* The Business of Software

A former community discussing the business of software, from the smallest shareware operation to Microsoft. A part of Joel on Software.

We're closed, folks!


» Business of Software FAQ
» The Business of Software Conference (held every fall, usually in Boston)
» Forum guidelines (Please read before posting!)


Andy Brice
Successful Software

Doug Nebeker ("Doug")

Jonathan Matthews
Creator of DeepTrawl, CloudTrawl, and LeapDoc

Nicholas Hebb
BreezeTree Software

Bob Walsh
host, Startup Success Podcast author of The Web Startup Success Guide and Micro-ISV: From Vision To Reality

Patrick McKenzie
Bingo Card Creator

5GL or scrap it?

Not a frequent contributor here, sorry, I'm a shy guy.

I'm developing a RAD tool for data driven Applications. Anything with tables I mean. Could never categorize my Tool properly though - why it is special and so on. Until recently.

There are two parts of it:

a) a Table builder/librarian - nothing new. My implementation is a WYSIWYG one, not the best probably, but it does what it supposed to do - Tables with views, forms and reports.

b) a way to combine those Tables into Applications. I assign kid-parent properties to  those Tables (plus a couple of more relationship types).  A few lines of text or a few connections on a graph.

If Tables prepared/saved as templates in a library (like Web Site templates), the process of choosing and clicking together a Web App with all the whistles takes  minute(s).

The price to pay for an Application  definition simplicity is complexity of run-time:
there must be a client and a server ones - in present incarnation both are about 30K source lines of Java.

The Table Builder is the most visible part of the tool. But the most important is the way to connect the Tables. I did not understand that myself and thus could not express either. The least verbose page with some screens is at moderngroupware.webs.com (just avoiding directing traffic to my website). The next best is fancydata.com/manifesto.html ...

OK, now the technical part is over.
I'm ending up with more questions then answers though - and asking you, Ladies and Gentlemen for advice.

Question 1) is there a technological value in the approach, has it been done before?

Question 2) if Q1 shows some value - what would be your thoughts on cannibalizing the trade?
How ethical is too even look it that direction? personally I enjoy 100-1000 productivity bust. (not selling much though - I'm a [very]bad marketer).

Question 3) an approach "develop first, sell later" did not work for me (surprise-surprise).
It most likely will not work in future either. So what would I do:
3a) scrap it and move on
3b) patent it (feels like patenting a periodical table in some sense - see pictures at modergroupware.webs.com)
3c) develop it again  - present version does not have that sexy tic-tac-toe builder (takes 2-3 years for just one API... say Java could be replaced by HTML5, ... by then)
3d) look for a VC/angel (did not work before)
3e) look for a team - that is the work for a team...
3f) go Open Source (or find a foster home for the idea)
3g) something else...

I like the Team approach (with all it's drawbacks).
What would be Your Judgment (Q1-Q2) and your advice Q3, please?

Thank you and best of day to you.
Alex Semenov Send private email
Thursday, March 07, 2013
Done before? - Of course it has! http://www.ironspeed.com/ comes to mind immediately but there has to be thousands of others. In fact I think every other programmer has come up with something like this at one point in there career. I've done one myself before I learned that its "a Really Bad Idea For A Business" TM.

Why so? Its really really really hard to sell. As you've found out.

Why so?
-Its hard to describe so traffic isn't going to come to you
-Your trying to sell to developers (who are a hard crowd as we can always do something better ourselves about 2 weeks)
- They have to develop considerable effort into learning your thing before we even see the value.
- Anyone who's been around the block knows that it will cause as many problems as it solves (yes you can do a demo in really quick but when it meets real world problems you will invariably find its hard to tweak - I know you will say yours is infinitely flexible but...)

So with that out of the way - *** Cut your losses and move on! ***

Don't mean to be rude with you I just feel very strongly that your wasting your time with it (just as I was with my pet project a decade ago)

Sincere good luck with whatever you do next and learn from this one - if you're doing it as a business rather than a hobby then figure out how to sell it first.
Ryan Wheeler Send private email
Thursday, March 07, 2013

Your post made me laugh because I had this same idea pop into my head a few weeks ago.  And I wasn't even aware it had already been done. 

The "difficult to tweak" thing is definitely true.  I have enough trouble with some of the current web application frameworks when I want to do something the original authors hadn't envisioned.
James A. Send private email
Thursday, March 07, 2013
" (not selling much though - I'm a [very]bad marketer)"
This is a really crowded market, and if you make any sales you are a better marketer than you give yourself credit for.  Your market is bigger than just your approach to DB application development, it is all and any approach to taking stored data and displaying it.

http://en.wikipedia.org/wiki/FOCUS has been on the market since 1975 although WebFOCUS didn't come out until 1997.  I didn't have to start using FOCUS myself until 1991, before that we had data analysts to do that for us.  Actually that was probably the beginning (for me at least) of downsizing meaning that I had to become my own administrative assistant. 

The world doesn't need a better tool for generating forms and reports.  The world doesn't even need a better tool for analyzing data, what the world needs is better data.  We have quantity of data approaching infinity and quality of data degenerating into worthless noise.

