As you have learnt there are a lot of questions about what should be kept for a person when you delete other persons. This is not an easy aspect to manage.  A similar problem occurs when importing data, what is redundant and what is in conflict? Who judges that persons are the same? Designing software tools to make this process easier is not easy. My first rule would be to keep all persons who are witnesses to a retention person's Event, even more strongly those persons that are involved more than one event that involve your prime retention population (1 degree of separation) - but how do you automate that? RobinL On 18-Feb-19 02:31 PM, Richard Damon wrote: > On 2/17/19 10:52 PM, Erik Hov wrote: >> Thanks for all help I have gotten with how to delete people from my large >> project!. >> I have not actually tried to move people yet. >> I am a bit concerned if moving a person also removes the person`s >> parent/parents ,so I will try to move two or three people first to see if >> it`s really true that the selected person`s parents also automatically are >> being moved. If that happens ,I must reconsider what to do....even if I >> really don`t see any other option for deleting those 121000 people than >> using the move people feature...... >> Erik >> > The parents shouldn't be removed as people, but the parental > relationship (the relationship tag between them) will be removed. > -- Robin Lamacraft, Adelaide, Australia --- This email has been checked for viruses by Avast antivirus software. https://www.avast.com/antivirus

<As you have learnt there are a lot of questions about what should be kept for a person when you delete other persons. This is not an easy aspect to manage.< I have worked on this setting of flags in thinking that those I am setting flags to are those who will be deleted. I am pretty sure I have not forgotten anyone. I didn`t know that possibly ALSO the parent of any selected person will be deleted when using the move feature.....(will try later today). <A similar problem occurs when importing data, what is redundant and what is in conflict? Who judges that persons are the same?< When doing an import (at least a gedcom which I haven`t done for many years) you don`t get much choice of determining if persons are the same. Not with the import genBridge into TMG either. Guess you have to look for doublegangers within the TMG for that. <Designing software tools to make this process easier is not easy. My first rule would be to keep all persons who <are witnesses to a retention person's Event, even more strongly those persons that are involved more than one <event that involve your prime retention population (1 degree of separation) - but how do you automate that?< Hmmmm.....not sure whow to answer this! Originally I had about 183000 people with the AAS=Y flag that I was able to make a new project of from the List of People report. A Similar list from LoP with 874000 of the AAS=N didn`t work. Anyway ,of those 183000 I set many of the people back to AAS=N ,exactly because they were married to people or had children that was married to people not belonging to the AAS=N people in the project that I wanted to keep. After doing that ,I am left with those beforementioned 121000 people that doesn`t need to be in the main project. I guess most of these 121000 are involved in more than one event on the ones I shall keep ,but since they are only descended from or married to AAS=Y people/relatives ,I don`t need them to be in the project anymore. If anything should go wrong. I have a backup of the project ,and also have all of it in another project with a different name in case the backup shouldn`t work.

