Dennis Lee Bieber replied: > VFP can only address bytes 0 to 2147483647. > At best, it will refuse to save any record that would require > exceeding the 2GB boundary. > At worst, it wraps around and corrupts data ... a negative offset Actually, as a retired computer geek who was involved in the Y2K issues, wrap around can be even worse. Rather than a negative offset it is more likely to truncate, overwriting your oldest data. 2147483647+1 => 0 2147483647+2 => 1 ... Yes, Erik, you really, Really, REALLY!! need to split your project into multiple projects. You worried that: > I can`t split the project since all of the people > are connected to me somehow. > The focus with that project has been starting with me and then > entering as many ancestors I knew ,and then also go on to all > I could find of descendants of all the siblings and halfsiblings > of my ancestors etc. Well, you are now "cursed" by your success at finding so many relations. I would suggest splitting based on some level of direct ancestor. I would first split off a "close ancestor" main project trimmed to a project that only goes back on the direct line from you to some set of selected direct ancestors. The level back does not have to be consistent, it could depend upon keeping the number of people which would be included in each of the split projects roughly the same. This main project would include all these direct-line ancestors descendants, and those descendants ancestors, but not the selected ancestor's own ancestors nor their siblings. I would then split off separate non-main projects based on these selected ancestors. For example selecting all 3*great grandparents could result in 16 projects, or 6*great could be 128+1 projects, whatever level you think you might need. Each X*grandparent project would include only that person, their direct ancestors and siblings, plus all these people's siblings, descendants and ancestors. But it would not include this X*grandparent's descendants and relations since they are in the main project. Yes, you may wind up with duplicates in multiple projects if there are cross marriages among descendants or ancestors. You will have to examine the projects to find these. Then I would probably create some custom tags which would tell me something like: "All the descendants (or ancestors) of this individual may be found in project XYZ." That could let you limit that duplicate descendant or ancestor line to a single project as well as help you locate them. Hope this gives you ideas, Michael
At 07:48 PM 11/21/2017, Michael J Hannah wrote: >Dennis Lee Bieber replied: >>VFP can only address bytes 0 to 2147483647. >>At best, it will refuse to save any record that would require >>exceeding the 2GB boundary. >>At worst, it wraps around and corrupts data ... a negative offset > >Actually, as a retired computer geek who was involved in the Y2K >issues, wrap around can be even worse. Rather than a negative >offset it is more likely to truncate, overwriting your oldest data. >2147483647+1 => 0 2147483647+2 => 1 ... Y2K wrap around was due to the use of an implied century with a 2-digit /decimal/ year. The 2GB limit is due to use of a /signed/ binary 32-bit integer rather than an /unsigned/ binary 32-bit integer (which would give a 4GB limit). The positive range of the integer uses 31-bits, it becomes negative when it overflows into the 32nd bit. {It's also a problem with anything written in Java -- which does not have an unsigned data type; and I don't know if it supports 64-bit integers either.} -- [email protected] Dennis Lee Bieber HTTP://home.earthlink.net/~bieber.genealogy/
Truncating and overwriting aren`t good. But what to do now? and after that, how splitting the project I have? What to do now?..... Before I think I can split anything I have another problem (which may be no problem??)....I am doing some maintenance on the people in my large project. (And to this job it is very convenient to have the IDs I have done various reports to find ,and use them in my large project where all the people are....): It is editing on the people such as deleting some name-var tags ,editing some divorces (which came double up after a gedcom import) ,changing some special letters in occupations etc ,confirmation without a year ,editing unknown sex, putting dates to burial info ,looking into people listed with no children ,go through notes on people (taken from an old genealogyprogram), editing marriages for people without marriagedates..... also have some records to go through from TMG 4 and get them into TMG 9 which includes......editing certain special letters after gedcom import ,children out of wedlock ,adopted ,stillbirths ,twins/triplets etc ,editing addresses ,burials. If I split my large project ,many of these people probably will not show up ,at least not with the same ID number ,and I will have to go through a dozen of projects to find where they are located for getting the chance to edit them. With that said ,I also realize it may not be advisable/safe to edit them in my large project either ,but it had been the easiest......so what should I do? Edit them in my large project and risk getting my project totally corrupted ,or split it and have a difficult time refinding all the people because of new IDs and not knowing for sure where to refind them??? When it comes to splitting ,I think maybe I should make a "close ancestor" main project that goes fom me and back to my greatgreatgrandparents ,and then a project for each of these` families. My father`s father and mother is not closely related. My mother`s parents are not related very closely either so should be no problem splitting in 2 x greatgrandparents there. My mother`s father`s parents were related from 200 years back. My mother`s father`s grandparents were 2nd cousins ,and so were two of their grandparents again ,so it would be some "double cousins" there....hopefully it would not be any problem?. Guess finding double people in 17 projects wouldn`t be too hard though....