Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Build Android flavor with -source 8 -target 8 (but still avoid Java 7+ APIs) #5269

Closed
cpovirk opened this issue Oct 9, 2020 · 5 comments
Closed

Build Android flavor with -source 8 -target 8 (but still avoid Java 7+ APIs) #5269

cpovirk opened this issue Oct 9, 2020 · 5 comments
Labels
P3 no SLO package=general platform=java8 type=other Miscellaneous activities not covered by other type= labels
Milestone
31.0

Comments

@cpovirk
Copy link
Member

cpovirk commented Oct 9, 2020 •
edited
Loading

Currently, our "Android" flavor of Guava supports 2 use cases:

  • Android users (API level 15 (Ice Cream Sandwich) and up)
  • Java 7 users

We are definitely not going to start using Java 8 APIs. Thus, we are not going to increase our minimum Android version to 24, nor are we going to require our users to enable library desugaring. In short: We'll continue to support older versions of Android with no changes on users' part.

However, we are investigating whether to stop supporting non-Android users who use Java 7. This would let us simplify some implementations (mainly by using lambdas), but more importantly, it would let us make some API improvements -- like default methods on interfaces and type annotations (such as for nullness).

This is part of a larger effort by many Google projects to consider dropping support for Java 7. Please let us know if you anticipate problems from this change -- mainly, if you are using Guava in your library or app and you support Java 7. Thanks.

(Please also spread the word. I will additionally announce this on our mailing list. Later, we may also try to introduce some warnings when Guava is run under Java 7.)
@cpovirk cpovirk added type=other Miscellaneous activities not covered by other type= labels package=general platform=java8 P3 no SLO labels Oct 9, 2020
@ejona86 ejona86 mentioned this issue Oct 9, 2020
Closed
@Saucistophe Saucistophe mentioned this issue Oct 17, 2020
Open
@ganeshbch
Copy link

Migrating from Java 7 to Java 8 is recommended because of New language features, security improvements, Optimum performance, Writing code more efficiently, Oracle recommendation to uninstall pre-8 versions to avoid security risks, Older releases no longer publicly supported, Third party Java libraries no longer being developed and supported.

To check the new features in JDK 8 by library - check https://www.oracle.com/java/technologies/javase/8-whats-new.html

To check incompatibilities between JDK 7 and JDK 8 - https://www.oracle.com/java/technologies/javase/8-compatibility-guide.html#A999387

To check incompatibilities between Java SE 7 and SE 8 - https://www.oracle.com/java/technologies/javase/8-compatibility-guide.html#A999198

To check Features removed in JDK 8 - https://www.oracle.com/java/technologies/javase/8-compatibility-guide.html#A999476

To check Features removed in Java SE 8 - https://www.oracle.com/java/technologies/javase/8-compatibility-guide.html#CHDGHIIH

I do not anticipate any change in Applications using Guava as long as the API's remain the same and return the same result as they used to when they were in Java 7.
The compiled app with guava libraries must be able to run on Java 8 platform.

Hope this helps resolve the issue.

@kluever kluever mentioned this issue Dec 11, 2020
Merged
kluever pushed a commit that referenced this issue Dec 11, 2020
@cpovirk @kluever
Log a warning if running under a Java 7 VM.
0357006 
More precisely, log a warning if lambda expressions or type annotations in our classes would produce an exception. If someone wants to use Retrolambda or a similar tool to rewrite our classes, that's fine with us if it works. And our support for Android is unchanged: The Android toolchain rewrites lambdas and removes type annotations.

This is a step toward removing Java 7 support entirely: #5269

RELNOTES=Introduced a warning log message when running under Java 7. This warning is not _guaranteed_ to be logged when running under Java 7, so please don't rely on it as your only warning about future problems. If the warning _itself_ causes you trouble, you can eliminate it by silencing the logger for `com.google.common.base.MoreObjects$ToStringHelper` (which is used _only_ for this warning). This warning prepares for [removing support for Java 7 in 2021](#5269). Please report any problems.

-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=346795766
kluever pushed a commit that referenced this issue Dec 11, 2020
@cpovirk @kluever
Log a warning if running under a Java 7 VM.
dc52e6e 
More precisely, log a warning if lambda expressions or type annotations in our classes would produce an exception. If someone wants to use Retrolambda or a similar tool to rewrite our classes, that's fine with us if it works. And our support for Android is unchanged: The Android toolchain rewrites lambdas and removes type annotations.

