Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]
Groups > comp.lang.clipper.visual-objects > #992 > unrolled thread
| Started by | "Jamal" <vodotnet_nospam@yahoo.com> |
|---|---|
| First post | 2011-04-29 19:04 -0400 |
| Last post | 2011-04-30 00:12 +0000 |
| Articles | 2 — 2 participants |
Back to article view | Back to comp.lang.clipper.visual-objects
bBrowser 2.0.8.0 RowSelect( ) "Jamal" <vodotnet_nospam@yahoo.com> - 2011-04-29 19:04 -0400
Re: bBrowser 2.0.8.0 RowSelect( ) "Geoff Schaller" <geoffx@softwarexobjectives.com.au> - 2011-04-30 00:12 +0000
| From | "Jamal" <vodotnet_nospam@yahoo.com> |
|---|---|
| Date | 2011-04-29 19:04 -0400 |
| Subject | bBrowser 2.0.8.0 RowSelect( ) |
| Message-ID | <ipfg5q$p4l$1@dont-email.me> |
I am using bBrowser 2.0.8.0 under VO 2837 and I faced a KID size limit exceeded messages. I traced the issue to RowSelect( ) callback method being called so many times during the Use(), Notify( ) and various other methods. In fact, the RowSelect( ) was called Number Of Records listed in the bBrowser upon initial load! Imagine if you have thousands of records being loaded! I used the RowSelect( ) to load related child data into another bBrowser window based on the row selected in the the first bBrowser window. In the second bBrowser clear scope and set scope based on data selected in first bBrowser. I initially thought based on the vague documentation that it is invoked when the user either clicks or selects the row via keyboard, but due to due to internal calls in various bBrowser methods, which I think are unnecessary calls, caused the KID stack to reach its limit. From the docs: "bBrowser:RowSelect() is a callback method that is called by the bBrowser whenever another row is selected." Since I have source code, I commented out all those unnecessary calls ( except in NotifyRecordChange( ) ), and loading the bBrowser is now faster and without the KID stack errors. The same goes for CellSelect( ) but I have no use for it at this time. Just thought to share this info. Are those issues resolved in 3.0 or 4.0? Jamal
[toc] | [next] | [standalone]
| From | "Geoff Schaller" <geoffx@softwarexobjectives.com.au> |
|---|---|
| Date | 2011-04-30 00:12 +0000 |
| Message-ID | <4dbb53eb$0$77724$c30e37c6@exi-reader.telstra.net> |
| In reply to | #992 |
Jamal. No - poor programming. You should re-enable the bBrowser functionality because it has other reasons for doing it that way. You are going to break other functionality. It is better to manage a 'state' flag to say your window is built and bBrowser has finished initialising. Now your "relation" mechanism can be enabled off that method. Another way to achieve the same thing is to turn off notification so that RowSelect does not fire. But the KID stack error is yours - you are not managing objects properly or you have regressive loops of some kind. Geoff "Jamal" <vodotnet_nospam@yahoo.com> wrote in message news:ipfg5q$p4l$1@dont-email.me: > I am using bBrowser 2.0.8.0 under VO 2837 and I faced a KID size limit > exceeded messages. > > I traced the issue to RowSelect( ) callback method being called so many > times during the Use(), Notify( ) and various other methods. In fact, the > RowSelect( ) was called Number Of Records listed in the bBrowser upon > initial load! Imagine if you have thousands of records being loaded! > > I used the RowSelect( ) to load related child data into another bBrowser > window based on the row selected in the the first bBrowser window. In the > second bBrowser clear scope and set scope based on data selected in first > bBrowser. > > I initially thought based on the vague documentation that it is invoked when > the user either clicks or selects the row via keyboard, but due to due to > internal calls in various bBrowser methods, which I think are unnecessary > calls, caused the KID stack to reach its limit. > > From the docs: "bBrowser:RowSelect() is a callback method that is called by > the bBrowser whenever another row is selected." > > Since I have source code, I commented out all those unnecessary calls ( > except in NotifyRecordChange( ) ), and loading the bBrowser is now faster > and without the KID stack errors. > The same goes for CellSelect( ) but I have no use for it at this time. > > Just thought to share this info. Are those issues resolved in 3.0 or 4.0? > > Jamal
[toc] | [prev] | [standalone]
Back to top | Article view | comp.lang.clipper.visual-objects
csiph-web