Wednesday, May 10, 1995 12:37:46 PM GenWeb Item From: Gary Hoffman,ghoffman@UCSD.EDU,Internet Subject: Re: Ownership and updating To: GenWeb Yes, I too like Scott's idea of using an HTML form with password as one of the fields. Bill is advocating a feature that enables anyone to add comments to an individual record. While I concur in the idea, I am very wary of a system that could become the equivalent of electronic grafitti. If any Internet Joe can add text to a record that I own, then they will. Someone will write a macro that will spam the entire GenWeb database quicker than you can say "Kevin Mitnik" and then we'd all have spam on our faces. I will continue to hold out for a system that lets only the "owner" update a genealogical record. If someone doesn't like what is there, they can put up an alternate version on their own server. 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, May 10, 1995 2:10:30 PM GenWeb Item From: Anders Andersson,andersa@Mizar.DoCS.UU.SE,Internet Subject: Re: Indexing and then linking To: GenWeb Gene Stark writes: >In the interests of flexibility, I think it would be good to take advantage >of the capabilities of DNS to avoid having to assign a standard name >such as "gendex.html" to the index files. We can also avoid having to >specify a standard location for it, which may be inconvenient or incompatible >with administrative policies on some servers. Rather, standard TXT records >in the subdomain "databasename.version0.genweb.org" would specify the name >and URL of the index file for the database "databasename". This would also >permit database maintainers to retain their Version 0 index files when >"Version 1" GenWeb decides on a different index file format. Just for the record, I'd like to second this point. We have to clear up some issues regarding the general availability of DNS TXT records, but apart from that I don't see any reason to start implementing a less flexible system with host aliases involving the cooperation of host system managers who may or may not be concerned with genealogy (such a system may end up as a burden that will be difficult to get rid of once we have employed it). I agree with Gary and Gene that DNS isn't really a good vehicle to index massive amounts of genealogical data. DNS comes in handy for locating significant resources such as entire databases and their servers on the Internet without relying on a single "index server" that must be available at all times, but once that initial step towards accessing a piece of information is taken, other tools are much better suited at looking up individual records. We could use a good distributed indexing system, but this is beyond our current abilities to design. DNS is distributed, but not good for such a demanding task. We have access to good indexing tools in HTTP and WAIS software, but they aren't distributed in design. Let's use an adequate combination of these existing tools while working towards the ultimate solution. *** To help separating the threads in this discussion, could we assign simple names to the individual goals of the GenWeb project, rather than having to identify each subgoal by means of frantic handwaving and spurious references to technical peculiarities? I want to see something like "GenWeb Resource Identifiers", "GenWeb Goal #17" or whatever, as long as the name is unique and comes accompanied by a definitive description (preferrably in the form of an HTML page) that can be updated by means of mutual agreement. Maybe we can use for an index of these subgoals, in order to give structure to the project and encourage interested parties to submit their contributions? -- 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, May 10, 1995 3:44:50 PM GenWeb Item From: steele@census.works.ti.com,Internet Subject: Document References To: GenWeb Gary said: > Well, that same source could be cited by GenWeb pages on several servers. > Of course, there would be no need to trplicate the entire citation, as long > as we had a link to it anywhere on the network. Gary went on to suggest 1) using computerized facilities of the LOC and/or 2) that we users begin creating a distributed database of genealogical source materials that will shadow the GenWeb. A distributed database of genealogical source materials sounds great until you examine what that means. One example that comes to mind is material I find in a published genealogy book that in itself then cites a primary source such as a county court record. I now want to cite the book that I used, but somewhere researchers would like to know the primary source. How does that relationship between sources get expressed? Material we read in sources still can be interpreted in different ways. There will be disagreement on the interpretations. For example, I have the statements that Hiram Kirkpatrick was b. 1809 in TN, his wife was Ann A. (1850 & 1860 census ) and their first child was b. 1837 in AL. (verified by both census) For the purposes of research, you might record that Hiram and Ann were married circa 1836. Some people would not record that relationship, they would wait to discover some primary source that can be cited. In over 10 years of research there seems to be NO primary record, but by preponderance of evidence I have concluded that there was a marriage probably in 1836. How do I cite that the conclusion was drawn from examining other sources. (Right now I do it through surety levels on the Dates, people, and locations and recording all sources) Different methodologies for interpreting the data are used by each researcher. The Silicon Valley PAF Users Group guidelines are an excellent start toward attempting to standardize data entry, but to generate a distributed database of sources to share, I feel we need to do some work on how and why sources are to be cited in our Web genealogies. Another question is: Will having distributed sources complicate the configuration of GENWEB to the point of making the Web unusable? (Net & System Loading, complications of installation and maintenance) Jeri Wednesday, May 10, 1995 6:58:47 PM GenWeb Item From: Bill Minnick,svpafug@rahul.net,Internet Subject: Re: Ownership and updating To: GenWeb GARY HOFFMAN WRITES: >Yes, I too like Scott's idea of using an HTML form with password as one of >the fields. Bill is advocating a feature that enables anyone to add >comments to an individual record. While I concur in the idea, I am very >wary of a system that could become the equivalent of electronic grafitti. Gary, I can't agree with your conclusion -- that we can't allow people to enter updates into a data base! Please re-assess my points of argument. First the system I advocate would place the public inputs on a separate page from the GENWEB official page for an individual; no edits would be made on the "official" page; only a link added into the official page to permit immediate public access to the newly submitted material. There would be a link created on the official page for each publicly submitted page (up to some max number of added public pages) . The "owner" could limit to perhaps a max of two, three, even 6 or 10 added pages before "closing the hatch" to public input on a given individual. If the site gets more than, say, X entries into the same data base in less than a minute, or more than N submissions in a data base in a week, then that submitter can considered an attacker, and be blocked. There are probably better ways yet to automatically prevent a "grafitti binge". I won't roll over and play dead on this issue, and lose a key benefit the Web can provide to Genealogy. >If any Internet Joe can add text to a record that I own, then they will. >Someone will write a macro that will spam the entire GenWeb database >quicker than you can say "Kevin Mitnik" and then we'd all have spam on our >faces. I will continue to hold out for a system that lets only the "owner" >update a genealogical record. If someone doesn't like what is there, they >can put up an alternate version on their own server. And my suggested approach assumes that the "Owner" makes the final changes to the "top" or "primary" page for that individual - - after a healthy public debate and hopefully, submission of new and better information. If we create an attractive research environment, people will be motivated to contribute. "IF WE BUILD IT, THEY WILL COME" -- you may recall the quote from the movie, "Field of Dreams". Put the word "right" just after the word "it" in the quote, and we have the GENWEB moto. I suggest we ask our world-wide GENWEB technical experts how to solve problems you have posed so we can provide the mechanism to host the research debate. As an added note, we have had several automatic e-mail forms in place for months, have received more than a hundred excellent inputs, and have not received one intentionally bogus submission. Yes, I think we need to safeguard against a coordinated attack, but giving up on a powerful system for immediate public research exchange is not the solution. I will not accept a solution where my schedule gates immediate accessability by the public to every public input entered in good faith relating to the 50,000 people in the data base I am managing. I hope this clarifies for you where I am coming from on this very important aspect of GENWEB. Oh, thanks for the report on the NGS meeting. Your GENWEB snowball is clearly rolling and growing. Regards, Bill Minnick Wednesday, May 10, 1995 8:17:46 PM GenWeb Item From: Annelise Anderson,ANDRSN@HOOVER.STANFORD.EDU,Internet Subject: Helm's Genealogical Tool Box To: GenWeb The location is: http://ux1.cso.uiuc.edu/~al-helm/genealogy.html It works most of the time. Annelise Wednesday, May 10, 1995 10:15:45 PM GenWeb Item From: Annelise Anderson,ANDRSN@HOOVER.STANFORD.EDU,Internet Subject: Citing Sources To: GenWeb It seems to me that a record ought to be complete with its sources, that is, a record for an individual or a family group. This would involve repeating a reference to, say, particular church records, as such records are likely to relate to several people in a database. If I find a person in another data base and want the information, I'm going to copy it or save it or print it or whatever my browser lets me do--it would seem rather awkward then to follow a link to some reference and have to either copy (by hand) the reference or copy the entire document. An alternative is for a database to have a reference list that can be copied or saved if the people in the database are of interest. Then the citations in the individual records could be of a short form, e.g., an author or title and a page or line number. I hope you won't scare people off with standards for citations and such that seem to high to meet--actually the LDS succeeded in intimidating me, and I have never sent my data to the ancestral file because I did not think it would be satisfactory to them, and I wasn't quite sure how to fix it up. GENSERV was not so intimidating, nor did I feel reluctant to put my own data on a WWW page. Nevertheless some of the citations are so peculiar that I cannot understand them myself and often there is no clue as to the location of a complete reference when only a title and page number are provided. Some time I will fix this. Some time. Meanwhile, it's there as is or not at all.... Have you looked at Helm's Genealogical Toolbox surname index? A work in progress, but working quite well, I think. I was surprised to discover that my own database was indexed there...one gets a feeling that this whole thing is in motion on its own, sort of out of control already.... I enjoyed Gary Hoffman's reports from the NGS meeting-- Annelise Wednesday, May 10, 1995 10:15:45 PM GenWeb Item From: Annelise Anderson,ANDRSN@HOOVER.STANFORD.EDU,Internet Subject: Citing Sources To: GenWeb It seems to me that a record ought to be complete with its sources, that is, a record for an individual or a family group. This would involve repeating a reference to, say, particular church records, as such records are likely to relate to several people in a database. If I find a person in another data base and want the information, I'm going to copy it or save it or print it or whatever my browser lets me do--it would seem rather awkward then to follow a link to some reference and have to either copy (by hand) the reference or copy the entire document. An alternative is for a database to have a reference list that can be copied or saved if the people in the database