Learning Revolution

[Sarah Reichelt]

Hi Mikey,

Although your email doesn't actually include any questions or ask for 
advice, I thought I would jump in anyway as I have had a similar 
experience converting a large HyperCard stack to Revolution :-)

In HyperCard, searching through multiple cards was extremely fast. 
Something like:
   mark all cards by finding "stuff" in bg fld "Data"
would return the results almost instantly even with thousands of cards.

Revolution doesn't do that in anything close to the same speed. My 
stack was (I think) 3000+ cards. Search speeds on HyperCard were less 
than 1 second. On Revolution they went to about 20 seconds.

However, Revolution has one enormous advantage over HyperCard as 
regards data storage - fields can hold up to 4GB data instead of HC's 
32K. In my case, I restructured the data to be stored in a single field 
in a one card data stack. Where the original stack had fields, I used 
columns.

Make an interface stack that takes the data from a single line and puts 
it into the relevant fields, then have a save routine that assembles it 
all back into a single line and puts it back into the data field.

Searching this data field can be incredibly fast, using filter, offset, 
lineOffset, matchText or whatever.

As  general rule, I avoid importing HyperCard stacks directly. I find 
that Revolution's enhanced feature set usually gives me better methods 
for doing the same task, so a re-design is worth the effort, unless it 
is a stack I only need to run a couple of times.

Sarah

On 09/03/2005, at 2:26 PM, Mikey wrote:

> The (mission critical) app. that I'm going to be moving from HC to
> another environment has thousands or cards in it.  Actually, it has
> tens of thousands of cards in it, but I've separated them into
> multiple stacks because there are logical reasons (time based) to do
> so.
>
> So, one of the first things I tried in RR was to create a stack with
> around 50,000 cards, put the id of the card into a field, then
> randomly pick a card in the stack, change the value in the field, go
> back to the first card and do a find for the value I was looking for.
>
> Searching through thousands of cards for a value in a field, or to
> pick a string out of a text field that is several lines long is part
> of what this stack does.