Have you ever wished to be able to look at all the facts in your family tree database that happened within a day’s horseride of a certain location? Until RootsMagic contains such a report, this adaptation of the LifeLines query helps you view your events for any geographic area in addition to looking at the lifeline of any person in your database.
Edit the SQL file to set the coordinates of your target location and the distance from it you want included. Set the values to (0,0,12450) to include all facts for everywhere. As coded near the end of the query, the results will be sorted by RIN and SortDate so that each person’s events are all together sorted by timeline. With SQLiteSpy et al, you can override that sort by clicking on a column header. Click on SortDate to mix people’s events in a pure timeline sort that may reveal some interesting connections.
Compare EXIF, IPTC, XMP metadata stored in image files to the Caption, Description and other metadata stored in a RootsMagic 5 database for image files linked to its Media Gallery and copy between the files and the database.
The principles and procedures outlined here, while based on the way Picasa uses metadata stored in image files, are applicable with appropriate modifications to relating other photo managers to RootsMagic 5.
Compare RM5 and Image File Metadata
Introduction
I use Picasa to manage my collection of photographs and other media and often write captions in Picasa’s picture editor for photos that I upload to Picasaweb online albums. Of course, working in RootsMagic, I get to write captions and descriptions for some of the same photos and for many others that I may not have captioned in Picasa. LessTX and others have asked that RM have some capability of using metadata fields stored in the image files as a way of loading values into RootsMagic or storing them in the image (e.g., Insert Image Metadata into Database and some of the results in this search of the RootsMagic Forums). This sounded attractive on the surface and so I have looked into it further.
Description
One result can be seen in this screenshot of a query in SQLite Expert Personal (which displays the thumbnail images, unlike most other SQLite managers):
Screenshot comparing RootsMagic media captions and descriptions to those stored by Picasa in the image file itself.
Under the left red box is the media caption and description metadata stored in the RootsMagic 5 database in the MultiMediaTable, along with the thumbnail image, image file name and path. Under the right hand box are the values extracted by ExifTool from the Caption-Abstract and Description metadata fields stored in the image file by Picasa. With this convenient comparison of the two sources of the image metadata, I can readily see discrepancies between the two and take any needed corrective action in the corresponding application.
Some Observations and Reservations re Automatic Synchronisation
I’ll describe how I developed this query after discussing some of my observations and reservations about automatically synchronizing the captions and descriptions between the RootsMagic database and the images.
Picasa ver 3.8.0 and RootsMagic 5.0.1.0
Picasa stores the same value in both the Caption-Abstract and Description tags in the image (but I cannot account for one example above of several that have a Caption-Abstract value but not Description – perhaps it was written by an earlier version). Therefore, there cannot be a 1:1 relationship between the RootsMagic metadata fields and the image’s metadata fields.
Neither RootsMagic nor Picasa have a low limit on the length of the string that can be stored in their respective places. However, the RootsMagic caption should be constrained to be quite short because so little space is provided for printing it in the one place it is used, a Scrapbook report. If the Description is also included in the Scrapbook, then both Caption and Description values must be severely limited. Picasa displays captions most generously in its slideshows, less so in the Photo Editor and to 16-60 characters under its variably zoomed thumbnails.
Given these significant differences, I cannot see a way of automatically copying from the metadata stored in the image file to the RM fields, without following up to edit them into appropriate sizes. The reverse may be more successful: concatenate the RootsMagic fields and copy the same value to both Caption-Abstract and Description tags in the image file.
That said, the use of the image file in a Picasa Album might dictate a quite different caption from that used in a RootsMagic Scrapbook report (just as some users were dismayed to discover that in upgrading from RM4 to RM5 they had lost the ability to have a different caption for the same image when associated with different persons, events, etc. – the usage dictates the caption).
Therefore, caption copying must be manually controlled, except for copying to blank fields which, nonetheless, is liable to require manual follow-up.
Although these observations are specific to RootsMagic 5-7.6 and Picasa, the RM constraints also apply in the general case and the conclusions may not be significantly altered, except for the possible maintenance of the separation between Caption and Description.
I have not explored whether there is any appropriate image metadata tags that may correspond with RM’s media Date and Reference Number.
Installation
Copy the windows command line shell sqlite3.exe from the sqlite-shell-win32-x86-3nnnnnn.zip file downloaded from the Precompiled Binaries for Windows section of sqlite.org to a folder of your choosing. The batch file below expects it to be in C:\Program Files (x86)\SQLite but that can be revised to suit.
Install exiftool.exe from ExifTool in a folder along with all of the related batch files and sql scripts. It could be located elsewhere if you revise the batch file accordingly.
Create a shortcut to the batch file RM5comparePicasa.bat with the Target property [“C:\yourpath\RM5comparePicasa.bat” “C:\yourpath\yourRMdatabase.rmgc”], enclosing the database path in double quotes. Name the shortcut something like “RMcomparePicasa yourdatabase“. Make such a shortcut for each database you want to review.
When you click on the shortcut, it launches the batch (or Windows command) file with the path to your database file passed to it in the %1 register. The batch file deletes the old files it created in a previous run and then proceeds to build a short sql script GetImageListFromRM.sql. Then it calls sqlite3 to open your database file and run the script which outputs the results of the query to the ExifTool arguments file RMimagefiles.args, to which it then appends the -execute argument. The args file contains the list of JPG files with full pathnames recorded in the RM5 database Media Gallery. [1]
Next the batch file calls ExifTool to go through the list of image files, find them if it can, and extract the values from the Caption-Abstract and Description tags along with the paths and file names to a new file, ImageDescriptions.txt. There’s a pause to review any possible error reports in the command window, e.g., file not found. [2]
Now the batch file calls sqlite3 again to open or create the local database RM-Media.db3, creates PicasaTable with four columns matching those in the tab-delimited ImageDescription.txt and imports the latter into the table. [3]
That’s it for the batch file, except to generate instructions as to what to do next: Open RM-Media.db3 with your SQLite manager (SQLite Expert Personal if you want to see the thumbnails), copy and run some SQLite statements the batch generated. [4]
If everything worked as intended, you should see results like this screenshot (adjust column widths to suit, of course).
[Inline comment: ve3meo Jan 22, 2012
There is a flaw in the above to be corrected if there are multiple image files with the same file name in different paths. It has been correctly done in the RM5copyMetaDataToRM query.]
Copy RM Image Metadata to Image File (Picasa compatible)
Description
The batch, SQLite, ExifTool procedure copied the RM5 caption and description to the Caption-Abstract metadata tag in the image file, the result of which is seen as a caption in the Picasa Photo Editor.