are of interest. Then the citations in the individual records could be of a short form, e.g., an author or title and a page or line number. I hope you won't scare people off with standards for citations and such that seem to high to meet--actually the LDS succeeded in intimidating me, and I have never sent my data to the ancestral file because I did not think it would be satisfactory to them, and I wasn't quite sure how to fix it up. GENSERV was not so intimidating, nor did I feel reluctant to put my own data on a WWW page. Nevertheless some of the citations are so peculiar that I cannot understand them myself and often there is no clue as to the location of a complete reference when only a title and page number are provided. Some time I will fix this. Some time. Meanwhile, it's there as is or not at all.... Have you looked at Helm's Genealogical Toolbox surname index? A work in progress, but working quite well, I think. I was surprised to discover that my own database was indexed there...one gets a feeling that this whole thing is in motion on its own, sort of out of control already.... I enjoyed Gary Hoffman's reports from the NGS meeting-- Annelise Wednesday, May 10, 1995 11:08:30 PM GenWeb Item From: Anders Andersson,andersa@Mizar.DoCS.UU.SE,Internet Subject: Avoiding graffiti To: GenWeb Gary Hoffman writes: >If any Internet Joe can add text to a record that I own, then they will. I agree in principle. You don't want irrelevant graffiti associated with your records, and you don't want irrelevant graffiti to fill up your physical disk. The second goal is relatively easy to achieve. Simply don't allow unauthorized data to be stored on your disk. You could still allow links (in the form of URLs or other pointers) to the disks of other people, meaning that any graffiti artist would have to bring their own concrete wall. However, bogus links would themselves constitute unwanted garbage, and you may want to avoid storing them as well. The first goal, that of avoiding association with garbage, is more complicated. It's obvious that you don't want the graffiti to turn up on the same page as your family records. What about links to other pages, successively `distanced' from your authoritative data? Ultimately, you are going to link to the pages of other GenWeb contributors, and I suppose it will be up to them to decide what unauthorized material to allow links to (if any) from their own pages. Now your records may be indirectly associated with garbage, although the GenWeb reader would have to manually follow a number of links to get to it from your pages. Do you want to avoid this situation too? If we want the ability to automatically link to unauthorized data for the purpose of speeding up the research process, then I think part of the solution may lie in reliable identification of garbage contributors. If an intentionally bogus link (or one million of them) is discovered in some otherwise useful index that accepts being updated by anyone, it should at the very least be possible to identify the originating host, preferrably also the individual user if there are many of them. Index management tools should make it easy to remove bogus entries originating from a particular host or user once they are identified. Another key approach may be to make the GenWeb environment boring and unattractive to vandals. The possibility exists today to submit unauthorized articles to several thousand Usenet newsgroups and have them automatically distributed across the globe. Why doesn't this happen on a regular basis? Because it's trivial! Doing it involves no intellectual challenge at all, and only the most naive computer jerks would consider such spamming `fun' (the most famous of the few major spammings we have seen actually originated with a US law firm). Thus, don't waste time trying to develop bogus `security' features to prevent improper material from being entered, since they merely beg to be circumvented. Burglars are more often found in locked rooms than in open meadows, to the advantage of the meadow owners. Those are my comments on solving some of the problems mentioned. Whether we want the ability to have unauthorized data entered at all remains an open question, irrespective of potential vandalism. -- 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 Thursday, May 11, 1995 3:21:05 AM GenWeb Item From: Mickey Lane,MLANE@csi.compuserve.com,Internet Subject: Re: ROOTSBOOK web server To: GenWeb Matthew Helm writes: >One thing that you might want to think >about adding is a source field for the data. Given the various forms that >the GEDCOM data comes to you, I don't know how feasible that is. However, >it is useful in putting the information that the user is seeing into some >sort of context. In theory, each entry page is supposed to have a "" indicator immediatly following the "Entry N" label at the top left of each page. All but one or two books are missing this indicator due to a programming error that has since been fixed. It's my intention to have the "" indicator be a hypertext link to a page produced by the software that displays as much information as possible about the book. In my own work, this page was hand written and inserted at the beginning of each book. I never implemented the software to do anything like this although I've had lots of ideas over the years. Given the GEDCOM files I've used, there's not much available. Some of them contain the name and address of the author which is probably the most valuable part. The computer can calculate all sorts of basically useless information like number of entries, min, max and mean date, primary locations and so on but since people don't think in terms of books, there's not much point to it. Certainly something to think about.... Mickey. Thursday, May 11, 1995 3:33:38 AM GenWeb Item From: Cliff Manis,cmanis@progcons.com,Internet Subject: Some problems with security To: GenWeb Anders and GenWeb: Ref this part of your message on Wed May 10 21:02:16 1995 > > From: Anders Andersson > > To: genweb@UCSD.EDU, ghoffman@UCSD.EDU > > Subject: Avoiding graffiti > Another key approach may be to make the GenWeb environment boring > and unattractive to vandals. The possibility exists today to submit > unauthorized articles to several thousand Usenet newsgroups and have > them automatically distributed across the globe. Why doesn't this > happen on a regular basis? Because it's trivial! Doing it involves > no intellectual challenge at all, and only the most naive computer > jerks would consider such spamming `fun'...... > ....Burglars are more often found in locked rooms than in open > meadows, to the advantage of the meadow owners. Burglars and Jerks will always be around either trying to give us a headache or steal something. My comments here are mostly about the JERKS. When I first started the GenServ in 1991, they seemed to come out of the walls. Some of them would attempt most anything at all to break the "access_code" being used by the GenServ system. They would try most anything at first but after a while, and several attempts by several jerks to obtain access without contributing a GEDCOM file. No names here, but I do have a private "jerks.txt" textfile with a few of the names and email addresses which are valid addresses. These attempts to gain access to whatever you setup will happen and hopefully the GenWeb system of access 'at that time' will handle it with ease. The GenServ system here is still using the exact same system as in 1991. That is, we give an access_code. We assign the persons name and the access_code. The programs are looking for these strings FIRST and if not found - any further request is dumped to the bit bucket. If these are found, processing of a request with be initiated, but the request may only be sent to an email address which is known by the Genserv. So even if you obtain my access_code and send a valid request to the Genserv system, it only sends the report to me, the valid and verified user. We capture the complete message (with headers) of all requests, and save them for about one month in case of problems. As you can imagine, we have been through many of these same trials and tribulations which the GenWeb will face in the near-future. I have been reading with much interest all the articles posted about the proposed standards. Best of luck with your proposals and during all this - please do not forget the people who only have a few names. Many of them have not yet produced a GEDCOM file, and at least 80% of them have only read the word 'GEDCOM' but have yet to see a GEDCOM file. Many of these same people actually think we have a standard data format which can be read by any other program used by the genealogical community. If I have more keep time and a faster access the www would be much more exciting. Best of 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/ - Thursday, May 11, 1995 9:25:05 AM GenWeb Item From: Rik Vigeland,rikv@wv.MENTORG.COM,Internet Subject: Updates vs. Grafitti To: GenWeb I believe that Anders Andersson and Bill Minick have hit the nails right on the head. I think a simple summary would help: 1) Allow only the creation of links on a page of data, not new pages. 2) The links will point to an alternate record of data; they might appear simply as "See Also in ". People would get into the habit of checking for "See Also"s. 3) Limit the number of links created. Multiple rules could apply: Number of links per record (page) Number of links created in a certain time period or rate of time Number of links in overall database Perhaps the database server can distinguish between permanent links that the owner has set up and those that are created by someone in the public at large. 4) The owner of the database decides what to do with the links as his/her schedule permits and how critical the change is. In the end, the owner may choose to retain the links as pointers to data better than his/her own, and avoid duplication of records. But that is another topic. Regards, Rik Vigeland - rikv@wv.mentorg.com Anders Andersson replied to Gary Hoffman: >I agree in principle. You don't want irrelevant graffiti associated >with your records, and you don't want irrelevant graffiti to fill up >your physical disk. >The second goal is relatively easy to achieve. Simply don't allow >unauthorized data to be stored on your disk. You could still allow >links (in the form of URLs or other pointers) to the disks of other >people, meaning that any graffiti artist would have to bring their >own concrete wall. However, bogus links would themselves constitute >unwanted garbage, and you may want to avoid storing them as well. >The first goal, that of avoiding association with garbage, is more >complicated. It's obvious that you don't want the graffiti to turn >up on the same page as your family records. What about links to And Bill Minnick replied to Gary Hoffman: >First the system I advocate would place the public inputs on a separate page >from the GENWEB official page for an individual; no edits would be made on >the "official" page; only a link added into the official page to permit >immediate public access to the newly submitted material. There would be a >link created on the official page for each publicly submitted page (up to >some max number of added public pages) . The "owner" could limit to perhaps >a max of two, three, even 6 or 10 added pages before "closing the hatch" to >public input on a given individual. If the site gets more than, say, X >entries into the same data base in less than a minute, or more than N >submissions in a data base in a week, then that submitter can considered an >attacker, and be blocked. There are probably better ways yet to >automatically prevent a "grafitti binge". I won't roll over and play dead >on this issue, and lose a key benefit the Web can provide to Genealogy. Thursday, May 11, 1995 9:41:33 AM GenWeb Item From: Rik Vigeland,rikv@wv.MENTORG.COM,Internet Subject: Duplication vs. Departure To: GenWeb OK, folks, here's one that's been on my mind. Example case: I have my ancestry on-line. Another distant cousin does the same. We have 120 ancestors in common, and no arguments about correctness. The data is identical. Next, some poor user runs a query and comes up with (at least) two sets of identical matches. The user may then have to go through quite a bit of work to determine that my cousin and I are in total agreement and that the data is, in fact duplicate. So -- I could take