Jump to content

'makerphile

Members
  • Content Count

    13
  • Joined

  • Last visited

Community Reputation

0 Neutral

About 'makerphile

  • Rank
    novice

FileMaker Experience

  • Skill Level
    Novice

Platform Environment

  • OS Platform
    Windows
  • OS Version
    Win 10
  1. I have two sets of data: rankings (Name, Rank) and partner_selections (Name, Name_of_Partner if have one yet). I want for each person to calculate the best-ranked potential partner among people who do not already have partners (i.e. among records in which Name_of_Partner field does not contain a value, sort for the best associated ranking). I can do this by using a calculation field to hold the pairing_status (1 if have partner, 0 if don't), create a global field "Zero_placeholder" that holds the value of 0, and then do a self-join where names are not-equal and Zero_placeholder = pair
  2. The inability to use the self-join with not-= to look up the previous serial number appears to be based on the timing of when the self-join match criterion is evaluated relative to when the serial number field data is committed after creation. Here is an even simpler example database to exhibit the problem. The example database consists of one table and one field: serial_number, which is auto-entered upon creation. A self-join is done where serial_number is not-equal to serial_number from the duplicate table. Upon creation of the first record, as long as the cursor stays in the serial_numb
  3. I have represented a bug in a simple exercise: a field "previous_serial_number" in a table looks up the auto-entered serial number from the just-previous record in the table. Record 2 should have 1 in the previous_serial_number field, for example. Something is not working: record 2 has 2 (it's own serial number) in the previous_serial_number_field. I'm doing a self-join, with the match field being the serial number and the relationship being not-equal, sorting the matching records on the serial number in descending order. Record 2 is not supposed to be matching with itself, but it is
  4. Suppose I have have two fields: "double" and "half". For a given record, the value of double will be twice that of half (e.g. double = 4, half = 2). Sometimes it will be convenient to enter data for half (e.g. 2), in which case one wants data to be calculated and entered automatically for double (e.g. 4). At other times, entering data into double will be more convenient, in which case one wants half to be calculated automatically. What are some approaches to providing flexibility like this in choosing among fields for data entry?
  5. Thanks for your thoughts. I had been hoping for something like a portal to an itineraries table that would show ALL the acceptable routes between specified termini (e.g. each field representing a node on the route). Your first approach suggests to me a script that systematically works its way through all the possible routes branching out from the origin until limits are met or exceeded and further branching is thus cut off. Doing this kind of thing in C is easy, using "for" loops. Would scripting be the right approach here? I admit I have not done scripting with much ambition. The sec
  6. I hope more tractable. This is like an airline flight search, where the number of stops/changes and overall cost and/or time of travel can sharply limit the itineraries considered. It's not yet an optimization problem (though one could sort by overall distance, etc.) and doesn't require traveling a path through all locations. Hope that shrinks the problem enough to make a difference.
  7. What would be a good database design strategy for something like a road trip planner? One starts with data in a LEG OF TRIP table whose fields are: START LOCATION, FINISH LOCATION, DISTANCE BETWEEN START AND FINISH. When a user inputs an overall start location, an overall finish location, a minimum and maximum number of trip legs, and a minimum and maximum distance of travel, the database would display a list of possible acceptable itineraries connecting the overall start and finish locations, perhaps sorted based on some user-specified parameters such as by number of trip legs. Let us sup
  8. Thanks for the response, Soren, which helps to clarify the motivating issue. I'm guessing Soren's question is why I use a Set Field script step rather than a Lookup to enter data into a field in the portal row, since Soren's examples all involve using Lookups to get portal row field data from previous portal rows. The answer is that in these examples (and in my example back when I tried using only Lookups rather than a script), new populated portal rows don't get created automatically. You have to enter data into at least one field in a portal row and then commit the record to activ
  9. Hi, I think my script ("test committing script") incorporates your suggestion, but still requires clicking outside the portal rather than tabbing out. Here is the script: If [group payments::Excess Amount > 0] Go to Portal Row [select; Last] Go to Portal Row [select; Next] Set Field [payments for individuals::amount for individual; group payments::Excess Amount] Commit Records/Requests [] End If More details. The parent table is "group payments" and the child table is "payments for individuals." group payments::Excess Amount is a calculation field that inc
  10. Hi folks, I use an event-based script trigger to populate the next portal row with looked up or calculated data when data for a field on the previous portal row gets edited, provided that one first clicks outside the portal. But clicking outside of the portal to trigger a script isn't much easier than clicking a button to trigger the script. It would be nice to be able to tab out of the edited field (an action that now tabs me into the next portal row, which stays blank) and have the tabbing trigger population of the next portal row. Is this possible?
  11. Sorry for the late thank you, I didn't open your attachment correctly so that I couldn't see what you had done. Now that I've got it working, it looks very nice. Is it possible to get the thing to supply possible values without clicking or tabbing out of the boxes, so that you can see whether continued keystrokes are narrowing entries down to a unique entry? That's the nice aspect of autocomplete of which I was trying to take advantage. Thanks again
  12. I am having trouble using autocomplete within a portal using a value list drawn from a matching field of a different table. Is there a workaround or should I be using a different strategy? I hold classes with variable attendance, charging per class attended. For each class date, I want to enter a list of attendees onto an attendance list. Then I want that attendance list to generate line items for billing purposes, one line item per student per class. Here is the tricky part. I want to enter, on each line of the attendance list, a couple of keystrokes representing initials or parts
×
×
  • Create New...

Important Information

By using this site, you agree to our Terms of Use.