I agree - it's easy to cherry-pick examples that are well-suited to ORDA.  The 
ORDA engine is definitely fast, but our tests show that the classic commands 
are still faster overall.  You just have to write a lot more code.

In my tests in client/server mode (and 4D always demos ORDA in single-user 
mode), ORDA is usually about 10x faster than equivalent SQL, but still several 
times slower than classic (and OMG 10x faster! .. but we're talking 
milliseconds usually). If all you care about is raw speed, then classic 4D is 
probably still the fastest... but ORDA is very elegant, and it makes 
development easier and fun.

If you get into tokenization vs. strings, then the picture gets muddier... but 
once 4D gives us the option to have a text file-based structure then a whole 
new world of tools open up for maintaining and refactoring code (with a steep 
learning code). Times are changing. 4D is joining the rest of the programming 
language world (and not forcing us to rewrite....).


> On Apr 11, 2019, at 8:16 AM, Jim Hays via 4D_Tech <4d_tech@lists.4d.com> 
> wrote:
> I'll disagree a bit on the value of ORDA.  JPR showed some examples of how
> much faster ORDA can be than the old "classic" way.
> I'm really interested in putting these to the test in our vertical market
> app.   We don't have huge data files, but we do a lot of cross-table number
> crunching in a normalized structure that may really benefit.

4D Internet Users Group (4D iNUG)
Archive:  http://lists.4d.com/archives.html
Options: https://lists.4d.com/mailman/options/4d_tech
Unsub:  mailto:4d_tech-unsubscr...@lists.4d.com

Reply via email to