I have been involved in partners and sales at App-DNA™ for a few years. I therefore meet a lot of system integrators and customers who are actually planning and delivering Win7 transformations. In fact thanks to the appeal of our great piece of software, we are involved in the majority of major projects taking place at the moment. That gives us THE broad view of what’s working or not on the Win 7 front - on a real time basis.

I will try to share some of this insight and opinion in my posts, to hopefully open the debate with all the other professionals working in this space. Below is my classic migration process scheme:

And my first post will be on the most commonly ignored first trap: Discovery

Customers’ portfolios are often not exactly well organised...The good news is that it is messy in almost every customer organisation, but the most IT centric.
 i.e. nooo.....don’t feel guilty about your ramshackle backyard, your neighbour has exactly the same concerns 

More often than not some Central applications are well managed but the customer has only a (very) rough idea of the total number of apps in the company. They would also like to take this number down aggressively as it is a key cost driver.

To compile a list of all the .exe on a variety of machine is easy. ACT will do that or you can use a number of more sophisticated tools to achieve this Inventory stage. Ok, that was easy.

Now you have a 10,000 exe list for your 30,000 user organisation.... A quick clean up to eliminate redundant records and you are down to 6,000... A deeper cleansing would take you to 1-2,000. But how can the IT Dept understand what these apps are for? And which to keep or dump? 

They’ll need to ask the application owners for what the apps does, where it goes, who use it, how it install, the prerequisites, etc.... That is the Discovery phase. And customers systematically underestimate the effort and time needed for this task. Just because it is not rocket science does not mean it is not time/resource consuming.

One has to get every apps owner to fill a complex form and someone from IT to sit next to every owner in order to record how the install is performed. Assume ½-1 day per apps to contact the user, set a meeting time, get the info, record, the install, compile all that in a recordable form, etc... So ½ day of junior resources carried by the IT dept budget, and ½ day of the application owner time carried by the different BU in the company.

If you have 1,000 apps, you now have a minimum IT dept charge of 500 days. And an equivalent non IT department cost of the same on the company. Loss of other employee time does not show on the IT  budget but it is still a cost on the company. And it does not increase IT popularity internally. 

Assuming the IT dept has a few people on the bench. Maybe 5 people could be tasked to do this job full time? (not on top of their current job) ... We still face a minimum 100-200 days duration. That is 5 to 9 month elapsed.... Sounds too long, does't it? And it does stop everything else from progressing: can’t budget properly, can’t RFP in the best condition, can’t start testing...and all that costs money as well as putting the project at risk.

So my advice: use brute force to solve the problem. Call the SWAT team in.

Bottom line:

Get working on Discovery and rationalisation ASAP.

Do not underestimate the nuisance potential of this task. Plenty of smart Enterprise customers have seen their project stall and fail due to the lost momentum on that first hurdle.

•      Discovery takes a lot of effort and potentially time
•      Discovery will condition the real beginning of your project
•      It makes the transformation  more manageable by reducing the number of apps
•      It reduces wasting money on apps that should be retired today

The good news: watch this space as App-DNA changes the game.

I've been a bit surprised by the volume of queries we've had about organizations' ability to deploy 64bit versions of Windows client and server platforms. In a previous post I provided some stats derived from our benchmark portfolio of around 5000 apps showing that around 17% had critical issues which meant that the apps had components that would not run on a 64bit platform.

 This gets varied responses from organizations, some think this is low and others feel this is high. An interesting corollary to this dataset was extracted to answer the question: "Does the age of the application indicate likelihood of 16bit components". Essentially is an older portfolio likely to have more 16bit components that will need addressing?

 Storing the portfolio DNA in the way we do makes digging answers like this out really easy. The graph below takes a sample portfolio and identifies apps which have 16bit components, groups them according to the year in which the application installer was created, and compares the number of apps with 16bit components with those which do not.

* the above is derived from a mixed sample of 6581 applications

The interesting takeaways are:

• Number of apps with 16bit components is a relatively flat trend but decreasing
• In general, older portfolios are likely to have apps with a higher prevalence of 16bit components.

