diff --git a/howTo/sve.md b/howTo/sve.md
index 2618f541b4d2565ceb5cb09764e7d53f27a17f3a..fc876f2cd45b655a2b911ef360dd8675c793a2db 100644
--- a/howTo/sve.md
+++ b/howTo/sve.md
@@ -4,15 +4,24 @@
 
 A windows [virtual desktop](https://sotonac.sharepoint.com/teams/IT/SitePages/Services/SouthamptonVirtualEnvironment.aspx/)
 
-**Use the Win 10 Student service** as it has research & academic software suites such as RStudio etc. The Win 10 Staff service is a generic staff service intended for admin & professional staff use (for now).
+The SVE offers two services:
+
+ * Win 10 Student service - this one has research & academic software suites such as RStudio etc. But it does **not** have persistence. This means that any packages you install or any repos you clone into your local space will vanish when you log out. The only expection is if you:
+    * clone the repos to your MyDocuments/OneDrive account
+    * work out how to use your Windows profile to 'host' the packages
+ * Win 10 Staff service - this is a generic staff service intended for admin & professional staff use (for now). While it **does** have persistence, it does **not** have RStudio installed...
  
-## Why would I use it?
+## Why would I use the SVE?
  
 1. It hosts most of the applications you'll need including RStudio etc
 1. It offers easy access to [data folders such as J://](keepingData.md) as you are effectively working 'on campus'. This makes data loading fast. Well... faster than doing it over your home broadband.
 1. You can easily access your oneDrive folders.
 1. The virtual PC instance you log in to has reasonable memory allocation so unless you have huuuge datasets you should be OK.
 
+But:
+
+1. No persistence (see above)
+1. It does not seem to be able to 'mount' Sharepoint/Teams folders - so you cannot easily load data held in them
 
 ## Git
 Git is installed - yay!
@@ -22,6 +31,8 @@ You will need to create your RSA key so that git (inside or outside RStudio) can
 ## RStudio in SVE
 
 ### Packages
+See note above re persistence.
+
 If you need to add new packages use the install.packages() function. This seems to be able to bypass a permissons issue in C:/Apps/RLibraries which causes the normal the RStudio GUI/tab method to fail.
 
 Better yet use the [loadPackages() function](https://git.soton.ac.uk/SERG/workflow/-/blob/master/R/loadLibraries.R) developed at SERG :-)