Contact us Blog

Technical information

Apsona for Salesforce is an add-on application that makes you significantly more productive when using salesforce.com.

Software components

The software architecture of Apsona for Salesforce consists of

The browser component. The browser component displays the user interface within the salesforce.com window, and manages all the user interactions. When necessary, it interacts with the data transport layer, which in turn uses salesforce.com's Apex API to communicate with the server. The browser component uses pure HTML, JavaScript, CSS (Cascading Style Sheets) and Ajax techniques for its display, and does not rely on any plugins.

Salesforce components. On the server side, Apsona for Salesforce uses the salesforce.com's native VisualForce technology. It comprises:

There are no triggers, validation rules, processes or other Salesforce components except the ones listed above.

Data access and access rights

Apsona for Salesforce relies heavily on the metadata aspects of the API calls (describeGlobal and describeSObjects) that Salesforce provides. This API includes information about what fields are accessible to and modifiable by the current user. This means that all of the profile information, sharing rules and access rights configured for the user are automatically inherited when the user accesses data via Apsona for Salesforce. In other words, the data and access rights available to the user via Apsona for Salesforce are exactly the same as those available directly via salesforce.com's native user interface. The user experience, however, is better (in our opinion).

Object visibility and access

Your Administrator can specify exactly which objects are visible and available via Apsona to a particular user profile, via the Apsona configuration mechanism. Moreover, if an object is unavailable to a user's profile in Salesforce, it remains unavailable in Apsona, even if the Administrator has marked it visible in the Apsona configuration for that profile. Thus Apsona cannot be used to subvert Salesforce's security mechanisms.

Software delivery

Our software is structured to deliver the browser-side components from our servers whenever the browser needs them, i.e., at the very moment when the browser requests the Apsona for Salesforce tab from salesforce.com. This architecture offers several benefits: Conceptually, this delivery model is virtually identical to the Software-as-a-Service (SaaS) model that salesforce.com itself uses, except that it is applied to browser-side software.

Technology platform

The core of our offering is a web application UI platform that constructs a rich Ajax-based user interface entirely based on metadata (i.e., descriptions of the objects such as their names, labels, field names and types). Our platform produces a user interface that addresses the most common data management use cases for web applications, such as the standard CRUD operations, import, export, reporting and analytics facilities. The metadata-driven architecture enables it to fully implement strong data validation at the individual field level (such as dates, phone numbers, email addresses, currencies and pick lists) as well as reference fields across multiple metadata objects. In addition, it exploits the reference field descriptions across the metadata objects to produce desirable features such as smart querying, filtering with arbitrarily many terms, reporting with multiple visualizations (lists, charts, pivots, and the like), child tabs, calendar views based on arbitrary dates, and lookup of cross-object references when importing or exporting data. The user interface also offers commonly-sought benefits such as retention of context, high responsiveness, lack of screen refresh, and a pleasing look and feel.

A large part of the platform is coded in JavaScript. It can run mostly in the browser, but we have also successfully deployed critical parts of it on the server side via the Rhino JavaScript engine for Java.

Because of its reliance on metadata, the platform can be configured for almost any existing web application. Integrating it into a new application is a matter of creating two components: the metadata description in JSON, and a data service object that can talk to the back end service for the application. In the case of salesforce.com, we are able to offer a similar integration. We rely on the metadata API (describeGlobal and describeSObjects) to dynamically create the metadata description for consumption by the platform. We have created a data service object as a wrapper around the sforce.connection object included in the AJAX API. This architecture enables us to defer to salesforce.com's metadata API for field-level access information (for data validation) and user security information (e.g., that a certain field is read-only or inaccessible for a certain user).

Data security and logging

An important consequence of this architecture is complete data security: customer data transmission is entirely restricted to between Salesforce and the browser, i.e., customer data never passes through any third-party servers. Even the generation of Word and Excel documents by the Document Generator add-on happens entirely within the browser. The only case where any form of customer data is transmitted to an external server is when a user invokes a Document Generator action to produce a PDF document. In this case, the Word document must be converted to PDF format (since that cannot currently be accomplished in the browser), so the Word document is sent over to our servers, which respond with the corresponding PDF version. To be clear: This external server is only used if (a) you have a license for the Document Generator add-on, and (b) you use the add-on to generate PDF documents.

There are two other types of information that we do track. One is user information, containing the user's full name, user name and email address, so that we can provision licenses for users. The second is usage information, such as the actions that users take within the application. For example, when clicks the "Add to Campaign" button, a call is made to the Apsona servers to log that action. Our usage logs enable us to quickly determine and proactively fix bugs in the software, as well as to focus our development and performance enhancement plans on the most-used and most-useful aspects of our applications.