This is basically a query in RM-Media.db3 (created and updated by the Compare utility with the RM5 database attached) to extract the list of files for which RM has a caption and the image file does not, along with the concatenation of caption and description values, and getting those into a file in the format required for ExifTool arguments with the appropriate arguments to tell ExifTool to write the value into the Caption-Abstract tag of the image files, and then run ExifTool against the args file. All this is managed by a batch file as was done for the Compare procedure. Here is an example showing how the RM5 Caption + Description end up being displayed as a Picasa caption.
This was a very tricky procedure to finally get going and resulted in considerable hair loss! The arguments for ExifTool must be one per line so carriage return/line feed controls (CR/LF) in the RM5 values must be replaced. Thus the white space above the URL in the RM Description disappears in the Picasa caption, the CR/LFs being replaced by periods.
This same difference applies in using ExifTool to copy the metadata from the image to RM: ExifTool itself substitutes a period for CR/LF embedded in the metadata values. Thus metadata values on either side containing CR/LF cannot be exactly replicated on the other, at least with ExifTool as the intermediary.
Installation
sqlite3.exe and exiftool.exe are already installed as above
[Inline comment: ve3meo Jan 22, 2012 There is a flaw in the above to be corrected if there are multiple image files with the same file name in different paths. It has been correctly done in the RM5copyMetaDataToRM query.]
/* RM5copyMetadataToImageFileExifToolArgList.sql
*/SELECT'-Caption-Abstract='||REPLACE(CAST(RM.Caption AS TEXT),x'0D0A','.')||' '||REPLACE(CAST(RM.Description AS TEXT),x'0D0A','.')||CAST(X'0D0A'AS TEXT)||
RM.MediaPath
||
RM.MediaFile
||CAST(X'0D0A'AS TEXT)||'-execute'|| Pic.ROWID
||CAST(X'0D0A'AS TEXT)AS Args
FROM MultiMediaTable AS RM
INNERJOIN PicasaTable AS Pic
ON RM.MediaFile COLLATE NOCASE = Pic.MediaFile
WHERE(RM.Caption ISNOTNULLOR RM.Description ISNOTNULL)AND(Pic.Caption LIKE'-'AND Pic.Description LIKE'-')
;
Create a shortcut to the batch file RM5copyMetadataToImageFile.bat with the Target property [“C:\yourpath\RM5copyMetadataToImageFile.bat” “C:\yourpath\yourRM5database.rmgc”] and name it something like “RM5copyMetadataToImageFile – yourdatabase“. Make such a shortcut for each database you want to copy captions from.
ECHO OFF
REM RM5copyMetadataToImageFile.bat
REM
REM Build sql file
ECHO ATTACH DATABASE '%~1' AS RM ; > RM5copyMetaData.sql
TYPE RM5copyMetadataToImageFileExifToolArgList.sql >> RM5copyMetaData.sql
"C:\Program Files (x86)\SQLite\sqlite3.exe" RM-Media.db3 < RM5copyMetaData.sql > RM5copyMetaDataToImageFile.args
ECHO .
ECHO Check RM5copyMetaDataToImageFile.args before proceeding to write to the image files
PAUSE
exiftool -@ RM5copyMetaDataToImageFile.args > RM5copyMetaDataToImageFile.log
ECHO .
ECHO Finished copying captions to image files
PAUSE
END
Procedure
First run the Compare utility against your database to ensure that the RM-Media.db3 is up-to-date with the current metadata from all of the images used by the database. Then click on the shortcut “RM5copyMetadataToImageFile – yourdatabase“. The Windows Command console will open and, almost immediately, you will asked to
Check RM5copyMetaDataToImageFile.args before proceeding to write to the image files
Open the file (in the folder common to all of the .bat and .sql files you have installed as instructed above) with a non-wrapping text editor and note that each file is addressed by a triplet of three lines and an empty line after as:
-Caption-Abstract=Obituary - Colin xxx North Bay Nugget - Aug 13, 1956
C:MyDocsFamilyTreeHolxxxScrapbookxxxxxxn, Colin - obituary - 1956.jpg
-execute2
-Caption-Abstract=Registration of birth of Helen xxx 13
C:MyDocsFamilyTreeHolxxxScrapbookxxxxxxth, Helen Elizabeth - Birth Register xxx13.jpg
-execute3
...
There should be nothing else in this args file than the triplets of single lines as above. The number following -execute is the PicasaTable rowid, which may be useful in tracking down a problem if ExifTool encounters a bad image file (I think it will report that number in its error message).
Then, press any key to launch ExifTool to work on the list of files. ExifTool copies each file to a .jpg_original file and re-writes the .jpg file with the new Caption-Abstract value added.
After some time, proportional to the number of files being processed, you will see the message “Finished … Press any key to continue.” Because the batch file is really basic with no error-trapping, you will see this message regardless of errors. It’s possible that no file may have been processed. Inspect the file RM5copyMetaDataToImageFile.log for any error message.
Assuming no errors, if Picasa is set to monitor the folders where these image files are located and was running at the time they were being revised by ExifTool, you will have seen the little pop-up message from Picasa as it updates its thumbnails and its internal storage of captions for searching. The ability to search through captions is one of Picasa’s strong features, along with its facial recognition and multiple views: Albums, People, Folders.
Copy Image Metadata from JPG to RM
This procedure is trivial compared to the opposite direction. The heavy work has been done in the Compare… procedure, which should be carried out before attempting to copy metadata to the RootsMagic database as the Copy… relies on the RM-Media.db3 database created and updated by Compare… The only operation that makes sense to me that would benefit from such a procedure is to copy multiple image files’ Caption-Abstract values to empty RootsMagic captions for those files. Copying onto existing captions is necessarily a manual process and could be done with copy/paste between the applications.
This is essentially a SQL query against both the RootsMagic database and the RM-Media.db3 database. However, it’s complicated by the fact that I cannot figure out a way to make a single UPDATE statement work for multiple captions so a batch procedure is needed to generate a sql script file with multiple UPDATE statements, one per caption to be copied, and run that.
