Strategies For Software Services/product Companies next decade




These requirement are going to stay for next decade:Strategy-Small1Where can Software services/product firms lay emphasis for next stage of development. Or the areas which will see maximum amount of work coming in future..

Or What areas of knowledge should software companies develop manpower on:
1. Game development and Gamification:


2-7. Each of the Seven areas in development:


As you read you realize software which can take advantage of multiple processor available on the devices None of sotware present in market today is written to take advantage of this fact. It may be possible an new language may come up to take benefit of this fact of we can still use old java/C++ threads more offen or we can distribute load on server by more specific COM/ DCOM or Distributed Common Request broker Architecture CORBA to processor level at server.. We have virtual switches and VM ware or Zen virtualisation which can exploit maximum benefit from it.
8. More virtualised network stack: this I wrote 2 yrs back still valid to quote here:

private and public cloud new API will emerge:

9. from SDLC V model to Agile and now to lean Agile ..use of six sigma to control process is just one part of mathematics being used for quality control but there would be new data model which will be tested based to mathematical modelling like probability distributions new model industry specific models would keep emerging.
like how for security project how security user stories are plugged into model
or read

10.  BI would be Everyware:
parallelism , map reduce algorithm and cloud

Ubiquitous Computing is were everyone is moving now

Ubiquity in next frontier where software is moving what are important characteristics of ubiquitiy

If we see here how different stack are built over a period of time For instance: Oracle Stack from storage using sun technology and data base oracle in middleware: Oracle fusion middleware, Operating system solaris, and ERP solutions like peoplesoft, Sielble, and Oracle financials and retail apps..On all these areas solutions should work across what was missing was communication piece for which also Oracle acquired lots of communication companies…Now Same way

Microsoft Stack: Windows OS server /networking , HyperV hypervisor,SQL server database, biztalk middleware,MSBI Bi, dynamics as ERP with financial/CRM etc module..there is PAAS which can leverage this all across Called software are cutting these boundaries..

If we take definition of Ubiquitous computing it collective wisdom of moving toward miniaturization, inexpensive, seamlessly integrated and wireless networked devices working on all daily use items and objects like watch to fridge etc..same vision on which long back

all models of ubiquitous computing share a vision of small, inexpensive, robust networked processing devices, distributed at all scales throughout everyday life and generally turned to distinctly common-place ends.We have ambient intelligence which are aware of people needs by unifying telecom,networking and computing needs creating context aware pervasive computing. On back hand where we have all the data stored in cloud storage ..we have integrated stack..not every component of stack needs to talk to this new ubiquitous computing devices and software.

what technologies are colliding there:

Data communications and wireless networking technologies: moving towards new form of devices sensitive to environment and self adjusting , without wire connecting to each other creating meshup network. drive towards ubiquitious computing is essential to networks drive towards wireless networking.
Middleware: We have PAAS PlAform As Service in cloud mere all miniaturized device have limited storage will store data. To leverage this data as well to work all across the virtualization like we have Microsoft azure as discussed above and Oracle fusion middleware
Real-time and embedded systems: all real time messages needs to captured using Real time OS RTOS and passed to devices to interactivity with outside world dynamic.
Sensors and vision technologies: Sensors sense and pass information important part of ubiquitous computing.sensors in fridge senses out of milk and starts interacting with mobile to sent information to retail store to send delivery (its a typical example).
Context awareness and machine learning: device is aware whether its near to bank or near to office or police station and start reacting to relevant application this is geolocation..going deep watch when we go inside water start beaming depth from the river ded comes out and shows time..on same display context aware..still when it goes near to heat heat sensor sends temperature to display.
Information architecture: huge data will be generated from this network now this data needs to be analysed depending on its type its storage ans retrival architecture varies..big data will not stored same way RDBMS is stored.
Image processing and synthesis: and bio metric devices needs to get image of the to authenticate and send information. Image processing algorithm like edge detection algorithm will run over this huge data to get satellite data captured and fed into edge detection algorithm to find water bodies using huge variation in reflectance level as we move from sand to water..

There wold be huge usage of there in next generation BI systems.

So tools like uBIquity will make difference in future:

As BI becomes pervasive everyone would surely want to use it.. its natural evolution process for and user to get attracted to BI system where user can create his own query to find it become pervasive ti would enter into every device and here were it will strat interacting with ubiquity…ubiquity is future in BI.

Cloud innovation heating up network protocol stack and telecom stack

As cloud adoption picks up it will stir up networking stack..not only that telecom stack…precisely reason the great visionary bill gate picked up skpe for acquisition…and unified computing is at play again.

Here are Cisco back up plan how its affected.: Cisco’s imediate threat from software driven networking

How Data center market is affected?

Through compute virtualization decoupling the operating system from the underlying hardware, compute virtualization provides an extremely flexible operational model for virtual machines that lets IT treat a collection of physical servers as a generic pool of resources. All data center infrastructure, including networking, should provide the same properties as compute will unlock a new era of computing more significant than the previous transformation from mainframe to client-server.

The advantages of this shift bear repeating: once infrastructure is fully virtualized,any application can run anywhere, allowing for operational savings through automation, capital savings through consolidation and hardware independence, and market efficiencies through infrastructure outsourcing models such as hosting and“bursting.”But, data center infrastructure is not yet fully virtualized. While the industry has decoupled operating systems from servers, it has not yet decoupled the operating system from the physical network. Anyone who has built a multi-tenant cloud is aware of the practical limitations of traditional networking in virtualized environments. These include high operational and capital burdens on the data center operators, who run fewer workloads, operate less efficiently, and have fewer vendor
choices than they would if their network was virtualized.

Problems with non virtualized network stack in data centre:

#1. Hardware Provisioning: Although VM provisioning is automated  to run on any server.But creation of isolated network (and its network policy) is done manually by configuring hardware often through vendor specific APIs.effectively data centre operations are tied to vendor hardware and manual configuration.So upgrades are difficult.

#2. Address Space virtualization:

VM’s next hop is Physical router in network.There are 2 problems which arise

i) VM share same switch or L2 network limiting there mobility and VM placements.In multi-tenant environment it leads to downtime

ii) Sharing of same forwarding tables in L2 or L3 so no overlapping IP address space.In multi tenant IP adresses should be as desired by customer.virtual routing and forwarding (VRF) table limits and the need to manage NAT configuration make it cumbersome to support overlapping IP addresses or impossible to do at scale.

#3. Network services tightly coupled with hardware design cycle:

Due to long ASIC design development times,so organizations that operate the largest virtual data centers don’t rely on the physical hardware for virtual network for provisioning or virtual network services. Instead they are using software-based services at the edge, which allows them to take advantage of faster software development cycles for offering new services