At 2/18/2019 06:47, Erik Hov wrote >After doing that ,I am left with those beforementioned 121000 people that >doesn`t need to be in the main project. I guess most of these 121000 are >involved in more than one event on the ones I shall keep ,but since they are >only descended from or married to AAS=Y people/relatives ,I don`t need them >to be in the project anymore. If anything should go wrong. I have a backup >of the project ,and also have all of it in another project with a different >name in case the backup shouldn`t work. If your project is like mine, most people are connected in some way with the rest of the project. That connection may be direct in the sense of parent/child through multiple generations including cousins and aunts/uncles. Or the connection may be by marriage. The third type os connection is as a Witness - minister/best man.maid of honor at a wedding, a business associate, or any other event in which multiple people are involved. Thus, trying to identify persons that have no connection with others in the project is almost like hunting for the proverbial needle. In many cases, you almost need to review each candidate individually for removal from the project. If you are lucky, that person will have no connection and can easily be deleted or marked for deleted (Flag set). For many others, there will be a connection. In such cases, you need to decide whether they will be deleted/marked for deletion and, if so, the best option would then be to remove the connection first. This would then resolve any potential problem that might occur during the deletion. Note that if you attempt to delete a person with direct connections then TMG will notify you that there are certain connection and dis-allow the deletion. Then, if you still want to delete that person, you will need to remove that connection (Tag) first. Then, try again. During the creation of a new project via the Secondary Output of the List of People report, there is no such warning since you aren't really deleting anyone from the original project, but copying the persons to be retained to the new project. Finding persons with direct connection is fairly easy. No Relatioship Tags to parent, and no Marriage Group Tags will find the great majority. Other connections are less easy to identify via a filter. A thorough knowledge of the project will suggeswt other poaaible connections and ways to filter for them. Still other may need to be individually reviewed. You have, so far, deci8ed that 121000 persons are not longer needed in your project, The only way to remove them (other than one at a time) is via the List of People=>Secondary Output=>Create New Project option. In this method, there may be connections lost since the connected persons are not copied to the new project. In my experience, those not copied would not be that many. My guess would be that those mpt copied would not be more than a few hundred. In your case, it might be somewhat larger -- approaching a thousand. Many of these can be identified quickly and easily added to the new project. Still others might not be noticed for a long time. In regard to your problem of getting errors running the List of People report =>Secondary Output option, let me note what I recall of what you have had happen. 1. Your main project is at or nearing the 2GB limit imposed by FoxPro (the underlying program that TMG uses). 2. You have created one or more new projects from that "old" project, with the new project containing a relatively small number each (three/four thousand people). There was no problems doing this. 3. You have tried creating a new project by copying the old project minus the 121000 people that you no longer want in the project. This has failed. Since #2 above worked bu #3 did not, it suggests to me that you are running up again the 2GB limit in #3. In such a case, it appears to me that you will not be able to create a new project via the #3 method. Thus, you are limited in how you can reduce the size of your project to well below the 2GB limit. There is the possibility of someone very knowledgeable of the TMG file structure working with each project file (there are about fifty excluding the index files) and deleting the desired records. The problem with this method is that even a very knowledgeable person may miss deleting certain record or that too many records are deleted. The inter-relationship of the fifty files makes such work very hard. If everything gos as planned, it would work, but the odds are that something will go wrong. The only loss would be of the person's time as you would still have your original project. Note that the person doing the work here would have to know exactly who to delete and the work would probably be more extensive than the only other option. This only other option (in my opinion) is the safest and best. But this is like saying that a dentist drilling on your teeth without ansthesia is easily withstood. As I see it, your only real option is to individually delete each of the 121000 persons. As I said before, at some point in this process, the project (following Optomization) will be small enough that you can continue the rest using the List of People report method. This is not a happy solution, but I think it is what you will have to do. Lee

