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. Monday, April 3, 1995 11:28:05 AM GenWeb Item From: Brian Randell,Brian.Randell@newcastle.ac.uk,Internet Subject: Re: genweb.org established To: GenWeb Fred: >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. Right on! If you look at the Web pages we are creating for uk+ireland you will find a discussion of how and why our server is structured the way it is which you might find interesting. It's at URL: http://midas.ac.uk/genuki/ Cheers Brian Dept. of Computing Science, University of Newcastle, Newcastle upon Tyne, NE1 7RU, UK EMAIL = Brian.Randell@newcastle.ac.uk PHONE = +44 191 222 7923 FAX = +44 191 222 8232 URL = http://www.cs.ncl.ac.uk/~brian.randell/ Monday, April 3, 1995 7:53:38 PM GenWeb Item From: ANNELISE ANDERSON,ANDRSN@HOOVER.STANFORD.EDU,Internet Subject: Re: Ancestor Data Base To: GenWeb The number of "end of line" ancestors is not so large....these are the ancestors for whom no parents have yet been found. Once parents are found, they are no longer "end of line"--they are descendants. Note that people who had no children and people who disappeared without a trace do not become part of the data base. Smiths and Schmids and Schmidts and Andersons and Andersens are always a problem. Presumably birth and death dates help narrow the search. It just seems to me that if you predict the direction in which this is going to go, given the capability of the Internet for communications and on-line data retrieval, this alternative is a real possibility. The problem is merging GEDCOMs into descendant data, or rather merging descendant data from descendancy data developed from individual GEDCOMs, and of course dealing with all the errors in the data. Annelise Monday, April 3, 1995 7:56:49 PM GenWeb Item From: Rik Vigeland,rikv@wv.MENTORG.COM,Internet Subject: (Fwd) Ancestor Data Bank To: GenWeb A few thoughts on Annelise's post. 1) The total number of people who lived in Europe between 1600 and 1800, inclusively, will be a LOT more than 2 million. I would guess it's actually somewhere between 20 and 50 million. 2) If it were easier to start early and trace forward, more people would have done it by now. If I have an ancestor surnamed "Smith", it is not practical to look at all the Smiths in 1700 and trace their descendants forward, even if it were possible. A name like Twisleton or Shoentahlmeier or something else quite rare may be a different matter, but this is the exception. And I still might trace them forward and not find ME and MINE! There may also be any number of dead ends, families that have no more descendants after 3 - 4 - 5 generations! How much time are we willing to expend on them before we find this out? How do we know when we run into a dead end? You could document ancestors of people today who could care less about their roots (but hopefully somebody would some day?) 3) A more practical method would be to encourage people to enter MORE names into their GEDCOM files. I, for example, do not have ALL the children of ALL my ancestors entered in. Some families I do, most I do not because I just don't have the time. How many nth great aunts, uncles and cousins do you (all) have on paper that you haven't entered? I'd bet at least as many as direct ancestors. There is already plenty of work to do for the average amateur genealogist, BUT, if you or I took five or ten extra minutes when we find an ancestor in a book to record at least all the siblings and then enter them into a GEDCOM, there would undoubtedly be a lot more "matches" returned when someone submits a query. For that matter, just think of all the additional GEDCOM databases that we could create if we did nothing more that computerize every family history we could get our hands on! There must be a few million KNOWN ancestors who do not reside in ANY database. Go get them first! In summary, I suspect that there are already far more projects that we could take on with a known value (i.e., someone would benefit from it), that we do not need to think up even more projects. Rik rikv@wv.mentorg.com *** My Family Tree is Lost in the Forest! *** ============================================================================= ANNELISE ANDERSON (ANDRSN@HOOVER.STANFORD.EDU) wrote: >It strikes me that in terms of European ancestors, we are all looking >for the same people--perhaps about a million of them. These would be >the "end of line" ancestors living in Europe between 1600 and 1800 or so, >perhaps a few earlier. > >Is it possible that eventually we will identify these 1 million (or 2 >million) people and provide a million names with descendants. Descendants >would appear on more than one list, of course. A descendants list would >involve merging separate data bases. This does not eliminate the pro- >blems of overlapping GEDCOMs. It just works from a different direction-- >on the assumption that these ancestors are hard to find and, once found, >they should stay found. > >Comments? > >Annelise Anderson --- End of forwarded mail from ANNELISE ANDERSON Monday, April 3, 1995 7:58:12 PM GenWeb Item From: ANNELISE ANDERSON,ANDRSN@HOOVER.STANFORD.EDU,Internet Subject: Ancestor Data Bank To: GenWeb It strikes me that in terms of European ancestors, we are all looking for the same people--perhaps about a million of them. These would be the "end of line" ancestors living in Europe between 1600 and 1800 or so, perhaps a few earlier. Is it possible that eventually we will identify these 1 million (or 2 million) people and provide a million names with descendants. Descendants would appear on more than one list, of course. A descendants list would involve merging separate data bases. This does not eliminate the pro- blems of overlapping GEDCOMs. It just works from a different direction-- on the assumption that these ancestors are hard to find and, once found, they should stay found. Comments? Annelise Anderson Tuesday, April 4, 1995 9:15:59 AM GenWeb Item From: Rik Vigeland,rikv@wv.MENTORG.COM,Internet Subject: Re: Ancestor Data Base To: GenWeb >The number of "end of line" ancestors is not so large....these are the >ancestors for whom no parents have yet been found. Once parents are >found, they are no longer "end of line"--they are descendants. Note >that people who had no children and people who disappeared without a >trace do not become part of the data base. Perhaps the intended scope of Annelise's original post was not clear to me. If one starts with "end of line" ancestors who are ALREADY in GEDCOM databases somewhere, then yes, they should never be "lost" again, and may not be in the *multiple* millions. >It just seems to me that if you predict the direction in which this is >going to go, given the capability of the Internet for communications and >on-line data retrieval, this alternative is a real possibility. The main problem is that someone still has to do all of the work. We have a long way to go to get information on-line in the first place, and, as I stated before, that's just for DOCUMENTED ancestors. Many people are still sifting through original records (which may someday be available on-line?). In another 20 years we may be the last generation of genealogists who have to pay their dues in dusty courthouse basements. But I do think that it will take years of work. If we save someone a year of research some day by having GENWEB crunch out detailed pedigrees, hopefully that person is willing to spend a year repaying the system by digging out something not in the system yet and entering it in. The ratio of on-line research to dusty papers will increase exponentially. >The problem is merging GEDCOMs into descendant data, or rather merging >descendant data from descendancy data developed from individual GEDCOMs, >and of course dealing with all the errors in the data. I couldn't agree more. Let's get busy. Rik. rikv@wv.mentorg.com (That's not West Virginia, it's Wilsonville, Oregon) Thursday, April 6, 1995 8:36:01 AM GenWeb Item From: Mickey Lane,MLANE@csi.compuserve.com,Internet Subject: a few testers needed To: GenWeb Greetings, I'm getting close to putting ROOTSBOOK on the web and I need a couple people to help test it prior to turning it loose. (First release will have 100,000+ people :-) If willing to spend some time testing (and letting me know what's busted!) please send mail to mlane@csi.compuserve.com Thanks, Mickey. Thursday, April 6, 1995 8:52:24 PM GenWeb Item From: LaMont Bankson,montypi@indirect.com,Internet Subject: Re: a few testers needed To: GenWeb >I'm getting close to putting ROOTSBOOK on the web and I need a couple people >to help test it prior to turning it loose. (First release will have 100,000+ >people >:-) > >If willing to spend some time testing (and letting me know what's busted!) >please send >mail to mlane@csi.compuserve.com > >Thanks, >Mickey. Mickey, Let me know, I'd be happy to test for you montypi@indirect.com Friday, April 7, 1995 4:56:01 AM GenWeb Item From: Mickey Lane,MLANE@csi.compuserve.com,Internet Subject: re: a few testers needed To: GenWeb Yesterday, I said: >I'm getting close to putting ROOTSBOOK on the web and I need a couple people to >help test it prior to turning it loose. Today, it's: Yikes! Enough, already! I've got so may responses that it'll take two days just to read 'em all. To those who have responded, please hang in there. It may take a while but I'll get back to you... Thanks, Mickey. Monday, April 10, 1995 8:11:29 AM GenWeb Item From: Gene Stark,gene@starkhome.cs.sunysb.edu,Internet Subject: ged2html GEDCOM to HTML converter, version 2.0 available To: GenWeb Version 2.0 of my "ged2html" GEDCOM to HTML converter program is now available via my home page at: http://www.cs.sunysb.edu/~stark You can also view my family history document, which shows what the output of the program looks like, from that link. I have used ged2html to completely preprocess my own data into HTML files. However, several people are using version 1.x of ged2html together with LifeLines to construct HTML pages on the fly. The main improvements over the previous (1.x) versions are: * Executables for DOS and WINDOWS that correspond to the current sources. * The YACC parser has been replaced with a new parser that will read "any" GEDCOM. It complains only about the most grossly malformed GEDCOMS and will generally be able to make some sense out of anything that looks remotely like a GEDCOM. * Option-controllable distribution of individual files into subdirectories. This improves the performance of the program for GEDCOM's containing more than 100 individuals, especially under DOS and Windows * Some support for automatically including source information in the output. I have used this program to process GEDCOM's of well over 100K lines, containing thousands of individuals. The program will process all the GEDCOM's on Yvon Cyr's Acadian/French Canadian CD-ROM. - Gene Stark Sunday, April 23, 1995 6:01:54 PM GenWeb Item From: Annelise Anderson,ANDRSN@HOOVER.STANFORD.EDU,Internet Subject: Data Base On-Line To: GenWeb I've converted my GEDCOM to HTML format; it's available at: http://hoover.stanford.edu/~andrsn/index.html Tim Doyle has included it in his listing of genealogy data bases. I used Gene Stark's program for converting GEDCOMs to HTML--the DOS/Windows version, which I got by FTP from genealogy.emcee.com. (I got PKUNZIP.EXE, a 1994 version, somewhere else). It worked beautifully; no editing of file names was necessary. This is a small data base (~250 names) and took 15 seconds to process. I have read the postings here on domain names and have not