Topics started by Erwin Ekkel
Looking at the process flow design screen you have no clue if a process action has added functionality. To make the design screen easier to read it would be really helpful to add some indicator that functionality is added to a process action. Either changing the colour of the action step or add an indicator to it.
Multiple open projects and how to handle themNext release
In the SF it is possible to open multiple projects or project versions. For example you can open functionality for version 1.00 and version 1.10 of your branch. However you are able to edit both. Sometimes if you need to check something with older versions/different branches this can become very confusing and there is the risk of editing the wrong branch or project.Multiple ideas: Do not switch the active project upon switching tabs Add some kind of colour (red) or a warning sign (red exclamation mark) to show the current tab is not part of the active project Make any tab that is not the active project read-only. Make it possible to switch all opened tabs to the latest version, with the click of a button
Currently when adding a user group to a user in IAM the begins on is set based on the begins on of the user. When the begins on period lies in the future this makes sense. When the begins on lies in the past this is incorrect. I want to be able to see when the group was added to correctly trouble shoot rights issues. I would suggest this logic: if begins on >= getdate()then begins onelse getdate()
Currently there are no limitations on constraints regarding the datatype. It's possible to create constraints on domains that make no sense, see picture below. Please add a validation for constraints that are incorrect and need to be corrected. These constraints cause all kinds of problems when deployed on the database.
When copying a table in the SF the references are not copied, however the indexes that are used for the references are. These indexes no longer make sense since you don't have the reference on the new table. So please remove these when copying a table.
Can someone explain to me how this is not considered to be a merge conflict? The column was renamed both in the trunk and in the branch. Should it not give a merge conflict since it was renamed in both? Trunk version 1.22 → branch 1.0 created Trunk v 1.23 column x renamed to xxtBranch v 1.01 createdBranc v 1.01 column x renamed to xxbb
During development (especially during the user test) you sometimes have to make quick changes to functionality. However executing the code from the functionality - result screen will do a drop and create, messing up any rights set to the procedure. It would be very helpful if instead the code did a if exists alter , if not exists create. Since this is probably not wanted since some people use the result screen to edit their code I would also be happy if there was a second button (generate as alter procedure). This way any rights set to the object will remain intact. This would save a lot of time during user tests and would decrease the change of errors due to overwriting the current procedure.
In Code history the first entry is the trunk project where the code was set to reviewed. I understand you don't want each iteration in code history. I would however very much like to have the starting point or the baseline. So if the code was created in version 12 but reviewed in version 15. you get two lines: update originated in version 15. --the version where the code was approvedInsert originated in version 12. --the version the code was first created Currently you get this, which is not the truth since development was started in version 12, not in version 15.: Insert originated in version 15.
When copying a table or a view a lot of things are copied with it, some wanted some unwanted. When copying a view the old template is assigned to the new view, in most cases I do not want this to happen. Same goes for tasks, when copying a table more often than not I do not want the tasks assigned to the new table. To make the copying more user friendly (and to give some insight of what it is actually copying) i would like a checklist with objects it is going to copy with the option to deactivate some. This way I would know copying a view also copies the tasks, table parameters, assign the template and copy the filters.
Currently if a control procedure is in development the severity given to it in validation is Error. In the current version this will stop the execute complete creation if you have abort on error selected. In reality I want to abort the creation on errors like database errors or major model errors, not on a procedure that is not yet set to "waiting for review status". So I propose to change the severity for control procedures in development to warning instead.
When executing source code there is a pop up that asks if a new project version should be created. This popup does not show on the initial execution fora new database. In this case the same rules apply, it is still best practice to start in a new project version if you want to do model changes or other big changes. So please also show this popup on first execution of execute source code. It is still a question so it can be denied if needed.
Process flows have an output tab with the general and task parameters tabs. The general tab is the first tab which only holds the status code. I think the most used option should be the first tab since this can save a lot of clicks on a day basis. I propose to move the task parameters in front to be the first tab since this is the one most used.
Assign rights based on objectDuplicate
When assigning rights to new objects you have to select the role's manually. It would be very helpful if there was an option to set the rights based on a different object. Adding this functionality would remove the manual labour to find all roles that have access to the reference object. example:task A on table T1 has full rights for the roles role1, role2, role5 now there is a new task: task B on the same table. I want to set the rights on the same roles as task A. So task B would be full access for roles role1,role2, role 5
Look up references also show up in Model Rights - New Objects. However the reference itself does not need any rights. If you do go to the object you get a screen which implies there are no roles in the project. Since there is no point in having this object appear in New Objects. Please hide these objects behind a pre filter.
