August 2, 200718 yr I have a database that has matured to the size of 120MB. The file includes the following and I was wondering if this is normal. 53 Tables 4 tables have 1000+ fields ea. 10 tables have 200-800 fields 250+ scriptes 140+ layouts 500 records in 1 layout 2700 records in 2 layouts 3650 records in 1 layout I have used the maintenance feature almost monthly to reduce by about 10MB each time. We have only been using the database for 8 months and it has grown to this size. Please advise
August 2, 200718 yr I'd say you have nothing to worry about. See here . AFAIK you should not be doing file maintainance and you should read this before doing it anymore. Michael
August 2, 200718 yr 4 tables have 1000+ fields ea. 10 tables have 200-800 fields Others might not raise the yellow flares here? However does this seem like a sub-optimal use of structure, not as if there nessersarily would be the big savings in the filesize, but the file in it self might prove pretty inefficient??? Could you explain the denomalized approach you've chosen ...sometimes could there indeed be valid courses ...but the number of denormalized fields points in direction of a pretty casual approach to structure - in my humble opinion. Tables holding beyond 20 fields max. would make me start wonder. Could you descripe you fields nature and their utilization, if say a fair share of them are lookups, would there big savings in filesize be instantly noticable, when structurlization kicks in ...similar might a bias toward unstored or refernced fields yield bytes upon bytes Honestly - I would have a hard time imagining relevant fieldnames of this scope which seems to be in the vicinity of 5000 ...if I weren't to name them (deliberatly making naormalization errors) would it be via numric postfixes. This means that the structure by all means already have gone vinegar leaving considerably space for some kind of normalization, if I were too tempted to neglect.... --sd
August 2, 200718 yr I was going to say that the question shouldn't be "Is my database normal?" But rather, "Is my database "normalized?", but the OP seemed concerned only about the physical size of the db. But since Soren brought it up... As AudioFreak already pointed out, the size of the db is no problem for Filemaker. But the db probably has other, major underlying problems that represent poor choices in design. Hard to say without know what the database is supposed to accomplish. The best question would be, "Does this database accomplish what it is supposed to in an efficient manner?"
August 2, 200718 yr This reminds me of a post on TechNet, something about initializing 400+ global variables in the Open script. I can't imagine the need for 400+ variables. How can you possibly keep track of that. I need a table... I agree that this certainly sounds like the data model isn't the best it could be.
August 2, 200718 yr Yeah 400+ on open is a little much, Load FM go for coffee kinda deal. I have a script which sets about 400 Globals mainly cuz it was written 4 years ago and I didn't know any better : and the boss insisted it look like an Excel sheet. I'm in the process of fixing that file.....lol
August 2, 200718 yr This thread reminds me of this awesome screen shot that Allan Hunter once posted on TechTalk.
August 2, 200718 yr I didn't know any better and the boss insisted it look like an Excel sheet Say you're bit of a trainspotter, then would a visit to NRM in York be manditory ... http://www.nrm.org.uk/home/home.asp ...but what the point here is, that if you attend the dept. for "carriages & wagons" will you realize that the earliest wears idiosyncatic similarity to horse drawn stage coaches. A metaphoric rub off you can't avoid completely, even R2D2 is such an association. The field vs. record ratio is in what this thread is about, obviously falling under such an idiosyncracy! --sd
August 2, 200718 yr ... and the earilest motorised farm tractors were "steered" using reigns with the operating walking behind the plough.
August 2, 200718 yr "A metaphoric rub off you can't avoid completely, even R2D2 is such an association." Astrodroids? Why not protocol droids as well? Edited August 2, 200718 yr by Guest
August 3, 200718 yr I have no idea what his last post meant, honestly. I think the recent "why can't this be Excel" posts have pushed him over the edge. Are you OK, Soren? How come you haven't helped with this thread?
August 3, 200718 yr I think the recent "why can't this be Excel" posts have pushed him over the edge Indeed! ...please note that in my genetic code isn't there much baldness to spot, however am I now left to be plucking mustage and eyebrows. What I however was on about, is the difficulties of escaping a design discourse, where some indeed makes sense to follow, while others are hopelessly cyclopic - but all this is much more eloquently put in this text than I ever could hope of doing - here: http://acm.org/cacm/AUG96/antimac.htm --sd
Create an account or sign in to comment