I finally got around to trying out the latest and greatest version of App-V. I’m very impressed at the efforts to improve the tool set, although, there is still some areas which require work but it’s nice to know that there is at least some progressive thinking around important areas. I actually had screenshots using 4.60.20200 which I was going to use for a blog on the basics of sequencing for beginners. I may still upload that also as I think it’s a more basic break down and more useful to someone who is just starting out. But for this post, I’ll go through sequencing with 4.6 SP1 and maybe touch on some of the nuances between this new version and the previous version for anybody familiar with App-V, that has maybe just not had the time to review the new sequencer.
For the purposes of this post, lets go ahead and choose to ‘Create a New Virtual Application Package’. We can see on the right side that Microsoft have introduced links to helpful sites for tips, recipes etc.
On the next screen we are going with the default option to just create the package. A new option for 4.6 SP1 is the use of a Package Accelerator which is the possibility of using a template from a previous application sequence. This is very useful as others can create Accelerators for tricky applications and make them available via the community sites. A great and clever introduction which could save time and money, if leveraged correctly.
The next screen is also new. A useful introduction is that on this screen if you have something such as the Windows Defender service running on your sequencing machine, which might cause sequencing issues, it will be flagged to you on this screen so you can disable the service. It’s also a nice little reminder. “Do you have your machine ready because we are about to start our capture/sequence”
Another modified screen here, it actually allows you to choose the type of application you are sequencing. This is because all of these options would mean a different method to be used. A standard application obviously being a straight forward sequence. An Add-on or Plug-in will require to be DSC’d or pointed to the parent application and the Middleware may be DSC’d to the dependant application which requires it.
This is pretty straight forward. If you have an installer file to point to, do that here.
The next screen is quite interesting as it suggests that Microsoft are shying away from how they use to handle the Root Package Name. It appears the 8.3 is no longer the best practice as it suggests that by Default the application name will be used. I like this introduction myself as the management of the 8.3 seemed a little pointlessly annoying. The option is still there however to define your own unique name that does not just use the Package Name.
For my example I’m just doing Mozilla Firefox as it’s easy to sequence. So you can see by default it’s populated the Root folder name. I’m using my virtualized firefox right now to do this blog so, so far so good. It worked out well, I was trying out 4.6 SP1 when I was also downloading Firefox 4.0 😛
When you are ready than move on and go through your install. You should as a rule of tumb launch the application maybe three times after it’s installed just to ensure you’ve captured any additional configurations required. It takes a little application knowledge to know what the good and bad things here are e.g. what might be a bad thing to capture
Carry out your stop monitoring which will analyze the files/reg etc. put down during install and virtualize them for you. Then you should have the option to run the application for capturing the Feature Block 1. The Feature Block 1 of course being the initial blocks of the application loaded down to appear to the user. This optimizes performance as the user see’s whatever menus or options we trigger during the testing once the application launches, in the background the rest of the application loads down what should be seamlessly to the user. It’s important to know the applications main functionality to capture that in here to ensure the correct options and menus appear for use when first loaded otherwise on a poor network the user may not see the menus or options they require until the application loads the additional blocks down.
Another funky new feature is the warning if something is detected to be possibly wrong during the sequence. It’s a nice feature however what I have found is it’s throwing some flags that may not be an issue as in the example below it’s saying it may not have included files that are important to the install. That is because the folder location was in my exclusion list as it was merely noise I did not want included anyway so I wasted my time looking. I can’t get that time back!!
So once all this has been configured the next step is for these changes to be saved as per the below.
Next I would go to customize to change the OSDs and other customizable options. The Default will just save whatever your default settings are.
Firstly Edit the shortcuts. Here you can change the icon, the shortcut location and File Type Associations. Also the shortcut name and OSD name.
I noticed that the sequencer picked up version 2.0.0 also I modified my OSD name to also reflect the release version and remove the spaces. I guess that is a bit of an old best practice but it’s still one my employer goes by because you never know what deployment method may have an issue with spaces in the path. None should but some do.
Next is for restricting the OS. As a best practice you should really restrict only to the OS you test the application on otherwise you could allow the app to be deployed to an OS on which it may not work.
The next screen is for the old comments section which ironically in old versions of the sequencer back in Softgrid days appeared on the first screen! Anyway it’s a useful field to use as I see it as my own little private properties which could then be parsed through and reported against later by using a common delimiter such as “;”
Package is complete and you may get some warnings flagged.
I would now like to offer my own little critique with this version and I know I might be a little negative considering the nice new changes featured such as LOCAL INTERACTION being included in the OSD’s by default now, the option of what kind of app it is, the community section, the two warning screens for things which may be a conflict to your successful sequencing, the 8.3 being changed etc. But I’m not sure I like the sequence of the dialogs. It seems now that if you want to say for example modify your path or port number, you need to go the Edit screen after this. What I’ve noticed is if you are using the increment system for your SFT file, by doing the edit you are going up to sft_2. Also, I thought it made sense doing this right after the monitoring anyway. I guess the idea is to prompt streamlining and to have this taken care of through the Default.sprj but I kind of liked the option.
I also find it a little odd that Microsoft seemingly went back on it’s own best practices for enforcing an 8.3 root folder name structure and the suggestion of streaming on the target platform. Now both changes do make sense to me so I get why they have made them but the fact they made the changes now makes me question the push behind having at least the sequence platform changed in previous best practice docs.