entirely understood them. However I'm willing to change the directory structure and if I redo the HTML files, which I will do if there's enough new data, I can, I believe, change the name "index" and all references to it in all files with relative ease; my wordprocessor will search and replace across files. (At this point I'm not willing to ask my systems administrator to do anything.) This data base includes my notes, which were not orginally prepared with the idea that they would be read by others. These notes include references, and often they are incomplete--a shortened title or just the author's name and date. Usually the complete reference exists _somewhere_, i.e., in the notes for another person. Now I imagine anyone really interested could get in touch with me by e-mail and ask for the full reference. Meanwhile there are perhaps two ways to handle references--a complete reference in the notes (or some other field) every time, so that the data on an individual are complete; and a separate reference list or bibliography of the type that appears at the end of academic papers. Any thoughts on this? Thanks Annelise Monday, April 24, 1995 4:46:47 AM GenWeb Item From: Gene Stark,gene@starkhome.cs.sunysb.edu,Internet Subject: GEDCOM sources (was: Data Base On-Line) To: GenWeb >Meanwhile there are perhaps two ways to handle references--a complete >reference in the notes (or some other field) every time, so that the >data on an individual are complete; and a separate reference list or >bibliography of the type that appears at the end of academic papers. >Any thoughts on this? The GEDCOM 5.3 standard provides for "source records", which can be cross-referenced from individual or family records. My ged2html conversion program handles these cross-references to a certain extent, so that the cross-references are followed and the full source information presented in each individual page. Ideally, one would have one source record for each reference, and then cross-reference them from the other records, so that the complete source information would only have to be entered once. Unfortunately, most genealogical database programs do not seem to support this feature at the present time. - Gene Stark Tuesday, April 25, 1995 3:19:18 AM GenWeb Item From: Peter and Ann Floyd,sunshine@acslink.net.au,Internet Subject: Re: Data Base On-Line To: GenWeb OK, My surname database is getting to the point where I'd like to put it (and some other research) on the Web for all to see and comment. I have just under 500 names, some dating back to 1200, but mostly in the 17-1900's. Not a huge database, but big enough for me. My problem, To put it on the web, I need some way of showing the data in HTML format. My choices appear to be as follows, but each have drawbacks: 1. Use that utility GED2HTML (the author of which I can't recall right now). ADV: Very nice format, includes an index, full information and links to the next person. DIS: 501 files, each of a reasonable length, but still too many files for my SysAdm to like it. 2. Use a straight text file/GEDCOM file and include a GEDCOM viewer for users to download. ADV: Small number of files, easily updated. DIS: Casual browsers won't download all necessary files just to see what I've got. 3. Place a GIF of the tree online for all to see. ADV: Easy to see what I've got. DIS: Loses a lot of information (locations, full details, notes etc). Can anyone tell me if there are other ways of doing this? I've heard of the Lifelines project and another person who wrote a UNIX 'c' program to overcome the many files problem of '1' above ... but I'm no 'c' programmer, and don't know where to start. Cheers:-------------------------------Peter and Ann Floyd\ |Email:Sunshine@acslink.net.au Revelations 14:6-7.| \Genealogical Research: WHITBY anywhere,anytime. Mail me!/ Tuesday, April 25, 1995 5:28:57 AM GenWeb Item From: Mickey Lane,MLANE@csi.compuserve.com,Internet Subject: Re: Data Base On-Line To: GenWeb Peter and/or Ann Floyd wrote: >My problem, To put it on the web, I need some way of showing the data in >HTML format. My choices appear to be as follows, but each have drawbacks: A fourth option is to send the GEDCOM file to me. If you do, I'll add it to the 178 files containing 78,000 people currently being served by the ROOTSBOOK web server. This server is being used by a few people who are helping test it. When someone, somewhere gets thier act together and actually ships me my copy of the Alpha C compiler, I'll put the last few pieces in place and announce it's URL here. The entry count should be well over 100,000 by then. I have 114 databases I haven't even processed yet... Mickey. Tuesday, April 25, 1995 6:01:28 AM GenWeb Item From: Tim Doyle,tdoyle@netcom.com,Internet Subject: Re: Data Base On-Line To: GenWeb > My choices appear to be as follows, but each have drawbacks: > > 1. Use that utility GED2HTML ... still too many files ... I agree. > 2. Use a straight text file/GEDCOM file and include a GEDCOM viewer for > users to download. I think that you might be confused about how this type of method works. If you are referring to what I call the Abell method or the IGM method, then the program which extracts the information from the GEDCOM file is located on the SAME system as the GEDCOM. When a user requests information on a person in the GEDCOM, the program extracts the information from the GEDCOM and sends only the data needed. The entire GEDCOM is never sent, and the program is also never sent. Read below for more information on the method that I have been working on. > 3. Place a GIF of the tree online for all to see. I wouldn't recommend this... > Can anyone tell me if there are other ways of doing this? The problem, as I see it, is to have an automated process which will present a user with information from our genealogical databases. The trick is to do this without having a huge number of files on the system as well as to minimize the amount of extraneous data sent to the user. Having a single HTML file for each individual is not an option for my 7,000+ person database. On the other hand, expecting the user to download my entire 3 meg GEDCOM is just as bad. Several people have created scripts which work in conjunction with the LifeLines program to extract data from the database when requested. I personally don't want to have to load LifeLines just to be able to get my data on the Web. Vic Abell wrote a program which extracted key information from a GEDCOM and created a new, special format file, as well as several indexes to it. When a user requested information, another program would use the indexes and then extract the information from this special file and send it to the user. To me this was a great way of doing it. But why the special middle-man file? I decided to write my own. The result is what I call the Indexed GEDCOM Method. The script IGMMake, which creates indexes to the GEDCOM, does its job in under a minute for my large database. The scripts IGMGet and IGMFind get an individual from the GEDCOM and search the GEDCOM for any text, respectively. IGMFind is still being debugged and is not yet implemented for public use. For anyone who is interested, I have an index of all known GenWeb databases, organized by the method used to present the information. There are currently over 30 databases available. IGM is described and my Doyle test database is here if you'd like to see what IGM can do. I will soon (hopefully) be to the point of needing a few beta testers to try out IGM. Anyone interested?? GenWeb Site Index: http://www.doit.com/tdoyle/genweb/genweb.html Tim Doyle / tdoyle@netcom.com / tdoyle@doit.com WWW homepage: http://www.doit.com/tdoyle/ ftp directory: doit.com pub/tdoyle Tuesday, April 25, 1995 6:39:55 AM GenWeb Item From: Scott McGee,smcgee@microware.com,Internet Subject: time for an index? To: GenWeb With all the new genweb type pages poping up, I start to wonder if it is time for some sort of index. What I am thinking is a surname index. Basically, a list of surnames, and the url's of sites with info on that surname. It seems that if someone were willing to set up such a site, and take the time to make additions, that we could each send a sorted list of surnames (preferably with a count attached) to such a site. Software could be written to help automate the adding of new entries. That way, for the little bit I have off my homepage (as opposed to my full genweb site which is still not accessible from outside but has 4000+ entries) I could list the following: SURNAME Count URL Firth 1 http://www.cc.utah.edu/~sam8644/homepage.html Johnson 2 http://www.cc.utah.edu/~sam8644/homepage.html Jones 2 http://www.cc.utah.edu/~sam8644/homepage.html McCarthy 1 http://www.cc.utah.edu/~sam8644/homepage.html McGee 8 http://www.cc.utah.edu/~sam8644/homepage.html For the web availible version, encode the URL as a link to that URL and then when someone finds a name they want to look up, they just select the RUL portion and wind up at that site. So, what do you all think? Anyone want to start such an index. I can help write some code to help maintain it. Scott |- I never actually SAID that! -|- email: smcgee@microware.com -| |------------ http://www.cc.utah.edu/~sam8644/homepage.html ------------| Programming in assembly language is a lot like digging a post hole with a teaspoon. Of course, you DO have complete control over every bit of dirt! Tuesday, April 25, 1995 8:09:12 AM GenWeb Item From: Brian Randell,Brian.Randell@newcastle.ac.uk,Internet Subject: Re: time for an index? To: GenWeb Scott: >With all the new genweb type pages poping up, I start to wonder if it is >time for some sort of index. What I am thinking is a surname index. Basically, >a list of surnames, and the url's of sites with info on that surname. > >It seems that if someone were willing to set up such a site, and take the >time to make additions, that we could each send a sorted list of surnames >(preferably with a count attached) to such a site. Software could be written >to help automate the adding of new entries. That way, for the little bit I >have off my homepage (as opposed to my full genweb site which is still not >accessible from outside but has 4000+ entries) I could list the following: > >SURNAME Count URL > >Firth 1 http://www.cc.utah.edu/~sam8644/homepage.html >Johnson 2 http://www.cc.utah.edu/~sam8644/homepage.html >Jones 2 http://www.cc.utah.edu/~sam8644/homepage.html >McCarthy 1 http://www.cc.utah.edu/~sam8644/homepage.html >McGee 8 http://www.cc.utah.edu/~sam8644/homepage.html I'm a great believer in avoiding needless re-invention. Though I have to declare an interest, as they say, since I was so impressed with GenServ as to offer to edit the manual, I'd suggest that an obvious model for the searching and reporting facilities you might provide, and perhaps the method of providing them, is in fact GenServ. If I'm looking for Randells, say, I'd like to get reports on all meeting the search criteria I provide, not just a set of addresses of starting points to various heterogeneous home pages. And I'd like to have a common format for browsing all the various GEDCOM databases and extracting information from them - even if it was one of a number of a different such formats (just as there are a number of different Web browsers). Another analogy is to the various different information retrieval services which each provide a common interface to multiple different databases, library catalogues, etc., and allow you to do searches regardless of the fact that multiple such catalogues and databases are