Subscribe for automatic updates: RSS icon RSS

Login icon Sign in for full access | Help icon Help
Advanced search

Pages: 1 ... 7 8 [9] 10
 on: May 28, 2020, 02:16:50 am 
Started by Reuben B. - Last post by Reuben B.

Can you clarify what options / style settings you are using as your question is a little confusing
what is in your .4st

You refer to "side panel of the Startmenu".  The SidePanel refers to an Application in a GBC
it doesn't belong to StartMenu

The point I was trying to make is that by using a Tree container /  DISPLAY ARRAY, you can do whatever Genero allows as it is regular 4gl code, and not something special like the StartMenu implementation.  So your question, can you make a TREE container disappear, so if you have

then you can use ui.Form.setElementHidden to show/hide it

or SPLITTER to allow the container to be resized or not


 on: May 27, 2020, 03:45:21 pm 
Started by Reuben B. - Last post by Candy M.
Thank you Reuben.    We had just asked Support in the last couple of weeks if there was a way to have a webcomponent or fglsvgcanvas with a StartMenu.

So I have 2 questions for this TREE method.

1)  We like how the side panel of the StartMenu can disappear.   In fact, we have the side panel hidden as the default now.   The user must press the Hamburger icon to see it.   So is there any way to have the Tree dialog disappear like the side bar?

2)   We also like the tabbedContainer attribute for the window, so each application shows as a folder tab.  Is there anyway to get that behavior with using the TREE?

Thank you and hope you are doing well!


 on: May 27, 2020, 09:34:16 am 
Started by Reuben B. - Last post by Reuben B.
When answering a question that involves StartMenu, I tend to find that my first response is “Have you considered replacing StartMenu with equivalent functionality using a TREE container?” 

The reason I favour TREE over StartMenu is that it allows you to combine your TREE with the rest of what is in the Genero developer toolbox.  The StartMenu is very simple but with that simplicity comes restrictions.  The TREE and its underlying DISPLAY ARRAY can be combined with other dialogs in a multiple-dialog solution,  and you get access to the built-in functionality of the TREE container allowing you to respond in more ways to the actions of the user.


 on: May 26, 2020, 06:54:53 pm 
Started by Jos? V. - Last post by Jos? V.

Im having an issue with the xmlnillable property.

The WSDL Im using explicitly says that the tag can be nillable:
<xsd:element name="processingMessageType" nillable="true" type="xsd:string">

The code produced by fglwsdl does not include the XMLNillable property on the record and then when serializing the xml using  xml.Serializer.StaxToVariable it throws the following error:
Element 'processingMessageType': nil not allowed

This seems to be solved by manually adding the XmlNillable property on the 4gl field but the manual says this shouldn't be needed:
When getting an XML transformation into BDL, the Web Service engine will not raise an error and the variable is set correctly whether the variable tag is missing or has xsi:nil="true" set. In either case, the BDL variable is set as NULL.

Is the manual wrong, or am I doing something wrong?

Has fglwsdl any flag that allows XmlNillable to be put in the correct fields?

I attached a small 4gl program that serializes two tags.
* Record ns1addV1Response has the xmlnillable property and works fine.
* Record ns1addV2Response throws the error.
* The xml file has both tags with nil="true"

The test was done with Genero 3.10.

Thank you,
José Virtuoso

 on: May 25, 2020, 08:21:58 am 
Started by Reuben B. - Last post by Reuben B.
With Front Call’s, there are three similar Front Calls, execute, shellExec, launchURL to do something on the front-end, the equivalent of a RUN on the back-end.  Which one should you use? and what can you pass as parameters to these Front Calls?

Only one, launchURL is available on all our front-ends so that should be your preference.  This is the equivalent of typing something into the address bar of a browser, or clicking on a hyperlink, and is capable or more than just opening a web page.

shellExec is the equivalent of double-clicking on a file in Windows Explorer, but can also do the equivalent of right-clicking on a file in Windows Explorer and selecting an action.

execute is the equivalent of executing a command at the Command Prompt application prompt.


 on: May 23, 2020, 03:43:50 pm 
Started by Efisio S. - Last post by Leo S.
Hi Efisio, you might explain what exactly you want to achieve by replacing the build-in message with a custom one.
If its more than colors/borders , so positioning and on screen time this might scratch the new feature area.
Are you perhaps after positioning the MESSAGE/ERROR near the current field ?
Regards, Leo

 on: May 22, 2020, 08:37:13 am 
Started by Reuben B. - Last post by Reuben B.
Occasionally you have a requirement to read or write a complex variable from or to a file.  This could be during debugging, you want to read a complex variable or perhaps compare it to an earlier value to find the difference.  This could be during QA where you want to populate some variables with some known values or again compare the difference with an earlier saved result.

There is a handy little technique you can use that involves JSON and the TEXT datatype to read/write any given variable from/to a file.  Use the util.JSON.stringify method to turn a complex variable into a JSON string and write this to file.  Going the other way read a file that is a JSON string and parse that into a complex 4gl variable.


 on: May 21, 2020, 07:00:37 am 
Started by Reuben B. - Last post by Reuben B.
One of the great things about our support portal is it gives great visibility to all the support cases from around the world.  I have observed a number of questions relating to what you can and can't do with TABLE’s.

This article goes behind the scenes and looks at the AUi Tree and Front End Protocol around TABLE’s, and provides a background as to how a TABLE can be used to render the result set of an SQL statement.

The key points being ...

  • A TABLE has the characteristic that the number of columns is fixed whilst the number of rows is variable.
  • The properties of a widget are the same for every row of a TABLE.
  • The Front-End Protocol only sends the visible number of rows for a TABLE to the front-end client.
  • The behaviour to send more rows when the user resizes the Window or Table vertically is only active when a DISPLAY ARRAY or INPUT ARRAY is active.


 on: May 20, 2020, 06:35:07 am 
Started by Reuben B. - Last post by Reuben B.
A frequent request with the COMBOBOX is to alter its appearance to make it more obvious that it is active.    Particularly in Windows 10 where its background color is the same as an active BUTTON, which is not the same as an active EDIT widget.

You could follow the advice in an earlier Ask Reuben and use Pseudo Selectors to override the native appearance when a widget is active and/or inactive IF the native widget allows it.  For ComboBox you can use another technique where we render an alternative ComboBox that we have made available via presentation styles.


 on: May 19, 2020, 11:02:25 am 
Started by Reuben B. - Last post by Reuben B.
A review of the fgl_zoom repository code reveals that the library has no form files?  What form is being used to display the list of values that a user can select from?  The answer is that the forms are being created in memory dynamically as required for each individual zoom window.  How is that possible?


Pages: 1 ... 7 8 [9] 10
Powered by SMF 1.1.21 | SMF © 2015, Simple Machines