-- RM5copyMetadataToRM.sql-- generates an UPDATE statement for each blank caption in RM for which there is a caption extracted from the image file-- requires RM-Media.db3 with the RM5 database ATTACHedSELECT'UPDATE MultiMediaTable SET Caption=(SELECT Caption FROM PicasaTable WHERE ROWID='|| PicID ||') WHERE MediaID='|| MediaID ||';'AS Statement
FROM(SELECT MediaID, Pic.ROWID AS PicID
FROM MultiMediaTable AS RM
INNERJOIN PicasaTable AS Pic
ONREPLACE(RM.MediaPath || RM.MediaFile,'', '/')
LIKE Pic.MediaPath || '/' || Pic.MediaFile
WHERE Pic.Caption NOT LIKE '-' AND RM.Caption IS NULL
)
;
RM5copyMetadataToRM.bat.bak Download to the common folder: (Remove .bak extension added due some systems’ security)
ECHO OFF
REM RM5copyMetadataToRM.bat
REM Copies to blank RM5 captions the values from Caption-Abstract metadata stored
REM in the corresponding image files
REM Uses PicasaTable in RM-Media.db3 created by the related Compare routine
REM and the MultiMediaTable of the target RootsMagic 5 database.
REM Requires the sqlite3.exe command line shell
REM
REM Build and run sql script file that generates UPDATE statements for each caption to be copied
ECHO ATTACH DATABASE '%~1' AS RM ; > RM5copyMetaData.sql
TYPE RM5copyMetadataToRM.sql >> RM5copyMetaData.sql
"C:\Program Files (x86)\SQLite\sqlite3.exe" RM-Media.db3 < RM5copyMetaData.sql > RM5UpdateCaptions.sql
REM
REM Build and run the sql script file that UPDATEs the target RootsMagic 5 database captions
ECHO ATTACH DATABASE '%~1' AS RM ; > RM5copyMetaData.sql
TYPE RM5UpdateCaptions.sql >> RM5copyMetaData.sql
ECHO Check RM5copyMetaData.sql before running the caption UPDATEs
PAUSE
"C:\Program Files (x86)\SQLite\sqlite3.exe" RM-Media.db3 < RM5copyMetaData.sql
ECHO Blank RM5 Captions Updated
PAUSE
END
Create a shortcut to the batch file RM5copyMetadataToRM.bat with the Target property [“C:\yourpath\RM5copyMetadataToRM.bat” “C:\yourpath\yourRM5database.rmgc”] and name it something like “RM5copyMetadataToRM – yourdatabase“. Make such a shortcut for each database to which you want to copy captions.
Procedure
First run the Compare utility against your database to ensure that the RM-Media.db3 is up-to-date with the current metadata from all of the images used by the database. Then click on the shortcut “RM5copyMetadataToRM – yourdatabase“. Follow the on-screen instructions. Use RootsMagic to check the new captions.
Further Considerations
The two unidirectional copy caption to blank caption procedures could be combined into one batch and made to standalone without the need for the Compare… procedure to be run in advance.
A comparison between RM’s Caption + Description and the image file’s Caption-Abstract by substituting the CR/LF’s in the former with periods might be feasible.
If MetaData comparisons and copying/editing were part of a future RM feature or outboard utility, it would be desirable to have a directional choice per image where there are non-blank values in both.
What relationship should there be between RM’s Date and Reference Number fields and any EXIF, IPTC, or XMP tag?
What changes may be needed for compatibility with media managers other than Picasa?
There is a flaw in the above to be corrected if there are multiple image files with the same file name in different paths. It has been correctly done in the RM5copyMetaDataToRM query.
I’m going to write this up as a free standing utility program that performs only one function. But of course I would prefer that it be included in a bundled and comprehensive utility program. The proposal is for a Named Group and Color Coding Manager for RM5. This intent is to supplement the Named Group and Color Coding capabilities that are already in RM5. For example,
It is possible to create a Named Group from a collection of individuals that are color coded. But it is not possible to create a color coding scheme from Named Groups.
The criteria used to create a Named Group or to color code a collection of individuals does not apply to any persons or facts or changes made to the database after the criteria are applied.
It is not possible to save the criteria that were used to establish a color coding or to save the criteria used to create a Named Group and later to reapply those criteria.
It is not possible to document the purpose of a Named Group or of a color coding with a comment that is associated with the Named Group or color coding.
It is not possible to establish a default color coding scheme for a database, to temporarily change some of the color coding in the database, and then automatically to reset the color coding back to the default.
There are not timestamps maintained about a Named Group or a color coding to indicate when they were established or when they were last reapplied.
It is not possible for the criteria used to define a Named Group or a color coding to include complete Boolean logic (AND, OR, NOT, and parentheses).
It is not possible for the criteria used to define a Named Group or a color coding to apply two or more tests to the same fact. For example, the test “census date equal 1850” and the test “census place contains Tennessee” are not guaranteed to be applied to the same census fact.
There are a number of criteria that would be useful to include in the definition of a Named Group or in the creation of a color coding that are not supported by RM5. Examples are that it is not possible from within RM5 to search for number of parents or number of children, and there are a number of source and citation fields that cannot be searched.
The intent will be to address all these issues.
Here follows a mockup of a proposed “main screen” for the utility program. The mockup assumes that an RM5 database has been opened and that the database already includes a number of named groups.
Note that one of the groups was created from within RM5 itself. Such groups will not include additional metadata needed by the proposed Group Manager.
The proposed Group Manager will need two additional tables in the RM5 database. A table called the GroupDefTable would contain the following data elements.
GroupDefID – a numeric primary key that has no other purpose than to be a unique primary key.
OwnerID – a unique foreign key that can be joined to RM5’s own GroupTable and LabelTable.
CreationDate – the date the Named Group was created.
EditedDate – the date the Named Group was lasted edited.
Comment – Descriptive text for the group (the area in yellow).
Color – the color code to be applied to all the members of the group (if any). This data element is on the Edit Group screen below.
ColorDefaultFlag – a flag to indicate whether this group and its color is a part of the default color scheme for this database. This data element is on the Edit Group screen below.
In order to edit the group criteria for an existing group, the user would double click one of the groups in the list, or would single click or scroll to one of the groups in the list and click the Edit Group button at the top of the screen.
In order to delete the group criteria for an existing group and to delete the group, the user would single click or scroll to one of the groups in the list and click the Delete Group button at the top of the screen.
In order to create a new group, the user would click the New Group button at the top of the screen, and the screen would look something like the following.
After entering the data for the new group, click Edit Group to bring up the Edit Group screen where the group criteria are entered.
This screen and the underlying GroupCriteriaTable are not fully formed in my mind just yet. Because this note is becoming so long, I’ll return to it later and fill in more details of how this screen would work for entering the group criteria and how the underlying table will work. I’ll also follow up with some more details of how the group definition process would interact with color coding.