my data out and create a pointer or two into my cousin's data. But I worry - what if my cousin won't always be able to on-line? And I may have a few dozen other cousins to monitor. If I remove my copy, and my cousin goes off-line without my knowing it, then the user who originally ran the query would have come up blank. If I were that user, I suppose I would rather sort through duplicate data than no data at all. But if there are two copies of the data, there could certainly be five or ten. And, for this argument, I would as soon NOT consider the case of conflicting information for now. That has been handled somewhat already. An extreme case my be the royalty data. But still, in the general case, I may get some satisfaction of knowing that my own data is handy should I want to browse through it with a Web server, and having it locally is certainly much faster than finding it remotely. Call it selfishness. To summarize, the question would be, how do we avoid too much duplicate data while still guaranteeing at least one copy exists? Thoughts, please! Rik Vigeland rikv@wv.mentorg.com Thursday, May 11, 1995 11:45:13 AM GenWeb Item From: Mickey Lane,MLANE@csi.compuserve.com,Internet Subject: ROOTSBOOK web server To: GenWeb Update: Now serving 176,000 entries to the web. The partial URL problem has been fixed. The code has been modified so that living people won't be displayed. I am slowly re-doing the files so you may find quite a few still present. At least you won't find any new ones. http://mlane2.inhouse.compuserve.com:8000/GenWeb.htm Mickey. Thursday, May 11, 1995 2:49:29 PM GenWeb Item From: Bill Minnick,svpafug@rahul.net,Internet Subject: Re: Updates vs. Grafitti To: GenWeb Rik Vigeland wrote on May 11, 1995: >I believe that Anders Andersson and Bill Minick have hit the nails >right on the head. I think a simple summary would help: > >1) Allow only the creation of links on a page of data, not new pages. > >2) The links will point to an alternate record of data; they might > appear simply as "See Also in ". > People would get into the habit of checking for "See Also"s. > >3) Limit the number of links created. Multiple rules could apply: > > Number of links per record (page) > Number of links created in a certain time period or rate of time > Number of links in overall database > > Perhaps the database server can distinguish between permanent > links that the owner has set up and those that are created by > someone in the public at large. > >4) The owner of the database decides what to do with the links as > his/her schedule permits and how critical the change is. > >In the end, the owner may choose to retain the links as pointers to >data better than his/her own, and avoid duplication of records. But >that is another topic. Rik: YES ! Regards, Bill Minnick Thursday, May 11, 1995 5:00:23 PM GenWeb Item From: Bill Minnick,svpafug@rahul.net,Internet Subject: Re: Indexing and then linking To: GenWeb (GENWEB, 10 May 1995) Anders Andersson writes: >To help separating the threads in this discussion, could we assign >simple names to the individual goals of the GenWeb project, rather >than having to identify each subgoal by means of frantic handwaving >and spurious references to technical peculiarities? I want to see >something like "GenWeb Resource Identifiers", "GenWeb Goal #17" or >whatever, as long as the name is unique and comes accompanied by a >definitive description (preferrably in the form of an HTML page) >that can be updated by means of mutual agreement. > >Maybe we can use for an >index of these subgoals, in order to give structure to the project >and encourage interested parties to submit their contributions? TO: Gary Hoffman: Anders makes a good point here. Is this something you can take on and set up? Regards, Bill Minnick Thursday, May 11, 1995 11:47:50 PM GenWeb Item From: Herbert Stoyan,Herbert.Stoyan@informatik.uni-erlangen.de,Internet Subject: Rootsbook does not work To: GenWeb Is there anybody who was able to use the Rootsbook url? Friday, May 12, 1995 6:08:25 AM GenWeb Item From: Anders Andersson,andersa@Mizar.DoCS.UU.SE,Internet Subject: Re: Duplication vs. Departure To: GenWeb Rik Vigeland writes: >Next, some poor user runs a query and comes up with (at least) two >sets of identical matches. The user may then have to go through >quite a bit of work to determine that my cousin and I are in total >agreement and that the data is, in fact duplicate. This is a general problem, which we already can see in the archives of software and other stuff available via FTP. Some archives are considered more or less `official' in that they provide a central clearinghouse for everything within a particular field, and other archives are set up as official mirrors in order to distribute the load on the server hosts. Then you have a lot of sites which set up FTP archives holding selected copies of individual packages, meaning that they will probably turn up in Archie searches, but it may be difficult to determine whether these copies actually represent the latest version of the software you are looking for. Uniform Resource Names (which are still being designed) or any similar record identification scheme will come in handy here. The important thing is to consider identity and location as two separate issues. Even if you and your distant relative agree 100% on the identities of your ancestors and all data you have on them, you may both want to store a private copy each of your common records. A single URN would then refer to an identical record stored in two (or more) locations, and a user retrieving such a record should be satisfied with any single copy, so we need not worry about any particular server being up, as long as at least one is available. In the case when you and your relative disagree on the data for a particular individual, then you would create one record each for this person, but by definition, these records won't be identical, and thus they will have different URN's. Now, how do we link all these records together? This is a problem which is specific to the GenWeb context, where the links themselves will remain disputable items of information for as long as the research continues. Even if two researchers agree on the data for a particular individual, and thus can refer to the same record, they may disagree on the data for this person's parents. Thus, putting the parent link *inside* the record would be a bad idea, because then any dispute anywhere in a tree would split the entire tree into two mostly identical sets of records with different URN's. Instead of ... --> Record #1 <-- parent/child link --> Record #2 <-- ... where a dispute over the contents of Record #1 would force a split also of Record #2, I'd suggest something like Record #1A | ... --> Link box #1A <-- parent/child link --> Link box #2A <-- ... | Record #2 | ... --> Link box #1B <-- parent/child link --> Link box #2B <-- ... | Record #1B where the researchers A and B would each maintain their own tree of `link boxes' which themselves aren't part of the records they refer to. Please note that the above is only an abstract schematic diagram intended to show the conceptual links. An actual implementation may go to great lengths optimizing the structure in various ways. However, in a multi-researcher environment we need to maintain the kind of information represented here, while it's the duty of the user interface to hide the complexity of the structure and only point out alternatives when there is disputed information. -- Anders Andersson, Dept. of Computer Systems, Uppsala University Paper Mail: Box 325, S-751 05 UPPSALA, Sweden Phone: +46 18 183170 EMail: andersa@DoCS.UU.SE Friday, May 12, 1995 10:00:47 AM GenWeb Item From: Anders Andersson,andersa@Mizar.DoCS.UU.SE,Internet Subject: Project organization To: GenWeb Bill Minnick writes in response to me: >>Maybe we can use for an >>index of these subgoals, in order to give structure to the project >>and encourage interested parties to submit their contributions? > >TO: Gary Hoffman: > >Anders makes a good point here. Is this something you can take on and set up? I believe we may have to provide some more specific input first, like what subgoals we can identify at the moment. These should be mutually exclusive, meaning that a specific topic or question belongs in one place or another, but not in two or more at the same time. I have a few candidates (in no particular order): - Software (libraries of common code, genealogical data standards) - Resource Identifiers (URN, URL, domain names, cross-database links) - User Interfaces (HTML, visual layout, data presentation, feedback) - Indexing (local and global indices, search facilities) - Data Maintenance (server availability, archives, backup services) - Legal and Ethical Affairs (copyright, access, responsibility, privacy) - Operations (GENWEB.ORG domain maintenance, mailing list organization) I believe that some people will concentrate on some of these but not on others (as for myself, my primary concerns now are with Resource Identifiers and Operations). In a sense, each problem area can be seen as a workgroup, although GenWeb participants can of course move freely between the groups without having to register or something. Anyway, I find the workgroup concept convenient for the discussion. By the nature of the project, all these fields will interact heavily with each other. However, the structure should help us to put each question in a proper context. The Software group will rely on the Resource Identifier group to establish precise syntax conventions for external references, and so will probably the Indexing group. The Legal and Ethical Affairs group may have to dictate to the Data Maintenance group what can be stored where. The Operations group should serve the current operational needs of established services. Each of the above (if adopted) should be accompanied by a verbose description of the problem area (defining what the goals of the GenWeb project are within that field) and a list of proposed solutions or other comments as needed. I hope that someone can take on the task of providing a WWW page for each area and acting as editor (and perhaps co-author) for it. Discussion regarding the contents of these pages can preferrably take place on the GenWeb mailing list (corrections of minor typos and such should of course be sent directly to the editor alone). This should form the foundation for the continously updated documentation of the GenWeb project, stating what arguments have been presented and which decisions have been made, and perhaps thus chronicling the history of the project. Remember, this was just an example. Use as you see fit. -- Anders Andersson, Dept. of Computer Systems, Uppsala University Paper Mail: Box 325, S-751 05 UPPSALA, Sweden Phone: +46 18 183170 EMail: andersa@DoCS.UU.SE Friday, May 12, 1995 12:34:45 PM GenWeb Item From: Mickey Lane,MLANE@CSI.compuserve.com,Internet Subject: ROOTSBOOK info To: GenWeb Several people have found interesting links in the ROOTSBOOK web server. Their first comment is "Who did this work?" and they specify some person's name or perhaps a family name. I'm working on providing an answer but it's not going to be easy. Most of the GEDCOM files I've looked at don't contain anything along these lines and what's there is formatted all sorts of different ways. (Ain't GEDCOM wonderfull? Snort!) I'd like to request that people use at least the book name or book name and entry number when asking questions. Thanks, Mickey. Friday, May 12, 1995 5:52:33 PM GenWeb Item From: Mickey Lane,MLANE@csi.compuserve.com,Internet Subject: ROOTSBOOK info To: GenWeb I said: >Several people have found interesting links in the ROOTSBOOK web server. >Their first comment is "Who did this work?" and they specify some person's >name or perhaps a family name. Gene Stark says: >Couldn't ROOTSBOOK provide users with contact addresses for the persons >who submitted the data? Seems like a glaring omission. Uh...yeah, it is a "glaring omission." If I'm not mistook, that's what I said I was trying to fix. Gimmie a break, will ya? There's about 500 dif- ferent GEDCOM files I'm working with and each one of 'em has some little quirk that either the author of the database or the writer of the software package thought would be "a neat way of doing things." I'd like to apply the same level of attention to detail that people have been talking about over the years and cite references for each piece of