involved. Cheers Brian Dept. of Computing Science, University of Newcastle, Newcastle upon Tyne, NE1 7RU, UK EMAIL = Brian.Randell@newcastle.ac.uk PHONE = +44 191 222 7923 FAX = +44 191 222 8232 URL = http://www.cs.ncl.ac.uk/~brian.randell/ Tuesday, April 25, 1995 8:25:26 AM GenWeb Item From: Tim Doyle,tdoyle@netcom.com,Internet Subject: Re: time for an index? To: GenWeb On Tue, 25 Apr 1995, Scott McGee wrote: > With all the new genweb type pages poping up, I start to wonder if it is > time for some sort of index. What I am thinking is a surname index. Basically, > a list of surnames, and the url's of sites with info on that surname. > > So, what do you all think? Anyone want to start such an index. I would be willing to work on such an index. If you have a database online, please email a list of surnames and the URL to your database. If you are already listed on my GenWeb Site Index, I'll be able to get the URL from that. I know that there have been discussions about an every name index and linking people between databases, but this surname index should fill the gap until these projects are actually ready. Tim Doyle / tdoyle@netcom.com / tdoyle@doit.com WWW homepage: http://www.doit.com/tdoyle/ ftp directory: doit.com pub/tdoyle Tuesday, April 25, 1995 9:49:22 AM GenWeb Item From: Matthew Helm,helm@alexia.lis.uiuc.edu,Internet Subject: Re: time for an index? To: GenWeb Scott, For some time I have been wanting to set up a site that was a surname index to all genealogy information on the WWW. I've laid the groundwork for it in my Genealogy Toolbox, but I want to make it its own site. So, I'm willing to take on the task. My job will certainly be a lot easier if a sorted list of surnames can be generated by a program. Unfortunately, I'm not a programmer, and until a program exists, I'll try to build the index by hand. Thanks. Matthew Helm helm@alexia.lis.uiuc.edu Genealogy Toolbox - http://ux1.cso.uiuc.edu/~al-helm/genealogy.html On Tue, 25 Apr 1995, Scott McGee wrote: > With all the new genweb type pages poping up, I start to wonder if it is > time for some sort of index. What I am thinking is a surname index. Basically, > a list of surnames, and the url's of sites with info on that surname. > > It seems that if someone were willing to set up such a site, and take the > time to make additions, that we could each send a sorted list of surnames > (preferably with a count attached) to such a site. Software could be written > to help automate the adding of new entries. That way, for the little bit I > have off my homepage (as opposed to my full genweb site which is still not > accessible from outside but has 4000+ entries) I could list the following: > > SURNAME Count URL > > Firth 1 http://www.cc.utah.edu/~sam8644/homepage.html > Johnson 2 http://www.cc.utah.edu/~sam8644/homepage.html > Jones 2 http://www.cc.utah.edu/~sam8644/homepage.html > McCarthy 1 http://www.cc.utah.edu/~sam8644/homepage.html > McGee 8 http://www.cc.utah.edu/~sam8644/homepage.html > > > For the web availible version, encode the URL as a link to that URL and > then when someone finds a name they want to look up, they just select the > RUL portion and wind up at that site. > > So, what do you all think? Anyone want to start such an index. I can help > write some code to help maintain it. > > Scott > > |- I never actually SAID that! -|- email: smcgee@microware.com -| > |------------ http://www.cc.utah.edu/~sam8644/homepage.html ------------| > Programming in assembly language is a lot like digging a post hole with a > teaspoon. Of course, you DO have complete control over every bit of dirt! > Tuesday, April 25, 1995 9:50:39 AM GenWeb Item From: T.T.Wetmore,ttw@beltway.att.com,Internet Subject: Re: Database On-line To: GenWeb Tim Doyle (>): >The problem [extracting HTML files from genealogical databases], as I see >it, is to have an automated process which will present a user with >information from our genealogical databases. The trick is to do this >without having a huge number of files on the system as well as to >minimize the amount of extraneous data sent to the user. Agreed, and this is one of the basic issues of genwebbin' that has been discussed a lot here. >Having a single HTML file for each individual is not an option for my >7,000+ person database. On the other hand, expecting the user to >download my entire 3 meg GEDCOM is just as bad. Agreed. >Several people have created scripts which work in conjunction with the >LifeLines program to extract data from the database when requested. I >personally don't want to have to load LifeLines just to be able to get my >data on the Web. If you have UNIX, if your current program can extract GEDCOM, you can have a LifeLines-based, on-demand, HTML generating WWW system that meets your needs up and running in short order (there are detailed instructions around). Besides that, you can have one of the most powerful genealogical systems around as a free bonus. I have no vested interest in this (though I am the author of LifeLines, which is free), but I would dislike someone taking your advice about this simply because you "don't want to." Use of LifeLines to manage a WWW gen web site is an excellent alternative that is proven and available now, and has significant advantages that the two other methods you outline below do not. >Vic Abell wrote a program which extracted key information from a GEDCOM >and created a new, special format file, as well as several indexes to >it. When a user requested information, another program would use the >indexes and then extract the information from this special file and send >it to the user. Sounds like a good solution. >To me this was a great way of doing it. But why the special middle-man >file? I decided to write my own. The result is what I call the Indexed >GEDCOM Method. The script IGMMake, which creates indexes to the GEDCOM, >does its job in under a minute for my large database. The scripts IGMGet >and IGMFind get an individual from the GEDCOM and search the GEDCOM for >any text, respectively. IGMFind is still being debugged and is not yet >implemented for public use. Sounds like an excellent system also; sounds like both methods depend on "middle men," which I have nothing against. Afterall, the LifeLines "middle man" is an entire genealogical database system, much bigger than Vic's file or your index! Thanks for info on your web site. Tom Wetmore, 4/25/95, ttw@beltway.att.com Tuesday, April 25, 1995 11:03:42 AM GenWeb Item From: LouPero@aol.com,Internet Subject: what is this? To: GenWeb I am not sure how I got on this mailing list, but it sounds interesting. Unfortunately, I have no idea what most of it is about, but I would like to learn. Is there a way to get some sort of explanation? Tuesday, April 25, 1995 11:43:12 AM GenWeb Item From: Brian Randell,Brian.Randell@newcastle.ac.uk,Internet Subject: Re: time for an index? To: GenWeb Scott: >I don't know much about GenServ. Can you tell me more. I'm sure I've heard >of it, but can't pull specifics out of my memory right now. Is it availible >on the net or what? The manual, and samples of a number of the various types of report it can produce are at http://boulmer.ncl.ac.uk/genuki/GenServ/ A paper describing it appeared in the September issue of Genealogical Computing. In summary, it is a system which has been populated by a large number of GEDCOM files submitted by users, who then have the right to submit emailed queries, to which automated replies are sent. (It is implemented using LifeLines, of course! :-) The owner Cliff Manis is contemplating putting a Web interface to it, but at the moment is (rightly IMHO) concentrating on the service itself. Cheers Brian Dept. of Computing Science, University of Newcastle, Newcastle upon Tyne, NE1 7RU, UK EMAIL = Brian.Randell@newcastle.ac.uk PHONE = +44 191 222 7923 FAX = +44 191 222 8232 URL = http://www.cs.ncl.ac.uk/~brian.randell/ Tuesday, April 25, 1995 2:01:15 PM GenWeb Item From: Scott McGee,smcgee@microware.com,Internet Subject: Re: time for an index? To: GenWeb Right, GenServ is what I was thinking it was. An excellent idea too, but not quite providing the function I was advocating. Web surfers now have to go to one of the sites cataloging all the different genweb sites, and then visit each site individually looking for surnames of interest. What I had in mind was one site that would tell you which other sites had the surname you were looking for. I have corresponded with Tim Doyle a few times, and he is well on his way now to providing this capability. I have gotten mail (some I haven't ansered yet) from a couple of others interested in doing the same. I also got a note from the author of Rootsbook concerning his software. It too is excellent, but like genserve, does not fill the role I had in mind. It will allow a common interface to all databases it contains and even has (will have) a web presence, but again, I look at the existing web sites and would like an index the surnames on them. I see one BIG disadvantage and one big ADVANTAGE to services like GenServe and RootsBook. The advantage is the huge volume of data that is all accessible via a common interface. This makes assesing the data much easier. The dis- advantage I see is that they lack (inherently, IMHO) the vast diversity of ideas that individual genweb pages present. Every time I check out a new genweb site, I tend to see a new idea that others hadn't thought of. Some get incorporated into my own software, or at least marked as possible additions. Others are interesting or even intriguing but limited to specific applications. I think it would be a loss if we didn't have this dirversity. My feelings on the subject are that the two methods can coexist side by side. I have about 2000 people in my database and a variety of reports that I can generate. As soon as our sysadmins make the server accessible to offsite access I hope to share this. This will not stop me from adding my data to GenServe and/or RootsBook. I will likely add my data to both services and still maintain my own server. I will take advantage of other services too. Things like the ROOTS Surname List, the soc.genealogy.surnames newsgroup and others. Scott |- I never actually SAID that! -|- email: smcgee@microware.com -| |------------ http://www.cc.utah.edu/~sam8644/homepage.html ------------| Programming in assembly language is a lot like digging a post hole with a teaspoon. Of course, you DO have complete control over every bit of dirt! Tuesday, April 25, 1995 2:23:20 PM GenWeb Item From: Annelise Anderson,ANDRSN@HOOVER.STANFORD.EDU,Internet Subject: Re: time for an index? To: GenWeb It seems to me that a surname index alone is useful only for really unusual surnames, and as hopeless as the telephone book for common ones. The two methods that seem to have been developed to further define a surname in time and space are the Tiny Tafel and the ROOTS-L for- mats, each of which provides a line for a surname. ROOTS-L provides an associated e-mail or S-mail listing; in this case the association would be to a URL. Perhaps something could be developed that's even better than the two existing formats. Annelise Tuesday, April 25, 1995 3:51:28 PM GenWeb Item From: Gene Stark,gene@starkhome.cs.sunysb.edu,Internet Subject: Re: Data Base On-Line To: GenWeb >Having a single HTML file for each individual is not an option for my >7,000+ person database. On the other hand, expecting the user to >download my entire 3 meg GEDCOM is just as bad. A reasonable operating system (*not* DOS) has no problem handling 7000 individual files. On