This year was fun!

As 2008 draws to a close, I thought we’d take a look back at some of Data-Driven Web Applications at Work’s most popular stories. These aren’t necessarily the most important pieces of news I reported this year, but simply the posts drawing the most attention from my readers:

Is Adobe Flex the next silver bullet?

Zoho Creator and Google App Engine. So what?!

Santa, you'd better watch out!

I would like to thank all my blog visitors for attention and all feedbacks, for it really helped me to better understand each product I’ve tested and gave a better perspective what each system was capable of.

I know that many vendors of web-based data driven products also review my testing results. So, if somehow I didn’t recognize it was worth mentioning and you consider it to be a powerful tool in any issue resolving, please, let me know. And I will definitely test it and review it in my blog.

If you have any specific feature which you consider to be unique and I have not mentioned it in my blog, but it is really of a help for your users, feel free and contact me. I would be glad to take a look at it.

Thank you once again and have really great holidays!

Santa, you'd better watch out!

So, holidays are coming up and web-based services are preparing for that.

Jared Leavitt from Access Communication (QuickBase PR agency) dropped me a link to an interesting case study: SantaBase - Holiday Management Application. But I must say if this case study was for real, our holidays would be in danger. Why? Just consider this:

Here I’ve found out in 2006 there were 73.3 million children under the age of 18 in the United States.

Well, is QuickBase capable of such data amount processing within one table? I guess no. Unfortunately I can’t check it using my trial account and it would be really great to get an official response from QuickBase.

But since I check everything in practice I tried to import in my app the maximum quantity of records it was possible. I imported it as 20 000 records at one setting, and trying to do it the 4th time I’ve got such an error:


This is my Application Statistics:


So, can Santa afford such an application?

The case study doesn’t show us the exact way how kids register their orders in QuickBase:

"Thousands of bags of mail containing letters to Santa could be eliminated (saving millions of trees) as children logged directly into QuickBase to update their gift wish list."

For better understanding of what is going to happen, you can simply check QuickBase pricing here.

If each child is registered user in this SantaBase App, Santa should get ready to pay not less than:

$249 / Month + 36500 packs of 2000 users * $5,000 / Month = $182,500,249 / Month

Even in case all kids log their requests from Santa's official site and he won’t register such quantity of users, it is neccessary to pay for 73.7 million of records saved in QuickBase anyway:

$249 / Month + 2920 packs of 25000 records * $100 / Month = $292,249 / Month

I don’t think Santa can afford it, especially now, in such difficult times.

I bet no product I’ve tested can solve Santa’s problem right now and offer him an appropriate solution.

Who can argue with that?!

Now I am materialized in 12 Seconds

Like many folks out there I decided to actually show myself to the world too, and 12 seconds is a great tool for doing that!

From now on anyone can check out on my channel what I am going to talk about next and find out if it’s of any use just in 12 seconds.

I am really glad to share my experiences with common users to make the app roll out a bit easier and faster. So, see ya!

Get through with unlimited users in Coghead

Is the statement "unlimited users" really true for some web-based services out there? And what is meant by this in Coghead?

Trying to figure out how this can be done and the peculiarities of the implementation I came across the following:

Actually, Coghead offers 2 ways for licensing an unlimited user’s access, that can really add to the value of your app. Here it is what is that all about and how it can be implemented in real-life case:

Coglets

A Coglet allows you to take a piece of Application and place it in your Web page or Google Gadget. Coglets can also be configured to allow anonymous users to save their information as a record in your Coghead Application without having to be invited to your Account.

Coghead API

Coghead provided an API that enables you to consume and publish data to your Coghead applications programmatically. Using that API, you can use your Coghead applications to power rich, dynamic websites and even integrate into other applications.

It is also necessary to buy Limited User Option Pack ($50/mo) which supports an unlimited number of users to access data via Coglets or API-based applications.

Conclusion:

This is a pretty good and comparably not that expansive solution to secure an unlimited users access to your application. If the other Coghead functionality features seem to satisfy your needs, of course.

Zoho Creator and Google App Engine. So what?!

I’ve been impressed of the latest hype concerning Zoho Creator deployment to Google App Engine. I just can’t grasp what that is all for?! Really...

I mean the very point of web-based data driven products deployment is that your application evolves simultaneously with your business and the way your business does (for it evolves all the time and reflects all the changes of its structure). This is why I can’t understand, what is the point?

To turn your flexible application into a set of source code on python for deploying it in Google App Engine? And what should the user do when the requirements are changed?

Contradictions of Caspio approach to support unlimited users

Check out Caspio Bridge approach to users licensing policy. The good news is Caspio Bridge doesn’t restrict the quantity of users (anonymous or authenticated). You can review the pricing here.

There is a "but" though..

On the other hand Caspio Bridge doesn’t have any build-in tools to work with authenticated users. Yes, there are some mechanisms to tune web user authentication and basic record level security but I think that won’t suffice for real-life app roll out, I mean with complex multi level access, not only to records, but fields as well.