This is a step toward removing Java 7 support entirely: #5269

RELNOTES=Introduced a warning log message when running under Java 7. This warning is not _guaranteed_ to be logged when running under Java 7, so please don't rely on it as your only warning about future problems. If the warning _itself_ causes you trouble, you can eliminate it by silencing the logger for `com.google.common.base.MoreObjects$ToStringHelper` (which is used _only_ for this warning). This warning prepares for [removing support for Java 7 in 2021](#5269). Please report any problems.

-------------
Created by MOE: https://github.com/google/moe
MOE_MIGRATED_REVID=346795766
@ijuma
Copy link

ijuma commented Feb 12, 2021

Out or curiosity, what's the downside of requiring library desugaring?

@cpovirk
Copy link
Member Author

cpovirk commented Feb 12, 2021

For most users, it's probably a reasonable thing to enable. However, there are downsides in some cases. As I understand them, the downsides are:

@copybara-service copybara-service bot mentioned this issue Mar 9, 2021
Closed
copybara-service bot pushed a commit that referenced this issue Mar 9, 2021
@cpovirk
Move Java-7-VM warning from MoreObjects to Preconditions.
d6fc364 
RELNOTES=Increased the aggressiveness of [Guava 30.1](https://github.com/google/guava/releases/tag/v30.1)'s warning log message for running `guava-android` under a Java 7 VM. (Android VMs are unaffected.) If the warning _itself_ causes you trouble, you can eliminate it by silencing the logger for `com.google.common.base.Preconditions` (which is used _only_ for this warning). This warning prepares for [removing support for Java 7 in 2021](#5269). Please report any problems. We have tried to make the warning as safe as possible, but anytime a common library logs, especially as aggressively as we do in this new release, there is the potential for [`NullPointerException`](https://stackoverflow.com/a/41017717/28465) or even [deadlock](https://stackoverflow.com/a/48009613/28465). (To be clear, Guava will not log under Java 8 or Android, but it will under Java 7.)
PiperOrigin-RevId: 361604103
copybara-service bot pushed a commit that referenced this issue Mar 9, 2021
@cpovirk
Move Java-7-VM warning from MoreObjects to Preconditions.
00c25e9 
RELNOTES=Increased the aggressiveness of [Guava 30.1](https://github.com/google/guava/releases/tag/v30.1)'s warning log message for running `guava-android` under a Java 7 VM. (Android VMs are unaffected.) If the warning _itself_ causes you trouble, you can eliminate it by silencing the logger for `com.google.common.base.Preconditions` (which is used _only_ for this warning). This warning prepares for [removing support for Java 7 in 2021](#5269). Please report any problems. We have tried to make the warning as safe as possible, but anytime a common library logs, especially as aggressively as we do in this new release, there is the potential for [`NullPointerException`](https://stackoverflow.com/a/41017717/28465) or even [deadlock](https://stackoverflow.com/a/48009613/28465). (To be clear, Guava will not log under Java 8 or Android, but it will under Java 7.)
PiperOrigin-RevId: 361700569
This was referenced Mar 13, 2021
Merged
Closed
Merged
Closed
@copybara-service copybara-service bot mentioned this issue Mar 23, 2021
Closed
copybara-service bot pushed a commit that referenced this issue Mar 23, 2021
@cpovirk
Allow usages of java.util.Objects.requireNonNull in the backport.
03f03cd 
Relevant to #5269 in that it's an exception -- a case in which we *are* willing to use a Java 7 API because we know this particular API is safe to use under Android.

RELNOTES=n/a
PiperOrigin-RevId: 364549818
copybara-service bot pushed a commit that referenced this issue Mar 23, 2021
@cpovirk
Allow usages of java.util.Objects.requireNonNull in the backport.
d1c3865 
Relevant to #5269 in that it's an exception -- a case in which we *are* willing to use a Java 7 API because we know this particular API is safe to use under Android.

RELNOTES=n/a
PiperOrigin-RevId: 364567470
@copybara-service copybara-service bot mentioned this issue Mar 23, 2021
Closed
copybara-service bot pushed a commit that referenced this issue Mar 23, 2021
@cpovirk
Set -source 8 -target 8 in the open-source build, and remove Java8U…
d7058a4 
…sage.

This is largely a rollback of the original Java8Usage change, but I've modified it in a few ways:

- I set `-source 8 -target 8` in the backport.
- I kept the code to make includes/excludes fully work with maven-compiler-plugin, since that could save us some confusion down the line.
- I incorporated a rollback of the changes to Preconditions from CL 361700569, in which I moved the warning from MoreObjects to Preconditions.
- I added errors to our internal release scripts so that we don't accidentally make a release that drops Java 7 support before we're ready.

Fixes #5269

PiperOrigin-RevId: 364405695
@hiranya911
Copy link

We started seeing this in Firebase Java SDK builds:

WARNING: Java 7 compatibility warning: See https://github.com/google/guava/issues/5269
java.lang.Exception: Guava will drop support for Java 7 in 2021. Please let us know if this will cause you problems: https://github.com/google/guava/issues/5269

We would certainly like to drop Java 7 support, but not before the larger Google Cloud client libraries ecosystem decides to do so. AFAICT cloud libraries like Firestore and Storage are still Java 7 compatible. Has a decision been made to migrate all these libraries to Java 8?

@cpovirk
Copy link
Member Author

cpovirk commented Jun 16, 2021

Hi, thanks for reporting that. We have been in contact with the Google Cloud people, and we won't be dropping support unless they are. Part of the reason for this warning was to share any information we could find with them.

@cpovirk cpovirk mentioned this issue Jun 29, 2021
Closed
eperret added a commit to eperret/omakase that referenced this issue Jul 22, 2021
@eperret
chore: Bumps [guava]([200~https://github.com/google/guava) from 29-jr…
775abf3 
…e to 30.1.1-jre

Updated to addres an identified issue
-   `io`: Deprecated  `Files.createTempDir()`. ([fec0dbc](google/guava@fec0dbc)) ([CVE-2020-8908](https://nvd.nist.gov/vuln/detail/CVE-2020-8908); continuing discussion in  [#4011](google/guava#4011))