Jerry
Discussions & comments from Wikispaces site
ve3meo
Great!
ve3meo
22 January 2012 13:47:42
What a great surprise this morning, Jerry! Your proposal and detailed outline sound very thorough and useful. I think Named Groups is long overdue for enhancement and your tool will be most welcome by many. I look forward to your further description and progress and would be happy to assist in any way I can.
Are you developing in Visual C++?
Tom
snowathlete
snowathlete
24 October 2012 12:42:33
i also think this looks like an excellent idea. i would definately use this. does it have an ETA?
ve3meo
Manual Mark/Unmark
ve3meo
23 January 2012 23:02:21
I wonder if you have given thought to building and maintaining a list of individual Mark/Unmark settings in addition to the algebraic rules. Your Group Editor might be an adequate user interface with one row taken for each person. Alternatively, I could envision a Mark > Persons or Unmark > Persons opening up another dialog window with the list of persons in the database (as in RM Explorer) with checkboxes – maybe a common list with two mutually exclusive checkboxes each, one for Mark, one for Unmark (or maybe the proper term is Exclude). At its simplest, the dialog interface could just be a list of RINs that the user copies from RM.
I showed a very crude Mark/Unmark of individuals in
This may warrant another table with three fields, GroupID, PersonID and Mark or perhaps a 4th, Unmark. Three would be adequate but the 4th might be easier to work with.
thejerrybryan
thejerrybryan
24 January 2012 05:22:16
Yes, I envisioned a mark/unmark capability on an person by person basis. I don’t think an additional table would be required. But whether an additional table would be required or not, the trickier part would be the user interface. As you suggest, the two basic options for the user interface would be to have row after row of “mark/unmark individual nnnnn” (i.e., by RIN number), or to have a RM Explorer style of marking capability
I’m trying to stay away from what I think is RM’s excessive clickiness. So for example, to enter comments about a group, a note window would not open up. Rather, the user would type directly into what I’m describing as the “yellow area”. I would like to do the same for mark/unmark on a person by person basis, but it may be necessary to do it RM Explorer style.
I have recently added a new drive to my pc and wanted a quick way to update the media path file. Whilst there is standard tool within RM5 to fix broken media links, it searches through my entire pc which now has 5TB of data – it picks up files from backup directories and generally takes a long time. I also have the option to update the path directly in the MultimediaTable. Some examples below:
I want to review the records with a particular path
SELECT * FROM MultimediaTable
WHERE Mediapath like ('J:\FAMILY_HISTORY\PAXTON\CERT%');
I want to update this path to the new path
UPDATE multimediatable
SET Mediapath = 'S:\New_Family_History\PAXTON\CERT')
WHERE mediapath LIKE ('J:\FAMILY_HISTORY\PAXTON\CERT%');
If i want to replace only some of the path and not the entire string i use the REPLACE function
UPDATE multimediatable
SET Mediapath = REPLACE (Mediapath, 'J:\FAMILY_HISTORY', 'S:\New_Family_History')
WHERE mediapath LIKE ('J:\FAMILY_HISTORY\PAXTON\CERT%');
Cheers,
Sean
Discussions & comments from Wikispaces site
ve3meo
Inline comment: “a quick way to update the media path file”
ve3meo
04 September 2018 03:34:47
ve3meo Jan 12, 2012
Good examples of Search, Update and Replace. Also see the page Search & Replace for more examples. Of course, RootsMagic own Search & Replace function on “Multimedia filenames” in the “Field to search” selection does much the same, with the ability to confirm each replace.
One report that would be a great boon to RM users is one that would look at the recored information for each individual and work out where Census entries were missing. Of course this would vary by country of user but I am approaching this from the point of view of a UK User where 95% or more of his people records are for UK family.
For example we can assume at the moment that potential Census records exist for viewing for every tenth year between 1841 and 1911. The logic require is, therefore, to look at the birth/baptism/christening and death/burial dates of the individual as well as the existing Census entries on the Rootsmagic database and determine which Census entires are likley to be missing. We would also need some logic to cope as well as it could with individuals for whom no birth or death dates are known. The following “selections” would pull together what I want:
a) Individual has Birth Date earlier than 2nd April 1911 and Death Date no earlier than Sunday 6th June 1841:
Look through the Census entries on RM and if any are missing over the period the person was living between 1841 and 1911 then highlight this fact.
b) Individual has Birth Date earlier than 2nd Paril 1911 but no recorded Death Date:
In the absence of any other data – proceed as for a) above from date of Birth to 1911.
c) Individual has no Brith Date and a Death Date recorded after 1841:
In the absence of any other data proceed as a) above from 1841 to date of death.
d) Individual has no birth date and no death date:
Clearly we do not want to report on the many people who were “obviously” born after 1911 or “obviously” dead before 1841. One solution might be to look at Parents or Children of these individuals and try and determin whether they should be included in the report. Or perhaps just an option to include everyone in this category or exclude everyone in this category. Or to see if there are any Events of any type recorded for the individual between the period 1841 and 1911.
Or perhaps others have a better idea?
In a) b) c) and d) above I think the code should determine a birth date from the presence of either a Birth entry or in its absence a Baptism entry or a Christening entry. Similarly the Death entry should be used or in its absence a Burial entry.
I would envisage the report looking something like:
Name of Individual Birth/Chr/Bapt Death/Burial Potential Missing Census References
xx xxxxxxxxxxx xx/xx/1837 xx/xx/1902 1851, 1861, 1881
xx xxxxxxxxxxx xx/xx/1794 xx/xx/1852 1841, 1851
xx xxxxxxxxxxx xx/xx/1878 N/K 1901, 1911
Clearly for RM Users in other countries, where a different ranges of Census data isavailable, the start and end year parameters would have to be different.
MVS.
Discussions & comments from Wikispaces site
ve3meo
Inline comment: “where Census entries were missing”
ve3meo
04 September 2018 01:48:57
ve3meo Jan 16, 2012
In some respects,
Census Needed – Named Group does what you want. The qualifying rules are simpler but may be enhanced. Once the group is created, you could generate reports restricted to it, or, just select the group and work through it. Periodically, re-run the query to refresh the group membership and take satisfaction from seeing it shrink!
Inline comments
ve3meo
Comment: In some respects,
Census Needed – Na…
ve3meo
16 January 2012 18:21:38
In some respects, Census Needed – Named Group does what you want. The qualifying rules are simpler but may be enhanced. Once the group is created, you could generate reports restricted to it, or, just select the group and work through it. Periodically, re-run the query to refresh the group membership and take satisfaction from seeing it shrink!
OK, this is a little misleading as it will only work for images (i.e. not applicable for text files or pdf files) ~ if anyone can show me how to check for valid paths from sqlite, i would be stoked!
SELECT *
FROM Multimediatable
WHERE Thumbnail is null
AND Mediafile like '%.jpg';
Discussions & comments from Wikispaces site
ve3meo
Comment: “how to check for valid paths from sqlite”
ve3meo
03 September 2018 21:49:34
ve3meo Jan 15, 2012
I have done this indirectly, by exporting the full path of each file to a text file to be processed by a command file (.bat or .cmd): see
Backup Media with Database – 7Zip
Backup Media with Database – RAR
Also, I have copied the results of a query to Excel and created hyperlinks (either in the query itself or by formula in Excel): see
Media List Query.
It requires more than SQLite itself to test that the stored path leads to the target file.
At the very least, your query lists those image files that have not been opened by RM, hence nothing in the thumbnail field.
RootsMagic 5 and below support for Ahnentafel numbering is limited to 32 bits, or 32 generations, even though the report allows the user to select a larger number. If the report does run over 32 generations, the Ahnentafel number starts over and is therefore incorrect. This query demonstrates that SQLite itself can support a 64 generation Ahnentafel report.
The results of this query look very similar to those from the Ancestors Query, with the addition of both a binary and decimal Ahnentafel number for the last person in a direct ancestral line from the starting person, expanded to 64 generations. It is a very large and slow query that cries out for help from a high level language because SQLite itself does not support loops nor binary-decimal conversion. If your database is large, I strongly recommend using a SQLite manager that supports run-time parameters so that the results are limited to the ancestry of just one person (SQLite Expert Personal or SQLite Developer), not that it is faster but rather the processing or results may exceed the software’s capacity to manage memory.
To build and revise it many times, I used Excel to generate the 64 lines from each of around 8 formulas, sorting and unsorting a large block of interleaved phrases for each revision. As such, the query is not very readable, having many very long lines and lacking indentation.
There must be a more efficient way of coming up with the lineage than the two similar methods I have used in this and its parental query (despite the similarity in the appearance of the results, Ahnentafel-64 uses a significantly different relationship in building the lineage). It is simply too slow to be attractive to run on databases larger than a few thousand persons. So there’s a challenge! Come up with a better one!
— the Excel spreadsheet on which the query was developed using formulae.
Having torn my hair out trying to reconcile differences between RootsMagic 5’s Ahnentafel list and the results I was getting with my query, I discovered that RM5’s follows the parental line selected in the Pedigree View. I had one person shown with her adoptive parents and compounded with other issues I was having with incorrect Ahnentafel calculations, missing the last parent, etc., it took me quite a while to understand that some longer lines I was getting that were not showing up in the RM5 report were not because of an error in my query but because of this undocumented behaviour in RM5! My query follows the Birth parental line only; RM5 appears to follow whichever set of parents you choose.
This page is intended to stimulate some discussion about developing utilities for RootsMagic in a high level programming language. Please add your thoughts. Would you be interested in:
using utilities that do things RootsMagic does not currently do?
developing such utilities?
testing?
Development Platform?
I had been setting up an old Borland Delphi 4 with DiSQLite3, thinking it was time to move some of the SQlite scripts into some kind of menu driven collection of utilities to make them more amenable to persons flummoxed by working directly with queries. I had just compiled a fake RMNOCASE extension for SQLiteSpy on it which would also work with DiSQLite3, and thought that this would be a logical platform because RootsMagic is also developed on contemporary Delphi and DiSQlite3.
Then along came Steve Turley who started the ball rolling with RmSplit on the page SplitTree, using Visual Studio C# 2010, because, as he said:
I chose Visual C# because I find it's a quicker development environment for projects like this.
The advantage of C++ is that it has a more direct connection to the base sqlite library which
is written using unmanaged code. For other projects, C++ tends to be more portable, but an
application written specifically for Windows like this one is tied to a single platform anyway.
Mostly, it's a personal preference. I am a physics professor and do a lot of research work with
students. I have found that I can bring them up to speed faster on C# programming than C++ so I
tend to gravitate towards that for simple windows programs.
That said, if I was writing a big production code or if performance were at a premium, I'd
probably use C++.
I installed Visual C# 2010 Express and Steve’s project package and was able to generate my own version of RmSplit with minor changes in the user interface. I’ve been a program ‘dabbler’ since before DOS, just enough to whet my appetite and maybe accomplish something odd for work or play but never on an ongoing basis. Never got into object-oriented programming – I’m a procedural guy and not sure if C# is the way to go for me.
Another option for me is Tcl/Tk. Tcl was used to develop SQlite and I have some passing acquaintance with it from about a decade ago. And it’s totally free, unlike DiSQLite3 and Delphi, and, for that matter, anything above the Express level in Visual Studio products. On opening Visual C# Express, it says it is for Evaluation Purposes Only; I cannot find the EULA so am unsure of the restrictions, and am seeing a 28 day countdown which suggests that there may be features that get turned off when that period expires.
PHP has SQLite built-in, is well suited for server-side web applications, and is free. Never used it and don’t know how suited it is to a stand-alone app.
And I just had a brief look at SQLite for Excel. Took me a while to understand what the sample demo was doing. Basically adds VBA macros to a workbook via which a SQLite database can be worked on. The demo seemed pretty slow as does RmSplit so I wonder about the number of layers between the user and the data.
The worrisome thing is the learning curve on any of these. Peter Norvig, author of Teach Yourself Programming in Ten Years says his “recommendations for a first programming language would be Python or Scheme. SQlite is included in recent Python distributions; links for Scheme wrappers on sqlite.org are dead.
Peter also advises to Use your friends, i.e., use what they use as there is greater strength in a community. If you are interested in collaborating on the development of utilities for RootsMagic, what would you use? What are you using?
(Added by Jerry Bryan): I’ve been following this discussion with great interest. I agree with the opinion that as long as SQLite remains unbundled where a user has to download an SQLite manager and also download an SQLite script and figure out how to operate the two together, SQLite scripts will only reach a very small niche of the RM4/5 user base. Being able to download a bundled program which is fully self contained (an EXE file, if you will) should enable useful utility functions to reach a much greater audience than the unbundled approach will ever reach. So I’m excited about the possibility of bundled utility applications.
