domingo, julio 15, 2007

¿Es "viejo" el ISeries (AKA AS/400)?

John Ghrist, en System I Network, se pregunta sobre la persistente idea de que el ISeries "es anticuado". Idea que se extiende a medida que los nuevos agentes de decisión pertenecen a generaciones más y más recientes. Más allá de su desarrollo del tema, (orientado a cuestionar a aquél nivel de "expertos" que obran por lo que la última ola proponga), resultan interesantes algunos de los comentarios que otros le agregaron:

Stroud Custer:
How "modern" the System i is is totally irrelevent if it doesn't have a sizeable development community willing and able to keep up with the times. Most of the RPG programmers I've met in 20 years on the platform are hopelessly clinging to RPG III and have no idea how to use the system APIs. Many are hoping they can retire without having to learn anything new. I've moved on to the Linux platform, which has a very lively development community.

Natural 5250 to graphic without extra pay/charge is what System i needs.
We have HATS, which is too expensive for many organizations and works badly.
We have WebFacing, which works better though has many bugs and now since Wdsc7 there's a mandatory payment via WDHT runtime.
IBM is cutting off its own leg by trying to get too much money out of modernization, which is vital to the life of System i. Staying "greenscreen" means slow death. IBM knows this but still it seems to me that it wants every bit money it can get from those loyal customers which have stayed on System i.
Instead of giving cheap tools to ISVs IBM should be able to sell 'cheap-enough' solutions to (new) SMB customers with a modern look and feel.
WDHT w WebFacing and WebQuery with mandatory replacement (new costs to customers) of Query/400 are only 2 examples. And yes, I do understand the need of IBM to fund its own actions but that should be done via growth, not taking more from the existing customer base via mandatory payments, which will only lead those customers to other platforms.


I am in my third consecutive position where the "AS/400" is perceived as antiquated technology and due for retirement/replacement. I am constantly arguing that the AS/400 is antiquated technology as they replaced it with the iSeries and now System i some time ago.
The problem is that these companies have come to believe that the only UI available with this box is "green screen" and that's the main reason it needs to go and then stop investing in these applications. Unfortunately I've witnessed, on more than one occasion, these tried and true apps being replaced by enterprise apps written in C, VB, or JAVA that are actually slower, more difficult to use, and less stable than what was replaced and ultimately more expensive to maintain. The old rules regarding the replacement of software no longer seem to apply. These being that you need to replace an app when it no longer meets your functional requirements, has performance problems, or resides on outdated technology.
I think the word that needs to be communicated to executive management is that the tools needed to be successful exist on System i technology and that continuous improvement should be the driving force.


My current job is riding the iSeries into the sunset. The decision was made years ago, by the incoming CEO because he wanted a specific software suite. Nevertheless, the enterprise is large enough to use the iSeries in several other jobs. It will never be considered for the following reasons.
1. No one knows what the platform can do. The application suite to be retired was written in RPG III, and didn't get modernized until after the 2003 version. The system operators aren't certified, and don't even know that there is an "iSeries Navigator." The last RPG programmer preferred RPG III and had never used CODE400 or WDSc. I added an e-mail notification to an application, and everyone in the department was amazed.
2. The box costs too much. The VIP program is an "IF" structure. IF you don't have to pay a platform surcharge, because the application was already written for i5/OS, AND IF the application is database intensive so the included DB2 replaces purchasing a database, AND IF there are enough additional applications that can be installed that will result in removing existing hardware or forcasted hardware requirements, AND IF enough applications are ported to the i5, so that the i5 system support uses fewer job positions than the replaced platforms, AND IF all of the above doesn't push the i5 into a higher processor group that eats up any savings, THEN an i5 is cost justified. TCO is not static, and neither is availability. Improvements in both in other platforms has removed a lot of the cost justification in paying extra for an i5.
3. A major portion of the IT world hates the i5. There is an entire generation of IT workers whose education and experience would be discounted were the i5 to grow in any significant manner.
I personally own a i5 520. I love it, but when I went to integrate applications with Lotus Notes I ran into the same cost problems sited above. It was cheaper even with TCO to use another server for Domino, than to upgrade the 520.

No hay comentarios.: