Bob7

Based on recommendation from a FM teacher of mine, I bought: "Database Design for Mere Mortals: A Hands-On Guide to Relational Database Design" by Michael J. Hernandez, and it is suiting my needs well. Bob

Thanks for the feedback. I'd seen something quite a while back on another forum about using GetURL and opening a PHP page, copying and pasting the result, to make this work in some other situation, but that it didn't work in 7 (and the guy was going to try it in 8 but hadn't posted about it). I have no idea what .php page he was referring to. I'll try to find that post and ask him, reporting back any news. Bob

I realize that same-named fields (e.g., "NameFirst") in different (related) tables have their own context. Does giving them different names, like "NameFirstEmployee" and "NameFirstVendorContact" have much value to developers (an extra, but not necessary identifier as to which field/table is being referred to)? I'm thinking the overhead of the extra length might not be worth it, but I'm not sure. Thanks. Bob

I have some good FM books and training materials, but would like to buy one general RDB theory / design book (relationships, data normalization, ERDs, etc.). I've seen a few that look interesting, particularly: "Relational Database Design Clearly Explained, Second Edition" by Jan L. Harrington Any recommendations for an intermediate FM developer appreciated. Bob

Great ideas, thanks. I've posted the current ERD: ERD Link I'll attempt a quick and dirty explanation (short on time now). A resource center has employees who work with coordinators (they are teachers too, btw) to bring in new resource assets (audio files, etc.) to be used by students. Courses and coordinators are linked with departments. We list the content of each resource asset member. We tie the resource assets to course sections (bc not all course sections necessarily use the same asset as the course). Each course instance occurs during an academic quarter, e.g., "Spring 2006". Assets are derived from a "Publisher" (whomever made the source resource asset). Bob

Let me eleborate. I'm working on a db design, and so far my ERD has 19 tables, 10 of them join tables. The reason is I initially have lots of many-to-many relationships, like coordinators being 'in charge of' one or more courses at any given point in time. If the courses are thought of as being only ones offered only during any given academic quarter (e.g., Spring 2006), then one coordinator has zero, 1, or many courses. But if the course is thought of as a course instance, meaning course X has a record (in a 'course' table) for every academic quarter it is offered, then the relationship is many to many. Based on the latter scenario, I sketched out a join table which is 'course instance'. The other approach might be to leave out the join table, and create multiple course records (for each instance of each course) in the course table. I'm leaning toward using the join table. My entire (current ERD) is full of this kind of situation, so whatever route I choose may affect the entire design. Any thoughts on this appreciated. Bob

For primary key values, using auto-entered (and incremented) serial numbers, where the numbers utilize an alpha prefix, I'm wondering what the pros and cons are of using the form of: "abc_1" vs. "abc_000001" are. One thing I like about the latter is that the numbers are of fixed length. Just a visual preference really on my part (while scanning through multiple records, comparing them, etc.). The disadvantage seems to be that you need to make an arbitrary decision as to the most numbers the system would ever need, both for actual records, and also accounting for record deletions. I believe that the latter number example, even if exceeded in terms of leading zeros, would still increment; it would just make the length longer (and therefore no longer a fixed length across all record). Correct me if I'm wrong on that. Anyway, any feedback on a preferred method is appreciated. BTW, I'm aware of more advanced methods of creating primary keys (using time and date stamps in combination with serial numbers, etc). Thanks. Bob

My bad... I was thinking about having multiple files and having to associate the Value List to a particular file -- but my solution will be one file, many tables, so, as you write, there is no table association. Thanks for straightening me out on that -- it makes the everything much easier. Bob

Instructors work with technical employees to create learning resources. We want a record of these creations, including the instructor and tech employee involved with the creation of each resource, which year and quarter it was created (e.g., 2006_Spring), as well as the instructor and tech employee currently responsible for each resource. Resources are often linked with (used by) current course sections (there are many sections per course). This linkage is represented by a join entity (table), which includes the year and quarter of each resource usage by the sections. Bob