(lot of cross checking....) I know my ancestors very far back in history ,and have for curiosity taken it very far back ,that would say way before Charlemagne (only for fun and not for use..). My parents were 16th cousins I believe ,but will not be able to find their relationship with splitting the project because they will be in probably several different projects. Anyway ,I agree that it is much better to split the project I have in 16+1 different projects than having all in one with over a million people. But first ,I have to know what to do with that "editing problem"....!!. Should I keep on with editing in the large project which is the easiest or split up and letting it be harder to edit the ones I want to edit? Thanks Erik -----Opprinnelig melding----- Fra: TMG [mailto:[email protected]] På vegne av Michael J Hannah Sendt: 22. november 2017 01:48 Til: TMG-L submission address <[email protected]> Emne: Re: [TMG] 2Gb file limit Dennis Lee Bieber replied: > VFP can only address bytes 0 to 2147483647. > At best, it will refuse to save any record that would require > exceeding the 2GB boundary. > At worst, it wraps around and corrupts data ... a negative offset Actually, as a retired computer geek who was involved in the Y2K issues, wrap around can be even worse. Rather than a negative offset it is more likely to truncate, overwriting your oldest data. 2147483647+1 => 0 2147483647+2 => 1 ... Yes, Erik, you really, Really, REALLY!! need to split your project into multiple projects. You worried that: > I can`t split the project since all of the people are connected to me > somehow. > The focus with that project has been starting with me and then > entering as many ancestors I knew ,and then also go on to all I could > find of descendants of all the siblings and halfsiblings of my > ancestors etc. Well, you are now "cursed" by your success at finding so many relations. I would suggest splitting based on some level of direct ancestor. I would first split off a "close ancestor" main project trimmed to a project that only goes back on the direct line from you to some set of selected direct ancestors. The level back does not have to be consistent, it could depend upon keeping the number of people which would be included in each of the split projects roughly the same. This main project would include all these direct-line ancestors descendants, and those descendants ancestors, but not the selected ancestor's own ancestors nor their siblings. I would then split off separate non-main projects based on these selected ancestors. For example selecting all 3*great grandparents could result in 16 projects, or 6*great could be 128+1 projects, whatever level you think you might need. Each X*grandparent project would include only that person, their direct ancestors and siblings, plus all these people's siblings, descendants and ancestors. But it would not include this X*grandparent's descendants and relations since they are in the main project. Yes, you may wind up with duplicates in multiple projects if there are cross marriages among descendants or ancestors. You will have to examine the projects to find these. Then I would probably create some custom tags which would tell me something like: "All the descendants (or ancestors) of this individual may be found in project XYZ." That could let you limit that duplicate descendant or ancestor line to a single project as well as help you locate them. Hope this gives you ideas, Michael The TMG archive is found here: http://archiver.rootsweb.ancestry.com/th/index/TMG/ Instructions on how to subscribe to TMG: http://lists.rootsweb.ancestry.com/index/other/Software/TMG.html ------------------------------- To unsubscribe from the list, please send an email to [email protected] with the word 'unsubscribe' without the quotes in the subject and the body of the message
Erik, You might consider using the Reference field to store each person's ID numbers. The Reference field is searchable using TMG's Search function (from the menu: Edit / Find a Person / Search by Reference Number). Then once you split the database you can still find someone with that search. It's also possible to include the reference number in the picklist (pull up the picklist and select the Options button)... the only issue there is that the reference number is an alpha-numeric field and sorts as text rather than as a number, so you would get an order like 1, 11, 111, 2, 21, 211, 3, etc... no good for sorting but might be handy for your purposes. I'm pretty sure there's a way to use John Cardinal's TMG Utility to get the ID numbers loaded into the Reference number field. I think it involves outputting your ID numbers using a TMG List of People report, then importing using a TMG Utility function. I don't know if doing that will expand the size of the large table and corrupt your database or not. It might be safer to do the import after splitting the database, just to keep from corrupting the database, but it would take a better expert than me to guide you through that process if you need help with it. Good luck with your database! Joyce Rivette -----Original Message----- From: TMG [mailto:[email protected]] On Behalf Of Erik Hov Sent: Wednesday, November 22, 2017 4:34 AM To: 'The Master Genealogist Rootsweb Email List' Subject: Re: [TMG] 2Gb file limit Truncating and overwriting aren`t good. But what to do now? and after that, how splitting the project I have? What to do now?..... Before I think I can split anything I have another problem (which may be no problem??)....I am doing some maintenance on the people in my large project. (And to this job it is very convenient to have the IDs I have done various reports to find ,and use them in my large project where all the people are....): It is editing on the people such as deleting some name-var tags ,editing some divorces (which came double up after a gedcom import) ,changing some special letters in occupations etc ,confirmation without a year ,editing unknown sex, putting dates to burial info ,looking into people listed with no children ,go through notes on people (taken from an old genealogyprogram), editing marriages for people without marriagedates..... also have some records to go through from TMG 4 and get them into TMG 9 which includes......editing certain special letters after gedcom import ,children out of wedlock ,adopted ,stillbirths ,twins/triplets etc ,editing addresses ,burials. If I split my large project ,many of these people probably will not show up ,at least not with the same ID number ,and I will have to go through a dozen of projects to find where they are located for getting the chance to edit them. With that said ,I also realize it may not be advisable/safe to edit them in my large project either ,but it had been the easiest......so what should I do? Edit them in my large project and risk getting my project totally corrupted ,or split it and have a difficult time refinding all the people because of new IDs and not knowing for sure where to refind them??? When it comes to splitting ,I think maybe I should make a "close ancestor" main project that goes fom me and back to my greatgreatgrandparents ,and then a project for each of these` families. My father`s father and mother is not closely related. My mother`s parents are not related very closely either so should be no problem splitting in 2 x greatgrandparents there. My mother`s father`s parents were related from 200 years back. My mother`s father`s grandparents were 2nd cousins ,and so were two of their grandparents again ,so it would be some "double cousins" there....hopefully it would not be any problem?. Guess finding double people in 17 projects wouldn`t be too hard though....