I have read numerous posts from others deploying web solutions for their databases that are also used for Servoy solutions. It seems to me that one of Servoy’s great limitations is that it can only play with its own clients when it comes to data manipulation, at least without flushing clients and/or restarting the server.
In my case and I would suspect, in many others, this drawback is quite large when you consider the breadth of the internet and its uses - like taking orders through a web site or receiving form submissions from your web site - Servoy is the perfect solution except it doesn’t know that these new records exist so clients never see them without the above mentioned hassles.
So - would it be possible to have a method that basically looks for any PK’s that are higher in the external DB than those that are in the Servoy DB and then just insert those records into Servoy? Or would this confuse Servoy since the data already exists in the external DB? The suedo code would be something like…
find highest PK in Servoy for a given table
SQL query for any PK’s above the highest Servoy PK
Loop through dataset and insert those found records
Thanks for your input ![:D]( "Very Happy")
you seem to talk about that servoy uses is own database.
This is not the case, new records will be seen by servoy when you do a new search. New records aren’t a problem at all. Of course if you make new records from 2 places (servoy and a webclient) you must make sure that they use a db sequence value (auto increment column) as a pk. Because the pk’s have to be in sync if both are creating new records.
If you also update records in a webclient then servoy needs to be flushed somehow. Because we do cache as much as we can to get as much performance as possible.
Yes - I was under the impression that Servoy kept its own DB - good to know that it does not. I am letting MySQL keep track/do the PKs.
Okay - I think I get it - I was missing this for a while - so new inserts DO show up in new searches but updates do not because they are already cached. If this is the case then I am much relieved. I couldn’t understand how to utilize this tool for web dev now I feel much much happier. Thanks for the clarification.
Do portals also show new records when the form they are on has a record change - assuming the forms relationship causes the portal data to change - or do you have to do an explicit search on the data?
Thanks again for clarifying.
if you trigger a loadAll (ctrl-j) then related foundsets will also be flushed and searched for again, so new records do show up then.
In the further, we don’t know when this will be implemented, we are looking how much demand there is for this, we will build a headless client.. So that you can have X clients running on the server without having a GUI. But then you can update or flush records from the client running on the server.