<If your project is like mine, most people are connected in some way with the rest of the project. That connection <may be direct in the sense of parent/child through multiple generations including cousins and aunts/uncles. Or <the connection may be by marriage. The third type os connection is as a Witness - minister/best man.maid of <honor at a wedding, a business associate, or any other event in which multiple people are involved. Thus, trying <to identify persons that have no connection with others in the project is almost like hunting for the proverbial <needle.< Most of the connections are of the two first types. I don`t really know if I have any people entered as witnesses etc ,at least I haven`t entered any such witnesses as I can recall. In many cases, you almost need to review each candidate individually for removal from the project. If you are lucky, that person will have no connection and can easily be deleted or marked for deleted (Flag set). For many others, there will be a connection. In such cases, you need to decide whether they will be deleted/marked for deletion and, if so, the best option would then be to remove the connection first. This would then resolve any potential problem that might occur during the deletion. <Note that if you attempt to delete a person with direct connections then TMG will notify you that there are <certain connection and dis-allow the deletion. Then, if you still want to delete that person, you will need to <remove that connection (Tag) first. Then, try again.< Yes ,but this is if you use the delete person feature on the edit menu ,right? ,and not the move people feature?. Or is it both? <Finding persons with direct connection is fairly easy. No Relatioship Tags to parent, and no Marriage Group Tags <will find the great majority. Other connections are less easy to identify via a filter. A thorough knowledge of the <project will suggeswt other poaaible connections and ways to filter for them. Still other may need to be <individually reviewed.< I haven`t disconnected any of the AAS=Y to the AAS=N people. I marked the AAS flags with Y when going through them manually ,and also set a lot of the flag values back to N when going through them again to find out which people I will keep in the project. <You have, so far, deci8ed that 121000 persons are not longer needed in your project, The only way to remove <them (other than one at a time) is via the List of People=>Secondary Output=>Create New Project option. In this <method, there may be connections lost since the connected persons are not copied to the new project. In my <experience, those not copied would not be that many.< Now ,I am confused on two things. First I don't know if the move people feature actually also moves the parent of the selected people to be deleted. Second ,I thought you just said that the List of People.....create new project option doesn`t delete any people? And then you say above that "The only way to remove them (other than one at a time) is via the List of People=>Secondary Output=>Create New Project option."....... <In regard to your problem of getting errors running the List of People report =>Secondary Output option, let me <note what I recall of what you have had happen. < 1. Your main project is at or nearing the 2GB limit imposed by FoxPro (the underlying program that TMG uses). < 2. You have created one or more new projects from that "old" project, with the new project containing a <relatively small number each (three/four thousand people). There was <no problems doing this. < 3. You have tried creating a new project by copying the old project minus the 121000 people that you no longer <want in the project. This has failed. <Since #2 above worked bu #3 did not, it suggests to me that you are running up again the 2GB limit in #3. In such <a case, it appears to me that you will not be able to create a new project via the #3 method. Correct on all three!. At least I know that I can`t create any project of the 875000 I am going to keep. I am fully able to copy whole projects (File-Copy project) ,but not parts of projects on the List of People report. <This only other option (in my opinion) is the safest and best. But this is like saying that a dentist drilling on your <teeth without ansthesia is easily withstood. As I see it, your only real option is to individually delete each of the <121000 persons. As I said before, at some point in this process, the project (following Optomization) will be small <enough that you can continue the rest using the List of People report method.< I tend to agree with you that that is the safest way to be completely sure that nothing will be changed unless I know it. I did create a new project project from the main project and called it "TEST" it has all of the AAS=Y and AAS=N people in it. I created a dataset within that project and I am now trying the move people feature. When making the new dataset I chose to have all the places and other features in the new dataset as well. Then I tried to move one person to the other dataset to see how it went. So far TMG seems to be at a standstill at 27% on the placenamefile ,and being there for soon 10 minutes. I will try to create a dataset without including any datatypes and see if that is any clue.... _______________________________________________ Email preferences: http://bit.ly/rootswebpref Unsubscribe https://lists.rootsweb.com/postorius/lists/tmg@rootsweb.com Privacy Statement: https://ancstry.me/2JWBOdY Terms and Conditions: https://ancstry.me/2HDBym9 Rootsweb Blog: http://rootsweb.blog RootsWeb is funded and supported by Ancestry.com and our loyal RootsWeb community

