I'm having some performance issues in my database perspective in PDSOE. I believe there was a regression related to connecting to databases to view schema. Below is a screenshot with an explanation to follow.
My recollection is that, in prior versions of PDSOE, the filters for the database tables would not only be remembered from one instance of application to the next, but the filters would be *APPLIED* and the schema would always be limited based on those filters.
However, today the filters are not applied at the point when we first connect to the database in PDSOE 11.7.2. Instead I am forced to wait for a whole minute for the "DB Structure" view to refresh itself with a bunch of schema I don't care about. The filters won't take effect until after I proceed to open that filtering dialog and simply click the OK button. That will finally apply the filters that should have already been in effect to begin with.
Hope this is clear. I wonder if this filtering feature isn't used frequently. Perhaps most ABL developers are still using the old "data administration" or something like that. Any pointers would be appreciated.
Hi,
Seems it is not a regression. This is an existing issue from the old releases. I have verified the behavior on different releases and the filters are not applied as soon as we connect to the database. We need to refresh the database for the filters to be applied. This is a bug.
Thanks,
Ramadevi
Hi,
Seems it is not a regression. This is an existing issue from the old releases. I have verified the behavior on different releases and the filters are not applied as soon as we connect to the database. We need to refresh the database for the filters to be applied. This is a bug.
Thanks,
Ramadevi
Thanks for taking a look. It seems like this would be a fairly popular view in PDSOE, and any improvements would be appreciated.
I'm a bit surprised that other developers have not complained. Perhaps waiting a minute for schema is a relatively small problem. Other development activities, like compiling a project take vastly longer.
Or perhaps there is another factor involved that made the connection process slower in recent releases for large databases. In theory it should be instantaneous to get a list of tables and nothing more should be done until the user clicks on a desired table.
I have found a work-around that involves collapsing the tree in the DB structure window first, then connecting, then opening and closing the filters to make them take effect, then expanding the tree again. It is a bit of a pain ... but saves a minute of my time.
One of the reasons others have not complained might be because that this require an SQL connection (most probably using JDBC to fetch meta-data) and most are just using the 4GL engine, not all databases have SQL users defined.