information - but it's just not there. I'll be lucky if I can attach a name to who owns all these different databases. Mickey. Friday, May 12, 1995 6:53:17 PM GenWeb Item From: Gene Stark,gene@starkhome.cs.sunysb.edu,Internet Subject: ROOTSBOOK info To: GenWeb >Several people have found interesting links in the ROOTSBOOK web server. >Their first comment is "Who did this work?" and they specify some person's >name or perhaps a family name. Couldn't ROOTSBOOK provide users with contact addresses for the persons who submitted the data? Seems like a glaring omission. - Gene Stark Thursday, May 11, 1995 9:41:33 AM GenWeb Item From: Rik Vigeland,rikv@wv.MENTORG.COM,Internet Subject: Duplication vs. Departure To: GenWeb OK, folks, here's one that's been on my mind. Example case: I have my ancestry on-line. Another distant cousin does the same. We have 120 ancestors in common, and no arguments about correctness. The data is identical. Next, some poor user runs a query and comes up with (at least) two sets of identical matches. The user may then have to go through quite a bit of work to determine that my cousin and I are in total agreement and that the data is, in fact duplicate. So -- I could take my data out and create a pointer or two into my cousin's data. But I worry - what if my cousin won't always be able to on-line? And I may have a few dozen other cousins to monitor. If I remove my copy, and my cousin goes off-line without my knowing it, then the user who originally ran the query would have come up blank. If I were that user, I suppose I would rather sort through duplicate data than no data at all. But if there are two copies of the data, there could certainly be five or ten. And, for this argument, I would as soon NOT consider the case of conflicting information for now. That has been handled somewhat already. An extreme case my be the royalty data. But still, in the general case, I may get some satisfaction of knowing that my own data is handy should I want to browse through it with a Web server, and having it locally is certainly much faster than finding it remotely. Call it selfishness. To summarize, the question would be, how do we avoid too much duplicate data while still guaranteeing at least one copy exists? Thoughts, please! Rik Vigeland rikv@wv.mentorg.com Friday, May 12, 1995 7:05:32 PM GenWeb Item From: Bill Minnick,svpafug@rahul.net,Internet Subject: Re: Project organization To: GenWeb Anders Andersson writes in response to me: >I believe we may have to provide some more specific input first, >like what subgoals we can identify at the moment. These should >be mutually exclusive, meaning that a specific topic or question >belongs in one place or another, but not in two or more at the >same time. I have a few candidates (in no particular order): > > - Software (libraries of common code, genealogical data standards) > - Resource Identifiers (URN, URL, domain names, cross-database links) > - User Interfaces (HTML, visual layout, data presentation, feedback) > - Indexing (local and global indices, search facilities) > - Data Maintenance (server availability, archives, backup services) > - Legal and Ethical Affairs (copyright, access, responsibility, privacy) > - Operations (GENWEB.ORG domain maintenance, mailing list organization) To all GENWEBBERS: I like Anders' kernal of thought here, and have felt compelled to expand and rearrange his seven suggested points above. I now pass my thoughts back to you for further cut & paste by anyone who sees a need missed, a wrong turn, a typo, etc. I've taken the time to sift through many of my past GENWEB discussions, presentations, etc., to consolidate some of the thoughts and concerns about the GENWEB project that I/we've discussed (publicly and privately)in our local groups in Silicon Valley, and via e-mail. Please feel free to carve out pieces of this, tear it up and fix it up. I'll volunteer to edit this document until it stabilizes. Then I'll hand it off to Gary Hofffman to become a WEB page in xxx.genweb.org. _____________________________________________________________________________ GENWEB PROJECT REQUIREMENTS (an expansion and rearranging of Anders Andersson's Seven Points of Project Organization - by Bill Minnick 5/12/95 I USER INTERFACES A. Data Submission 1. Data Base Submission (list of sites - rules of submission) a. List of Sites accepting data - rules of submission b. Submission formats 1) GEDCOM files - Format Standards? 2) Other ? 2. Individual-Level Research discussion and debate a. Linking Dialog (e-mail) to individuals b. Storing/Archiving Dialog (list of sites - rules of submission) 3. New Information Feedback - corrections/updated information a. Linking new data to/from individual/family page b. Storing feedback data (list of sites - rules of submission) 4. Genealogical Data Standards a. Data Entry 1) Surname, Given Name, Titles, AKA, Name Case conventions, handling missing information 2) place names - order, standard abbreviations 3) place naming conventions 4) date estimation b. Source Citations 1) Separate source data base? or complete source citation with each individual? 2) Information required to define a source 3) Source Quality rating system and evaluation 5. Query Tools a. Indexing 1) Local indices (display format, content) (consider indexing by name, place, source, names with photos/videos, names with sound bytes, names with research conflicts, Other? Should every word be indexed to support efficient searches?) 2) Global indices (location, management, rules for submission) b. Search Functions: permit specification of Individual pages in a data base, links to publicly-submitted added info pages, sources, etc. 1) Literal searches 2) Genealogy-Intelligent search development: support specification of complex searches capable of he following: a) two words separated by a max of N words b) AKA Place searches (develop lists of place AKA Names) c) Date range search capability d) "Fuzzy" search for Names, Places (Superior to Soundex) 3) Useful search macros -intelligent search tools that are easy to use by the average person a) for example, devise an automated search sequence which first searches for a literal name/date/place combo; if no find, then widens aperture with fuzzy search, date range and AKA location passes. c. Possible Spider Functions (Automation of special searches) 1) automated identification and linkage of duplicate individuals 2) build listing of possible duplicate individual candidates 3) build listing of possible duplicate females: (for female spouses with unknown surname, correlate to female children based on given name, birth date vs. marriage date, birth & marriage location.) 4) Other? d. Access log 1) log all data base accesses 2) make log information available to public B. Data Presentation 1. Individual/Family page presentation a. preferred page format b. photo dispaly conventions c. display of links to feedback d. display of forms soliciting feedback 2. Form design for soliciting feedback 3. other reports (formats - pedigree, photo/audio album, etc.) 4. GEDCOM file Output (max - one family at a time) (perhaps other computer-readable outputs for private use upon request are needed?) II SOFTWARE A. HTML/HTTPD - Version control B. Public domain Web publishing software for genealogy (LifeLines, scripts, GED2HTML, etc.) C. Libraries of Commmon Code - executables, sources D. FAQs on GENWEB software tools and methods III RESOURCE IDENTIFIERS A. URN, URL, domain names B. cross-database links IV GENWEB MAINTENANCE, OPERATIONS and QUALITY CONTROL A. Data Maintenance 1. security issues 2. server availability 3. archives & backup services B. Operations 1. GENWEB.ORG domain maintenance 2. mailing list organization C. Quality Control 1. Devise ways to monitor public satisfaction with GENWEB performance 2. Assess the performance of GENWEB sites (devise simple measures of quality to rate GENWEB Sites) 3. Post periodic public reports of GENWEB Site performance (measure things like accessibility, response time to basic queries, measure and rate the quality of the data presented - by data base) V LEGAL AND ETHICAL ISSUES A. copyrights B. access C. responsibility D. privacy E. ownership F. international legal issues G. educational, non-profit status VI PUBLIC RELATIONS A. Goals and Objectives Statement B. Outreach Program: Show/Tell the Public how to use GENWEB 1. Prepare class materials (overheads, videos) a. What computer equipment, spoftware, service do I need to talk to GENWEB? b. How do I find an ancestor on GENWEB? c. How do I download a family from GENWEB and add it to my private data base? d. How do I locate,contact others researching an indiovidual? e. How do I submit my umpti-thousand ancestor data base to a GENWEB Site? f. What format must my data base be in? g. What if my research is not complete? Should I submit what I have? h. How do I input corrections and/or new sources/info to an individual on GENWEB? i. What is required to put a GENWEB server on the Web? 2. Recruit & Train volunteer GENWEB Advocates (individuals & genealogy organizations) world-wide 3. Recruit volunteer skilled computer people to help support site operations. VII FINANCIAL ISSUES 1. Address the need for financial support required (if any) to meet GENWEB Goals (resources needed beyond volunteer help and donated resources) 2. Identify and contact sources of funding (if required) _____________________________________________________________________________ Regards, Bill Minnick Friday, May 12, 1995 8:20:44 PM GenWeb Item From: Gene Stark,gene@starkhome.cs.sunysb.edu,Internet Subject: Re: Project organization To: GenWeb > GENWEB PROJECT REQUIREMENTS >(an expansion and rearranging of Anders Andersson's Seven Points of >Project Organization - by Bill Minnick 5/12/95 Yikes! There are lots of interesting issues and all, but it has a very rigid and bureaucratic sound to it. The beauty of the World-Wide Web is the way in which the available data is maintained and presented by those persons who have a direct interest in the data, according to the dictates of their personal preferences. Rather than construct a huge edifice for GenWeb, with lots of rules and regulations concerning how everyone cross their "t"'s and dot their "i"'s, I'd rather see it get started with a simple, well-selected set of standards that will permit the use of some automated search, indexing, and archiving tools. The whole "rules for submission" thing makes it sound very centralized. I don't see that there is ever going to be an adequately staffed and funded centralized administration that is going to process submissions and check them for compliance with rules and regulations. Rather, things should be set up so that anyone who wants to make their data available can do so with a minimum of fuss and bother. If they don't follow standards, their data will just not be very useful, that's all, because nobody will be able to find it. - Gene Stark Saturday, May 13, 1995 10:12:57 AM GenWeb Item From: dick@bgs.com,Internet Subject: Re: Project organization To: GenWeb One way to think about this is a combination of standards and conventions. The standards the things like naming and search rules so that the whole system hangs together. The conventions are some suggestions of how to format and what to include in your web pages so that people will get the most out of them. The idea of conventions is very important to people who use applications like PAF where all of the source information is in a large "text glom". Without these conventions it is very difficult to make any sense out of the data. The are also important to people developing programs to extract data from GEDCOM or from specific application databases so that they have some order to them. Besides providing order, the main point about conventions is that they can be violated when you have good reason to do so. Insufficient time to get it right might even be considered a good reason 8^{). Dick Schoeller | dick@bgs.com | "Es ist ein Narr, der meint, BGS Systems | 617.891.0000 x 374 | es sei nicht schad, das Kind 128 Technology Center | | auszuschu"tten mit dem Bad" Waltham, MA 02254-9111 | | - Thomas Murner 1512 Sunday, May 14, 1995 10:01:17 AM GenWeb Item From: Anders Andersson,andersa@Mizar.DoCS.UU.SE,Internet Subject: Re: Project organization To: GenWeb Gene Stark writes: >Rather than construct a huge edifice for GenWeb, with lots of rules and >regulations concerning how everyone cross their "t"'s and dot their "i"'s, >I'd rather see it get started with a simple, well-selected set of standards >that will permit the use of some automated search, indexing, and archiving >tools. What is simple and well-selected may be a matter of opinion. I think Bill's tree structure (or any tree of that kind) will be useful to describe what the GenWeb is all about, but maybe our first goal should be to define the major (top-level) issues and document them in order to clear up some confusion as to what belongs where, what is part of GenWeb and what isn't. Then we can allow each section to develop its own substructure independently while still learning from the others. I believe that making a decision about this project skeleton as a single unit is a bad idea. It shouldn't really matter to the overall structure what the different public relations issues are, for instance. We'll need to keep checking the entire structure for possible conflicts in the decision and implementation process, but most issues can be addressed in a context that is much smaller than the entire project. This isn't an experiment in networked genealogy only; it's an experiment in networked project organization as well. We are on our own. -- 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 Sunday, May 14, 1995 1:10:48 PM GenWeb Item From: Peter and Ann Floyd,sunshine@acslink.net.au,Internet Subject: Re: Rootsbook does not work To: GenWeb Works fine for me. I found that it takes Mickey about a day or two longer than he tells me to get the indexes running with new entries ... but the data is definately there, and working fine. I love it so much I am using it as the URL location for my own trees - rather than running my own online database. If you aren't using ROOTSBOOK - find out about it. I understand that some people want to control their own databases locally (fair enough ... keep working on genweb) - but until then, I think ROOTSBOOK is a great place to be. Herbert - Mickey will be able to help you out if you simply can't find the data you are after ... but if you can't see it at all - ask your sysadm why! Cheers:-------------------------------Peter (& Ann) Floyd\ |Email:Sunshine@acslink.net.au Revelations 14:6-7.| \Genealogical Research: WHITBY anywhere,anytime. Mail me!/ Sunday, May 14, 1995 2:48:28 PM GenWeb Item From: Anders Andersson,andersa@Mizar.DoCS.UU.SE,Internet Subject: Re: Project organization To: GenWeb My earlier suggestion: >> - Software (libraries of common code, genealogical data standards) >> - Resource Identifiers (URN, URL, domain names, cross-database links) >> - User Interfaces (HTML, visual layout, data presentation, feedback) >> - Indexing (local and global indices, search facilities) >> - Data Maintenance (server availability, archives, backup services) >> - Legal and Ethical Affairs (copyright, access, responsibility, privacy) >> - Operations (GENWEB.ORG domain maintenance, mailing list organization) Bill's proposal goes along with my line of thought. However, I'm not convinced that we should enter that level of detail quite yet. I'd like to concentrate on the main headings, trying to define what goes under each of them in a general sense, rather than itemizing specific subtopics (when I mention them, I provide them as examples of typical content). Therefore, at this stage I'll focus my comments around Bill's main headings: > I USER INTERFACES > II SOFTWARE > III RESOURCE IDENTIFIERS > IV GENWEB MAINTENANCE, OPERATIONS and QUALITY CONTROL > V LEGAL AND ETHICAL ISSUES > VI PUBLIC RELATIONS > VII FINANCIAL ISSUES In summary, the following are the major differences between my and Bill's views of the project: - Indexing as a separate problem area. Bill makes it a part of User Interfaces (more specifically of Data Submission/Query Tools, which is a place in the hierarchy I don't quite understand). My concept of User Interfaces is probably more narrow in scope than is Bill's. I consider it mainly to be an issue of ergonomics, i.e. how should data be presented for human consumption, what information should go on a WWW page, what facilities for user feedback should be provided, what is proper use of HTML etc. Indices and search tools will certainly have their user interfaces, but the primary task of the Indexing group would be to define what mechanisms should exist, what services need to be coordinated for global searches, and how data must be organized to serve this need. If Indexing isn't found worthy of a section of its own, I'd put it under Software (a heading that, unfortunately, doesn't really tell the reader much about the contents of the corresponding section), or possibly Data Maintenance. - Scope of the Operations area. I put it last in my list, partly because I considered it different from all the others in the sense that Operations is something which is taking place right now, and more specifically handled by Gary Hoffman in maintaining the existing GENWEB.ORG domain and the GenWeb mailing list at UCSD ("behind the scenes stuff", to use Gary's own words). However, Bill probably has a point in that we need to plan for future operations as well, including quality control. I'd like to point out the difference between central operations focussed around the GENWEB.ORG domain, and the work that contributing database managers will perform. The task of planning for the latter is what I considered would be the job of the Data Maintenance group. - Public Relations, which I considered mentioning, but decided to silently include in the Operations group (so far, Gary seems to have been running the PR department almost by himself). Bill appearantly has a much broader agenda for this PR stuff, including a user help desk. Still, I'd put that under Operations. - Financial Issues, which I didn't consider specifically, but I'd probably put that under Operations too (seeking sponsors would be a job for the PR group). However, when we start talking about actual money, we'll probably need a better defined organization than just an Internet domain name registered by Gary Hoffman. Some of the financial stuff, like what transactions may have to take place between GenWeb users for access to the data, should probably be handled by the Legal and Ethical Issues group (as a policy issue) or Data Maintenance (paying for network service). I'd like to emphasize the experimental and developmental nature of the GenWeb project. I'm more interested in implementing an interesting idea than running a business. However, I admit that I'd be less afraid of spending voluntary work on this if I knew that someone will make sure that the project survives its own implementation, once we have a working design. That doesn't mean operational and financial issues should take up half of our present agenda. It's not too essential that we get this right from the beginning, if we take care not to rush ahead until we have resolved all disputes that may arise from an initial decision, and are prepared to change the agenda in face of good arguments. Having a single, non-optimal agenda to argue about (meaning that it stands if nobody comes up with a superior agenda) may be better than arguing about an unlimited number of proposals and not getting anywhere. I believe that our primary disability is that we have a hard time envisioning each other's point of view. -- 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 Sunday, May 14, 1995 5:50:41 PM GenWeb Item From: Bill Minnick,svpafug@rahul.net,Internet Subject: Re: Project organization To: GenWeb Anders Anderrson writes (14 May '95) >Bill's proposal goes along with my line of thought. However, I'm >not convinced that we should enter that level of detail quite yet. >I'd like to concentrate on the main headings, trying to define what >goes under each of them in a general sense, rather than itemizing >specific subtopics (when I mention them, I provide them as examples >of typical content). Anders, Before I comment on details, I want to thank you for writing a thoughtful, well reasoned response to my last letter on the subject of GENWEB organization. Letters of this caliber and depth take time and commitment to compose. >Therefore, at this stage I'll focus my comments >around Bill's main headings: > >> I USER INTERFACES >> II SOFTWARE >> III RESOURCE IDENTIFIERS >> IV GENWEB MAINTENANCE, OPERATIONS and QUALITY CONTROL >> V LEGAL AND ETHICAL ISSUES >> VI PUBLIC RELATIONS >> VII FINANCIAL ISSUES > >In summary, the following are the major differences between my and Bill's >views of the project: > > - Indexing as a separate problem area. Bill makes it a part of User > Interfaces (more specifically of Data Submission/Query Tools, which > is a place in the hierarchy I don't quite understand). My concept > of User Interfaces is probably more narrow in scope than is Bill's. > I consider it mainly to be an issue of ergonomics, i.e. how should > data be presented for human consumption, what information should go > on a WWW page, what facilities for user feedback should be provided, > what is proper use of HTML etc. My background is system design, and I tend to focus on what the user wants and needs. That's why I put INDEX and SEARCH functions under Query tools which are under User Interfaces. For the purposes of getting the work done, I would go along with putting these functions under software, but the first task is to define the user's query needs and how they translate into index and search software requirements. > Indices and search tools will certainly have their user interfaces, > but the primary task of the Indexing group would be to define what > mechanisms should exist, what services need to be coordinated for > global searches, and how data must be organized to serve this need. > If Indexing isn't found worthy of a section of its own, I'd put it > under Software (a heading that, unfortunately, doesn't really tell > the reader much about the contents of the corresponding section), > or possibly Data Maintenance. I think Indexing and Search tools are one of the most important aspects of GENWEB. If we are organizing the outline according to importance, I say put Index/Search functions at the top level. > > - Scope of the Operations area. I put it last in my list, partly > because I considered it different from all the others in the sense > that Operations is something which is taking place right now, and > more specifically handled by Gary Hoffman in maintaining the > existing GENWEB.ORG domain and the GenWeb mailing list at UCSD > ("behind the scenes stuff", to use Gary's own words). > > However, Bill probably has a point in that we need to plan for > future operations as well, including quality control. Anders, if we are ever to have the genealogists endorse GENWEB, we'll have to place information quality at the top level of our plan. > I'd like to > point out the difference between central operations focussed around > the GENWEB.ORG domain, and the work that contributing database > managers will perform. The task of planning for the latter is what > I considered would be the job of the Data Maintenance group. Good point! > > - Public Relations, which I considered mentioning, but decided to > silently include in the Operations group (so far, Gary seems to > have been running the PR department almost by himself). Bill > appearantly has a much broader agenda for this PR stuff, including > a user help desk. Still, I'd put that under Operations. OK by me (PR under operations) as long as we identify the need and fill it. I've already joined the PR department along with Floyd Nordin. Together, we've put on detailed GENWEB presentations to over 500 people in the San Francisco Bay area during the past year. We'll be presenting and demonstrating the GENWEB concept in a formal presentation at the Genealogy Conference in Seattle, WA in September, '95 sponsored by the Federation of Genealogical Societies (FGS). > - Financial Issues, which I didn't consider specifically, but I'd > probably put that under Operations too (seeking sponsors would be > a job for the PR group). However, when we start talking about > actual money, we'll probably need a better defined organization > than just an Internet domain name registered by Gary Hoffman. Agree; put under operations. > Some of the financial stuff, like what transactions may have to > take place between GenWeb users for access to the data, should > probably be handled by the Legal and Ethical Issues group (as a > policy issue) or Data Maintenance (paying for network service). Perhaps Legal