At 2/18/2019 15:46, Erik Hov wrote >>Note that if you attempt to delete a person with direct connections then >>TMG will notify you that there are <certain connection and dis-allow the >>deletion. Then, if you still want to delete that person, you will need to >><remove that connection (Tag) first. Then, try again. > >Yes ,but this is if you use the delete person feature on the edit menu >,right? ,and not the move people feature?. Or is it both? Yes, the warning is only when using the Edit=>Delete Person(s) feature. >>Finding persons with direct connection is fairly easy. No Relatioship Tags >>to parent, and no Marriage Group Tags will find the great majority. Other >>connections are less easy to identify via a filter. A thorough knowledge of >>the project will suggeswt other poaaible connections and ways to filter for >>them. Still other may need to be individually reviewed. > >I haven`t disconnected any of the AAS=Y to the AAS=N people. I marked the >AAS flags with Y when going through them manually ,and also set a lot of the >flag values back to N when going through them again to find out which people >I will keep in the project. Regardless of the Flag Setting, if a person connected to someone of the opposite setting is copies to a new project then the connection is lost. ><You have, so far, deci8ed that 121000 persons are not longer needed in your >project, The only way to remove them (other than one at a time) is via the >List of People=>Secondary Output=>Create New Project option. In this >method, there may be connections lost since the connected persons are not >copied to the new project. In my experience, those not copied would not be >that many.< > >Now ,I am confused on two things. >First I don't know if the move people feature actually also moves the parent >of the selected people to be deleted. No, only the moved person(s) are moved. A moved son will lose his father unless the father is also moved. >Second ,I thought you just said that the List of People.....create new >project option doesn`t delete any people? And then you say above that "The >only way to remove them (other than one at a time) is via the List of >People=>Secondary Output=>Create New Project option."....... No, the List of People=>Create new project does not delete anyone in the _original_ project. Only the persons copied to the new project will be in the new project -- effectively makeing the people from the old project that are NOT moved to appear as having been deleted from the new project when compared with the old project. For example, you have a project with ten people numbered #1 through #10. You create a new project based on a Flag set to Y for the even numbered persons (the even numbered people are copied to the new project, The new project will have the following: Name Old# New# A 2 1 B 4 2 C 6 3 D 8 4 E 10 5 >I did create a new project project from the main project and called it >"TEST" it has all of the AAS=Y and AAS=N people in it. I created a dataset >within that project and I am now trying the move people feature. When making >the new dataset I chose to have all the places and other features in the new >dataset as well. Then I tried to move one person to the other dataset to see >how it went. So far TMG seems to be at a standstill at 27% on the >placenamefile ,and being there for soon 10 minutes. I will try to create a >dataset without including any datatypes and see if that is any clue.... The problem with moving persons to a different data set is that you are still in the same project and still subject to the 2GB limit. However, this may or may not be an issue if you periodically run the Maintenance routibe especially the Optionze routine. Note that when a person(s) is/are moved to another data set (say from data set 1 to data set 2), while that effective removes (deletes) the person(s)from the original data set (#1), the physical deletion from data set @1 does not take place until the projecty is Optimized. Thus the project files will increase in size until the project is Optimized. So, starting with a project that has a size of 4096318 bytes and moving say ten persons, you will have a project size of say 4098525 bytes. If you then Optimize, the size of the project will again be 4096318 bytes (or within a few bytes of that as there may be some adminstrative data that is involved that hasn;t been taken into account. And at 2/18/2019 15:46, Erik Hov wrote > made a test with the TEST project moving people to another dataset within >the project. I tried on 5 people at first ,and after some time it worked. >I got them over in the other dataset ,but after deactivating the main TEST >dataset and just having up the other dataset ,it took a very long time >waiting for example the F2 picklist to appear. >Of those 5 people ,I had parents on 4 of them (and those parents wasn`t >moved). Only the 5 I chose to be moved was moved ,no one else...... I assume that the TEST project is a copy of your original project. As I noted beforre, being so close to the 2GB limit of FoxPro can have a large impact on this. That would be what I noted, that only the five people are moved. Keep in miond here what I said above that "moved persons" from data set #1 remain in data set #1 as well as being copied to data set #2 UNTIL you Optinmize the project. That is, the persons that were moved are MARKED as deleted and thus can't be seen anymore in TMG but are still in the files. Optimize actually removes records that are MARKED as deleted. Lee

<Regardless of the Flag Setting, if a person connected to someone of the opposite setting is copies to a new project <then the connection is lost. Correct ,I found out when I tried to move people in the TEST project. By having to set flags manually because of the size ,I have also been able to set flags to the right people ,so that I can be quite confident that the AAS=N people are the ones that I shall keep plus also run several reports to find people I might have missed. <No, only the moved person(s) are moved. A moved son will lose his father unless the father is also moved.< That was what the move person feaure in the TEST project showed me. I selected 5 people ,and it was only 5 who was moved. I am now trying a larger groupt just to see how it seems... <For example, you have a project with ten people numbered #1 through #10. You create a new project based on <a Flag set to Y for the even numbered persons (the even numbered people are copied to the new project, The <new project will have the following: <Name Old# New# < A 2 1 < B 4 2 < C 6 3 < D 8 4 < E 10 5 How does this add up? Does this mean that by creating a new project from the List of People-Create new project ,the new project gets new Idnumbers?? <The problem with moving persons to a different data set is that you are still in the same project and still subject <to the 2GB limit. However, this may or may not be an issue if you periodically run the Maintenance routibe <especially the Optionze routine.< That was how I did the deleting of people when doing my first split. I did optimize all the time after deleting about 3000 people at a time. I have to do the same now ,optimizing al the time. I am very aware of the 2 GB limit ,and I should have split my project a long time ago. Playing around with the TEST project now to find out about things. <I assume that the TEST project is a copy of your original project. As I noted beforre, being so close to the 2GB limit of FoxPro can have a large impact on this.< Yes ,it is a copy of my original project. After moving about 2500 people from the TEST project to the other dataset within the project ,I deleted the dataset and I have just now started optimizing. It now has about 119800 AAS=Y people.