Thanks, that is what I was leaning toward. Since more than one table will need access to it, I'm not sure which table to place it in. SInce the tables are related, I suppose it doesn't matter technically -- it's more an issue of placing it in a table that sort of makes sense. Bob

x-posted to fsa tech_talk (no response there) Dymo File Folder Labels 30576 in a Dymo LabelWriter 330 Turbo, using the recommended 30327 File Folder settings. I can print fine using one line of text, but I want to print two lines (Monaco 12 point), which fills up the height of the label with a few pixels of breathing room, after I've adjusted the field size. When printing (and previewing), the top half of the first line is significantly cut off. If I make the body part larger and slide the field down, both lines print, but it is printing partially off (too low) the label. I don't know how (if even possible) a top margin can be adjusted in a preset. I next created a custom paper size, but I can't then select it -- it is grayed out. Mac OS X 10.3.8, FMD7.0v3. See attached pdf showing layout, preview, browse, and print preview (OS X). Suggestions? Thanks. Bob

x-posted to fsa tech_talk (no response there) Dymo File Folder Labels 30576 in a Dymo LabelWriter 330 Turbo, using the recommended 30327 File Folder settings. I can print fine using one line of text, but I want to print two lines (Monaco 12 point), which fills up the height of the label with a few pixels of breathing room, after I've adjusted the field size. When printing (and previewing), the top half of the first line is significantly cut off. If I make the body part larger and slide the field down, both lines print, but it is printing partially off (too low) the label. I don't know how (if even possible) a top margin can be adjusted in a preset. I next created a custom paper size, but I can't then select it -- it is grayed out. Mac OS X 10.3.8, FMD7.0v3. See attached pdf showing layout, preview, browse, and print preview (OS X). Suggestions? Thanks. Bob

x-posted to fsa tech_talk (no response there) Dymo File Folder Labels 30576 in a Dymo LabelWriter 330 Turbo, using the recommended 30327 File Folder settings. I can print fine using one line of text, but I want to print two lines (Monaco 12 point), which fills up the height of the label with a few pixels of breathing room, after I've adjusted the field size. When printing (and previewing), the top half of the first line is significantly cut off. If I make the body part larger and slide the field down, both lines print, but it is printing partially off (too low) the label. I don't know how (if even possible) a top margin can be adjusted in a preset. I next created a custom paper size, but I can't then select it -- it is grayed out. Mac OS X 10.3.8, FMD7.0v3. See attached pdf showing layout, preview, browse, and print preview (OS X). Suggestions? Thanks. Bob dymo_cut-off.pdf

I started out to whip up a personal-use database that would track computer network wall ports, network devices (mainly switches, potentially hubs, routers, and bridges) attached to them, and hosts (computers, network printers, etc.) My goals are: to easily find what is connected to what for network troubleshooting, host identification (when I'm given a report of an infected computer off a wall port, and where I may not have the MAC address yet in a database); and for bandwith planning. My initial thinking was that each wall port would connect to one uplink of a switch, each switch would have 0, 1, or many active downlink ports, and each active downlink port would have 1 host. Seemed simple enough, then the fun began. 1) I remembered that some hosts were connected directly to a wall port (no in-between switch, etc.). 2) A network device could have another downstream device (a cascaded switch, even though in practice I'm not doing that). 3) A wireless access point (WAP) has an indefinite number of downlinks, but this shouldn't be too hard to design I hope (using virtual downlink 'ports'). Note: since wireless hosts tend to move between WAPs, I'm not trying to show a relationship between WAPs and specific hosts, just to show that a wireless host has a potential uplink to an unspecified WAP. 4) Hosts can also be network devices, e.,g., an ethernet-connected host with a wireless card can serve as a WAP, so it is then both a host and a network device (a bridge, specificially). Should I simply represent it as two records, each depicting the specific role? After review, I thought that although wall ports are simply extensions of downlinks of 'corporate' switches (in closets) that I don't have any control over, perhaps I should design the database so that a wall port instead is just considered a downlink port of another network device (switch in this case) -- and then I can have a field that contains the value of 'corporate' or 'departmental' responsibility. I'm trying to define the entities, attributes, and relationships. Initially I was thinking that uplink and downlink ports would be attibutes of network devices, and uplink ports as attributes of hosts (unless a host should have a downlink attribute if serving as a WAP). My first go through is as follows: network_devices - attributes (I'm only listing some of the important ones) -device_id -uplink_ports (1 or many) -downlink_ports (1 or many) hosts- attributes -device_id -uplink_ports (1 or many) -(no downlink ports, unless I want one record of a host to depict it as a WAP) Each uplink and downlink port of network devices and hosts is either inactive (a 0 relationship?) or is active (a 1 relationhip; not: 1, many). So for active connections there is a 1:1 relationship between a downlink port and an uplink port. For WAP downlinks, I think I should think of it as having potential multiple virtual ports, each connecting to only one host. Any feedback in further defining the ERD would be very much appreciated. thanks FileMaker Version: Dev 7 Platform: Mac OS X Panther