HTML5_Logo_2561

Developers should not believe the HTML5 hype

As an Architect and Lead developer, I have to sit down with the non-techie and convince them why I think a technology or framework is the right one for a project and then I build my team to deliver the solution. The past two years or so, the marketing engine of technology companies have been spinning at full throttle. HTML5 is a the future, everything will run in the browser and more blah blah blah.
Don’t get me wrong, HTML5 brings some good technologies to the web application developer. Nevertheless, I think there is a lot of hype as HTML5 doesn’t really bring anything new to the table.
First of all, there are multiple type of applications that we are all well-aware of;

  • Consumer
  • Business
  • Enterprise

Consumer Application

If you are building a consumer oriented application which doesn’t use any of the native functionality of OS, then I would recommend the application to be built in HTML5 and all the RIA fanfare that comes with it. You can best view this in the mobile space where developers build either native or web apps depending on the application requirement. Remember that web application (HTML5 ) do not have direct access to printers, USB port or any other hardware such as Bluetooth and network services. Simple word processing application that can be developed in HTML5 such as Google Doc (where this blog is typed from). I haven’t come across any serious application written in HTML5 or the likes yet (JavaFX, Flash and Silverlight are not HTML5).

Business Application

Critical business application such as POS which requires access to barcode scanner as an example can’t be written using HTML5. You can have a native application delivered through the browser such as Java Applets ( or JavaFX) using webstart. Java applets can access OS features and hardware and provides another layer of security. Business application are delivered in controlled environments, for example, the application can be deployed on Linux desktop in company “a” environment only. For as much “fanfare” one might create around HTML5, these type of application will not cease to exist.

Enterprise Application

Enterprise applications come in various forms, from desktop to servers. As this is a comparison to HTML5, I will only focus on desktop application. First, let take a financial company such as a stock brokerage firm. There is a reason why stock trading application run as closely to the OS as possible ( and also to the exchange), in one word, PERFORMANCE. Web browsers performance sucks regardless of which one you are using, JavaScript is just too slow to implement some of the logic. In the trading business, a millisecond is all that is required to lose millions of £. Can you for one second imagine building a Bloomberg trading platform using web technologies, that’s laughable. There is a reason why the finance industry are still using Java Swing as their desktop technology of choice.

Conclusion

HTML5 is a promising step in the right direction to building scalable robust web application but it will not obliterate desktop applications, not today, not tomorrow, not ever (really!?). We can already see it in the mobile space where developers rather write native application so that they can utilize OS features and hardware. Web applications cannot access you local files directory (I am not talking uploading a file to a site) to read or write.  A simple operation such as reading available space in a directory or writing a log to a local directory is not possible (again I am not talking about downloading or saving a page). Therefore, developers should not believe the hype. HTML5 is not the silver bullet and it is a shame when companies such as Adobe sends mix-messages by discontinuing their Flex offering. Anyway, Adobe has never been a big player in the enterprise desktop application market.

If you disagree with my points, feel free to share your thoughts.

Cloud computing vs Open Source – Most CIO cannot tell the difference

Few weeks ago, I posted a discussion here in a CIO forum on LinkedIn and frankly some of the comments are very interesting to say the least. Now I decided to bring some technical guys to the discussion table. My argument was and still is as follow
“I have been reading many threads about Cloud computing and I came to conclude that most CIO (maybe because of their business background) do not know the difference or real benefits of Cloud computing or Open Source software.

Hi, I am not expecting everyone to agree with me but here is my argument. Most CIOs are interested in Cloud computing because of cost. COST!!? That’s very interesting because to my humble opinion, this only applies to IaaS (Infrastructure as a Service). If you do not want to spend money on hardware and be left with redundant system then IaaS is a good approach but that’s about it. I work with Cloud computing (specifically SaaS) and this next one might hurt my business but it has to be said; SaaS is not cost saving when you comparing it Open Source alternatives that you can host on your IaaS system…. Read more here