Back to question 3:  my suggestion is g), keep the IP rights to whatever you have done so far and whenever you are working for someone on a project and you can utilize your prior work, sell your client a license to incorporate it into whatever you are working on.
Howard Ness Send private email
Thursday, March 07, 2013

Thank you for suggestion :-). I'm aware of IronSpeed(listed in *brothers of arms* section somewhere at my fancydata.com)

How IronSpeed (or anybody else) automate say this:  I have a DB with 2 tables - Departments and Employees, which are related in one-to-many relations. Will IronSpeed generate all the necessary SQL statements if I try to delete a specific Department? What if Employees  is related to Employment Records, which in  turn could be related to Supporting Documents and so on?

From what I see (prove me wrong) - they will forward me to secluded *customize the code* section to hand code consistency part of hierarchical DB. And that is a simplest example.

I define arbitrary CRUD logic on arbitrary set of related tables - not UI (which is trivial).

Makes any sense or I'm missing something?
Alex Semenov Send private email
Thursday, March 07, 2013
> I define arbitrary CRUD logic on arbitrary set of related tables - not UI (which is trivial).

should read like this:

I define arbitrary CRUD logic on arbitrary set of related tables - not UI (which is trivial) by clicking on tic-tac-toe matrix. 

All "hard to tweak " CRUD logic is included. Sure I cannot completely automate first person real-time shooter - it took me 3K lines of code to tweak - mostly animation. Anything tabular  comes included.
Alex Semenov Send private email
Thursday, March 07, 2013
Simplify your message or scrap it.

For example, can you say the following:

Hey Ruby on Rails developers, would you rather spend two weeks on yet-another-CRUD-website or define the tables and be done with it in an afternoon while still billing the same amount? I thought so. Here is your secret weapon. Our product:

 - Uses ActiveRecord and efficient queries
 - Applies best practices in security (see <testimonial from security expert>)
 - Cloud ready: deploy on Amazon, Heroku or YoMammasCloud, no sweat!
 - Extensible without modifying any generated code
 - Automated

Here, look at this really cool demo that I generated by defining these 10 tables! If you can hack it, you get a free license. (or something).

You've got to hook people right away and I go to your website and I'm like "shit, this is something for IBM."

And yeah, your market is definitely developers who develop the same CRUD stuff over and over and over. Anything else probably gets too complicated. Cascading deletes like you described are about as difficult as eating a delicious piece of pie.
Bring back anon Send private email
Thursday, March 07, 2013
@ Bring back anon

There are two original questions:
Q1) is a technical one and I'm asking it on the technical forum:
is there any value in the technology?  has it been done before?

Q2) how to sell (announce, declare?) the value if any exists

Mr. anon more reflects on the second one - non technical and more practical - how'd you sell it if the Q1's answer is "yes". He's absolutely right. And I appreciate Mr. anon suggestion very much and sincerely - because it gives an advice and direction (too bad I cannot write like that, even if  I *borrow* Mr. anon great declaration - I'd spoil something else on the rest of the website). Should I hire somebody to do the whole site? and marketing? Likely yes, if the Q1 answer is Yes.

So, tech aces, please answer the Q1, if you will...

And correction:
"yet-another-CRUD-website " - the CRUD-website is done by just clicking  - means pretty much free (although HTML/JS front end is yet to be implemented).

The technology extends beyond building CRUD apps/websites.  CRUD app is a skeleton. The meat could be any small(or big) scriptlets on any language (I implemented only Java API, sorry).

The value I think is in the universal way to describe an App - CRUD and beyond. The secondary value is in the architectural draft (current implementation) on how to do it and how it might look or feel.

Thank you for input. I'm no salesman, too bad...
Alex Semenov Send private email
Thursday, March 07, 2013
The market is certainly crowded. Checkout SQLMaestro, they do have similar offerings. Also checkout jqGrid which is an open source project but over the years they came up with a nice revenue model.

Marketing seems like an uphill battle and I think either open source it and also provide a web based cloud platform where you can charge for a small fee.

ZOHO also offers something similar.

If you don't feel like open sourcing then is it possible to host it and let users signup and play with it or create projects with it ?

Once they are done you can let them download the complete project with generated code.
nilesh Send private email
Thursday, March 07, 2013

I appreciate all responses, looks like I cannot deliver a message:

> SQLMaestro, they do have similar offerings
I do not let a developer/user to taste any SQL - none. All generated on the fly, all executed on the fly, disposed on the fly - like any interpreter screens the actual *byte codes* (first approximation). There is a next abstraction level - another language, a single  statement of which is expanded into  dozens of  SQL statements. Nah, mine is not an SQL tool. More like those mysterious tools by Scopio from recent thread http://discuss.joelonsoftware.com/default.asp?biz.5.851346.64
where managers build Enterprise Apps without IT support.
BTW - a class of Enterprise Apps is not automated yet -  well, any App where grids could be replaced with 2D-3D environments (All the Dispatches and beyond). Grid is a difficult to pass paradigm. But that is another story.