my 33MHz 486 system running FreeBSD I can create 7000 files in a few minutes. The average file size for my database is less than 512 bytes, so one file/one disk sector. - Gene Stark Tuesday, April 25, 1995 6:04:21 PM GenWeb Item From: Annelise Anderson,ANDRSN@HOOVER.STANFORD.EDU,Internet Subject: Re: Data Base On-Line To: GenWeb From: IN%"gene@starkhome.cs.sunysb.edu" "Gene Stark" 25-APR-1995 17:25:06.77 To: IN%"tdoyle@netcom.com" "Tim Doyle" CC: IN%"genweb@ucsd.edu" Subj: RE: Data Base On-Line >Having a single HTML file for each individual is not an option for my >7,000+ person database. On the other hand, expecting the user to >download my entire 3 meg GEDCOM is just as bad. A reasonable operating system (*not* DOS) has no problem handling 7000 individual files. On my 33MHz 486 system running FreeBSD I can create 7000 files in a few minutes. The average file size for my database is less than 512 bytes, so one file/one disk sector. - Gene Stark It would be interesting to consider how many of us have "reasonable" operating systems. My 486 50Mz running MSDOS 6.2 had no trouble creating 250 files but the Hoover VAX had trouble when I downloaded them; a "RED DISK FULL" problem that slowed operations on the disk. It is not certain I was responsible for this but it is likely. Fortunately I was deleting old mail files shortly after and the disk recovered, but will undergo major overhaul of six hours Saturday morning. --Annelise Anderson Tuesday, April 25, 1995 9:03:40 PM GenWeb Item From: Deanna Weymuth,dew@tyrell.net,Internet Subject: Re: time for an index? To: GenWeb Matt do you have any idea how I got on this list or how I can get off. I belong to a teachers list and Iit takes almost all of my computer time. This looks interesting but I don't have online time for it. Thank you Deanna Tuesday, April 25, 1995 10:18:21 PM GenWeb Item From: Michael A. PattonĄ genealogy on the web,MAP=GenWeb@BBN.COM,Internet Subject: Re: time for an index? To: GenWeb We discussed this quite a bit in the early days, but it sort of fell by the wayside as some of the other details of initial deployment happened. A couple of people (me included) outlined some techniques that we should think more about. That earlier discussion should be in the archives. As the recent discussion of how to do the indexing has shown, there are a number of ideas on what type of index is the most useful. I think most of these are valid ideas, and the right approach is to define a common way that suitable info can be extracted from each GenWeb site so that various people can experiment with putting together indexes in various forms. This would want to be some sort of summarization, rather than dumping a whoile DB. One idea I presented earlier is to have a URL (or set of URLs) for each GenWeb DB which is a text/plain document designed for machine parsing. These could pretty easily (for fixed definitions) be done in most of the schemes I've seen people using. Then through the central registry at GenWeb.Org, an indexer can find all these and index the info in any way they want. The indexes themselves would also register, thus making it possible to also build meta-indexes automatically. The most difficult part is coming up with definitions for the summary documents. I find it hard to believe that we could apriori define a format (or small set of formats) that actually managed to cover all cases. One idea that I've bandied about is making the URLs allow for specifying what's desired. This also has the advantage of being more efficient, but the disadvantage that it essentially requires that the data be provided with CGI type service which some people don't have. I think I may have a way of addressing that problem. Now on to some specifics... One way that I have my local site (sorry, not yet accessable, I'm working on it :-) indexed is with a [Virtual] page per surname (partial URL ".../surnames/.html"), listing each person in the DB with that surname. I could then have a central page similar to what Scott describes which might contain: McCurley (3) (a link to McCurley.html) Meade (4) (a link to Meade.html) Orr (11) (a link to Orr.html) Patton (29) (a link to Patton.html) Plumley (1) (a link to Plumley.html) Richards (1) (a link to Richards.html) (hmmm that's a pretty good idea, I think I'll do it, it'll obviously be ".../surnames/index.html", easily generated with awk/sort/uniq :-) Then a type of index that I'm particularly interested in could accumulate these from many servers, and offer a simple correlation index and a ranking of servers with each surname. This could be augmented with an RSL-like localization of each entry (although I'm not thinking hard about that, yet, because I don't see any way to automate the generation of this info), although for indexing and lookup purposes, a somewhat different form might be better. -MAP Wednesday, April 26, 1995 1:26:33 AM GenWeb Item From: Annelise Anderson,ANDRSN@HOOVER.STANFORD.EDU,Internet Subject: Re: time for an index? To: GenWeb The indexing question is obviously complicated. I would like to suggest that ultimately it is desirable not only to be able to use an index to find the data bases one wants to search, but also to have some kind of central registry that says: Person x in this data base is the same as person y in that data base. ^^^ ^^^ If this is going to be possible, it would seem to me that e|`ach person in every data base has to be given a unique identification number. Otherwise the individual records have to be identified by location or URL, and it seems that these would change over time. Does this make sense, or not? If a unique identification number (probably of nine digits) is really necessary--ultimately--someone, that is some single entity, has to assign these numbers, or assign a batch of numbers to a data-base creator to use, and these numbers should stay with the individuals to whom they are assigned. I'm not sure how the matches would eventually be made and recorded, although one could speculate. But isn't that the ultimate in being able to search--to find another data base with some of the same people? Annelise Anderson Wednesday, April 26, 1995 2:32:33 AM GenWeb Item From: Brian Randell,Brian.Randell@newcastle.ac.uk,Internet Subject: Re: time for an index? To: GenWeb Annelise: >Brian, I think GENSERV is terrific. ROOTS-L and Tiny Tafels only put >one in touch with people who might have information to share; GENSERV >actually provides information. > >I don't know the programming problems well enough to know how many >different ways there might be to do something comparable. I can't >think of anything comparable right now--I wonder about the Mormons' >IGI? Certainly a lot of names there. Tedious to follow up on. > >Then the next step (I just posted the message) is in my view to >make a central register of identical people, after first assigning >every individual in every data base a unique ID number. (Doesn't >appeal to me much for living people, but for the dead I don't mind.) There are considerable problems with that - unless you are willing to accept that such numbers may change, as people do further research. (You might belatedly find out that Great-aunt Wilhelmina and Great-uncle William were one and the same - in fact a famous drag artist! :-) This is the very sensible solution used by Genserv. You can find people in various databases by various search techniques, and then specify a particular individual either using the attributes that you used to find him/her, or you can use the index number that Genserv will have reported to you. But the GEDCOM files can be updated, and you are warned that such updating could well invalidate the number, which therefore shoud be used with caution. (Thus the index number is being used essentially as what is sometimes called a "hint" in the computer science literature.) As regards my Genserv analogy, the point is that Genserv contains some sophisticated solutions to the problems of searching and reporting on a set of separately stored GEDCOM files, which remain the property of the owners, and can be updated by them. (Sounds like Genweb doesn't it?) The difference is that in Genserv the files happen to be stored on the same (UNIX) computer. In fact, it is possible, using existing solutions, such as NFS, to glue the files stores of a number of geographically separate UNIX systems together (to form what is called a "transparently distributed system") so as to appear to Genserv like a single file store. More generally, the Web provides an alternative technology for glueing a large number of machines (running various operating systems) together to provide the appearance of a single (huge) hypertext document. The relevance of Genserv's sophisticated solutions to searching and reporting on a set of separate Gedcom files to the problem of indexing a set of Genweb sites therefore seem self-evident to me. For example, it would be fantastic to have something of the power of the Genserv "match" command applied to the whole Genweb world. (However, I do not claim that there are no significant implementation issues remaining to be solved, given the heterogeneity of the machines involved in Genweb and the inadequacy of many network links, or that Genserv has nothing to learn from Web browser interfaces.) I hope this has made things clearer. Cheers Brian PS The bottom line is that the easiest way to solve a problem is usually to steal a good idea solution someone else has already produced! :-) (As Tom Lehrer once sang: "Plagiarize, plagiarize, let no-one else's work evade your eyes.") Dept. of Computing Science, University of Newcastle, Newcastle upon Tyne, NE1 7RU, UK EMAIL = Brian.Randell@newcastle.ac.uk PHONE = +44 191 222 7923 FAX = +44 191 222 8232 URL = http://www.cs.ncl.ac.uk/~brian.randell/ Wednesday, April 26, 1995 5:21:36 AM GenWeb Item From: Anders Andersson,andersa@Mizar.DoCS.UU.SE,Internet Subject: Re: Data Base On-Line To: GenWeb Gene Stark writes in response to Tim Doyle: >>Having a single HTML file for each individual is not an option for my >>7,000+ person database. On the other hand, expecting the user to >>download my entire 3 meg GEDCOM is just as bad. > >A reasonable operating system (*not* DOS) has no problem handling >7000 individual files. On my 33MHz 486 system running FreeBSD I can >create 7000 files in a few minutes. The average file size for my >database is less than 512 bytes, so one file/one disk sector. Even though a 3 MB GEDCOM database containing 7,000 entries indeed makes for approximately 450 bytes per entry, a typical HTML file prepared for WWW presentation would be much bigger than that. First, you have the overhead of descriptive field labels, headers and all HTML mark-up to make the data readable to a human user. Second, each HTML page would probably contain data from the records of persons directly associated with the subject individual (parents, spouse, and children). I don't have any real figures, but I'd say doubling your 512 byte/file would be a conservative estimate. It depends partly on how "fluffy" you want your HTML pages to appear. If your machine has no problem handling 7,000+ files, or storing some 2-3 MB of redundant data, fine. I have no problem with that either. Still, I think pre-generating every HTML page a user may ask for is a poor design, unless you have a serious problem generating them fast enough on the fly, and I don't think it scales well. You may want to trade CPU power against disk storage. Ultimately, it's a choice each database provider will have to make for themselves, based primarily