-   Increased the aggressiveness of  [Guava 30.1](https://github.com/google/guava/releases/tag/v30.1)'s warning log message for running  `guava-android`  under a Java 7 VM. (Android VMs are unaffected.) If the warning  _itself_  causes you trouble, you can eliminate it by silencing the logger for  `com.google.common.base.Preconditions`  (which is used  _only_  for this warning). This warning prepares for  [removing support for Java 7 in 2021](google/guava#5269). Please report any problems. We have tried to make the warning as safe as possible, but anytime a common library logs, especially as aggressively as we do in this new release, there is the potential for  [`NullPointerException`](https://stackoverflow.com/a/41017717/28465)  or even  [deadlock](https://stackoverflow.com/a/48009613/28465). (To be clear, Guava will not log under Java 8 or Android, but it will under Java 7.) ([00c25e9](google/guava@00c25e9))
-   `cache`: Fixed compatibility between  `asMap().compute(...)`  and a load. ([42bf4f4](google/guava@42bf4f4))
-   `cache`: Added  `@CheckReturnValue`  to some APIs. ([a5ef129](google/guava@a5ef129))
-   `collect`: Added  `@DoNotCall`  to the mutator methods on immutable types ([6ae9532](google/guava@6ae9532))
-   `hash`: Removed  `@Beta`  from  `HashCode`. ([2c9f161](google/guava@2c9f161))
-   `io`: Removed  `@Beta`  from  `CountingOutputStream`.
([d394bac](google/guava@d394bac847467039530f514f880ecca27263d0ff))i
-   If you use guava-android in an Android project (as opposed to from a Java VM), you will need to  [enable desugaring of Java 8  _language features_](https://developer.android.com/studio/write/java8-support.html#supported_features)  if you have not already done so. (And if you are releasing an Android  _library_, then anyone who uses that library will also have to enable desugaring.) We expect for nearly all Android projects to have already enabled desugaring. But if this causes problems for you, please let us know on  [issue #5358](google/guava#5358). The purpose of this change is to detect potential problems for users now so that we can plan to use Java 8 language features in our implementation later this year.
-   Introduced a warning log message when running  `guava-android`  under a Java 7 VM. (Android VMs are unaffected, aside from the need to use desugaring, described in the previous bullet.) This warning is not  _guaranteed_  to be logged when running under Java 7, so please don't rely on it as your only warning about future problems. If the warning  _itself_  causes you trouble, you can eliminate it by silencing the logger for  `com.google.common.base.MoreObjects$ToStringHelper`  (which is used  _only_  for this warning). This warning prepares for  [removing support for Java 7 in 2021](google/guava#5269). Please report any problems. We have tried to make the warning as safe as possible, but anytime a common library logs, there is the potential for  [`NullPointerException`](https://stackoverflow.com/a/41017717/28465)  or even  [deadlock](https://stackoverflow.com/a/48009613/28465). (To be clear, Guava will  _not_  log under Java 8 or Android, but it  _may_  log under Java 7.) ([dc52e6e](google/guava@dc52e6e))
    -   Note that we subsequently made this warning more aggressive in  [Guava 30.1.1](https://github.com/google/guava/releases/tag/v30.1.1), including changing the logger it uses to  `com.google.common.base.Preconditions`.
-   `base`: Deprecated  `StandardSystemProperty.JAVA_EXT_DIRS`. We do not plan to remove the API, but note that, under recent versions of Java, that property always has a value of  `null`. ([38abf07](google/guava@38abf07))
-   `net`: Added  `HttpHeaders`  constants for  `Origin-Isolation`  and  `X-Request-ID`. ([a48fb4f](google/guava@a48fb4f),  [8319d20](google/guava@8319d20))
-   `reflect`: Added  `ClassInfo.isTopLevel()`. ([4106272](google/guava@4106272))
-   `util.concurrent`: Added  `ClosingFuture.submitAsync(AsyncClosingCallable)`. ([c5e2d8d](google/guava@c5e2d8d))
-   [Guava types can no longer be sent over GWT-RPC.](https://groups.google.com/d/msg/guava-announce/zHZTFg7YF3o/rQNnwdHeEwAJ)  Even the earlier, temporary way to reenable support (`guava.gwt.emergency_reenable_rpc`) no longer has an effect. ([0cb89dd](google/guava@0cb89dd))
-   `cache`: Fixed memory leak in  `LocalCache`  under  [j2objc](https://developers.google.com/j2objc). ([5e519d9](google/guava@5e519d9))
-   `collect`: Added two-element  `min`  and  `max`  methods to  `Comparators`. ([958186c](google/guava@958186c))
-   `collect`: Removed  `@Beta`  from  `Multimaps.toMultimap`. ([b6b4dc4](google/guava@b6b4dc4))
-   `collect`: Made the set returned by  `ImmutableMap<K, V>.keySet()`  serializable as long as  `K`  is serializable, even if  `V`  is not (and similarly for  `values()`). ([f5a69c3](google/guava@f5a69c3))
-   `collect`: Fixed bug in  `powerSet.equals(otherPowerSet)`  would erroneously return  `false`  if the two power sets' underlying sets were equal but had a different iteration order. ([215b1f0](google/guava@215b1f0))
-   `collect`: Eliminated  [j2objc](https://developers.google.com/j2objc)  retain-cycle in  `SingletonImmutableBiMap`. ([0ad38b8](google/guava@0ad38b8))
-   `eventbus`: Prevented  `@Subscribe`  from being applied to a method that takes a primitive, as that will never be called. ([554546c](google/guava@554546c))
-   `graph`: Made  `Traverser.breadthFirst()`  lazier, and optimized  `Traverser`  more generally. ([32f2d77](google/guava@32f2d77),  [b5210ca](google/guava@b5210ca))
-   `graph`: Added  `@DoNotMock`  to  `Traverser`. ([6410f18](google/guava@6410f18))
-   `io`: Deprecated  `Files.createTempDir()`. ([fec0dbc](google/guava@fec0dbc)) ([CVE-2020-8908](https://nvd.nist.gov/vuln/detail/CVE-2020-8908); continuing discussion in  [#4011](google/guava#4011))
-   `io`: Upgraded  `ByteStreams.copy(InputStream, OutputStream)`  to use the faster  `FileChannel`  if possible. ([a1e9a0b](google/guava@a1e9a0b))  [update: My mistake: This was  [rolled back](google/guava@e839f94), so it did not make 30.0.]
-   `math`: Added  `roundToDouble`  to  `BigDecimalMath`,  `BigIntegerMath`, and  `LongMath`. ([bee4f3c](google/guava@bee4f3c),  [2b5c096](google/guava@2b5c096),  [633abf2](google/guava@633abf2))
-   `net`: Added  `MediaType`  constants for several font/ types. ([571cf66](google/guava@571cf66))
-   `net`: Added  `HttpHeaders`  constants for  `Cross-Origin-Embedder-Policy(-Report-Only)?`. ([c3bf731](google/guava@c3bf731))
-   `testing`: Made  `EqualsTester`  test that non-`String`  objects are not equal to their  `String`  representations. ([c9570ea](google/guava@c9570ea))
-   `util.concurrent`: Added  `ClosingFuture`. ([52e048e](google/guava@52e048e))
-   `util.concurrent`: Removed the deprecated 1-arg  `ServiceManager.addListener(Listener)`. Use the 2-arg  `addListener(Listener, Executor)`  overload, setting the executor to  `directExecutor()`  for equivalent behavior. ([dfb0001](google/guava@dfb0001))
-   `util.concurrent`: Changed  `AbstractFuture.toString()`  to no longer include the  `toString()`  of the result. ([2ebf27f](google/guava@2ebf27f))
-   `util.concurrent`: Added  `awaitTerminationUninterruptibly`. ([f07b954](google/guava@f07b954))
@eperret eperret mentioned this issue Jul 22, 2021
Merged
eperret added a commit to salesforce/omakase that referenced this issue Jul 22, 2021
@eperret
chore: Bumps [guava]([200~https://github.com/google/guava) from 29-jr…
f317eed 
…e to 30.1.1-jre (#28)