In a perfect world you name every column correctly in the first try. However, more often than not software needs to be revised and columns need to be renamed. The only options I now know of to rename a column is to create an expression column or rename the actual column on the data model to create a new translation subject. Both have some disadvantages and feel very redundant for just a simple column rename action. What would help is the option to create an alternative translation object for a specific object (for example a column). Or choose a different existing translation object.
After creating a new branch or project version the code is not automatically generated. However a lot of functionality will not work until you generated the code. Please add a tickbox generate code groups to the tasks create new project version (also when a new project version is created from the creation process and to the create branch task. If you ask me this tickbox should be activate by default, but I like to hear your opinions on it as I mostly work on single person projects.
When the subject was changed (prefilters added, columns added, column order changed etc.). The way the screen was intended to be used can be changed as well. In this case the user preferences should be rebooted to default settings. The user is not always aware of a screen changing.To make this more user friendly it would be great if when a new project version is activated a popup will be shown on changed subjects after first opening the subject. The popup would give a message indication the subject was changed and it is advised to reset the screen to the default settings. This has 2 advantages: it gives the user a notification the screen was changed, and it gives the user the option to reset after a change to the screen.
Now that there is the option to save connections to target IAM databases/servers there is also the opportunity to introduce a proper OTAP environment. When setting up a target IAM you should have the option to select if this is a production/test/develop/acceptance environment. Each type should have it’s own set of options, for example: - Activate default system flows (you only want to do this on production in most cases). - Environment specific File storage locations.- Deactivate certain tasks that are not allowed to run on test/acceptance.
When renaming a column the column is treated as a new column. Meaning it will get the default value and not the old value of the previous column. To make the rename complete you have to also setup the data conversion to the old column. Please make the data conversion automatic (set the old column) on a rename since this is the expected behavior of a rename action.
The current deployer checks several things for indicium if IIS is present, .net core hosting etc. It would be great to also have a check if the licensing page is accessible. More often then not this is not the case. It would spare a lot of time during deployment if this could be checked with the deployer.
System flows are still quite hard to debug. It would be very helpful if there was a system flow log that logs each step with input and output parameters + flow parameters. Since this is only necessary when trouble shooting an issue by default the system flow is not logged. To activate the logging there would be a new checkbox in IAM on the system flows to activate additional logging.
Remember connection settingsCompleted
At the last step of the creation process (execute source code) the user has to connect to a database. One setting is remembered, I believe this is the connection set up in the runtime config? This is already helpful but it would be even better if all connections would be remembered and you could select one from the drop down list, just like how SSMS connections system works but then include the database. This would save the hassle of having to type the connection every time you deploy against a test, training, acceptance or production environment.
Custom system flow schedule in IAMCompleted
Currently if you want to have a scheduled process flow (system flow) you have to setup the schedule in software factory and then synchronize this to IAM. However if you want to deviate from this schedule you have to go back to the software factory create a new schedule or amend the existing one and synchronize to IAM. It would add a lot of flexibility to have the option to create a custom system flow schedule in IAM. Otherwise for larger applications you would have to create a schedule for each possible iteration. For this to work there needs to be an option to check whether a system flow schedule can be set to a custom schedule or not. So a tickbox in the software factory should be added "custom schedule allowed.”.
With the release of 2021.3 you have the option to run the complete creation instead of executing the deployment step by step (generate definition, validate definition, generate source code, execute source code, run unit tests). You have three options complete, selected and selected start with none. In my opinion a fourth option is missing. This option would be custom. For example some project always use smart upgrade method having this option selected by default would be helpful. Second. If there is a custom schedule this schedule should be selected automatically when opening the execute complete creation task.
Add templates with modification options to Code OverviewOn the backlog
Changing and finding certain objects currently is a hassle. If the template is not the object name you have to go to Code Overview find the object, go back to functionality find the control procedure and change the template. When multiple templates are assigned you might have to repeat this process several times. Or when you have to change multiple objects on the same main object (table/task etc.) Code Overview is a perfect way to get insight in the available objects. Why not give the option to modify the templates directly on this screen, or have a link to the functionality screen? This would save a lot of time and clicking between screens. This also makes programming easier since I often have to change the trigger, ctx, layout etc.. on the same object.
Login to the Thinkwise Software Community
Enter your username or e-mail address. We'll send you an e-mail with instructions to reset your password.