I’m an old C/C++ programmer from way back and I know nothing about C# except in as much as C# may have common roots in C. I’m not interested in advocating for C/C++ as the platform of choice, but I am interested possibly in writing a RM4/5 utility application myself in C/C++. And I’m interested in seeing a “baby steps” example of what’s required. Exactly what C/C++ library for SQLite do I have to download and from where? How do I do the compile? And would it be possible to have a really simple examplar to look at, even if it were written in C# instead of C++? For example, I would love to see a C++ or a C# program that would do something so simple as SELECT * FROM NameTable, and gather up and print the results. In other words, the example program would need no real logic at all. I would just like to see how to put the pieces together. If I could get even one such simple program to compile and execute, I think I could easily go from there to developing much more complicated and useful applications.
(Added by Steve Turley)
Here is an example in C++ to supplement the C# version posted elsewhere. My two easy choices are using gcc under linux or Visual Studio. I’ll start with Visual Studio since I’m guessing an RM user is probably running on a Windows platform. By the way, there is one issue that doesn’t show up in a query like SELECT * FROM NameTable. If you want to modify a table in the database, you may have to worry about the collation. Many of the tables are index with an RMNOCASE collation. I’ll save that issue for a little later.
Unzip this file to the course code directory for the above program. (An easy way to find this directory is to right click on the tab for this code and choose open containing folder.)
Add the sqlite3.def and sqlite3.dll to your project directory. (Right click on the project name and choose Add/Existing item…)
Create sqlite3.lib by executing the VS command LIB /DEF:sqlite3.def.
Add sqlite3.lib and sqlite3.def to the list of library files and Module definition files used by the linker.
Right click on the project name and choose properties.
Select the Input item in the Linker Group.
Add sqlite3.lib to the list of files with a terminating semicolon.
Unpack the file sqlite3.h and put it in the header folder for the project.
Here is a zip file with my project in it as I built it. RmCppConsole.zip
(Added by Jerry Bryan)
First a procedural question/comment: Simply editing one of the wiki pages seems not to be the best way to carry on what is essentially a threaded conversation. Is there a better way to carry on such conversations within the confines of the wiki architecture? Like most people, I’ve read wikis for years, but this RM/SQLite wiki is the first time I’ve ever joined a wiki and posted to it. I’m more familiar with threaded conversations in the context of forums and bulletin boards, blogs, and even old fashioned E-mail.
Second: Much thanks to Steve Turley for posting the sample code. It’s exactly what I was looking for, even down to the fact that it’s a console application instead of a Windows application. I’m a computer science guy, but I do IT for a living and seldom have an opportunity to do real computer science. I do teach a course occasionally, but it’s been several years since I’ve even had time to do that. Being a college administrator is not as much fun as teaching or as doing real computer science. So developing some object oriented code on sort of a hobbyist basis sounds like a lot of fun.
Third: As to my experience with C++, it’s strictly as a hobbyist. Our college still uses C for its intro and data structures classes, which are the ones I usually teach if I teach. The question of “first language to teach” has long been a hot topic of debate among computer science, and our college has chosen to stick with a procedural language as the “first language to teach”. I find that my C students always ask what the difference is between C and C++. It’s a curiously difficult question to answer in a way that’s both brief and meaningful. You really need to experience object oriented programming by doing it, and most people learning an object oriented language have a major “aha moment” when they “get it” about what object oriented programming really is. Having said that, I’ve probably written about 40,000 or 50,000 lines of C++ code in the last 10 or 15 years (just for fun). Until very recently, I’ve used the (now) very old Borland C++ compiler version 5. Serendipitously, I’ve just started the switchover to Visual Studio Express. I haven’t decided yet if the Express version is good enough, or if I’m going to need to upgrade to a professional version. For my ongoing projects that are not RM/SQLite related, I still don’t need anything except for console mode, but I’m eventually going to need huge data spaces and 64 bit operation. It’s not clear to me if Visual Studio Express will do 64 bit operation, or if it’s limited to 32 bit. Of course, 32 bit is fine for any RM/SQLite needs. And by the way and in case anybody is interested, my big C++ project for years has been the investigation of the mathematics of Rubik’s cube. There are many interesting problems that are computationally intensive and that sometimes run for months on a PC.
Discussions & comments from Wikispaces site
romermb
VB.net
romermb
08 January 2012 05:25:52
What would be the advantages and disadvantages of VisualBasic.net?
rsturley
rsturley
08 January 2012 05:38:42
VB.net would have the same interface to SQLite and WindowsForms environment as C#. Personally, I think the C# environment is richer and better suited to collaborative projects (easier to define and enforce interfaces using classes). However, the community of programmers with an interest in this project may have a larger group with VB.net background. The end users of the utilities probably won’t see much difference in applications programmed either way.
rsturley
Bundled Utilities
rsturley
08 January 2012 05:34:22
The idea of bundled utilities intrigues me. My RmSplit was a bit of an experiment thinking about a larger package of utilities. Here are my comments on some of your ideas on the best development language.
C#: a proprietary environment that only works on Windows. It sounds like Bruce is working on expanding RM to be available on several mobile platforms and Mac OS. In principle, we could extend a project to work on other platforms using mono, but that would probably not be ideal.
Python: I’ve tried some of this kind of work on python before. It is a great tool for one-up kinds of things, but it doesn’t have strong tools for a GUI that would be a good interface for non-experts.
PHP: I have used this as a web programming language and a stand-along scripting language. I think I like this suggestion the best of the ones you suggested in your page. One nice thing about it is that the project could be hosted “in the cloud” and we don’t have to worry about distributing or updating the code. We could have a web-based package that incorporated an organized set up the SQL scripts appearing in this wiki process them on the server rather than the user’s local computer. I don’t think the community of users is so big that it would be a processing problem, but we could try it on a small set of scripts and see how much traffic we get.
Java (one you didn’t mention): This would either be stand-alone code (more portable than C# probably) or server based (.jsp files which similar organization to PHP). If we go with a cloud-based solution, I prefer PHP over java because server maintenance is a bit easier and servers supporting PHP are more widespread than those supporting jsp.
Anyway, if you decide to support a language and platform I have experience with I’m willing to help.
The first is a springloops site, and it has a link of some kind to basecamp, so I created one of those, too.