I am not sure that they understood the argument. I was not expecting anybody too agree with me and I wanted CIOs in the group to share their insight not sound like a copy of a Gartner report. Before making any conclusion, keep in mind that I am talking here about Software as a Service (SaaS), as defined here. In network architecture design, the external network  is represented by the “cloud” symbol therefore I would have to agree Wikipedia’s definition of cloud.  It is a software served over an external network regardless of a private network (private cloud) or public network (public cloud). YES – the internet was built on cloud technology. If you use any software over the internet; be it email, search, video or social network, the system is hosted by an external party which you do not have control over. The name cloud is just a marketing gimmick because hosted applications have been around for more than 20 years since Compuserve and Internet Relay Chat (IRC & BBS). So you see, “cloud” is definitely not new. Therefore software delivered over the internet such as Google apps and Salesforce can be also deployed using Open Source alternative. Cloud promises the following: scalability, peace of mind as it maintained by the providers and lower cost. CIO, I will tell you that your software, that it is deployed in the cloud or not will not be able to scale beyond the hardware and network capacity. I have another question for CIOs:
Do CIOs care or know where all their applications are running? 
That question is usually left to the technical team to decide within the budgets of the project and any other company politics. The biggest failures of IT projects is due to the fact that the CIO (or “the business” as it is usually referred as) do not know what they are trying to achieve (lack or misunderstandings of requirements) therefore it goes past the delivery and budget. 
So where does Open Source fits in, one might ask? Open Source software allows a business to take best of breed software and customised them to their own needs for free (as in free beer). If you do not want to support the Open Source software you can get support services but the software is hosted on your own system including your private cloud (Infrastructure as a Service). Let’s put it in to context; if a company requires a CRM system – they can download vTiger, customise it and either support it themselves or buy support. The same goes for an ERP system like Apache Ofbiz and MySQL for database. I am in favour of SaaS offering, after all I run a SaaS-based company. Nevertheless, CIOs do not confuse SaaS as being cheap, if you want something that has a low cost of ownership, then I suggest you go for an Open Source alternative if it exists. So why do I say that most CIO cannot tell the difference; SaaS is not cheap and does not have a lower cost of ownership when comparing it to Open Source alternatives. And by the way, you do not need to have an army of developers to build and customise your Open Source software. Do you really want the SaaS provider on a shared system to have access to your LDAP system? What happens when you want to integrate or port your existing data to your new fancy cloud based system? CIOs should not try to lie to me. I attended many meetings and conferences to know that most of them just want to please the board. What are the differences between cloud and virtualisation? What happens to your existing hardware once you moved your applications to cloud, have you thought of using them for virtualisation? So please do not make uneducated comments as “I don’t know a single CIO who can’t tell the difference between Cloud Computing & Open Source – maybe I move in different circles.” I am not talking about dictionary or Gartner definitions.  
CIOs, after reading this and before replying you should go and sit down with your technical team and ask them the difference, only after that you can provide a good educated comment. If you want to be on the “cloud” train then you are free to do so but if SaaS is not part of cloud offerings, please feel free to enlighten me as I am always willing to learn something new.
Who should make the decisions to go cloud or not? You are the CIOs, you tell me. Is it the developer, the architect, the project manager or the almighty CIOs from his ivory tower? There is no trickery here, every fool can define a word – just do this; think of a word then Google using the following “what is ‘the word’” then voila your definition. If you need someone to school you up then drop me an email. Based on the comments on LinkedIn, all I can say is WOW. Leave the technical aspect to the technical guys and you can focus on playing golf.
Top-10

10 Things all Java developers should know

Since Java (I know it’s not an acronym, but it stands out like that) was officially introduced in 1995, it has changed the way most of us look at the Operating System. Bill Gate (how ironic) once said that it was not about the hardware but the software which will be the future. A decade or more later, the fifth employee of SUN, John Gage said “The Network is the Computer”. Fast-forwarding to the 21st century and John seemed to be right. Anyway, Java was built not to depend on an Operating System and deployed through the network. Java through its applet technology gave birth to Rich Network Application aka Rich Internet Application (RIA). Java is not perfect; or we would not have various releases and more on the way, but Java has given birth to a wide range of programming language (just Google it to find out more).

Without further ado, I am going to get back to subject. This is a brief article on what I believe that every Java developers should know regardless of their experience. I do not personally believe that someone with 5 years experience is not as good as someone with 10 years experience. We all develop our own methods of working but as a developer you need to stay abreast of your technology. So, here are my top 10 not in order of importance (or?):

