Techcrunch just announced a new entrant into the DIY solution creation tools space:  Longjump

I ended up writing a long comment that I thought I would share here in the sandbox. 

-- 

I think many of these solutions are slightly off target with the way they are approaching the problem. DabbleDB might be the closest.

There are already enough “high level” languages to address the range of programmers from hobby coders to full time coders (Ruby/PHP -> Java/.net)

If solutions are to move to the next level of abstraction then the solutions need to be built at a higher level with the components broken down into stages:

New Information Acquisition ( Data Input/Import )
- New Information Forms
- Questionnaires
- Import of Data using standardized import templates with deduping/updating capabilities

Business Rules
- Simple Logic Statements that dictate how data is processed and when (Filemaker Pro’s Model is a good starting point)

Summaries and Reports
- Drag and drop creating of Reports (Very much like Crystal Reports)
- Report Scheduling
- Report Bursting (creating multiple reports from the same master report which have unique permissions and data for each)
- Ability to maintain the presentation pages for these which can be embedded into other CMS’ like Squarespace
- Export to various formats (embedded html/RSS/XML/RDF )

Workflow/Change Approval management
- Ability to pause for approval during different parts of the workflow through the solution

Note that none of these solutions approach the problem the way a coder would approach this problem - it is more like the way Quicken/Quickbooks approaches the problem of collecting, organizing, processing and reporting lots of information.

In a perfect world, users create their “app” by being guided through the stages of the solution and hopefully having more value on the other side of the experience.

I also think a great solution would try to encourage reuse of data and logic wherever possible by encouraging Yahoo! Pipes like connection of your ’solution’ which others. If all works well, niche solutions should develop, evolve and create more value for the ecosystem than any one person could create. This would let users focus on the piece that matters most to them - their unique data and business rules.

As for the people who think these solutions are not needed, I would say that there are lots of people like myself who understand how a solution should come together - but just don’t want to spend so much time in a traditional programming language and instead want to quickly build a proof of concept that can evolve into a workable solution and if the need arises can have parts of it replaced with more enterprise class code as scalability needs increase.