The Rain and The Shade

January 28, 2012

Mother of all useless User Interfaces

Filed under: Windows Azure — ovaisakhter @ 8:05 am

I recently was sent a phone from Nokia Connect for testing. When they sent the phone then sent an email with a DHL tracking number. You click this link and you all sort of information. I was very excited and was looking a this link quite often. I realized that this tracking page is a brilliant example of how clueless an engineer can be in capturing the user requirements.

The tracking entries started like Shipment picked up Hurray one useful information, then “Processed at Lambeth – UK” what? and after that there have been 10 similar entries. I got a glint of hope when I saw the entry “Departed Facility in LONDON-HEATHROW – UK” I thought super now it will be flying to Denmark but the next entry (after about 10 hours) was “Shipment on hold BRUSSELS – BELGIUM” ah ok so it so it goes to Belgium from UK but does it come to Denmark from there or it goes to Germany from there and then to Denmark I have no clue.

image 

I am thinking what would have been the requirement stated when this page was made. “User should be able to see the status of their shipment” and that is what this page is doing. But why exactly user wants to know/track the status of his shipment certainly to know when he will be able to get this shipment and that is one information that is missing from this page. Usually the installation wizards or data entry forms tell you that you are at the step 1 of 5 imagine a wizard telling you are at step 2 without telling you the total number of steps I am sure you will be as lost as I am.

The information in this tracking page may only be deciphered if you have in depth knowledge of DHL’s routes and the time it takes to go between each hop, otherwise you will be clueless. It is like showing the “Object null exception” as a message box to the user, it is like putting your log file on the user interface.

I am sure DHL spent a lot of money on this system and came up with this amazingly connected system which show step by step information to the user which has no practical use for him.

I can given few suggests on how these guys can improve the this page

  1. Put an expected arrival date
  2. Reduce the movements to only City level (if most cases)
  3. Add some type of progress bar on the page so that the movements can make sense

In the mean while this page will remain my reference point on how not to do a user interface.

January 5, 2012

Html In browser Database do we really need it?

Filed under: Html5 — ovaisakhter @ 9:31 pm

Last night I was watching Mel Gibson’s Payback on the TV. One interesting thing in the movie is that every one keeps on saying that Mel needs his 130000 bucks back and the poor man keeps on telling them its 70000 and in the end he gets the 130000 bucks.

I have been working on web technologies for last 12 years and in all this time I have never heard anyone saying “if only we can have a database in the browser”. We had discussed a lot of time the limitations posed by the limited client side storage available in the browsers, and I have heard things a number of things like “if only we could store a little more information on the client side”.

A number of technologies were created to over come the limitations of the traditional web technologies frameworks  like Flash and Silverlight are two good examples, but no one of these technologies tried to provide any features to provide a database on the client side.

We have gone leaps and bounds in terms of programming languages and tools that interact with databases. There are frameworks which will make your lives much easier. In .Net world we have technologies like Entity Framework which can simplify the data access code a great deal. Using type safe languages like C, Java, C# you can write highly maintainable and robust business logic code . Having database coded on the browser will make us a step in a back direction. JavaScript doesn’t strikes me as one of the most maintenance friendly language. Eventually we will see a lot of data access, business logic code written in a language not known for its maintainability which may result in a maintenance nightmare. I recently had an experience while making a prototype with two tables proved to be a nightmare. Every slight change caused me to change the code at a number of places and introduced bugs in the code more frequently then a code written in a more type safe language. Of course I would have used something like Entity framework if I was coding it for server which could have been 50 times efficient in coding and maintainability.

Usually the JavaScript developers are not the guys making a lot business oriented code, application written with these newly acquired power could suffer some week coding standards.   

The browser database is temporary in nature and is gone as soon as you wipe clean your browser history. This essentially means that the data has to be saved to a server based data source this means there will be a lot of code that will be written twice. 

New breed of client applications mean new breed of server side applications. Business applications that are built for connected access will not be compatible with the new breed of disconnected scenarios. The new applications made for these scenarios will have to take care of the problems which come with the territory of distributed databases hence increasing their complexity.              

After all this discussion one question comes to my mind is “Why now?”. Internet connectivity has been improving every day. There will be less and less business scenarios which will require disconnected access. For these rare scenarios there are already enough solutions. So feels rather absurd that connectivity is increasing every day and we are coming up with solutions to make disconnected business applications.

So as poor mister porter said we only need 70000 bucks which means we could have been satisfied with the offline storage and may not needed a full blown database in the browser.   

Blog at WordPress.com.