and ethical issues should go under operations, too. >I'd like to emphasize the experimental and developmental nature of the >GenWeb project. I'm more interested in implementing an interesting idea >than running a business. However, I admit that I'd be less afraid of >spending voluntary work on this if I knew that someone will make sure >that the project survives its own implementation, once we have a working >design. That doesn't mean operational and financial issues should take >up half of our present agenda. Good points. I am not as concerned about the final structure of the outline as I am that the topics I listed receive consideration in our planning efforts. I'd rather bring a topic up and have it soundly rejected, than have up fail to discuss an important subject we just plain didn't think was important. >It's not too essential that we get this right from the beginning, if we >take care not to rush ahead until we have resolved all disputes that may >arise from an initial decision, and are prepared to change the agenda in >face of good arguments. Having a single, non-optimal agenda to argue >about (meaning that it stands if nobody comes up with a superior agenda) >may be better than arguing about an unlimited number of proposals and >not getting anywhere. I believe that our primary disability is that we >have a hard time envisioning each other's point of view. Anders, please feel free to reissue a top level outline based on your best judgment at this point. Baring any other inputs, I think we can accept what you come up with. Regards, Bill Minnick Sunday, May 14, 1995 11:01:20 PM GenWeb Item From: Michael Simpson,simpsonm@NEONRAMP.COM,Internet Subject: Technical Specs Docment To: GenWeb Howdy, Reguarding the Technical Specifications Document: I do technical writing at work, updating procedures and writing new ones for computer operators to follow. I would be willing to take a look at the requirements for a Technical Specifications Document. Copywrite is no problem, as long as my name appears somewhere so that when I look for a job as a technical writer, I have a reference. Thanks, Michael Monday, May 15, 1995 9:17:28 AM GenWeb Item From: Gary Hoffman,ghoffman@ucsd.edu,Internet Subject: How to unsubscribe GenWeb To: GenWeb Just a reminder. If you would like to unsubscribe from the genweb mailing list, please send a message to listserv@ucsd.edu, not to the list. In the body of the message, please put the following: UNSUB genweb That's all folks. It's easy, it's simple, and it's quick. BTW, I consider it a violation of 'netiquette' to go on vacation and leave a vacation notice while subscribed to a mailing list. This causes your mail system to generate an automatic reply to the sender of every message posted to every list you subscribe to. Not only is this a lot of extraneous traffic, but It would be better to unsub the list when you leave for vacation and then catch up when you return by reading the archives. Thanks for your consideration. Monday, May 15, 1995 11:44:24 AM GenWeb Item From: Anders Andersson,andersa@Mizar.DoCS.UU.SE,Internet Subject: Re: How to unsubscribe GenWeb To: GenWeb Gary Hoffman writes: >BTW, I consider it a violation of 'netiquette' to go on vacation and leave >a vacation notice while subscribed to a mailing list. This causes your mail >system to generate an automatic reply to the sender of every message posted >to every list you subscribe to. Not only does this create a lot of >extraneous traffic, but it is bothersome to the person who sent the >original message. I'd like to put the blame on the vacation software rather than on its users. The Unix vacation(1) utility takes care not to return any messages in response to mail not directed specifically to that user. Using it is in no way a violation of 'netiquette', which is a concept that transcends operating system peculiarities. It may be optimistic to expect the same functionality from corresponding utilities on a Macintosh or Windows PC, but I wouldn't go so far as to call it a case of bad manners. >It would be better to unsub the list when you leave for vacation and then >catch up when you return by reading the archives. It depends of course on that user's interest in the lists vs. the need to inform coworkers about the leave. If I found myself stuck with such a vacation program, I'd inform the vendor about it and refuse to use it until the flaw was removed. Users should be informed about bugs in their software, but not blamed for unknowingly triggering 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 Tuesday, May 16, 1995 1:47:18 PM GenWeb Item From: brian.mavrogeorge@f30.n125.z1.fidonet.org,Internet Subject: Project organization/beginings To: GenWeb Anders wrote that our primary disability is that we have a hard time envisioning each other's point of view. What I see is simply a difference in approach to problem solving. I used to have a co-worker who was attached to his flipchart by his umbilical cord (grin). Every time we had a discussion he would start charting and outlining and categorizing. Drove me nuts. I was still in the exploration stage and he was already codifying. So some people approach the world (and genealogy I might add) in a scavenger mode -- take in as much information as I can get and find the patterns, find what is relevant and sift through it. Others go immediately to the outline - defining the problem, getting the subtopics, bringing structure to the issue. Neither approach better than the other as much as just an individual's way of solving problems. What is important, and Anders mentions it, is that we all agree that we may not get this right from the beginning and we all are conscious of the need to resolve disputes, and that each of us is prepared to change the agenda (outline) in the face of good arguments. So I am comfortable with Anders' and Bill's proposal of a structure for some discussions and decisions. If we have some item in the wrong spot it will become apparent as we flesh out topics. And I suggest that getting down every one's vision of GenWeb is a good start at understanding each other's point of view. As I go through the archives I will attempt to glean some of that information. And as everyone contributes their vision that doesn't mean they can't also start fleshing out some subtopics. Wednesday, May 17, 1995 3:24:48 PM GenWeb Item From: Anders Andersson,andersa@Mizar.DoCS.UU.SE,Internet Subject: Re: Project organization To: GenWeb Bill Minnick writes: >I think Indexing and Search tools are one of the most important aspects of >GENWEB. If we are organizing the outline according to importance, I say put >Index/Search functions at the top level. I don't think "importance" is an extremely useful measure on which to base the structure of the project components. Importance isn't inherent with any single component regardless of its surroundings. I'm concerned about having well-defined interfaces between the different components (and I'm not talking software components here, which would probably need a different structure). It should be possible to resolve most of the design issues locally within each group without interfering with the work of other groups. Still, I agree that the index/search issue is an important one, and I think this issue is well-defined enough to be placed at the top level. Further dissemination of it will tell us whether it's a wise decision. >Anders, if we are ever to have the genealogists endorse GENWEB, we'll have >to place information quality at the top level of our plan. I'm not sure how to interpret this. Information quality certainly has top priority, but that doesn't mean we gain anything from making it a topic of its own. I'd rather see information quality as an issue that has to be considered everywhere in the system design, just like software quality and documentation. However, we may need a Project Management group to handle these cross-group issues. >OK by me (PR under operations) as long as we identify the need and fill it. Certainly. I believe PR started with Gary Hoffman's proposal on the GEDCOM list one year ago, and appearantly it's going on as we speak. Since there is little about PR that can be designed in advance, I think this is part of what Operations is about. >Perhaps Legal and ethical issues should go under operations, too. Operations will probably need their share of "lawyers" to handle day-to-day legal matters. It's too early for me to say what they would have to concentrate on. However, I think of the Legal and Ethical Issues group as the GenWeb "politicians" drafting the GenWeb "constitution" (like, defining policies to be implemented by Operations). So, I'd like to keep them as a separate group, doing long-term development work along with the technical groups. >Anders, please feel free to reissue a top level outline based on your best >judgment at this point. Baring any other inputs, I think we can accept what >you come up with. Ok, how about this: - Project Management (overall project coordination and arbitration between groups, cross-group quality control and documentation) - Legal and Ethical Issues (copyright, access, formal responsibility, privacy, censorship, statutes and legal status of any formal organization) - Software [poor name; better suggestions wanted] (code libraries, GEDCOM and other standards for representation and transmission of genealogical data, research data semantics) - Data Maintenance (archive availability and reliability, backup facilities and service organization, network and server access, coordination of independent GenWeb database contributors, distributed resource allocation, load sharing) - Resource Identifiers (cross-database links, naming conventions, GENWEB.ORG domain contents, URN and URL specifications) - Indexing (local and global indices, topical indices, search facilities, "web crawlers") - User Interfaces (visual layout, data presentation, facilities for user contributions of data and other feedback, ergonomics, HTML and other hypertext issues, sound and image presentation formats) - Operations (financial control, public relations, user assistance, maintenance of central services like the GENWEB.ORG domain, project mailing list administration) As you see, this is basically my initial proposal with slightly extended topic descriptions, a little reordered, and with a new group added: Project Management. This all sounds pretty formal, but I hope the business-oriented terminology won't impede the innovative effort. Since English isn't my native language, feel free to improve on my terms. I have avoided enumerating the groups in order not to give them absolute positions (I prefer to see Operations last regardless of how many other groups we end up with, but this is just to suit my personal, intuitive taste). Referring to them by name should be sufficient. If we can start with these, we could write up a more detailed description of each group or topic, and see whether they hold together as a basis for further work and discussions. -- 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 Thursday, May 18, 1995 4:18:08 AM GenWeb Item From: Mickey Lane,MLANE@csi.compuserve.com,Internet Subject: re: Project Organization To: GenWeb Anders writes: > - Software [poor name; better suggestions wanted] (code libraries, > GEDCOM and other standards for representation and transmission > of genealogical data, research data semantics) How about Engineering? That's what you're asking for... You have a goal in mind, now you need some folks with plastic pocket protectors to "make it happen." :-) Mickey. Thursday, May 18, 1995 6:49:50 AM GenWeb Item From: Anders Andersson,andersa@Mizar.DoCS.UU.SE,Internet Subject: re: Project Organization To: GenWeb Mickey Lane writes: >Anders writes: > >> - Software [poor name; better suggestions wanted] (code libraries, >> GEDCOM and other standards for representation and transmission >> of genealogical data, research data semantics) > >How about Engineering? That's what you're asking for... You have a goal >in mind, now you need some folks with plastic pocket protectors to "make >it happen." :-) "Engineering" definitely sounds better than "Software", but on the other hand it seems a little too broad. I believe Data Maintenance, Resource Identifiers, Indexing, and User Interfaces all will contain corresponding amounts of engineering and plastic pocket protectors. "Software" is too broad also, which is part of my reason for wanting another name. Another thing