Providing automated architecture as a service, Splicer converts native enterprise stacks into a low code, rapid mobile app and web development platform. Splicer’s data model automation mechanism works in conjunction with various stable, open source java server technologies and mobile app development tools. This powerful combination enhances open source, enterprise mobile stacks — including high-performance, java servers (J2EE) running on the Play Framework or Spring Boot. Moreover, our development system supports heterogeneous web and native mobile clients for a comprehensive, cross-platform development solution.
This technical video gives a demo of these Splicer features: https://youtu.be/sVwSiU3YHi0
Low Code Analysis
In a review of the leading “low code” platforms David Ramel explains we need them “to meet the insatiable demand for apps amid a shortage of coding pros”1, and includes the new hpaPaaS category. Jason Bloomberg further refines that “low code … centers on professional developer”2, whereas “no code” is primarily for the “citizen developer”. I like how he takes a jab at these confusing, endless product categories that have been introduced for years in this space: “the big analyst firms can’t make heads or tails of this mess.”
Good. So it’s not just me.
Being a full-stack developer, none of these MBaas, RMAD, MADP, aPaaS, low code/no code, and now hpaPaaS solutions make much sense to me. For starters, each time I look at one of these “platforms”, I’m overcome with the same feeling of “oh no, not another thing to learn”. Each one expects us to learn their way of doing things. While “low code” might help expedite development, it does mean learning their way of doing things — meaning “lock in”. They to go to extravagant means to claim they don’t lock you in, but it’s just not true. Dig deep enough and you’ll find it. They talk about “open” and even “open source”, but look closely: perhaps they are not open-sourcing the actual server! Anyhow, I’ve looked at close to a dozen of these things. I only have time to review each one briefly, so please chime in if you have an opinion and can help shed some more light on the subject.
The first product Ramel mentions is Salesforce.com. (Imagine being able to offer force.com capabilities for your own enterprise! That is exactly the capability Splicer gives you: the ability to generate custom SDKs tailored to your server data along with the entire server stack including APIs.)
Anyhow, he mentions “Force.com is the only platform that lets you build powerful enterprise applications without writing a line of code”. Again, it seems these types of systems are best suited for “citizen” development, so as a developer, I disagree with “no code” for several reasons:
1. Code is good. Code allows us to leverage concepts such as composition, inheritance, and polymorphism. The trick is to reduce coding3, not eliminate it.
2. “No code” systems ultimately lock you into their architecture.
3. Historically, these types of systems have disappeared over the long run, and open source devours them (at least this seems to be the pattern to me).
Regardless, regarding developers, Bloomberg says “It empowers them to build more applications, and to make them more dynamic than they could have any other way, even with a finely tuned DevOps culture in place” and “support changing application capabilities”4.
The best way to do this — and to remain flexible — is with an “optimized” open source platform. Choosing the right platform is critical to provide “business agility”5. Ok, sounds good, but how?
Over the years, developers have evolved toward integrating open source software as a basis for unique solutions. In fact, this is how system architecture typically gets started. Splicer follows that tradition by generating a custom, mobile architecture that you can then build upon. Splicer also generates React.js forms and features a custom, model-specific, data admin console, which will give you a solid start if you need to build a web interface to your mobile data. We use well-established, well-known, stable, J2EE-compatible pieces such as Spring and Hibernate that are already being used in java-based enterprises — for ultimate compatibility — with the goal of a common model to support cross platform mobile development. And the open source CAM GUI uses simple stylesheets for customized form generation.
At Splicer, we automate the most complicated pieces — the data models, associated APIs, and Object-Relational fetching. Then we deliver an open source stack that populates these models with data. So instead of committing to a closed enterprise mobile app development platform as suggested in the articles, you can pick whatever “best of breed” tools you wish for components like web forms or even open source mobile app forms. Since integration can consume as much at 80% of the development effort, Splicer automates that expensive piece. This provides the most effective leverage while allowing your team to continue coding as they normally do. Splicer is a simple and effective way to dramatically increase your productivity without sacrificing control, nor learning a whole new set of development tools.
– Client data models are native and accessible from all client code
– Mix-and-match with other “best-of-breed” tools
– We provide source code for the client-server stack
“Optimized” Open Source
– Splicer is open source optimized to deliver data to mobile apps
– Automates data models for cross-platform mobile app and web development
– SaaS version available
For more information, including additional features and benefits, please see http://splicer.io – the first open source, low code, enterprise mobile app development platform featuring comprehensive mobile app development tools.
3 Leading Enterprise Low-Code App Development Platforms Compared. For an introduction to aPaaS, see What is Low Code Application Platform as a Service (aPaaS)?↩
A quick aside regarding Outsystems: Ramel mentions that there are some issues with code getting over-written on the clients. We do not have this issue with Splicer because we generate Model code that is orthogonal to the View and Controller code. So, we don’t wipe out any code when you change your Data Model. ↩
Bloomberg: “Low-Code/No-Code Is Far More Disruptive than You Think” ↩