>ZOHO also offers something similar
ZOHO is working with *not-related* tables - which is well, trivial.

I have a list of competitors at, well, fancydata.com/competition.html  - which are not competitors - we work at different layers.

> Marketing seems like an uphill battle
I can say - the battle is over. The intended commercial product is obsolete already - there are still no competitors there (Scorpio?) - but after I thought of the tic-tac-toe methaphor as a programming language - it is over (moderngroupware.webs.com - pics are there).
It is so powerful and beautiful... I'd have to redo the tool and runtime from (almost) scratch - 2-3 years of coding only - without a team it's a dead cat. +marketing +copywriting - definitely a dead cat.

What I'm saying - it looks like a silver bullet (again, I'm shy and humble) - but not for a lonely man with my skills set to develop.

Guys, it is a silver bullet - or I cannot deliver a concept? I do maintain that 3-5 smart developers could build a tool which will satisfy market demands on LOB Apps of such a country as China (which by site logs is more interested in my crappy site then any western country, one of which (Canada) I'm happy to belong to now... speaking about technological barriers BTW... )

Thank you again.
Alex Semenov Send private email
Thursday, March 07, 2013
"looks like I cannot deliver a message"
It doesn't matter how good the delivery is if the number of people who want to hear your message is too small to be economical.  In your case, the number of competing messages is huge.  This is not the formula for a successful business product.  If your code makes you 100 to 1000 times more productive, then go forth and be more productive.  Don't keep flogging a dead horse by trying to franchise your expertise.

I mentioned FOCUS for a reason, in the seventies and eighties there was a need for tools to get data out of datastores, but today we are swimming in a world of DB tools.  Data doesn't become more valid with better schema, and it certainly doesn't get better as we get more of it.
Howard Ness Send private email
Friday, March 08, 2013

I do believe there is business value in raising the level of abstraction where the existing level of abstraction is not appropriate - Exhibit A: development of business apps in Java (or any other 3GL). In a certain way, we have similar goals (even if our approaches are different). Take a look at:


I don't think that 4GLs are technical breakthroughs - this stuff has been done at exhaustion by many companies, and not long ago they were the preferred way of building business apps. The main challenge is getting developers to change their ways, given the momentum 3GLs currently enjoy in that area. My suspicion is that you need to make it feel obviously stupid for a developer not to use your tool. How you do that is yet to be determined - my plan is going to be to build realistic example apps and make it as easy as possible for any developer to come try and modify them as they wish - hence my choice of integrating a web-based IDE into Cloudfier.
Rafael Chaves Send private email
Friday, March 08, 2013
"I appreciate all responses, looks like I cannot deliver a message:"

No matter how cool the techniques behind a particular product maybe, the end users are just interested in getting their final outcome/results. They would less care about dynamic SQL generation, abstraction layer and so on..

As long as you give them a product that takes them from step 1 to step2 they will be interested. 

The optimist in me still thinks if you can commit 5 years as a side project, you will see a winner but for that don't get hung on the technical aspects

my 2c
nilesh Send private email
Friday, March 08, 2013
Looks reiteration of old http://discuss.joelonsoftware.com/default.asp?biz.5.844955.7 thread.

You still need a product that solves a specific problem and a website that communicates what your software. Website still looks more like some sort of scientific research software rather than a simple to use web application builder.
Sergey Kornilov Send private email
Friday, March 08, 2013
@ Howard Ness

I'm using it for 20 years for myself, just wanted to share - not selling anything. Wrong forum? Humans and rats using border species to try new things (wild West or a new food). Trailblazers often perish. General population under comfortable Gauss Bell either use the experience or not. I've dug a lot, found something promising, brought it up. The same way Ford, Sikorsky and Einstein could be directed to go use their innovations by themselves (not trying to get into those ranks of giants - I know my limitations). I'm too old to dug all that gold by myself - I found my samples and brought them here. Somebody will try that direction - that's what I want.

@ Rafael Chaves

3GL momentum and developers - right, no disruptive technologies are welcomed.  And I do have working apps which would make somebody look stupid - if they took a look.  I cannot make them and that is not my battle. Not a capitalist, just a *geologist*. But I'm willing to assist. Will check you site thoroughly. and thank you for the rear positive response.

@ nilesh

Thank you for optimistic assessment. Still would be nice to have a tech opinion. Not likely I'll spend 5 more years to digging alone. 

Here is a conclusion - the thread starting to go in circles.
a) Tech part - those who cared to assess (clips/screens/drafts) - keep it to themselves. But a majority do not try - appreciate the reason whatever it is.
Either way the main question is answered - is it possible and what  is the yield. Good news - nobody said it is nonsense.

b) Marketing part - has no sense to discuss without defining the value first. But that was the most engaging(unexpectedly for me).

I thank all the participants for interesting discussion. Those willing to explore more - please, contact. I'm willing to share.

Alex Semenov Send private email
Friday, March 08, 2013

This topic is archived. No further replies will be accepted.

Other recent topics Other recent topics
Powered by FogBugz