Maybe everything can be programmed on the level of specific data pages, but my point is the flexibility of such a system will be limited, let alone the difficulties of its support.

My impression is Caspio Bridge is oriented on such apps: The main part is developed as a front end (it can be any listing, from simple list of the company news/announces to real estate listing with search capability, pictures and other stuff) and with one or several backends for different categories of users which manage this list. You can check it yourself taking a look at the ready-made applications.

Quantity, quality and how TeamDesk combines it

After the decision to stick with definite service the first thing we are interested in is, of course, pricing and the quantity of users one can add for the given price.

I must say that’s quite a task to find the best quality-price ratio.

It takes time, but I’ll try to spare you some, giving a brief overview of what major vendors offer.

It won’t be out of place to clear up the pricing policy and what to expect from the service in further app roll out.

In TeamDesk each user with an access to the app is considered as licensed user. I've checked the pricing policy as well.

So, if you want to grant data access for many users (without perviousely specified quantity) TeamDesk suggests such options:

Web-to-record - allows generating a HTML code for any form and from any table, and displaying it on your web site or using for data input in the app (for example, you can create different variants of feedback/contact form). This is how it works on Wine Tracker app example from TeamDesk Application Library.


Let's assume I wanna create a public form for adding new wine items into my collection. First of all I should create the form with the fields I want to enter and tune web-to-record for it:


After that by pressing Generate HTML button you get a code, which you can use for adding records into table.


If you are concerned about privacy, but you don’t want to pay for SSL certificate, use resource of your app and display this web-to-record form right there.


This is how it functions in reality:


Of course, you can color it and add different amenities (header/footer/images) in any HTML editor.

If web-to-record is not enough, there is more sophisticated variant of doing this - SOAP API. This is the set of interfaces which allow you to get and modify your app data. This method gives you an opportunity to implement whatever you want – but it requires programming skills.

Here is an example of its implementation through javascript in a given app. For those who want to get even more advanced variant on .NET there is another way to implement this.

Conclusions

TeamDesk has tools for work organization of unlimited number of users. The developers thought it out, well, that’s good. What is not that good is the fact it requires the basics of HTML or programming knowledge. Personally I don’t like it that much, and this is exactly WHY.

My point is TeamDesk developers should think it over how to grant unlimited users access without any programming skills.

Get real! Offline web app use? No sense!

Many developers out there are beating up their heads how to make the web application available offline.

But what for?!

Its like trying to find out how to get the car moving as one runs out of gasoline.

Of course, users face definite problems while using web services. The truth is nothing is perfect.

Yes, easy access means easy lock-in...

You have no control of your data and content.

And the service may go down any time!

Or the worst: what if they cut you off?!!

I’d stop right here.

Of course its not a pleasure to realize it can happen any second. But its just how the whole system is working, what can you do about it?!

Come on, guys! It’s like being scared there won’t be electricity as you wanna grab some tea. So, it’s high time to own a personal power station or nuclear heating plant, just in case.

So, the world has simply changed and one should just keep up with everything it has brought.

Zoho Running Total. Where is it?

As I’ve made up my mind to share info on the functionality features of the most widespread (or not so popular, but still powerful) data driven apps, I’ve outlined some, that can turn the app to a business rocket.

Such functions as many-to-many relation, master-detail form, dependent dropdown, find and merge duplicates function, simultaneous record editing by multiple users option, as well as calculation of running total are really important for efficient app rollout.

Of course, no system can be labeled as good or bad just after trying its’ single feature. But the picture if it can solve your specific need is clearer anyway.

This time I got round to check Zoho Creator running total capability. Or I’d better say incapability. I was not the only one who'd like to do this. I've found here that the user requested to implement this, and yet another request here that the feature is needed. Since 2007 by the way.

What I’ve managed to find out here is that another Zoho product Zoho DB & reports is able to calcutate running total. This is something, but the way of Zoho DB and Zoho Creator integration is not found.

So, running total.. Where is it?!

Why go online? Why SaaS?

Talking about efficiency, leveraging all opportunities out there and stuff only folks which are blind to the obvious can pass up the chance I am going to speculate on.

It’s all about taking your own space ONLINE.

Go and search an online environment of your particular sphere of activity, I am telling you. The market is heading online with the light speed. And you’d better too. It doesn’t really matter if you’re selling socks or own a grocery store or whatever. The truth is the faster you get there the faster you will witness literally the transformation of your business. No kidding.

One of the benefits I have found testing web services, is that it helps focus on what eventually you gonna get. And if you make the choice really smartly (what I am trying to give a hand at) a tremendous value can be added to your business. This is what I especially like dealing with SaaS. Just apply this tool wisely and you'll definately dicover new revenue streams.

But to do this you have to be online in the first place. Your business definitely should! It’s really amazing what possibilities online space contains.

Go and get it!

Is Adobe Flex the next silver bullet?