At 2/18/2019 18:04, Erik Hov wrote >How does this add up? Does this mean that by creating a new project from the >List of People-Create new project the new project gets new Idnumbers?? Sorry, no, that was misleading. The ID#s remain the same. ><The problem with moving persons to a different data set is that you are >still in the same project and still subject <to the 2GB limit. However, >this may or may not be an issue if you periodically run the Maintenance >routibe <especially the Optionze routine.< > >That was how I did the deleting of people when doing my first split. I did >optimize all the time after deleting about 3000 people at a time. I have to >do the same now ,optimizing al the time. So, instead of say 874000 people in the project, you now have say 871000 people? That seems to be a step in the right direction toward your goal. Lee

I made a test with the TEST project moving people to another dataset within the project. I tried on 5 people at first ,and after some time it worked. I got them over in the other dataset ,but after deactivating the main TEST dataset and just having up the other dataset ,it took a very long time waiting for example the F2 picklist to appear. Of those 5 people ,I had parents on 4 of them (and those parents wasn`t moved). Only the 5 I chose to be moved was moved ,no one else......

At 2/14/2019 7:39 PM, Erik Hov wrote: >Of course!. >Then I have the same , Details, Children, Siblings (and possibly flag >window >,but don`t remember). >Doing a LoP report (not to screen...suppressing....) gives of course the >Report Definition Screen and a Progress window pop-up, followed by a >confirmation for copying to new project. The last window disappearing >when >the progress Window appears. > I'm out of ideas then (short of rebooting the computer, and opening the task manager to monitor how much memory TMG is sucking up while running -- as a 32-bit program, it will be limited to 2GB of user RAM)

When running TMG uses about 438748 kB memory (not working on any report or anything....just running). -----Opprinnelig melding----- Fra: Dennis Lee Bieber [mailto:bieber.genealogy@earthlink.net] Sendt: 14. februar 2019 20:42 Til: The Master Genealogist Rootsweb Email List <tmg@rootsweb.com> Emne: [TMG] Re: Error message trying to make a new project At 2/14/2019 7:39 PM, Erik Hov wrote: >Of course!. >Then I have the same , Details, Children, Siblings (and possibly flag >window ,but don`t remember). >Doing a LoP report (not to screen...suppressing....) gives of course >the Report Definition Screen and a Progress window pop-up, followed by >a confirmation for copying to new project. The last window disappearing >when the progress Window appears. > I'm out of ideas then (short of rebooting the computer, and opening the task manager to monitor how much memory TMG is sucking up while running -- as a 32-bit program, it will be limited to 2GB of user RAM) _______________________________________________ Email preferences: http://bit.ly/rootswebpref Unsubscribe https://lists.rootsweb.com/postorius/lists/tmg@rootsweb.com Privacy Statement: https://ancstry.me/2JWBOdY Terms and Conditions: https://ancstry.me/2HDBym9 Rootsweb Blog: http://rootsweb.blog RootsWeb is funded and supported by Ancestry.com and our loyal RootsWeb community

