Way many to long agos and far aways, I worked in controls systems. The particular object of my efforts was one of the "Atom Smashers" at a University research lab. The "powers that were" wanted (reather obviously) the maximum amount of flexibility in the set up and operation of the complex, and this 'dictated' the omission of a great number of what I considered 'reasomnablness' checks in parameters. On the other side, this same committe of "powers that were" also desired to minimize the operating expense (read salary) and thus intended to entrust the momentary operation of the complex to usually grad students. In the 'discussion' of these (somewhat conflicting) objectives, I stumbled on the somewhat obvious theorm that there is a simple trade off between 'ease' (read education and training necessary) of operation and flexibility of use.
Today, as a 'programmer', I see the same issue - for example with the selection of ADO vs DAO. ADO has, perhaps, some additional degree of flexability at the 'cost' of some additional knowledge requirements of those who implement it and certainly some additional loss of speed. On the other hand, DAO can -within limits- provide a simpler and faster interface for accesssing certain data objects.
To some degree, MS apears -to me- to continously searching for the single 'rosetta stone' of each process (e.g. Data Access). At the moment, 'we' are being urged to use ADO, and offered hints that DAO will (soon) become obsolete. I would wish that MS could get its' collective mentality that several or more approaches would be useful. At least one 'tricycle', one street bike (with optional training wheels) and one speed racer approach.
The choice -and its implications- would be readily available and perhaps come somewhat closer to achieving the "goal" of universal data access.
I would generally expect DAO to be somewhat more 'efficient' in access to structured data and ADO to mostly be slower in data operations. In ONE recent thread, there was even some discussion that DAO 3.5 was considerably faster than DAO 3.6 on a relatively significant database transform, and DAO 3.6 was REMARKABLY faster than the DAO (3.6) approach
MichaelRed
m.red@att.net
There is never time to do it right but there is always time to do it over