on what resources and facilities are available to them. -- 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, April 26, 1995 5:23:08 AM GenWeb Item From: Phlete Teachout,fteachou@eagle1.eaglenet.com,Internet Subject: Re: time for an index? To: GenWeb One of the problems I see in all this discussion is the very real possibility of a huge bottleneck if !the! index is placed on any one machine. Would it not be possible for individuals to 'register' their GEDCOM, place it on their own resource, and then have several sites check these resources periodically and produce an index a la 'veronica' or 'WebCrawler' or 'Lycos?' If this can be done, then the GEDCOM browser software could be developed as a 'client' that could be utilized via the index sites. As you can probably tell, I'm not overly familiar with the technological aspects of all this. - 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 > <<<<<<<<<<<<<<<<<<<<<<<<<<<<<<>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> Wednesday, April 26, 1995 6:15:52 AM GenWeb Item From: Mickey Lane,MLANE@csi.compuserve.com,Internet Subject: re: Time for an index? To: GenWeb Greetings, I don't know about y'all but I'm beginning to really enjoy this thread... It's kind of like watching them draw your lottery ticket number in r e a l s l o w m o t i o n. Annelise Anderson writes: >I would like to suggest that ultimately it is desirable not only >to be able to use an index to find the data bases one wants to >search, but also to have some kind of central registry that says: >Person x in this data base is the same as person y in that data >base. ^^^ ^^^ This would be a management nightmare. Why have two identical people stored at all? Why not have one database just link to the other? [...] >If a unique identification number (probably of nine digits) is >really necessary--ultimately--someone, that is some single entity, >has to assign these numbers, or assign a batch of numbers to a >data-base creator to use, and these numbers should stay with the >individuals to whom they are assigned. Nine digits? Why on earth nine? Here's how ROOTSBOOK has been doing this very thing since... oh...1986 or so. People are divided into groups using whatever logic seems to work. All who lived in a given county, a certain family, had blue eyes, whatever. This becomes known as a "book." It is one database file much like a GEDCOM file. (It also can have a whole mess of associated files but that's another issue...) Each person in a book is given a number. There arent really any rules but in practice, the main person in the book, if there is one, is No. 1. The parents of No. 1 might be No. 10 and No. 20. Or No. 100 and No. 200. You leave enough space between numbers to add people as time goes by. I like to keep the number of people in a book to about 50 with the entry numbers in the 1-500 range. When someone in one book is related to someone in another, a link is drawn between the two. The link is specified only by book name and entry number. Example: to link to entry 10 in the Smith book, someone in the Jones book would reference 10. When the software runs, it moves from database to database as the links direct. The software finds the different databases by means of a micro-database that is updated on the fly. I've done this micro-database a number of different ways over the years. Currently, it's just a text file in a known location. I have used (and maybe will again) a client-server mechanism where the servers network with each other and actively look for database files by searching their system's disk structure. When the genealogy part of the software runs, it becomes the client and asks for book so-and-so and some server provides the location. (Gee, can we say "DCE?") >I'm not sure how the matches would eventually be made and recorded, >although one could speculate. But isn't that the ultimate in being >able to search--to find another data base with some of the same >people? When I did the ROOTSBOOK database spec, I included a bunch of stuff in the headers that relate to "registering" the database. There were a couple of issues. Where does one register a database? Obviously someplace that's permanent and accessible. A registered database is assumed to be good. We all know that there is no such thing as "Good" so there has to be some mechanism for fixing things to make them "Better." Who does this? How? Assuming someone is using the database that's being fixed, how does one report this to the users? (ROOTSBOOK has a per-entry revision date and the processing software can be told to flag new stuff.) In my view of a linked database, each party in the link must acknowledge the other and they must agree. If you have a registered database and more than one researcher linking to it, you have a problem. You solve this problem with the concept of hard links and soft links. I put some verbage in the spec about this but never implemented anything. All registered databases must live within the same namespace. You can't have two registered Smith databases. Your genealogy software must learn to deal with temporary network outages. The obvious solution is to have all the stuff local but this sort of defeats the purpose of registered databases, doesn't it? I think that what is needed to make this all work is a committee to act as a registration agent, one or more places to distribute the registered databases (ie net sites) and some software that can distribute the locations of the databases to still more software (ie genealogy database stuff) than can function in this environment. The real problem is going to be determining who goes in which registered database. I think this will be the time-consuming task of the committee. Someone who is familiar with Gopher and/or Archie and/or ??? might want to comment. Using something existing and understood is almost always better than inventing new. Save the inventing for stuff that doesn't exist. Mickey. Wednesday, April 26, 1995 6:27:14 AM GenWeb Item From: Matthew Helm,helm@alexia.lis.uiuc.edu,Internet Subject: re: time for an index? To: GenWeb I think there have been several good posts looking at the problem of generating an index of surnames of GenWeb sites. I know that the scope of this list is GenWeb, but there is also a lot of genealogical data on the WWW that is not in database form. I'm talking about the many family histories and pedigree charts, that individuals who do not have the room on servers to place their marked-up gedcom files, have placed on the WWW and on some gopher servers. Is the scope of this project only to index GenWeb sites, rather than trying a comprehensive index of all surnames mentioned in genealogical sites on the WWW? I know that these "non-GenWeb" sites would make automation of the index difficult. However, I believe that these sites may contain valuable information to researchers (for example, the New Brunswick Genealogy site). So, I guess we need to decide what the scope of the project is, and how to make access to the greatest number of genealogical resources available with a common search interface. Thanks. Matthew Helm Wednesday, April 26, 1995 8:38:12 AM GenWeb Item From: Brian Randell,Brian.Randell@newcastle.ac.uk,Internet Subject: re: time for an index? To: GenWeb Matthew: >I think there have been several good posts looking at the problem of >generating an index of surnames of GenWeb sites. I know that the scope >of this list is GenWeb, but there is also a lot of genealogical data >on the WWW that is not in database form. I'm talking about the many >family histories and pedigree charts, that individuals who do not have the >room on servers to place their marked-up gedcom files, have placed on the >WWW and on some gopher servers. Is the scope of this project >only to index GenWeb sites, rather than trying a comprehensive index of >all surnames mentioned in genealogical sites on the WWW? > >I know that these "non-GenWeb" sites would make automation of the index >difficult. However, I believe that these sites may contain valuable >information to researchers (for example, the New Brunswick Genealogy >site). So, I guess we need to decide what the scope of the project is, >and how to make access to the greatest number of genealogical resources >available with a common search interface. Good point - and one that I argued in the Genserv context with Cliff Manis a number of months ago. (I believe that as a result he has actively explored means of making the Roots Surnames List available as though it were part of GenServ.) However a general solution will probably require extremely advanced so-called "inteligent agent" technology - and is probably best not thought of from just a narrow WWW viewpoint. Incidentally, I have often been faced with a choice of a crudely-interfaced system which contains a huge amount of valuable data, and a marvellously interfaced system that is relatively empty. No prizes for guessing which wins. This reality I feel should be born in mind in plotting GenWeb's future. cheers Brian Dept. of Computing Science, University of Newcastle, Newcastle upon Tyne, NE1 7RU, UK EMAIL = Brian.Randell@newcastle.ac.uk PHONE = +44 191 222 7923 FAX = +44 191 222 8232 URL = http://www.cs.ncl.ac.uk/~brian.randell/ Wednesday, April 26, 1995 9:35:56 AM GenWeb Item From: Chris Garrigues,cwg@DeepEddy.Com,Internet Subject: re: Time for an index? To: GenWeb At 8:11 AM 4/26/95, Mickey Lane wrote: > This would be a management nightmare. Why have two identical > people stored at all? Why not have one database just link to the > other? Because of issues of reliablity of information. You and I may share an ancestor for whom we disagree about some item of data (such as parentage). One of the problems with Genealogy from a database POV is that some of the data is unreliable, but still useful as leads for locating more reliable information. If someone is looking at my tree where I claim that John Smith is the son of Robert Smith and sees a link to your John Smith who is the son of James Smith, They'll want to know that we're both referring to the same person. Hopefully, along with the link from my database to yours, I'll have a note on why I disagree with your claims. Equally hopefully, you'll have a link back to mine with an explanation of why you disagree with my claims. Chris Chris Garrigues cwg@DeepEddy.Com Deep Eddy Internet Consulting +1 512 432 4046 609 Deep Eddy Avenue Austin, TX 78703-4513 USA http://www.DeepEddy.Com/~cwg/ Wednesday, April 26, 1995 10:15:15 AM GenWeb Item From: olsen@cs.byu.edu,Internet Subject: re: Time for an index? To: GenWeb >[...] >>If a unique identification number (probably of nine digits) is >>really necessary--ultimately--someone, that is some single entity, >>has to assign these numbers, or assign a batch of numbers to a >>data-base creator to use, and these numbers should stay with the >>individuals to whom they are assigned. > >Nine digits? Why on earth nine? > >Here's how ROOTSBOOK has been doing this very thing since... >oh...1986 or so. > >People are divided into groups using whatever logic seems to work. >All who lived in a given county, a certain family, had blue eyes, >whatever. This becomes known as a "book." It is one database file >much like a GEDCOM file. (It also can have a whole mess of >associated files but that's another issue...) > >Each person in a book is given a number. There aren t really any >rules but in practice, the main person in the book, if there is one, >is No. 1. The parents of No. 1 might be No. 10 and No. 20. >Or No. 100 and No. 200. You leave enough