is that I expect this group to deal with the definitions and semantics of various data structures, like the existance of partially conflicting relationship links and other data (researcher A believes that B is a child of C, which researcher D doubts because of a remark in document E, etc). Perhaps my mistake is to lump this complex issue together with the development and maintenance of common code libraries? I simply don't think the latter is an issue big enough to warrant its own group, but it has to be put somewhere, and data structuring seems close. However, for the purpose of settling on any name at all, I'd accept "Engineering" before "Software". Just like everything else, we can change it later if we want. -- 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 Thursday, May 18, 1995 7:30:54 AM GenWeb Item From: Bill Minnick,svpafug@rahul.net,Internet Subject: re: Project Organization To: GenWeb Anders Concludes: >However, for the purpose of settling on any name at all, I'd accept >"Engineering" before "Software". Just like everything else, we can >change it later if we want. We've used the term "Software Engineering" to resolve the ambiguity. -- Bill Minnick Thursday, May 18, 1995 9:05:28 AM GenWeb Item From: Anders Andersson,andersa@Mizar.DoCS.UU.SE,Internet Subject: re: Project Organization To: GenWeb Mickey writes: >Perhaps the departments you're looking for are Architecture and Engineering. No. Those labels indicate an entirely different structure. Maybe that's a better structure, but in order to employ it we would have to scrap most of my current proposal. You can't have "Resource Identifiers" and "User Interfaces" alongside "Architecture" and "Engineering", if you are to give every group a specific task. It would be like having two groups build a car: One group designs the passenger seats, and the other takes care of documentation. Great passenger seats. Great drawings. Great car. ;-) >The engineering group figures out how to do it, >and operations does it. I have specified an Operations group for the purpose of running central GenWeb services, but I consider the architecture and engineering process split up along component lines. If you think this won't work, please tell me why. >Reading these notes is a wonderful lesson in the difference between academia >and industry..... :-) Maybe we should decide whether the GenWeb project is a commercial or an educational experiment? -- 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 Thursday, May 18, 1995 11:52:12 AM GenWeb Item From: Mickey Lane,MLANE@CSI.compuserve.com,Internet Subject: re: Project Organization To: GenWeb Anders writes: >"Engineering" definitely sounds better than "Software", but on the >other hand it seems a little too broad. I believe Data Maintenance, >Resource Identifiers, Indexing, and User Interfaces all will contain >corresponding amounts of engineering and plastic pocket protectors. Perhaps the departments you're looking for are Architecture and Engineering. The architecture group figures out what to do and why it should be done that way, The engineering group figures out how to do it, and operations does it. Reading these notes is a wonderful lesson in the difference between academia and industry..... :-) Mickey. Thursday, May 18, 1995 4:38:57 PM GenWeb Item From: Bill Minnick,svpafug@rahul.net,Internet Subject: Re: Project organization To: GenWeb Anders Andersson writes (regarding GenWeb Project Organization: >Ok, how about this: > > - Project Management (overall project coordination and arbitration > between groups, cross-group quality control and documentation) > - Legal and Ethical Issues (copyright, access, formal responsibility, > privacy, censorship, statutes and legal status of any formal > organization) > - Software [poor name; better suggestions wanted] (code libraries, > GEDCOM and other standards for representation and transmission > of genealogical data, research data semantics) > - Data Maintenance (archive availability and reliability, backup > facilities and service organization, network and server access, > coordination of independent GenWeb database contributors, > distributed resource allocation, load sharing) > - Resource Identifiers (cross-database links, naming conventions, > GENWEB.ORG domain contents, URN and URL specifications) > - Indexing (local and global indices, topical indices, search > facilities, "web crawlers") > - User Interfaces (visual layout, data presentation, facilities for > user contributions of data and other feedback, ergonomics, > HTML and other hypertext issues, sound and image presentation > formats) > - Operations (financial control, public relations, user assistance, > maintenance of central services like the GENWEB.ORG domain, > project mailing list administration) > >As you see, this is basically my initial proposal with slightly >extended topic descriptions, a little reordered, and with a new >group added: Project Management. . . . . . > >If we can start with these, we could write up a more detailed >description of each group or topic, and see whether they hold >together as a basis for further work and discussions. ANDERS: In my opinion, your top level organization will suffice. However, I would prefer to see the topic, "Information Quality". identified at the top level. As a compromise, I would accept this as a subtopic under "User Interfaces" or even "Data Maintenance". Ahead of us, we have the daunting task of motivating the public (the many data base contributors) to document their sources. I've spent 5 years working this task with the Austin Families Association of America and the Silicon Valley PAF Users Group, and have achieved a slow, but measurable improvement in quality of submitted data bases. So if you see a spot for "Information Quality" in your proposed organization, it would please me. If not, we'll continue carry this task on as a separate endeavor. Regards, Bill Minnick Friday, May 19, 1995 5:36:53 AM GenWeb Item From: Anders Andersson,andersa@Mizar.DoCS.UU.SE,Internet Subject: Re: Project organization To: GenWeb Bill Minnick writes: > In my opinion, your top level organization will suffice. However, I >would prefer to see the topic, "Information Quality". identified at the top >level. As a compromise, I would accept this as a subtopic under "User >Interfaces" or even "Data Maintenance". > Ahead of us, we have the daunting task of motivating the public (the many >data base contributors) to document their sources. Yes, I believe you are right, and that Information Quality is a topic worthy of a separate discussion. Thanks for pointing it out to me. However, this should not be interpreted as relieving the other groups of the requirement to consider the quality aspects in every part of the design and implementation process. Let's say: - Information Quality (reliability of sources, source documentation, source comparison, user education) As for the ordered list, I'd put this after Legal and Ethical Issues, and before Engineering (Software), as a major topic in itself (making it a subtopic of either User Interfaces or Data Maintenance would severely distort the scope of those topics, I think). I believe the closest ties will be with the Engineering (Software) group due to the work on source documentation, but it's independent enough to have its own heading. I adopt this list of nine topics as my current proposal. How do we decide about it? Can we accept it based on the lack of maintained opposition? Only a few people have voiced their opinions on this issue, but I hope there is interest also among the silent majority. Gary, how many members does the GenWeb list have now, by the way? To summarize: - Project Management - Legal and Ethical Issues - Information Quality - Engineering (Software) - Data Maintenance - Resource Identifiers - Indexing - User Interfaces - Operations Anything else? -- Anders Andersson, Dept. of Computer Systems, Uppsala University Paper Mail: Box 325, S-751 05 UPPSALA, Sweden Phone: +46 18 183170 EMail: andersa@DoCS.UU.SE Friday, May 19, 1995 2:10:34 PM GenWeb Item From: Bill Minnick,svpafug@rahul.net,Internet Subject: Re: Project organization To: GenWeb Anders Andersson writes: >I adopt this list of nine topics as my current proposal. How do we >decide about it? Can we accept it based on the lack of maintained >opposition? Only a few people have voiced their opinions on this >issue, but I hope there is interest also among the silent majority. >Gary, how many members does the GenWeb list have now, by the way? > >To summarize: > > - Project Management > - Legal and Ethical Issues > - Information Quality > - Engineering (Software) > - Data Maintenance > - Resource Identifiers > - Indexing > - User Interfaces > - Operations > >Anything else? Anders, I concur with your list. Thanks for taking the lead on this phase of our discussions. Regards, Bill Minnnick Saturday, May 20, 1995 6:26:23 PM GenWeb Item From: T.T.Wetmore,ttw@beltway.att.com,Internet Subject: LifeLines 3.0.2 To: GenWeb Liners, Geders, Webers, LifeLines version 3.0.2 is released. It is available in source form at: ftp://hoth.stsci.edu/lines/lines.3.0.2.tar.gz ftp://ftp.cac.psu.edu/pub/genealogy/lines/lines.3.0.2.tar.gz Version 3.0.2 remains a UNIX only system. The release has a reference manual fully updated to version 3.0.2. It is in postscript form. Sorry for delays over release, and apologies that the source and event handling features are still not up to snuff. Tom Wetmore Sunday, May 21, 1995 12:41:49 PM GenWeb Item From: Robert Fieg,bfieg@gn2.getnet.com,Internet Subject: HTML genealogy To: GenWeb Dear GenWebbers, I have been reading this continuing GenWeb saga with interest, and I have also been doing some experimenting with Gene Stark's GEDCOM to HTML converter. It seemed to me that following a line by jumping from individual to individual with a web browser was a little confusing, and that a chart could be very useful for getting the "big picture." So I decided to put a pedigree chart in the HTML file for each individual. Furthermore, each person on the chart would be represented by a hyper-link providing an easy way view information about the various ancestors. To this end, I have written a program which adds these pedigree charts to the individual HTML files. A couple examples of the resultant genealogies, along with the program and instructions for obtaining and running it, have been placed on my homepage at: http://www.getnet.com/~bfieg I would appreciate feedback. I realize everyone is busy, but as time permits please check it out and let me know what you think. Bob Fieg Sunday, May 21, 1995 10:16:15 PM GenWeb Item From: Scott McGee,smcgee@microware.com,Internet Subject: My genweb software To: GenWeb A few months back, I made availible to some users, a copy of the files I was then using in my GenWeb server. (Keep in mind, our sysadmin has our www server set to refuse requests from offsite so you can't access my genweb server. It is NOT the meager bit I have off my homepage where I have too limited a quota and no cgi to do the full thing.) Since I have made some (IMHO) nice improvements in it since then, and talk has been made of how to implement some of them, I have put together an archive of the files I use currently and made it availible for downloading. To access it via web browser, just use the URL http://www.cc.utah.edu/~sam8644/genweb.tar.Z to download it. Please keep in mind that I do NOT consider this ready for a general release, and hope you will not distribute this archive elsewhere. You may freely use any of the files within it as a basis for developing your own. Installation instructions are not included, but feel free to ask. WHAT DOES IT DO. My current implementation allows you to display an individual page, a pedigree chart or a descendant chart on each individual in the database. The individual page provides name, birth, baptism, death, burial, spouses, children, and notes for the individual. The pedigree and descendant charts allow you to traverse the tree to other displayed individuals. The individual page can include an inlined picture if one is specifed using my proposed format in the database. (email for details) I would be very interested in suggestions concerning this implementation. If it is used to set up a site, please let me know so I can check it out too. Scott Web: http://www.cc.utah.edu/~sam8644/homepage.html | Why do I need a email: smcgee@microware.com (Scott McGee) | disclaimer? Nobody | Duct tape is like