Move Recommendations part down

This part was intended as a relatively brief overview of how to proceed,
in a manner similar to putting the conclusion at the top. It was clear
from the presentation that talking about "Oracle JDK" and "Oracle
OpenJDK" without attempting to establish those terms was confusing.
Perhaps talking about the distributions beforehand makes this easier.

Author: Mikkel Kjeldsen

notes/java-future.md

@@ -52,41 +52,6 @@ links-as-notes: true
     - Better understanding of impact of new strategy
     - Expected biggest obstacle: third-party dependencies (i.e. not JDK or JVM)
 
-# Recommendations\
-for next 18 months
-
-## Developers
-
-- Use distro\footnote{If not an option, use AdoptOpenJDK} OpenJDK
-- Build on JDK._LTS_\footnote{Expect to be stuck on Java 8 for a while}
-- Test build on
-    - JDK._LTS_
-    - JDK._stable_
-    - JDK._next_\footnote{\textit{Super} important}\footnote{Also really easy
-      in Travis CI}
-
-## Operations
-
-- Provide JDK._LTS_, JDK._stable_, **and** JDK._next_
-    - Use distro\footnote{If not an option, use AdoptOpenJDK} OpenJDK
-    - For _future option_ of paid support, use Azul Zulu
-- Use JDK._LTS_ in production\footnote{We still have to keep it updated}
-    - Consider shift to per-application JDK\footnote{This may happen
-      automatically as LTO becomes common}
-
-## Avoid
-
-- Oracle JDK in production --- requires license and we don't use commercial
-  support today anyway\footnote{Not because we shouldn't pay but because we
-  don't use what we'd be paying for}
-- Oracle JDK 8 in development --- legal but pointless
-- Oracle OpenJDK 8 in development --- legal but pointless
-- Oracle OpenJDK in production --- legal but irresponsible
-- Neglecting JDK in production
-- Relying on `--release=<platform JDK>`\footnote{We still need to specify it}
-  --- too superficial guarantees
-- JDK._stable_ in production --- third-party dependencies could trip us up
-
 # New release strategy
 
 ## Versioning
@@ -157,6 +122,41 @@ for next 18 months
 
 - Basically same story
 
+# Recommendations\
+for next 18 months
+
+## Developers
+
+- Use distro\footnote{If not an option, use AdoptOpenJDK} OpenJDK
+- Build on JDK._LTS_\footnote{Expect to be stuck on Java 8 for a while}
+- Test build on
+    - JDK._LTS_
+    - JDK._stable_
+    - JDK._next_\footnote{\textit{Super} important}\footnote{Also really easy
+      in Travis CI}
+
+## Operations
+
+- Provide JDK._LTS_, JDK._stable_, **and** JDK._next_
+    - Use distro\footnote{If not an option, use AdoptOpenJDK} OpenJDK
+    - For _future option_ of paid support, use Azul Zulu
+- Use JDK._LTS_ in production\footnote{We still have to keep it updated}
+    - Consider shift to per-application JDK\footnote{This may happen
+      automatically as LTO becomes common}
+
+## Avoid
+
+- Oracle JDK in production --- requires license and we don't use commercial
+  support today anyway\footnote{Not because we shouldn't pay but because we
+  don't use what we'd be paying for}
+- Oracle JDK 8 in development --- legal but pointless
+- Oracle OpenJDK 8 in development --- legal but pointless
+- Oracle OpenJDK in production --- legal but irresponsible
+- Neglecting JDK in production
+- Relying on `--release=<platform JDK>`\footnote{We still need to specify it}
+  --- too superficial guarantees
+- JDK._stable_ in production --- third-party dependencies could trip us up
+
 # _Or else..._
 
 ## What if we don't get in line?