space between >numbers to add people as time goes by. I like to keep the number >of people in a book to about 50 with the entry numbers in the >1-500 range. > I have been reading the WWW literature for the last several months as well as talking with people in the LDS family history department. What is needed here is a standard for electronic citations which is compatible with the WWW but handles the unique needs of family history. This standard would allow all of us to reliably reference each other's work. For it to work, however, will require a little more discipline in how we publish family history information. I hope to have a draft for such a standard up on the web within the next month at which point I expect lots of comments. ____________________ Dan R. Olsen Jr. Computer Science Department Brigham Young University Provo, UT 84602 (801) 378-2225 FAX 801-378-7775 Wednesday, April 26, 1995 1:04:31 PM GenWeb Item From: Mickey Lane,MLANE@csi.compuserve.com,Internet Subject: Copy of: re: Time for an index? To: GenWeb (If I do this right, I can get the response out before what I'm responding to show up ... :-) Scott McGee: >A couple of comments. ...all of which are valid. >I see a couple of problems with how some are envisioning things. Not with >their ideas, but with their assumptions. Please don't take my previous post as some hard and fast position on my part. No one has ever attempted to implement linked genealogy databases on a wide scale and before anything comes to pass, there's sure to be lots of discussion and compromise. I'm want to spout off about what I've done 'cause I think it's the way to go but I'm certainly willing to listen to others. >First, I don't feel it is correct to assume that people will necessarily >send their data to some other server agency. Many are likely to be willing >to make info availible but only if they retain control. Lots of people upload GEDCOMs to various places without any problems. The concept of registering just means that you upload to a specific place and the contents of the new file is made to fit with the existing material. It doesn't mean that you give up anything. >Second, people, especially those in the first note, are unlikely to want to >have someone else change their information. This means that if someone had >a John Doe in their DB and someone else finds that this John Doe was the >same as the J. Doe in their own database, it is quite likely that neither >will be sastisfied to just link to the other,... I see people linking several of thier own databases together and people linking to publically registered databases. I quite agree that people linking to other people's (private) databases is not going to work. >Third, In an ideal world there will never be mistakes. In real life there >will be databases where one person in each database are determined to be the >same, and then latter found to be different Maybe what we need to do is think in terms of a CD-ROM. Think of registering a database as the same as burning it into a CD - that many people can have a copy of. People link to the named databases. If someone changes one of the registered files, a new file, with a different name, is added to the CD. Once the news gets out about the new file, people can change their private databases whenever they want. >Fourth, again, in an ideal world, all duplicates would be know. In the real >world, you will always have people duplicated in multiple databases. You >will seldom know that they are the same! True. I'd like to at least try to devise a system that attempts to get rid of the dups, though. >Fifth, people disagree. If three of us have a John Doe, one of the other >two might think his John and my John are the same, but not the third. The >other might think his and the one are the same, but not mine, and I might >think they are all the same. There is no way to combine any of these three >Johns without upsetting somebody I guess people will have to make up their own minds and link to whatever they think is right. If enough people reject something in the set of registered files, maybe someone will get around to fixing it. >I don't know what the end meaning of all this is, but I am sure that these >points should be considered before we decide on anything intended to be a >final solution I don't think there is an end meaning. Something I should point out is that the discussion has sort of migrated from indexing web pages to linked databases. I subtle but significant shift! In the former, you're talking about a seamless interface to the results of private work. In the latter, you're talking of an endless thread running through private and public parts of the material used to generate the results - overlaid by the seamless interface to view.... etc. Mickey. Wednesday, April 26, 1995 7:54:24 PM GenWeb Item From: Cliff Manis,cmanis@progcons.com,Internet Subject: Some comments about www and GenServ To: GenWeb Dear Genweb: Some information and comments. I have been subscribed to the GENWEB list for the past 3 months, and have read 90+ % of the messages posted since that time. I still have copies of most. Many times I have wanted to reply and post some comments but have not had the time. How can one be SO busy ? Sometimes we learn more by reading and listening to what others think they want and need. During the last 2 months we have been completely re-doing the GenServ system. Within the next month the GenServ system (programs and databases) will on a new computer system with 2 gigs HD. The current GenServ system requires about 700 megs for programs and data. The GenServ system started in 1991, and has been LIVE on Internet since that time (with exception of about 5 months). Present Capabilities of the GenServ. At the present time, the system has over 600,000 names available from more than 400 different databases sent to me for this project. The system has more than 45,000 different surnames. The only way to get access to the system is to send me a GEDCOM file via diskette to be loaded into the system. Yes, this maybe a bad way to handle that but it works very well. At least this way, I get to load and check all the GEDCOM datafiles before they are put into the GenServ system. In todays world this sounds archiac, but it works well. We have automated the procedure for loading a new gedcom files into the GenServ system, and allow access. It works well. I can now send 10 or 15 GEDCOM (tested) files on a diskette back to the US, and it be loaded within a short time, and all administrative actions will be finished to give access to the providers of the GEDCOM databases. Since we are working on the new system, the new GEDCOM databases received during the past month are still waiting to be loaded, and I presently have more than 200,000 more names to be added. But during this time - have been giving access to tse sending GEDCOM files after their files having been tested. Last month alone, the GenServ has produced over 15 megs of output data for it's users which was emailed to them within 30 minutes on the day they requested the data. If you need more information or want to read the Documentation on the GenServ, please do not write me but do see my .sig in this message. I do not have the time over here to browse www, but I do have some access and have visited many places. Before I started this message - I had thoughts of quoting some of the messages I have read on the GenWeb during the last two weeks. I will not, this message to to long as it is. All the mechanics for adding the RSL (Roots Surname List) database to the GenServ have been finished, and that nice data will be added to the new system shortly. We are now working on an additional capability which will give a Data-Search along with a surname as a new feature for the GenServ. We do have lots of additional plans for the GenServ. We also have lots of Harddisk Space available for those of you wanting to send anysize GEDCOM file and be a part of the Genserv. If so, please read the documentation. I know at least some of the readers here on GenWeb are useers of the GenServ, and have requested data from it. I read in the GenWeb messages about some of the large 2000 names databases. IMHO, and since I have worked with so many different GEDCOM databases during the past 5 years, a database with only 2000 names is not a large one. Even in 1991, I loaded my 25,000+ names database in the GenServ project, and took approx 20 other gedcom databases to check what I hoped would be a valid test for a big system. At that time I also loaded up a database using LifeLines (UNIX) by Tom Wetmore with over 800,000 names and used it for testing to check if the LifeLines program could handle it, and it did without a problem. So I proceeded along with the GenServ putting it LIVE on Internet from my house until I was assigned to work in Korea. The GenServ system is now on a computer in system is now in Austin, Texas which belongs to my friend Ron McDowell. The GenServ system has several databases which have more than 50,000 names each. We presently have no REAL STANDARD for the GEDCOM format. What I mean here is that anyone around the world can presently say "my program will produce a GEDCOM datafile which can be read by other systems". For the past 2+ years, we have had little or no cooperation with the powers who write the standards, to settle on one, and say "ok, people let's get our act together". One of my biggest heartaches is the "so-called" GEDCOM files. I have spent 100's of hours since 1991 editing some of the ones received, but now not doing that now. I am using the GIM program (Genealogical Information Manager), PAF, and Brother's Keeper as standards If the data will load into one of these, I will use it, otherwise I notify the sender it did not load, and they do not receive access to the GenServ. If you could only see some of the "so-called" GEDCOM files I have received during the last FIVE years, you would be amazed. Some of the worst "so-called" GEDCOM datafiles have been produced by the biggest and best known genealogical programs. I will not mention any names in this message, but I have written (email) at least three of these major producers of GEDCOM datafiles, and they do not really want to hear about it. They are in the BUSINESS OF SELLING PROGRAMS and making money, and they do not want to hear they have made a mistake or their programs are not even compatiable with any of the standards. When I started the GEDCOM Mailing List (GEDCOM-L on the vm1.nodak.edu) we had just finished the GENTECH 93 conference, and discussed many of these same issues about the compatiability GEDCOM data. At the time we all hoped that some of the STANDARDS ISSUE would be resolved with the next year, and that the powers to be would be interested. Now two years later we still have many of the same problems with GEDCOM data being incompatiable with other programs. What can we do about the producers of genealogical programs who say they will import and export GEDCOM data ? Many of them do not even write a datafile which can be reaby any program, but they say "Gedcom Compatible". If we have this many problems with the GEDCOM standard, just imagine problems which will be encountered with a data standard on www. I have really seen come good ideas in the messages posted on the GenWeb, and I do appreciate the ability to read it. I am reading every message posted on it. Thanks. Good luck to all, Cliff Manis -- Cliff Manis cmanis@progcons.com Seoul, Korea GenServ "Genealogical Server" a service for making GEDCOM data available. For GenServ info, just send a message to: genserv-info@progcons.com WWW Genserv URL: http://www.cs.ncl.ac.uk/genuki/GenServ/ Wednesday, April 26, 1995 9:57:58 PM GenWeb Item From: William Kelley,wtk@panix.com,Internet Subject: Re: Copy of: re: Time for an index? To: GenWeb Scott McGee writes: >Mickey Lane write (in response to my message, >which due to my mistake probably came after his!): >>Maybe what we need to do is think in terms of a CD-ROM. Think of >>registering a database as the same as burning it into a CD - that many people >>can have a copy of. People link to the named databases. If someone changes >>one of the registered files, a new file, with a different name, is added >>to the >>CD. Once the news gets out about the new file, people can change their private >>databases whenever they want. > >But again, some people won't do that. I myself feel this desire to know >the data I make availible is the very latest I have. My current genweb >configuration is perfect for that. If I edit my database, you immediately >see the new information if you visit it by web. I am much less inclined to >"publish" the data when I know that tomorrow or the next day I may find >some new info that will make part of it wrong. I would want to imediately >recall all copies! I also know that there are other more this way than >I am. Publishing - meaning sending your database somewhere else - is definitely inferior (IMHO) to having private data *linked* to other private data. With links, private data at a particular site can be changed at will by the maintainer while the links to that data remain unchanged(excepting when the private data is physically moved to a different site/directory). How do we know when a certain database becomes out of date? If someone has been to your site, seen your data, and found nothing that links to their own data, they would have no reason to revisit your site, *unless* they knew that your data had changed. When you do make changes to your "private" local data, there should be a mechanism to make that fact "public" - the fact that your local data has changed. A message could be sent to the link or index manager site. William Kelley wtk@panix.com Wednesday, April 26, 1995 11:28:35 PM GenWeb Item From: Gary Hoffman,ghoffman@ucsd.edu,Internet Subject: GenWeb Update To: GenWeb Hello GenWebbers, This mailing list was quiet for about two weeks, then it all broke loose with one question. Now we have a good discussion going and we are moving the state of the art in a positive direction. Meanwhile, I have been busy behind the scenes on the GenWeb infrastructure. Here is an update: First, GenWeb is a topic of two presentations I'm giving at the National Genealogical Society's 1995 Conference on the States in San Diego next week. Again, I will attempt a live presentation using a WWW browser projected on a screen, explaining first the Internet, then World Wide Web, and finally how we can store and display genealogy data in WWW formats. This will be an update of the presentation I gave at GenTech 95 in Dallas in January. That presentation is available for others' use via URL http://demo.genweb.org/gentech95/title.html. Please let me know if you use this in a public presentation. Second, I am in the process of moving all my GenWeb pages from a Mac IIfx running MacHTTP to a '486 PC running BSDI Unix. If you have any links to genealogy pages on 'irpsbbs.ucsd.edu', please change your references to 'demo.genweb.org'. This is a CNAME or host alias that points to the new machine's real address. By using the CNAME, we can migrate files to increasingly more powerful computers as the need arises, but without requiring anyone to change their own reference pages (beyond this one change, of course.) Third, in response to popular demand, we are expanding the genweb.org with subnet designations that will assist in the growth of the network. The first subnet will be designated gen1.genweb.org and will be operational by May 1, 1995. Further details will be forthcoming. Finally, I have been compiling messages from this mailing list into an archive. You are welcome to browse the archive at URL http://demo.genweb.org/genweblist/genweblist.html. Eventually, I hope to have automated subscribing and unsubscribing capabilities by means of WWW forms plus automated archive updating. Just a reminder. If you want to unsubscribe from the list, please send your message to listserv@ucsd.edu and not to the list itself. Just put "unsubscribe genweb" into the body of the message, without the parentheses, of course. 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) 534-5727* *************************************************************************** Wednesday, April 26, 1995 11:44:52 PM GenWeb Item From: Annelise Anderson,ANDRSN@HOOVER.STANFORD.EDU,Internet Subject: Re: Some comments about www and GenServ To: GenWeb In response to Cliff Manis' post, I would like to say that I think GENSERV is wonderful. Second I would like to say that the founders and gurus of GENWEB (I am merely a participant/observer) are going to have the responsibility for establishing some standards at some point. I believe diversity and experimentation and competition are important, and many different ways of doing things have been discussed. But consider Cliff's frustration with GEDCOMs, as well as the frustrations of many users, not only those whose GEDCOMs are rejected but the many others whose messages we read who are having difficulty exchanging data. The GEDCOM standard was established (and has been updated) by the Church of Christ of the Latter Day Saints (I think that's the correct title,if not, my apologies), originally for more limited purposes than the purposes for which it is now being used by GENSERV, by programs that convert GEDCOMs to HTML, and so forth. So, the time is coming when the gurus of GENWEB are going to have to face their responsibilities and decide what's standard and isn't, and posting (and finally offering) certification of user programs on the question whether they meet GENWEB standards, perhaps on a whole series of criteria. "This program, when properly used, meets the standards GENWEB has established for 1) GEDCOM output suitable for Internet/WWW purposes; 2) ........etc." The interest in Internet genealogy is, I think, expanding rapidly enough, including the interest in being able to produce data that is suitable for various WWW purposes, that the power to influence the manufacturers of the programs is on its way. Yes, they want to make money; but they will find the sales of their products are hurt if they do not provide output that is suitable for the coming technology. While LDS may produce the best GEDCOM in the business, I do not think it is their desire or That's your job.responsibility to determine what is suitable for GEDWEB purposes. These points may apply to a range of issues far beyond GEDCOM. Respectfully, Annelise Anderson Wednesday, April 26, 1995 11:54:43 PM GenWeb Item From: Gary Hoffman,ghoffman@ucsd.edu,Internet Subject: GenWeb Indexing To: GenWeb Scott McGee's question, "Is it time for an index?" got us to review the huge problem facing us in indexing and linking the human pedigree. (That is the task we have set about, isn't it, when we decided to build the GenWeb?) I see a need for both an index and a linking function. Linking is going to be harder to do, so let's start with indexing. We seem to agree that we do not want a central index. But we could have a central index of distributed indexes. I explored this ideas somewhat in my original proposal (URL http://demo.genweb.org/gene/genedemo.html). Mike Patton has suggested a way to organize a site index by surnames. We do need to suggest some standards like these for site indexes. In addition, we have indexes of GenWeb sites being built by Tim Doyle and others. So what we need is a way to check each site index and build a master index of surnames available at each site. The central index lets me browse or search a surname list, then tells me which sites have records using these surnames. The master index would not have links to the actual records, but only to the surname index page at each indexed site. Following that link will let me see what records are available at that site, but the record index remains local to the site containing the records. What we need now: 1. a volunteer master index (or 'gendex') site which can be designated 'gendex.den1.genweb.org', and 2. a common format of organizing surname index pages along the lines Mike suggested. Perhaps someone can write a cgi or Perl script that updates the local indexes every so often. Updating the master index may be more difficult, but we can create ways of automating that, too. One suggestion I have made is that anyone carrying a GenWeb database should have a page placed at the root level of the http or WWW server with the name 'gendex.html'. The contents of the file should be standardized as well but should point to the local index of GenWeb records. This would allow anyone or any automated task to find GenWeb records without having to know the path in advance, just the site name. And the site name can be a genweb.org domain CNAME. The pace quickens. 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) 534-5727* *************************************************************************** Thursday, April 27, 1995 12:55:58 AM GenWeb Item From: William Kelley,wtk@panix.com,Internet Subject: Re: GenWeb Indexing To: GenWeb >What we need now: >1. a volunteer master index (or 'gendex') site which can be designated >'gendex.den1.genweb.org', and >2. a common format of organizing surname index pages along the lines Mike >suggested. >Perhaps someone can write a cgi or Perl script that updates the local >indexes every so often. Updating the master index may be more difficult, >but we can create ways of automating that, too. > >One suggestion I have made is that anyone carrying a GenWeb database should >have a page placed at the root level of the http or WWW server with the >name 'gendex.html'. The contents of the file should be standardized as well >but should point to the local index of GenWeb records. This would allow >anyone or any automated task to find GenWeb records without having to know >the path in advance, just the site name. And the site name can be a >genweb.org domain CNAME. What about having the 'gendex.html' file in a directory 'genweb' - we can't have *everybody* placing files at the root level of the server and it just makes for a cleaner setup. It would be nice if we could *also* use 'gendex.cgi' for those of us with automated surname lists. I suppose those of us with on the fly lists could name our CGI 'gendex.html', but here again it would be cleaner to do the former. It would mean anybody accessing the site separately from the master index wouldn't know the full URL, but hopefully most people will go through the master index and the correct URL will be listed there. Also, 'gendex.cgi' would allow for more specialized listings(i.e. listing the individuals with a given surname) and other expansion(for those who can use CGIs). Registering with the master index would require sending your site name, whether or not the local index is automated or not(something to allow for the above), and a list of surnames. Once you are registered, you should be allowed to update your entry in the master index - change automation info, surnames list, etc. This would allow for "distributed" updating of the master list. William Kelley wtk@panix.com Thursday, April 27, 1995 3:13:22 AM GenWeb Item From: Gene