  1. Remember the basic of Java language and OOP paradigm.
    Most experience developers seem to forget the theory behind the language. I am not saying that they are not good at their job but can they explain to junior developers why they have used interfaces instead of abstract classes or why implement a pattern over another one? As a programmer, you become very arrogant as you believe that you write the best code but in the real world, people work in teams with different skill set and experiences. It is important that you can backup your actions/ codes. A very simple question such as; when should I use a String object instead of a StringBuilder/ StringBuffer? You might take this question lightly but can you actually tell someone else the difference?
  2. Know your technology stack
    All developers have to know their technology stack. What does it mean? Java is not like other languages; Java has subsets such J2ME and superset such as Java EE. We have our own area of expertise but it is important to know the differences between the various sets of Java. Some basic questions such as the differences between SWING, Applet, Servlets, EJBs and JavaFX will boost your confidence. Most developers do not know how to tweak the JVM and the differences between the JRE and the SDK environment. Do you know why you need the SDK to be installed to run Tomcat but you only need the JRE to run an application?
  3. Experiment with various Java EE framework
    I am not asking you to be an expert in every single Java EE framework but it will make the difference if you are familiar with Spring and EJB. That should actually be the de facto framework that should be on every developers CV. Developers should know the difference between Java EE 5 (soon 6) and Spring. Hibernate is also brilliant and it’s used for data access but all developers should have moved to JPA by now. Hibernate also comply with JPA therefore there is no more excuses.
  4. Know a scripting language
    Java can be heavyweight for some simple tasks which can be implemented using a simple dynamic language such as Python, Perl(?) and others. I would also recommend to developers to learn shell scripting on their target OS.
  5. Know how to develop web services
    The network is the computer, therefore it is important to know the different web services framework available. Data are integrated through web services and opening your services to the “cloud”. SWING developers will probably not develop web services but I am sure that they will be connecting to data through web services clients. Understanding the difference between the standardised SOAP and non-standardised ReST will help you choose which is best to implement your services.
  6. Know how and when to multithread your application
    I have to put that in there. Developers should know when and why to multithread an application, thread inter-communication and monitoring. All developers, junior or not, should know how to write a multi-threaded application.
  7. Database development using JDBC and JPA
    This should be a development law. All developers should know how to write SQL queries and how to create databases. All enterprise applications store data in some sort of relational database systems and it is therefore imperative that this knowledge should be of second nature. Java EE 5 introduced JPA (JDO was there before) but it is not applicable to all situation. Hence, knowing the difference and when to implement one instead of the other is important.
  8. Know a client side scripting language and what is AJAX
    The network is the computer and Internet is the deployment platform. Java EE and its various framework are server side executiong which can put extra “load” on the server. If you are looking to move a cloud based system where the providers charges you per resources used, it might be wise to move some of the execution to the client side. AJAX has been buzzing the scene for the last 3 years and more. This is not a technology but a new way of doing something that already existed. There are numerous Java AJAX framework such as GWT and DWR which makes it easy to develop AJAX based application which are compiled to JavaScript. Developers should also know what is the AJAX theories.
  9. Know your competitors and do not take part on “what is the best IDE” discussion
    Java is not the only language that can do what it does. I think that Java is more mature and complete as opposed to other languages. Knowing the difference between Java and .NET or Java and Ruby is a good asset to have. You also need to know when and why to use one instead of the other. Please please please, do not get into “my IDE is better than x because…” discussion as it is good for the Java community to have multiple IDEs and framework available to use. Every tools have their place as for example JDeveloper is better than x if you are going to solely develop on an Oracle stack and etc…
  10. Know ANT (MAVEN?), TOMCAT and any other mainstrean application server
    ANT is the de facto build script for Java and its IDE-based development. Maven is becoming popular and soon can be as popular as ANT (not sure of its popularity in the financial sector). TOMCAT  should be immortalised as the based servlet container that all developers should be familiar with.

There are alot more to add but this is just some of the basics that I think all developers should have in their repertoire. Feel free to add to this list in the comments box. If I could had another one to this, would be; all developers not just Java, should know how to search the web and Google is your best friends (now support my advertisers by clicking on the links on the right 😉 ). I hope you enjoyed the entry and feel free to comments good or bad!!! are welcome.

36_1_37

London Start-ups TechWorld Event April 09

ETAPIX Global Ltd is sponsoring the London Start-ups TechWorld Event on 9 April 2009.

London Start-ups 2.0 TechWorld Event aims to build a platform where technology and Internet start-ups can gather together to exchange knowledge, network and reward innovations through funding partners.

If you are a Start-up company based anywhere on this globe and looking for partners, funding or just the right guys to transform your ideas into the real thing, then this event is for you. This is a chance to showcase your ideas, projects and or company in front of well selected panellists. There is only one catch; your business or idea has to be Tech or Internet related.

Our panellists have several years of investment and knowledge of Tech start-ups and can provide support through financing and workshops for your next BIG thing.

The videos will be broadcast live on USTREAM.com and all exhibitors will have their details forwarded to our network of partners for further promotion.

London Startup 2.0 TechWorld Event

Thursday 9th April 2009 at 18:00-22.30

EXPRESS BY HOLIDAY INN
LONDON – LIMEHOUSE see map

469-475 The Highway
London
E1W 3HN

We booked a conference room in Docklands London. We only have limited places available so hurry and get your ticket now @ http://www.urgoing.to/londonstartups
You can also join our hundreds of members on Facebook LinkedIn Smarta
or follow me on twitter @armelnene for an update.

Ticket : £15.90 (Exhibitors)
Ticket: £10.73 (Networkers and Guests)

TICKETS HAVE TO BE PURCHASED PRIOR TO THE EVENT.

NO TICKETS WILL BE SOLD ON THE DOOR.

Click here to see a list of the panellists including a short biography.

See you there and have your business cards ready.

I am not a geek… you are nerd