Please start any new threads on our new
site at https://forums.sqlteam.com. We've got lots of great SQL Server
experts to answer whatever question you can come up with.
| Author |
Topic |
|
matthewspare
Starting Member
3 Posts |
 Posted - 2005-08-25 : 15:29:24
|
| Here's a challenge I am facing right now. I need to develop a user interface allowing the user to be aware of 1.5 million distinct records that are Company names through a web app. Take a God's eye view perspective on this with me. Passing 1.5 mil recs to a drop down list for fast perusing is not preferable because of the 6 minute download time (I tried hehe). Anyone have any big ideas? Oh, by the way, multiply this question by 5. I have 5 fields with many records that I need to make the user aware of based on his selections.Thanks in advance |
|
|
spirit1
Cybernetic Yak Master
11752 Posts |
Posted - 2005-08-25 : 15:49:21
|
maybe paging? good search?Go with the flow & have fun! Else fight the flow  |
 |
|
|
Kristen
Test
22859 Posts |
Posted - 2005-08-25 : 22:03:27
|
| Hi matthewspare, Welcome to SQL Team!I'd go with the "good search" too. Even if it was instant scrolling down 1.5M rows in a dropdown won't help, and they only let you type the first letter to "jump to" one (I'm pretty sure that IE Drop Downs do this, whereas most Windows apps allow you to type the first few letter for a smarter "jump to").We put a [...] type button next to such fields, that springs up a popup with search criteria boxes, and then provides a results list - paged if you like, but TBH more than a few hundred and the user would be better to go back and refine the search criteria. In turn this can allow the user to use wildcards, or narrow the search using other criteria - Country or Town or somesuch.If you don't want a popup in your web app then you can use a hidden <DIV> within the page, or somesuch.Kristen |
|
|
|
AndrewMurphy
Master Smack Fu Yak Hacker
2916 Posts |
Posted - 2005-08-26 : 07:30:33
|
| Never mind it being a bad idea for 1 user, the problem would only get far worse with 'loads' of users generating such large lists...cache wouldn't keep up with the demand. just the same as a telephone book...nobody reads each and every page before seeking a specific item.maybe split the list into one keyed by 'A, B, C, D, E, F, etc"...ie each list at least 26 times smaller (on average). |
|
|
|
|
|
|
|
|
1.5 million recs with web based UI