Submitted by covingto on Thu, 06/02/2011 - 22:26
Red-R provides several mechanisms for saving and loading of data. As initially discussed data is analysed in pipelines and users can augment these with pre-made templates. Pipelines are associated with data, which means that the widgets are loaded with the data that was in them when they were saved. This is in contrast with templates which are not associated with data, even if there was data in the widget when the template was saved.
Pipelines are the major interaction that users will have with saving and loading. When loading a pipeline using the "Open" dialog, the existing schema is cleared and the old schema is reloaded as it was when the original was saved.
Templates are small collections of widgets that perform a specific task. For example, users may find that they connect a few widgets with the same settings often and it would be nice to load this mini-pipeline as it's own entity. Anyone can make templates but the way to do this is quite different from normal saving.
To save a template the user will first select the widgets that will be used in the template. Once selected, the user can click the Save As Template button. A dialog will appear into which the user can enter tags and a description. The description is used to help searching for the template later in the search-bar, so entering a concise message is a good idea. Templates may also be included in packages that users download from the repository.
To load a template the user can either use the search-bar or load using the "Open" dialog. The search-bar only indexes templates saved in the packages or in the Templates directory (Documents/Red-R/Templates) so users are encouraged to only keep templates in the Templates directory to get the most use of them.
Templates can be loaded several times into the same schema with no problem.
Pipelines can also be imported. This is similar to regular loading except that the new pipeline is loaded into the existing pipeline with settings and data. This allows users to include previous work into pipelines that are under development to compare data from several pipelines in a single pipeline.
A pipeline can only be imported once because otherwise there will be duplicate widgets in the pipeline which isn't allowed. Besides, the data and settings are already there so there would be little advantage in doing this.
Please post any comments or suggestions to the blog.