Perhaps somewhat as expected but interesting to see it validated.

My name is David Thornley and six weeks ago I joined the team here at App-DNA. I've been quite active  with many of our customers and partners but wanted to write a quick blog to introduce myself  to the community at large. I'm joining App-DNA™ with nearly 13 years of experience in the Installation and Packaging space where I was a leader on software development teams for the InstallShield, AdminStudio and Workflow Manager product lines. As the lead developer on InstallShield, I gathered requirements from enterprises and ISVs worldwide and worked with Microsoft engineering and product management teams to develop InstallShield's flagship product targeting MSI. As the Engineering Manager and Product Architect for AdminStudio I pioneered an integrated packaging approach which now allows one tool to build any package format, be it MSI, App-V, XenApp Streaming or ThinApp.

In addition to my various roles in Engineering  I've had the opportunity to speak at worldwide events as an evangelist for MSI and Application Virtualization.  Most recently I presenting a session at Microsoft MMS 2009 one preparing applications for application virtualization and gave a keynote at the PDS Packaging Event '09 on the topic of desktop management, desktop migrations, application virtualization and Windows Installer best practices.

After spending a long time at the same company it can be difficult to make the decision to move on. Beyond that, many who ultimately do make a move find themselves having a bit of buyers remorse; meet the new boss, same as the old boss. Maybe I've had a bit of luck, but App-DNA is an amazing company. Everyone here  is bright and energetic and has a passion for the work that they are doing. I've really been impressed with the technology; the Application DNA that we extract into our database allows us to perform extremely detailed analysis on any application and generate unexpectedly accurate results for our compatibility reports. Great people and great technology in one place, life is good.

Optimization in the datacenter with Server 2008 R2 and native support for x64 from App-V and now Citrix XenApp 6 means server farms can be consolidated and delivery of apps into server based x64 systems now able to take full advantage of the platform.

Citrix CTO Harry Labana recently announced the availability of XenApp 6 and this is great news for organisations looking to take advantage of the x64 platform. Whilst it will be some while until the majority of applications are converted to native 64bit, we see a relatively high proportion of application portfolios which will run in 32bit mode. The challenge for organizations is assessing which apps are suitable and then grouping them appropriately as part of a server consolidation process.

Below is a view on a 64bit summary of a 5000 app sample. While a significant number of this sample of applications contain incompatible 16bit code or 32bit kernel mode drivers, knowing which 66% of the portfolio can be deployed to 64bit platforms makes the planning and decision process a lot more straight forward.

 Its also worth noting that while 32bit apps cannot individually address more than 3Gb of memory, in a hosted, multi user environment where the OS can support large memory blocks, the serving of higher densities of 32bit applications per server becomes possible.

Loads of organisations are really getting into gear for their Windows 7 migration and looking to do things smarter than last time. A lot is being written about how to tackle the process Ruben Spruit recently posted a great article discussing a broad range of challenges associated with all of the choices facing organizations right now. Its not just about the new OS anymore as Virtualization has become a very disruptive part of the process.

Virtualization presents many great opportunities to optimize various aspects of application and desktop management but understanding how best to handle the applications remains  the biggest challenge.

Over the years we have been involved in helping many organizations through the process and have compiled a migration checklist for Windows 7 to help structure some of the thinking and planning that everyone will face when tackling the migration.

I hope this comes in handy for those looking towards Windows 7

Hands up everyone who thinks Windows 7 migration is going to be a nightmare.

It doesn’t have to be at all. It’s a migration, not a death march. It may have some difficult moments but it’s not going to be keeping you awake at night. At least not if you do it properly.

We’ve experienced dozens of major app migrations over the years, so we decided to put together a checklist of tips and resources to make life easier for enterprises. You might have already thought of some of them, others you might have missed.

Do your migration right and you’ll avoid an enterprise-sized headache. Do it wrong and… well, you can imagine.

Get hold of the checklist here.