[AccessD] ...for AD and anyone with an idea

William,

    Thanks for detecting the issue. Slow speed is found attributable to the custom function having to run more than once. Also, DCount() was used so as to retain editability.

    As seen from Drew's post, he has incorporated certain modifications, using collections. As an alternative, I have modified the sample by using an auxiliary table T_FixAdd (having just two fields, i.e. ID and FixAddress) which holds the results of custom function, ensuring that this function has to run only once. First option button displays the raw data (editable), while the second option button displays groups of similar addresses, duly highlighted in color (non-editable). Zipped copy has been sent to you directly

    During tests at my end (Access 2003 desktop on Win XP), the modified db took 5 sec for 50,000 records. You might like to test this as well as Drew's file. Based upon your feedback, I could update the sample at Roger's site, either as per my revision or as per the one suggested by Drew.

Best wishes,
A.D. Tejpal
------------

  ----- Original Message ----- 
  From: William Hindman 
  To: Access Developers discussion and problem solving 
  Sent: Thursday, February 18, 2010 01:12
  Subject: [AccessD] ...for AD and anyone with an idea


  AD

  ...I'm trying to implement your Match Similar Address sample (from Roger's 
  site) in a client app.
  ...I'm working against a linked table w/38k+ records and it is impossibly 
  slow
  ...I imported the relevant fields into a local temp table and replaced the 
  DCount with a standard replacement function
  ...its faster ...but only marginally ...still impossible to deploy to client 
  users

  ...any ideas on how I can speed this up?

  ...I considered splitting the records using some criteria but that 
  essentially eviscerates the primary purpose of the code in the first place.

  ...any ideas appreciated ...it's a sweet little piece of code and works well 
  on small tables ...but that's not what I have.

  William