That could be a place for us to start out…
the springloops site for the codebase stuff, and the basecamp for the visual planning, conversations, ideas, chat, milestones, etc
Are those sites and utilities that you think folks would find easy enough to use?
ve3meo
RMNOCASE
ve3meo
25 January 2012 18:45:03
Blundering around in Steve’s RMsplit project, trying to learn the rudiments of C#, OOP, Visual Studio and wanting to leap ahead to something useful, I did manage to run the SQLite Integrity Check with the fake RMNOCASE Steve came up with. The outcome is not good. It reports many index errors on databases that RM5’s integrity check reports as OK, just as does SQLiteSpy with the fake extension built with Delphi. The latter has many more reported discrepancies and the two do agree on most of those reported by C#.
This means that, until we can use the identical collation sequence as in RM, procedures that modify data so collated should be precluded from being used on RM4 databases and that, for RM5 databases, a strong warning that there may be unanticipated consequences and advise the user to use the RM5 Rebuild Indexes feature after any such changes by the utility.
ve3meo
Inline comment: “I would love to see a C++ or a C# program that would do something so simple as SELECT * FROM NameTable”
ve3meo
04 September 2018 03:39:33
ve3meo Jan 9, 2012
Jerry, download the project file that Steve posted on SplitTtree and look at RmSplitForm.cs for examples of the query. He also includes in the project the necessary SQLite files.
rsturley Jan 9, 2012
Jerry,
I created a c+ example program that does the simple thing you asked about (SELECT * FROM NameTable) using Visual C++. To keep it simple I didn’t put in any GUI stuff, but rather chose a simple console application. The example and a brief explanation of the process of creating it are in the wiki page.
ve3meo
Inline comment: “Is there a better way to carry on such conversations within the confines of the wiki architecture?”
ve3meo
04 September 2018 03:41:10
ve3meo Jan 12, 2012
Yes! Each wiki page has a Discussion “forum”, i.e., a page under the wiki page where you can post a topic (really any topic but preferably one related to the subject of the wiki page) and a threaded (sort of) discussion can flow. By sort of threaded, I mean that all posts under a given topic can be found under that topic but you cannot see which post a post has replied to and the reply system includes no quotes. For a straight serial conversation it’s fine. To open the Discussion page, you click on the double bubble button on the wiki page.
A variant of the Discussion page is to comment on selected text on the wiki page itself, as I am doing here. You have to Edit the page, highlight the text to comment on, click the Comment bubble button and write your comment in the form. The comment also appears on the Discussion page under the topic which is the highlighted text and replies thereto automatically show up as comments on the wiki page itself. To see the comments on the wiki page, you click on the button with the single bubble on the wiki page.
ve3meo Jan 12, 2012
OOps, got the way the topic works wrong, as you can see… The highlighted text on the wiki page is quoted in the post seen on the Discussion page; the topic of the post is “Comment added:”+ first few words of first sentence of comment.
As an external utility for the RootsMagic genealogy program, what would it do, how would it look?
Initially, it would take the many tweaks and modifications discussed on this site and roll them into a format that is usable for the average RM user.
Rationale
There are a lot of general database cleanup or modification items that users want to be able to do, but that RootsMagic doesn’t do on its own. Some of these things RM may do in the future, some maybe not, but since it’s possible to modify the database itself, it can be done. Users don’t know or need to care about the database structure or the commands needed to get it done. In order to make the many wonderful things on this site more widely used (and make maximum use of the efforts folks have gone into to develop and explain these things) a layer needs to be put on top of the SQL Queries, so that the words SQL and Query are really not even mentioned.
Example from TMG Utility
The best way to ensure success is to look at how others have achieved it. TMG is a genealogy program much like RM, it’s based on FoxPro, and there’s a wonderful utility program for it. Many TMG users benefit from the TMG Utility program, and very few know or care about FoxPro.
I have made a screencast of the utility, showing the program’s structure, the options provided to the users, and I’ve added notes of some of the parts of the program. It’s a 5 minute screencast, found here: http://screencast-o-matic.com/watch/clV6cZxnb (apparently deleted, accessed 2019-01-26)
You can, I think, watch the screencast via this frame, but for my notes, you probably need to go to the website itself.
Program Structure
The expanding folder view on the left is a very common, easily understood paradigm, and automatic “categorization” of all of the options. I highly recommend that RMutils use the same concept. The structure of TMG Utilities echoes the DBF files that are modified, for RMutils we’d use the database tables.
The function windows are pretty clean and straightforward.
The help “file” is an on-disk version of the program’s website, I’d be tempted to suggest making our help file go to the live website itself rather than on disk files. In that way, the help system is always as up-to-date as possible.
The program’s website, is, IMO, NOT a great example to copy.
Example from LTools
(added by ve3meo)
LTools is to Legacy Family Tree as TMG Utility is to The Master Genealogist, filling some needs that the main database program does not satisfy. Independently developed, it accesses the Legacy JET database (JET underlies Microsoft Access). Unlike the TMG Utility, it exposes the user to SQL queries, even supports user defined SQL queries, at least in the Deluxe (paid license) version. The Standard version is free. I downloaded the free version of LFT and LTools for the purpose of reviewing primarily the latter, insofar as we might also learn something from it that might be applicable to the rmutil design.
LTools also uses the expanding list feature. The grayed out area on the right in the Standard edition which shows the inaccessible Deluxe features is likely merged with the standard list in the Deluxe version, given the common root names.
Help is solely on-line.
As a SQLite user, I like the idea of being able to have an interactive SQL screen in the utility – I have asked for the same to be provided in RootsMagic itself. LTools appears to provide a way of storing and recalling user defined SQL queries – cool!
Discussions & comments from Wikispaces site
LessTX
Interface and Program Structure
LessTX
12 January 2012 18:40:56
There’s a lot to like in LTools. I like the idea of a “Standard and Advanced Mode” interface option. The help website looks superior, but reinforces the concept that “help on the web” is the way to go.
Standard doesn’t use the word query or SQLite and keeps things in plain English.
Advanced mode would enable the menu options to show, save, and modify the queries themselves.
Then also having more complex things be part of a paid product is a good idea, but allocating half of the app window doesn’t make a lot of sense. Having them as disabled items on the single list makes more sense to me.
ve3meo
Comment: “The expanding folder view on the left is a very common, easily understood paradigm…”
ve3meo
03 September 2018 20:00:25
ve3meo Jan 10, 2012
Nice! I have been grouping the existing SQLite queries to relate to the RM menu structure as that might lead to an easy way for users to find relevant ‘potions’. But some queries don’t readily fit or could tie to more than one category. In some respects, categorising by table name might be easier but many use multiple tables. There’s lots to play with, here. Thanks for the walkthrough and your comments.
