the future. And you can
participate in the process via the genweb list.
Cheers,
Gary
***************************************************************************
*Gary B. Hoffman, Computer/Language Lab Director e-mail: ghoffman@ucsd.edu*
*Graduate School of International Relations and Pacific Studies (IR/PS)*
*University of California, San Diego (UCSD) voice: (619) 534-7733*
*9500 Gilman Dr., La Jolla, CA 92093-0519 USA fax: (619) 534-3939*
***************************************************************************
Wednesday, March 29, 1995 1:47:52 PM
GenWeb Item
From: Nancy Sween,nsween@falcon.cc.ukans.edu,Internet
Subject: Re: Fwd: Re: Trial Hypertext for West Family
To: GenWeb
My 2 cents worth: Good answer on standards, Gary!
Nancy Sween
===================================================================
Genealogy site: http://falcon.cc.ukans.edu/~nsween/genealogy.html
Friday, March 31, 1995 10:06:45 AM
GenWeb Item
From: Robert W Lyon,lyon@cs.byu.edu,Internet
Subject: Request for Electronic Journals/Histories
To: GenWeb
I am a graduate student of computer science at Brigham Young University. My
thesis involves studying time references in texts, and algorithms to automate
finding such references. Currently, I am building a corpus of various journals
and histories to begin my research. If you know where I can download such
journals/histories, please send me the internet address. If you have any of
these in electronic form, I would appreciate you sending them to me through
e-mail. Our goal is to provide a time searching mechanism, to simplify the
searching through large quantities of media.
Thank you,
Robert W. Lyon
lyon@bert.cs.byu.edu
Friday, March 31, 1995 12:40:38 PM
GenWeb Item
From: Mickey Lane,MLANE@csi.compuserve.com,Internet
Subject: Request for Electronic Journals/Histories
To: GenWeb
Robert W Lyon writes:
>I am a graduate student of computer science at Brigham Young University. My
thesis involves
>studying time references in texts, and algorithms to automate finding such
references. Currently,
>I am building a corpus of various journals and histories to in my research. If
you know where
>I can download such journals/histories, please send me the internet address.
There seems to be some of this type of thing on CompuServe although I couldn't
guess at
the quanity since I wasn't paying much attention to it. If you don't get any
response anywhere
else, let me know and I'll look into it if I can.
The type of work that you're describing has long been an interest of mine as
part of a genealogy
software package I've been working on for entirely too many years. At one point,
I wrote some
code to parse freeform text such as you're describing to pick out names, dates,
places and such.
The object of all this was to feed text into one end of the program and get
genealogy databases
out the other.
The algorithm I used went roughly like this:
Scan the file looking for abbreviations. Convert to whole words.
Scan the file looking for sentences. Put in markers to delimit. Although the
English language
can be highly context sensitive, most of the stuff I ran into was expressed
in simple statements
that did not rely on adjacent text. By far the most troublesome constructs
are tables, charts and
lists as they use some measure of whitespace to convey information.
Convert to all capitals (or lower case).
Identify words as all alpha or all numeric or something else.
Locate numeric words.
In a progressive guessing type of thing, pass word strings surrounding,
between and including
the numerics to a date conversion routine. The routine returns the converted
date (if possible)
and a numeric value that approximates how good the conversion is. Example: "1
Jan 1844"
returns 1, "Jun 1922" returns 2, "1903-1904" returns 3, "1810-1830" returns 6
and finally,
"name was Fred" returns 51. The worse the string gets in terms of a date, the
higher the
number. The routine I had would take stuff like "summer of 1840" and return 1
Jul 1840 with
a value of 3 :-) I think you would be quite surprized with what it finds.
Last step, it deleted the converted text from the origional and replaced it
with the database
format for a date. Use some sort of delimitor so you don't keep parsing the
database fragments
you've added.
Successive passes of software looking for names continued the deletion of
text and the
addition of database fragments.
At some point, all the text would be gone and all that would be left was
database. I never
got that far, however.
I don't subscribe to every genealogy list but the one where this type of topic
seems to
be discussed from time to time is the GEDCOM list. All the list bickering has
pretty much
ruined usenet although I'd prefer to see this in that forum.
Good luck with your project. You're gonna need it - you've picked a big apple to
take a
bite out of. :-)
Mickey.
Friday, March 31, 1995 2:44:42 PM
GenWeb Item
From: Larry Maddocks Contract,LMADDOCKS@wpmail.code3.com,Internet
Subject: Request for Electronic Journals/Histories -Reply
To: GenWeb
There is a JOURNAL echo in the fido net that you could check out. I run a
genealogy
bbs that carries most of the genealogy fido echoes (801-466-5374).
I have been working on software that goes in the other direction: It takes a
genealogy database and creates a narrative from it. I've been looking for time
line
info to insert into the narrative. If you get a database of historical events
that
take place in various town in the world, I would be interested in it.
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
Robert W Lyon writes:
>I am a graduate student of computer science at Brigham Young University. My
thesis
involves
>studying time references in texts, and algorithms to automate finding such
references. Currently,
>I am building a corpus of various journals and histories to in my research. If
you
know where
>I can download such journals/histories, please send me the internet address.
Sunday, April 2, 1995 8:38:57 PM
GenWeb Item
From: Cliff Manis,cmanis@ProgCons.COM,Internet
Subject: GenServ GEDCOM Server (home page)
To: GenWeb
GenServ - Home Page
Genealogical GEDCOM Server
I am pleased to announce the GenServ has its WWW home page available
for anyone to see and obtain information about this system.
Some items of interest are:
* GenServ Information and some history
* How to get access to this Server
* Complete Genserv Documentation
The GenServ WWW server can be found at the URL:
http://www.cs.ncl.ac.uk/genuki/GenServ/
For those not having access to WWW, just send a blank message to the
following address for information. genserv-info@progcons.com
If you have any comments questions about this URL, please feel free
to contact Cliff Manis, Internet: cmanis@ProgCons.COM
GenServ "Genealogical Server" a service for making GEDCOM data available.