the FORCE! It has a light side | has ever beleived | and a dark side and holds the universe together. | me yet! Would you? Monday, May 22, 1995 9:16:30 AM GenWeb Item From: JimIsaak,isaak@ljo.dec.com,Internet Subject: HTML Table formats To: GenWeb The lastest HTML spec (3.0) (online via CERN) describes table entries; and the latests Netscape supports this format (available via FTP from netscape.com ... I suspect this will have a significant positive value for how we present Genealogical information in the future web. Jim Monday, May 22, 1995 10:39:50 PM GenWeb Item From: Annelise Anderson,ANDRSN@HOOVER.STANFORD.EDU,Internet Subject: Soundex Program To: GenWeb I thought some of you might be interested in knowing that there's a program for Soundex searching of textfiles in PC Magazine, May 30, 1995 issue, p. 261. The C source code is given; a C compiler is needed, or alternatively the program file with source code can be downloaded from PC MagNet's Utilitites/Tips Forum from Library 3 (Tutor) as PS.ZIP. This program will search a specified text file for a name (or text string, I suppose), which it first converts to Soundex and then finds all names with the same Soundex codes; as I understand it, it prints lines from the file containing the matches to the screen or to a printer. Annelise Tuesday, May 23, 1995 6:37:01 AM GenWeb Item From: Scott McGee,smcgee@microware.com,Internet Subject: My Genweb software archive To: GenWeb At least one person has had trouble getting my genweb software archive so I added a link in my homepage to let you download it. This is a little more public than I really wanted but it should make it easier. If you want to use it, go to my homepage (see my .sig for URL), select the genealogy section in the Features list. and it should be the second link and reads "Download my genweb software archive". Scott PS ANY feedback greatly apriciated. If at first, you don't succeed, | smcgee@microware.com (Scott McGee) go fry a hen. After all, fried | ----------------------------------------- chicken beats failure any time. | I was paid $5.00 to express these views! -------------> http://www.cc.utah.edu/~sam8644/homepage.html <------------- Friday, May 26, 1995 6:53:45 AM GenWeb Item From: Mickey Lane,MLANE@csi.compuserve.com,Internet Subject: ROOTSBOOK more-or-less on the web To: GenWeb Hi, ROOTSBOOK is back on line. The machine hosting the server crashes after about 20 hours of server use. I was away from the office for a week so I wasn't able to reboot it... As a systems software engineer, I'm a lot more interested in why the system crashes than I am in fixing it so y'all will have to put up with the off again, on again nature of the thing for a while longer. The server is now hosting 194,944 entries. My next objective with the ROOTSBOOK software is to make the source code available to whoever might be interested (Phase II for those who remember...) Mickey. http:/mlane2.inhouse.compuserve.com:8000/GenWeb.htm Friday, May 26, 1995 10:11:55 PM GenWeb Item From: Michael Simpson,simpsonm@NEONRAMP.COM,Internet Subject: Technical Specs Docment To: GenWeb Howdy, Reguarding the Technical Specifications Document: I do technical writing at work, updating procedures and writing new ones for computer operators to follow. I would be willing to take a look at the requirements for a Technical Specifications Document. Copywrite is no problem, as long as my name appears somewhere so that when I look for a job as a technical writer, I have a reference. Thanks, Michael Sunday, May 28, 1995 6:56:56 PM GenWeb Item From: T.T.Wetmore,ttw@beltway.att.com,Internet Subject: Re: GEDCOM Standardization To: GenWeb Terry, Got your note. I am not all that interested in GEDCOM standardization. My interest is in genealogy as a data intensive and data driven research activity, and in the forms that genealogical data takes as the genealogical research process unfolds. GEDCOM as a syntax is fine for holding all the types of data as I now imagine them (sources, evidence, events, persons, families, facts, what have you), but that we have far to go on the semantics. All the semantics for GEDCOM have been based on the person and family, which are only the "results objects" of genealogy. At present you can sense some attention begin paid to the event and evidence side of genealogy (eg, ROOTS IV, TMG), but we have a long way to go on that front. I don't know of any group of users/programmers doing anything serious about getting out a new standard. There is a mailing list, GEDCOM-L, that you are probably aware of, where some of these issues are discussed. There is also the GENWEB mailing list where there is some interest in standardization for WWW applications of genealogy. I don't see GEDCOM as a communication protocol as such, simply a format for carrying structured information; that is, I don't relate GEDCOM to the workings of a comm stack, except possibly to what bubbles out the top of it. When I imagine what it means to standardize GEDCOM, I don't imagine any OSI stackish stuff involved. I view GEDCOM and GEDCOM-like formats as much as formats for maintaining genealogical data as for exchanging it, a position I have to defend fairly regularly. I don't know what GENTECH is up to in the standards area. I gave a talk at their conference a couple years ago, and I correspond with a few of their members on an infrequent basis, but that's it. I don't feel I know enough about the genealogical research process to be advocating any standards. This is one of the goals I have with my LifeLines program -- as an experimental vehicle that I can use to help me learn what are the semantics of genealogical data. I view the syntactic issues as trivial issues, way down in the noise. But I've been forced to the conclusion that when most people start discussing genealogical standards, all they seem to think about are the trivial syntactic issues. I have no interest in that level of standardization. The semantics issues are the big ones, and frankly, very few users and developers are enough aware of what the genealogical research process actually is, to be even thinking about this issue, let alone have any competent opinions about it. Every time I say in a public forum that persons and families in genealogical databases are conclusions that are derived from evidence, and that real genealogical systems must deal with the processes of deriving these conclusions from evidence I stir up all kinds of emotion. The average user can't stand the notion that the persons in their databases are "just" conclusions, and the average developer does not seem much smarter. In such a world it seems unsafe to me to be developing standards. Call me old fashioned, but I've always thought it more important to understand an area before developing standards for it, and not the other way around. I hope whoever told you I was interested in this area also let you know I'm a crusty old fart, who, I fear, is getting crustier every day. Tom Wetmore Sunday, May 28, 1995 11:18:51 PM GenWeb Item From: Scott McGee,smcgee@microware.com,Internet Subject: Harvest To: GenWeb I got around to looking at the harvest info tonight, and they have a broker for WWW Home Pages. I did a genweb search on it and turned up a fair number of genweb pages! There is a form there to allow you to register your own homepage with the harvest database. I added my own, and thought I should encourage you guys to add yours too. It actually makes a nice index of genweb pages and can be used for a lot of other lookups too. The URL for the WWW Home Pages broker is: http://harvest.cs.colorado.edu/brokers/www-home-pages/query.html and has both the query form and the "add a url" form. Scott If at first, you don't succeed, | smcgee@microware.com (Scott McGee) go fry a hen. After all, fried | ----------------------------------------- chicken beats failure any time. | I was paid $5.00 to express these views! -------------> http://www.cc.utah.edu/~sam8644/homepage.html <------------- Monday, May 29, 1995 6:00:42 AM GenWeb Item From: Anders Andersson,andersa@Mizar.DoCS.UU.SE,Internet Subject: Re: Harvest To: GenWeb Scott McGee writes: >I got around to looking at the harvest info tonight, and they have a broker >for WWW Home Pages. I did a genweb search on it and turned up a fair number >of genweb pages! There is a form there to allow you to register your own >homepage with the harvest database. I added my own, and thought I should >encourage you guys to add yours too. It actually makes a nice index of >genweb pages and can be used for a lot of other lookups too. The URL for [GenWeb project groups: Data Maintenance, Indexing] There are actually several of these generic WWW indices around, each with its particular technical flavour. Does anyone keep an index of them..? The problem I see is that nobody can tell me which index to use in order to reach my audience, and I may find myself registering my pages with every index I stumble upon, just to make sure. It's as if the phone company didn't provide a phone directory, but you had to decide which out of several independent phone directories to advertise your phone number in. Not to mention the hassle of keeping the information up to date when you move, etc. If anyone is able to find their way to the GenWeb project by looking up the keyword "GenWeb" in the Harvest database, Lycos, Yahoo, Planet Earth, EINet Galaxy, Virtual Library, or any other index currently in operation, fine. That doesn't mean either of them (or all of them together) provides a service that is particularly useful to genwebbers for locating GenWeb pages, since we can easily set up an index for our particular purpose and demand that every GenWeb participant register with it. Rather than having Harvest or Lycos users skimming their way through thousands of references to randomly selected GenWeb pages (perhaps many of them duplicates), wouldn't it be better to send them a pointer to our own, supposedly authoritative, GenWeb index? Having every WWW information provider register every page of information with a central WWW index isn't going to scale well beyond the initial crowd of enthusiasts. We'll have to delegate. -- 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 Monday, May 29, 1995 8:15:43 AM GenWeb Item From: Tim Doyle,tdoyle@netcom.com,Internet Subject: Re: Harvest To: GenWeb On Mon, 29 May 1995, Anders Andersson wrote: > Rather than having Harvest or Lycos users skimming their way through > thousands of references to randomly selected GenWeb pages (perhaps > many of them duplicates), wouldn't it be better to send them a > pointer to our own, supposedly authoritative, GenWeb index? In the future, the GenWeb organization will undoubtedly want to set up its own database index. To fill the need for an index in the meantime, I have created and have made available a GenWeb Database Index, located at: http://www.doit.com/tdoyle/genweb/ I understand that this list may be superceeded by one created by GenWeb in the future, but for the time being, it is at least available. If you have any suggestions or know about databases which I have not included, please let me know. Tim Doyle / tdoyle@netcom.com / tdoyle@doit.com WWW homepage: http://www.doit.com/tdoyle/ ftp directory: doit.com pub/tdoyle Wednesday, May 31, 1995 8:26:22 PM GenWeb Item From: bmavroge@InterServ.Com,Internet Subject: FTM - World Family Tree project To: GenWeb Does anyone have details on Banner Blue's World Family Tree Project? From what little I have read they are encouraging their users to send in their family information for consolidation in one dataabase from which one can generate a family tree. Would like to know more info and if anyone has talked to them about integrating it with GenWeb? Wednesday, May 31, 1995 10:41:52 PM GenWeb Item From: rontay@wimsey.com,Internet Subject: Re: FTM - World Family Tree project To: GenWeb I like Banner Blue's product. That being said, they don't seem to have the slightest idea as to real Genealogy. Think about it - how would their approach compare wth Roots-L. My experience is that Banner Blue produces good software but their attempts to be genealogists are inadequate at best. Their view seems to be that the rest of us are conmmitted top making money for BB. Lots of peop;le are really trying to build worthwhile INTERNET resources. BB is simply trying to make money. Ron Taylor Vancouver Canada 7512 Murray Street, Mission, B.C. V2V 4B1 (604)826-1982 (H) (604)294-7734 (W) (604) 294-7710 FAX