At 2/14/2019 15:01, Erik Hov wrote >When running TMG uses about 438748 kB memory (not working on any report or >anything....just running). My tests running a full "copy" of a 66 thousand person project using the List of People report Secondary Output is that TMG uses aboiut 363 MB of memory and about 15% of the CPU. These numbers vary throughout the process, but not by much. Most of the time, the Task Manager reports TMG as "not responding" which means that that various Windows processes are taking place and TMG is awaiting those processes (e.g., the computer is thinking about things. It took about ten minutes to create the new project while nothing else was being done on the computer except background mail checks). For comparison, doing the same with a 10 thousand person project, the Task Manager showed much the same. The difference was that the new project was completed in about a minute. This is on a Win 10 -32bit system. A 64bit system would likely take less time and resource use would be different. The number of open windows appears to make no difference. Lee

Lee wrote: <My tests running a full "copy" of a 66 thousand person project using the List of People report Secondary Output is <that TMG uses aboiut <363 MB of memory and about 15% of the CPU. These numbers vary throughout the process, but not by much. <Most of the time, the Task Manager reports TMG as "not responding" which means that that various Windows <processes are taking place and TMG is awaiting those processes (e.g., the computer is thinking about things. It <took about ten minutes to create the new project while nothing else was being done on the computer except <background mail checks). <For comparison, doing the same with a 10 thousand person project, the Task Manager showed much the same. <The difference was that the new project was completed in about a minute. <This is on a Win 10 -32bit system. A 64bit system would likely take <less time and resource use would be different. The number of open windows appears to make no difference. When I tried to make a new project from the List of People report secondary output (from 874000 of the 996000 people) I have in my project (those with flag listed as AAS = N) ,the TMG was also most of the time not responding. 10 minutes to create the new project....well ,my report run for 1 1/2 day and when the error message came ,TMG stopped working at 11% on the namefile. (I have Win 7 ,64 bit system) TMG is also working now (maintenance/Validating) and task manager tells TMG to use about 729000 kB memory and abt 12-13% of CPU. Seems to me ,I have to look for other possibilities for deleting those ca 121000 AAS = Y people in my project.....

At 2/14/2019 19:09, Erik Hov wrote >TMG is also working now (maintenance/Validating) and task manager tells TMG >to use about 729000 kB memory and abt 12-13% of CPU. >Seems to me ,I have to look for other possibilities for deleting those ca >121000 AAS = Y people in my project..... The only other possibility to my thinking is to manually delete them 'one at a time'. But for 121000 people, that is somewhat extreme. Still, using that method to reduce the overall size of the project a small amount at a time until it is at a point where the List of People report method works may be about the only option. The ides would be to identify small family groups and delete them. Then at some point where you have deleted a number of these groups and optimized the project, try the List of People method. Theoretically, at some point, the project will be small enough to allow the satifactory completion of the List of People method. I suspect there is another faster method, but it would be very complex and would require a thorough understanding of the structural design of TMG projects and files and how they all tie together as well as the specialized software to do the work. Lee

Lee. Delete the 121000 one at a time would be extreme ,and take a long time. I agree that it would reduce the size of the project too. But I can`t identify small familygroups and delete them ,because some of the descendants of the people with AAS =Y are married to or have children with people with AAS = N flag people ,and I am going to keep the AAS N people ,so that`s not an option for me. Since all the people in the project have an AAS flag with value Y or N ,I have to somehow use the marked AAS Y people and delete them somehow. Any ideas on that would be most welcome!. If it had been possible to mark let`s say 50000 AAS Y people at a time ,make a new project of those ,and then delete that project with the 50000 ,and then continue with that until all 121000 are deleted?.....but just getting them to a project doesn`t delete them from the large project..... -----Opprinnelig melding----- Fra: Lee Hoffman [mailto:azchief@bellsouth.net] Sendt: 15. februar 2019 01:39 Til: The Master Genealogist Rootsweb Email List <tmg@rootsweb.com> Emne: [TMG] Re: Error message trying to make a new project At 2/14/2019 19:09, Erik Hov wrote The only other possibility to my thinking is to manually delete them 'one at a time'. But for 121000 people, that is somewhat extreme. Still, using that method to reduce the overall size of the project a small amount at a time until it is at a point where the List of People report method works may be about the only option. The ides would be to identify small family groups and delete them. Then at some point where you have deleted a number of these groups and optimized the project, try the List of People method. Theoretically, at some point, the project will be small enough to allow the satifactory completion of the List of People method. I suspect there is another faster method, but it would be very complex and would require a thorough understanding of the structural design of TMG projects and files and how they all tie together as well as the specialized software to do the work. Lee _______________________________________________ Email preferences: http://bit.ly/rootswebpref Unsubscribe https://lists.rootsweb.com/postorius/lists/tmg@rootsweb.com Privacy Statement: https://ancstry.me/2JWBOdY Terms and Conditions: https://ancstry.me/2HDBym9 Rootsweb Blog: http://rootsweb.blog RootsWeb is funded and supported by Ancestry.com and our loyal RootsWeb community