Stark,gene@starkhome.cs.sunysb.edu,Internet Subject: GenWeb Indexing To: GenWeb >One suggestion I have made is that anyone carrying a GenWeb database should >have a page placed at the root level of the http or WWW server with the >name 'gendex.html'. The contents of the file should be standardized as well >but should point to the local index of GenWeb records. This would allow >anyone or any automated task to find GenWeb records without having to know >the path in advance, just the site name. And the site name can be a >genweb.org domain CNAME. The idea of a standard, centrally locatable index file is a good one, but I think the requirements for its location should be somewhat more flexible so that it is not required to be at the root of a server. The reason is that many people may not be permitted to install file at will at the root of an HTTP server. In my case, I could set up an HTTP server at home and do whatever I want, but I would not be able to make the data available 24 hours if I did that. So, I use the HTTP server at work, but then there are certain restrictions imposed on where I can write files. I would suggest that the central registry contain, for each database, a URL indicating the base directory of the database. The "gendex.html" file could be located in that directory, as would e.g. "index.html" which would be the browser entry point to the database. - Gene Stark Thursday, April 27, 1995 3:54:11 AM GenWeb Item From: Cliff Manis,cmanis@progcons.com,Internet Subject: Some numbers for the lat two days To: GenWeb Readers: Just to give you all an idea about how many names and the sizes of GEDCOM files being received on diskette. These datafiles have been received (by U. S. Postal Mail, on diskette) during the last two days. Until today, I did not check the numbers but the GenServ system is close to having 1,000,000 names in GEDCOM data at this point. ALLEBA1.GED 4726 BRAYTO1.GED 3441 CORSBO1.GED 2369 EISKJA2.GED 49520 GOLDLO1.GED 1964 LANTCU1.GED 11985 LIGHSH1.GED 2244 MARLRU1.GED 4320 MCEAIA1.GED 2234 NEUMAL1.GED 6358 NORIBI1.GED 35260 OTTIKR1.GED 23607 WALLRO1.GED 54445 ====== 198163 names received and processed in the last two days. We have approximately 50 gedcom datafiles (not included the ones above) which have not been entered in the system yet because of upgrades to the system, and I want to add those to the new system. All GEDCOM files received are renamed to something like those seem above. i.e. if your name is Tom Brown, and this is the first gedcom you have sent to the system, I would probably name that file "browto1" if there was not already a database file by that name. If you send a second gedcom file - it might be named "browto2". I'm sorry but all the databases which contain genealogy on the BROWN family history cannot be called BROWN.GED in the system. There are lots of administrative things to be done when running a system and the GenServ has its chores which must be done also. These need to be done RIGHT the first time. I don't need mistakes in the system, from wrong databases names or data sent to invalid addresses. Good luck to all, Cliff -- Cliff Manis cmanis@progcons.com Seoul, Korea GenServ "Genealogical Server" a service for making GEDCOM data available. For GenServ info, just send a message to: genserv-info@progcons.com WWW Genserv URL: http://www.cs.ncl.ac.uk/genuki/GenServ/ - Thursday, April 27, 1995 5:31:09 AM GenWeb Item From: Brian Randell,Brian.Randell@newcastle.ac.uk,Internet Subject: Re: time for an index? To: GenWeb William: >>One of the problems I see in all this discussion is the very real >>possibility of a huge bottleneck if !the! index is placed on any one >>machine. > >Yes, and what happens if "the" index server goes down often or for extended >periods of time?! This is a known problem with at least one readily available solution at any rate in the UNIX world. I am involved in a research collaboration which has set up UNIX file servers in a number of different countries across Europe, connected together using the Andrew File System. This provides the appearance of a single standard file system, though in fact in the interests of availability and speed of access files are replicated on the various servers. We in fact have mounted Web servers on top of AFS, so that with no change to either AFS or the Web servers Web users get the performance and reliability benefits provided by AFS. Similarly, at some of the sites the AFS server is connected in via NFS, so as to be part of their local distributed file system. And I in fact access pages of information held in the AFS system using some protocol (whose name I don't even know) which presents the AFS system to me as though it is just another disk on my Mac. The lesson I draw from this is, of course, that wherever possible one should solve separable problems separately. For example, I regard such problems as (i) giving users the illusion that a distributed system is a centralized one, (ii) providing means of tolerating network and server failures, (iii) profviding sophisticated searching and reporting of GEDCOM databases, and (iv) providing nice graphical browsing interfaces, all as potentially separable problems. OK? Cheers Brian Dept. of Computing Science, University of Newcastle, Newcastle upon Tyne, NE1 7RU, UK EMAIL = Brian.Randell@newcastle.ac.uk PHONE = +44 191 222 7923 FAX = +44 191 222 8232 URL = http://www.cs.ncl.ac.uk/~brian.randell/ Thursday, April 27, 1995 5:35:10 AM GenWeb Item From: Brian Randell,Brian.Randell@newcastle.ac.uk,Internet Subject: Re: Some comments about www and GenServ To: GenWeb Annelise: IMHO, yours was one of the best postings on the topic in days! I hope it gets the attention it deserves. Cheers Brian PS I know it's wasting bandwidth, but just to make it clear to others who might have missed it what I'm enthusing about, the text of your message was: >In response to Cliff Manis' post, I would like to say that I think >GENSERV is wonderful. > >Second I would like to say that the founders and gurus of GENWEB >(I am merely a participant/observer) are going to have the responsibility >for establishing some standards at some point. > >I believe diversity and experimentation and competition are important, >and many different ways of doing things have been discussed. > >But consider Cliff's frustration with GEDCOMs, as well as the frustrations >of many users, not only those whose GEDCOMs are rejected but the many >others whose messages we read who are having difficulty exchanging data. > >The GEDCOM standard was established (and has been updated) by the Church >of Christ of the Latter Day Saints (I think that's the correct title,if >not, my apologies), originally for more limited purposes than the purposes >for which it is now being used by GENSERV, by programs that convert >GEDCOMs to HTML, and so forth. > >So, the time is coming when the gurus of GENWEB are going to have to >face their responsibilities and decide what's standard and isn't, and >posting (and finally offering) certification of user programs on the >question whether they meet GENWEB standards, perhaps on a whole >series of criteria. "This program, when properly used, meets the >standards GENWEB has established for 1) GEDCOM output suitable for >Internet/WWW purposes; 2) ........etc." > >The interest in Internet genealogy is, I think, expanding rapidly >enough, including the interest in being able to produce data that >is suitable for various WWW purposes, that the power to influence >the manufacturers of the programs is on its way. Yes, they want >to make money; but they will find the sales of their products are >hurt if they do not provide output that is suitable for the coming >technology. While LDS may produce the best GEDCOM in the business, >I do not think it is their desire or That's your job.responsibility to >determine >what is suitable for GEDWEB purposes. > >These points may apply to a range of issues far beyond GEDCOM. > >Respectfully, > >Annelise Anderson Dept. of Computing Science, University of Newcastle, Newcastle upon Tyne, NE1 7RU, UK EMAIL = Brian.Randell@newcastle.ac.uk PHONE = +44 191 222 7923 FAX = +44 191 222 8232 URL = http://www.cs.ncl.ac.uk/~brian.randell/ Thursday, April 27, 1995 6:30:04 AM GenWeb Item From: Ronald Cox,roncox@cam.org,Internet Subject: Re: Some comments about www and GenServ To: GenWeb On 27 Apr 1995 Brian Randell wrote >Annelise: > >IMHO, yours was one of the best postings on the topic in days! I hope it >gets the attention it deserves. > >Cheers > >Brian > >PS I know it's wasting bandwidth, but just to make it clear to others who >might have missed it what I'm enthusing about, the text of your message >was: > >>In response to Cliff Manis' post, I would like to say that I think >>GENSERV is wonderful. >> >>Second I would like to say that the founders and gurus of GENWEB >>(I am merely a participant/observer) are going to have the responsibility >>for establishing some standards at some point. >> >>I believe diversity and experimentation and competition are important, >>and many different ways of doing things have been discussed. >> >>But consider Cliff's frustration with GEDCOMs, as well as the frustrations >>of many users, not only those whose GEDCOMs are rejected but the many >>others whose messages we read who are having difficulty exchanging data. >> >>The GEDCOM standard was established (and has been updated) by the Church >>of Christ of the Latter Day Saints (I think that's the correct title,if >>not, my apologies), originally for more limited purposes than the purposes >>for which it is now being used by GENSERV, by programs that convert >>GEDCOMs to HTML, and so forth. >> >>So, the time is coming when the gurus of GENWEB are going to have to >>face their responsibilities and decide what's standard and isn't, and >>posting (and finally offering) certification of user programs on the >>question whether they meet GENWEB standards, perhaps on a whole >>series of criteria. "This program, when properly used, meets the >>standards GENWEB has established for 1) GEDCOM output suitable for >>Internet/WWW purposes; 2) ........etc." >> >>The interest in Internet genealogy is, I think, expanding rapidly >>enough, including the interest in being able to produce data that >>is suitable for various WWW purposes, that the power to influence >>the manufacturers of the programs is on its way. Yes, they want >>to make money; but they will find the sales of their products are >>hurt if they do not provide output that is suitable for the coming >>technology. While LDS may produce the best GEDCOM in the business, >>I do not think it is their desire or That's your job.responsibility to >>determine >>what is suitable for GEDWEB purposes. >> >>These points may apply to a range of issues far beyond GEDCOM. >> >>Respectfully, >> >>Annelise Anderson I second the thought. Following the string leaves the impression that a house is being built before the foundation is in place, or the cart is ahead of the horse. And don't forget all of the non technical individuals who may have a healthy database, but not the moxie to create a html page, or even the desire. And are maybe using an old 286 based machine, or are still using a quill pen. They might be just the person that has the name you need. Don't forget those of us who are neophytes to this wonderful world of modern technology, and what we have to offer. Ron Ronald Cox, roncox@cam.org, 74 Brunswick Dr., Beaconsfield, QC, H9W 5H2, CAN "He, who takes no interest in the history of his ancestors, does not deserve to be remembered by his posterity."