This page is intended to group existing SQLite queries and future functions
under a menu structure that parallels that of RootsMagic itself, with the
idea that similar naming, order and, possibly, locations on screen may
facilitate user navigation to the utilities most relevant to their area of
interest or concern. This is a first draft. RMutil-Menu.xlsx
Rmutils
(working name)
January 10, 2012
Main Menu
5:55 PM
SQLiteToolsForRootsMagic query or other description
Sort
Main
Sub1
Sub2
Query Name
Description
Page
Comment
1
File
1.1
File
Open
Open Windows Explorer browser on *.rmgc files.
Get database version, possibly do SQLite Quick Integrity Check
“PRAGMA quick_check(integer);”
1.2
File
Open Recent
1.3
File
Search for Files
1.4
File
Close
1.5
File
Copy
1.5.1
File
Copy
All
1.5.2
File
Copy
Less People
CustomDatabaseShell
Makes an empty Master from current database, preserving custom
fact types and source templates, Places, Master Sources and
associated Repositories and Addresses, and Place/Source Media
Clones a RM4/5 database without the RMNOCASE collation so that
we’re free to do what we want to the data. Even works with RM but it
might trip up on non-English alphabets.
Deletes unused rows from tables as reported by RMGC_Properties.
May need checkbox options to select what tables are cleaned. RM4
& 5 are different
Conceivably, Properties & Clean could be one form with
checkboxes for tables that have unused rows & a Clean button
1.6.3a
File
Clean
Delete Phantom Citations
Deletes citations of non-existent Sources from database that
manifest themselves in the Edit Persons screen as a citation
(counted and checkmarked) but return nothing when opened.
possible route to a control of Fact settings for GEDCOM export
1.8
File
Backup…
1.8.1
File
Database file only
1.8.2
File
Include media
RMfullbackup.bat
Backs up the database file along with all the media files
referenced by it to one file. Uses the command versions of SQLite3
(free), WinRAR ($fee) and 7-Zip (free). The 7-Zip version produces a
.rmgb file that RootsMagic can restore, media and database.
Backs up the database file along with all the media files
referenced by it to one file. Uses the command versions of SQLite3
(free), WinRAR ($fee) and 7-Zip (free). The 7-Zip version produces a
.rmgb file that RootsMagic can restore, media and database.
List and convert events having key properties in common to shared
events. Especially useful on imports from Legacy Family Tree in
which events were copied to multiple persons.
About as close as we can get to a RM4 Source List Report presented
in tabular form with which the results can be sorted and filtered
with relative ease.
Lists all citations in the database from which citations of
non-existent sources (‘phantoms’) and citations for non-existent
events or persons (‘headless’) can be found, along with other useful
information such as all citations per source.
Lists all source names cited for a person in descending order of
the count of duplicate citations – a help in finding and resolving
duplicate citations after merging.
Series of queries to list duplicate media file names in the Media
Gallery and list duplicate links to items in the Gallery. Repairs a
specific case of duplicate file names as an example; repairs all
cases of duplicate links.
Combo of SQLite queries and RM4 edits of Place List beats having
to edit every fact/event to split a Place into Place and Place
detail; otherwise, wait for the RootsMagician!
Manage, save, apply Fact settings for Export, Reports, Privacy
3.A
Lists
Source Templates
3.A.1
Lists
Source Templates
Sources by Template
SourceTemplateListWithCitationDetails2.sql
Lists Sources associated with Source Templates, as well as an
extension to Citation details. Free-Form templates are excluded
since they’re not included in SourceTemplateTable.
List Persons with specified Surnames. — Example of creating a SQL
View or Virtual Table and the explicit use of COLLATE NOCASE to
override the RMNOCASE collation defined for certain fields and
embedded in the RootsMagic application.
Removes CR/LF characters around custom fact sentences and
person/family/alt name/fact notes and adds CR/LF pairs to ends of
notes except the last, for first cut batch paragraphing.
This query strips extraneous Carriage Returns from the end of the
Footnote sentence template for the Source Template “Vital Records
(state-level, online derivatives)” which cause unwanted white space
in reports.
List all the Individual, Family and Shared Facts/Events for all
persons in a database. A complex query using UNION ALL, COUNT() and
GROUP BY, constants and NULL to assemble multiple SELECTs in one big
result.
Two queries that list the test results in a format suitable for
easy review and copying and pasting into other applications.
RM4.1.1.4 provides no DNA report.
Have you ever wished to be able to look at all the facts in your
family tree database that happened within a day’s horseride of a
certain location? This adaptation of the LifeLines query helps you
view your events for any geographic area in addition to looking at
the lifeline of any person in your database.
Lists all events for all persons whether in a database tree or
not, including shared facts, date, fact detail, site and place,
MRIN, other parties, and duplication indicator. Sorted by RIN and
Sort Date. With the right SQLite manager, can filter results for one
person.
This query helps you plot events from your RootsMagic database on
Google Maps, Google Earth, and Bing Maps, provided there are
geo-coded Places and Place Details (sites) in your database.
A Query for a To Do List in a Grid Format – using the REFN to
assign status for paragraphing and census. Versions for both SQLite
directly and MS Access via SQLODBC.
Lists duplicate name pairs with a weighted score indicating degree
of match. On a large database, produces results in 4.5 minutes
similar to what RootsMagic 4’s Duplicate Search Merge tool does in
45 min.
Copies all unmerged pairs from the results of RM4’s Duplicate
Search Merge Tool to RM4’s ExclusionTable so that they will not
reappear in subsequent runs of Duplicate Search Merge.
Builds or refreshes a group of persons whose lifetime probably
spanned a user-defined Census Year and who had some event in the
user-defined jurisdiction but not a Census fact for that year.
Warning: I don’t know what all of the current queries are, I
haven’t taken the time to look at the long list of stuff, it’s
currently too long and unstructure to make me believe that
it’s worth my time. In that, I’m probably like most folks who
come here, and am the kind of person this project is intended
to help.
Just be aware, my suggestions have NO relationship to whatever
work has already been done, just a relationship to my prior
program (TMG) and what I needed from a utility program, and
seem to still need from an RM utils program.
The utility program itself should have the standard
File/Edit/View/Tools/Help structure.
Users will come to the RMUtils program why? To either GET INFO
from their database or to CHANGE their database.
The pertinent areas are:
names, events, places, sources, repositories, media, others?
I absolutely see a structure similar to that of TMG being the
most intuitive for users.
ve3meo
ve3meo
21
January 2012 03:40:31
I have activated a Query Menu in the sidebar linking to a
new Query Menu page providing something of a structure that
hopefully helps newcomers navigate to a query of interest.