At 2/14/2019 21:02, Erik Hov wrote >Delete the 121000 one at a time would be extreme ,and take a long time. >I agree that it would reduce the size of the project too. I don't mean to delete all 121000, but a small number at a time -- say 100-200. Then Optimie and try creating the new project. That may work, but if not try another 100-200. Keep this up until the new project is created. Still not a happy option, but all I can think of. The group may not be parents and their children. It just means a group that tends to be connected in some way. The connection may only be that the AAS Flag equals Y. >But I can't identify small familygroups and delete them ,because some of the >descendants of the people with AAS =Y are married to or have children with >people with AAS = N flag people ,and I am going to keep the AAS N people ,so >that`s not an option for me. Wait a minute!! If the project creation you just tried had worked, those people would have been deleted anyway. You may need to reconsider what you had planned. You may want to go back and change some of those AAS=Y to AAS=N. So you may not have 121000 to delete after all. >If it had been possible to mark let`s say 50000 AAS Y people at a time ,make >a new project of those ,and then delete that project with the 50000 ,and >then continue with that until all 121000 are deleted?.....but just getting >them to a project doesn`t delete them from the large project..... Right. This was part of the design of TMG. The resoning is that "global" deleting of project members is frauight with possible problems -- primarily being that the wrong persons get deleted and that is not noticed until later after much more work has been done. Lee

Lee wrote: <I don't mean to delete all 121000, but a small number at a time -- say 100-200. Then Optimie and try creating <the new project. That may work, but if not try another 100-200. Keep this up until the new project is created. <Still not a happy option, but all I can think of. Not a happy option ,and it seems I have a hard decision to make about this....hmmm. Last time I split out a project ,I had the same problem with deleting the "Y" people. Then ,I had about 20000 to delete ,and that didn`t work sending to a project either (if I recall right). What I did back then was to send about 3000 to a dataset(!) ,and then delete the dataset after each time ,and then optimize (I don`t remember if I also run validate and optimize after that). I don`t remember correctly ,but I might have put the "Y" people into the focusgroup or project explorer ,marked about 3000 of them and then put them into a dataset. Just a thought...........Since I have several MUCH smaller projects.....would it be possible to delete all of those 121000 (or a portion at a time) by sending them to a dataset within ANOTHER much smaller project??? (That`s until my large project is small enough to send flagged people to a new project from the LoP) <Wait a minute!! If the project creation you just tried had worked, those people would have been deleted <anyway. You may need to reconsider what you had planned. You may want to go back and change <some of those AAS=Y to AAS=N. So you may not have 121000 to delete after all. They wouldn't have been deleted right away because they would still be in the large project while the AAS=N would be in the new project. But as you said....if the new project had been created correctly ,I would have deleted the old large one ,and "got rid of" those 121000. I could also as you said change some of those Y people to N ,but it would be some steps back. Also it had to be set back to AAS=Y anyway....and I would also need to write down Id numbers for all those I had changed into AAS=N in order to change them back to Y. <The resoning is that "global" deleting of project members is frauight with possible problems -- primarily being <that the wrong persons get deleted and that is not noticed until later after much more work has been done. Yes ,I can see that one.... Since I have an accentcolor on the AAS=Y people I guess I would notice if any AAS=N should be deleted if I took a look at all of the ones to be deleted in the focusgroup or project explorer.