Having tested a bunch of products I'd like to mention a difference of Coghead. Its interface is completely developed on the basis of Adobe Flex. One can understand the case when it is used in multimedia rich systems like YouTube, but why to apply this to data-driven services?!

And why not just make a rich application to Coghead backend?

Let me tell you few points why I dislike sites developed exceptionally on Adobe Flex.

1. Non-standard gadgets.


2. Blurry text.


3. Useless context menu.


4. Browser's keyboard shortcuts and extensions don't work.


5. I hate "Loading" screen.

My opinion what if one is going to use flash, it should be in small elements around a framework of web standards such as HTML & CSS. Fast, clean, simple and effective.

Dabble DB not too strong on dependent dropdown?

Don’t wanna disappoint you, but this time I have no choice. Having spent some time within Dabble DB in order to clear up things with dependent dropdown function I’ve got deplorable results.

There is no such option at the present time, what is confirmed also on forum.

What I would like to mention, it’s only my personal impression, of course, but I think that Dabble DB is much more developed for data analysis, rather than for data input. I state this after reviewing different options of Dabble DB functionality. If you don’t take my words for this, you can simply check out actual Dabble DB testing showcases.

But if you need just to analyse the data there is another system I would recommend you to take a look at Good Data. I guess in future Good Data may be a worthy competitor of Dabble DB.

Why hate HTML? And who really should?

Undoubtedly, these questions won’t bother anybody until the user will have to deal with actual HTML script himself. But, should he?!

When reviewing services it comes down to scripting my enthusiasm and esteem of the program goes way down.

Just seeing that common user is offered to write an HTML script I can’t help but conclude that this solution can only add to customer concerns rather than simplify his daily routine and data processing.

The thing is that at this stage of application rollout nobody informs the user of all the amenities of forthcoming data utilization using HTML.

And I bet you can hardly find at least one vendor out there, which will let you know from the very start about all the concerns while dealing with HTML scripting. You can make sure yourself and try the program that requires even modest knowledge of HTML. As soon as at least a line of HTML script is written the user is left alone to handle all the issues with this if any changes. No wonder that some developers would like to replace this issue to someone else but themselves.

If you aim is to turn to an actual programmer I think you’ll be totally ok with this.

But if you are looking for the best way to smartly utilize the data and get the most of it I’d advise you to pass such option by.

Amenities of Coghead find and merge duplicates function

If you want to keep your business running like a well oiled machine you have to make smart choices. This you can do only in case you are informed of what’s going on out there, what’s new and what exactly to expect from the service you are going to stick to.

Looking forward to define the practical value of program through the capability of each particular one to find and merge duplicates I am heading to Coghead testing.

As long as Coghead doesn’t have any build-in tools to find & merge duplicates, to implement this, developers offer to use Action Builder for creation of custom business logic which you can then call from within your Coghead application.


I tried to figure out the best way how to tune find and merge duplicates function, so I asked Coghead support to give me a hand and point me in the right direction.

This is what Jonathan McKibbin, Coghead support representative suggested. You can check steps one should take in order to program the function in Action Builder:

"It looks like you'll need to at least use a Retrieve Record step to find all 'duplicate' records based on some sort of filter which defines what a duplicate record is (two fields with the same 'name' value?), a For Each loop to perform a set of steps on each record found in the search, then some sort of Set Value step to change values, and a Perform Action step to update/delete the records themselves. The action will vary depending on if you're trying to merge values from a duplicate record and the original or if you're trying to switch linked records from a being linked to a duplicate record to being linked to the original record."

I wanted to program Find & Merge action in my Invoice app I used for testing of Coghead master-detail function.

Unfortunately, my experience is not enough to implement this Action on my own. But I am pretty sure it can be done through Coghead, looking back on my running total calculation test.

If you don't want to waste your time with all this set up there is always another option Application Development Sessions, if you can afford this, of course.

Many-to-many relation matters

Choosing the next topic for reviews I made up my mind to clear up things with many-to-many relation of different services, as this functionality is used pretty often.

One of good examples of many-to-many relations lies in students registering for courses: one student can attend one or more courses, and one course can be attended by one or more students.

Or a customer may have multiple reasons for making a purchase, and a purchase reason can be associated with multiple purchases.

So, in the upcoming posts I am going to analyze the approach of various systems to many-to-many relation implementation.

Getting results that count

I can already tell not only common users can find reviews of product testing beneficial, actually the vendors of tested products also do. Having completed some work yet, I must tell it keeps many on tenterhooks and one can understand why.

If any program functionality sucks, and users are aware of that, the rate of the service goes a way down no matter what. I think it really spurs developers to upgrade the system and get rid of existing shortcomings.

Sometimes even I am surprised with results of such an impact. And I am not kidding here. Just take Zoho Creator and Dabble DB cases, which are conspicuous enough. To add missing functionality in a couple of days (like Zoho did) just to meet the expectations… It’s really something, I must say.

It’s yet another proof that the testing results really added value to products in all respects.

So, even picking on system failures is actually of a help for vendors, for now they know where to plug the leaks.