(lot of cross checking....) I know my ancestors very far back in history ,and have for curiosity taken it very far back ,that would say way before Charlemagne (only for fun and not for use..). My parents were 16th cousins I believe ,but will not be able to find their relationship with splitting the project because they will be in probably several different projects. Anyway ,I agree that it is much better to split the project I have in 16+1 different projects than having all in one with over a million people. But first ,I have to know what to do with that "editing problem"....!!. Should I keep on with editing in the large project which is the easiest or split up and letting it be harder to edit the ones I want to edit? Thanks Erik -----Opprinnelig melding----- Fra: TMG [mailto:[email protected]] På vegne av Michael J Hannah Sendt: 22. november 2017 01:48 Til: TMG-L submission address <[email protected]> Emne: Re: [TMG] 2Gb file limit Dennis Lee Bieber replied: > VFP can only address bytes 0 to 2147483647. > At best, it will refuse to save any record that would require > exceeding the 2GB boundary. > At worst, it wraps around and corrupts data ... a negative offset Actually, as a retired computer geek who was involved in the Y2K issues, wrap around can be even worse. Rather than a negative offset it is more likely to truncate, overwriting your oldest data. 2147483647+1 => 0 2147483647+2 => 1 ... Yes, Erik, you really, Really, REALLY!! need to split your project into multiple projects. You worried that: > I can`t split the project since all of the people are connected to me > somehow. > The focus with that project has been starting with me and then > entering as many ancestors I knew ,and then also go on to all I could > find of descendants of all the siblings and halfsiblings of my > ancestors etc. Well, you are now "cursed" by your success at finding so many relations. I would suggest splitting based on some level of direct ancestor. I would first split off a "close ancestor" main project trimmed to a project that only goes back on the direct line from you to some set of selected direct ancestors. The level back does not have to be consistent, it could depend upon keeping the number of people which would be included in each of the split projects roughly the same. This main project would include all these direct-line ancestors descendants, and those descendants ancestors, but not the selected ancestor's own ancestors nor their siblings. I would then split off separate non-main projects based on these selected ancestors. For example selecting all 3*great grandparents could result in 16 projects, or 6*great could be 128+1 projects, whatever level you think you might need. Each X*grandparent project would include only that person, their direct ancestors and siblings, plus all these people's siblings, descendants and ancestors. But it would not include this X*grandparent's descendants and relations since they are in the main project. Yes, you may wind up with duplicates in multiple projects if there are cross marriages among descendants or ancestors. You will have to examine the projects to find these. Then I would probably create some custom tags which would tell me something like: "All the descendants (or ancestors) of this individual may be found in project XYZ." That could let you limit that duplicate descendant or ancestor line to a single project as well as help you locate them. Hope this gives you ideas, Michael The TMG archive is found here: http://archiver.rootsweb.ancestry.com/th/index/TMG/ Instructions on how to subscribe to TMG: http://lists.rootsweb.ancestry.com/index/other/Software/TMG.html ------------------------------- To unsubscribe from the list, please send an email to [email protected] with the word 'unsubscribe' without the quotes in the subject and the body of the message The TMG archive is found here: http://archiver.rootsweb.ancestry.com/th/index/TMG/ Instructions on how to subscribe to TMG: http://lists.rootsweb.ancestry.com/index/other/Software/TMG.html ------------------------------- To unsubscribe from the list, please send an email to [email protected] with the word 'unsubscribe' without the quotes in the subject and the body of the message
At 11/22/2017 09:42, Joyce wrote >You might consider using the Reference field to store each person's ID >numbers. The Reference field is searchable using TMG's Search function (from >the menu: Edit / Find a Person / Search by Reference Number). Then once you >split the database you can still find someone with that search. It's also >possible to include the reference number in the picklist (pull up the >picklist and select the Options button)... the only issue there is that the >reference number is an alpha-numeric field and sorts as text rather than as >a number, so you would get an order like 1, 11, 111, 2, 21, 211, 3, etc... >no good for sorting but might be handy for your purposes. Joyce has a good idea here and also a good warning about how numbers are treated in the Reference field as opposed to how numbers are normally treated. You can get textual numbers sorted in the proper sequence IF all numbers have the same number of characters. You would do this by adding zeros prefixing the numbers -- like 001, 011, 012, 002, 021, 023, 003, etc. This would allow sorting to be normal. Of course, this makes it harder to automate plugging the ID# into the Reference Field. John Cardinal may be able to suggest something that would mitigate this problem using his TMG Utility. Lee