Updated to addres an identified issue
-   `io`: Deprecated  `Files.createTempDir()`. ([fec0dbc](google/guava@fec0dbc)) ([CVE-2020-8908](https://nvd.nist.gov/vuln/detail/CVE-2020-8908); continuing discussion in  [#4011](google/guava#4011))

-   Increased the aggressiveness of  [Guava 30.1](https://github.com/google/guava/releases/tag/v30.1)'s warning log message for running  `guava-android`  under a Java 7 VM. (Android VMs are unaffected.) If the warning  _itself_  causes you trouble, you can eliminate it by silencing the logger for  `com.google.common.base.Preconditions`  (which is used  _only_  for this warning). This warning prepares for  [removing support for Java 7 in 2021](google/guava#5269). Please report any problems. We have tried to make the warning as safe as possible, but anytime a common library logs, especially as aggressively as we do in this new release, there is the potential for  [`NullPointerException`](https://stackoverflow.com/a/41017717/28465)  or even  [deadlock](https://stackoverflow.com/a/48009613/28465). (To be clear, Guava will not log under Java 8 or Android, but it will under Java 7.) ([00c25e9](google/guava@00c25e9))
-   `cache`: Fixed compatibility between  `asMap().compute(...)`  and a load. ([42bf4f4](google/guava@42bf4f4))
-   `cache`: Added  `@CheckReturnValue`  to some APIs. ([a5ef129](google/guava@a5ef129))
-   `collect`: Added  `@DoNotCall`  to the mutator methods on immutable types ([6ae9532](google/guava@6ae9532))
-   `hash`: Removed  `@Beta`  from  `HashCode`. ([2c9f161](google/guava@2c9f161))
-   `io`: Removed  `@Beta`  from  `CountingOutputStream`.
([d394bac](google/guava@d394bac847467039530f514f880ecca27263d0ff))i
-   If you use guava-android in an Android project (as opposed to from a Java VM), you will need to  [enable desugaring of Java 8  _language features_](https://developer.android.com/studio/write/java8-support.html#supported_features)  if you have not already done so. (And if you are releasing an Android  _library_, then anyone who uses that library will also have to enable desugaring.) We expect for nearly all Android projects to have already enabled desugaring. But if this causes problems for you, please let us know on  [issue #5358](google/guava#5358). The purpose of this change is to detect potential problems for users now so that we can plan to use Java 8 language features in our implementation later this year.
-   Introduced a warning log message when running  `guava-android`  under a Java 7 VM. (Android VMs are unaffected, aside from the need to use desugaring, described in the previous bullet.) This warning is not  _guaranteed_  to be logged when running under Java 7, so please don't rely on it as your only warning about future problems. If the warning  _itself_  causes you trouble, you can eliminate it by silencing the logger for  `com.google.common.base.MoreObjects$ToStringHelper`  (which is used  _only_  for this warning). This warning prepares for  [removing support for Java 7 in 2021](google/guava#5269). Please report any problems. We have tried to make the warning as safe as possible, but anytime a common library logs, there is the potential for  [`NullPointerException`](https://stackoverflow.com/a/41017717/28465)  or even  [deadlock](https://stackoverflow.com/a/48009613/28465). (To be clear, Guava will  _not_  log under Java 8 or Android, but it  _may_  log under Java 7.) ([dc52e6e](google/guava@dc52e6e))
    -   Note that we subsequently made this warning more aggressive in  [Guava 30.1.1](https://github.com/google/guava/releases/tag/v30.1.1), including changing the logger it uses to  `com.google.common.base.Preconditions`.
-   `base`: Deprecated  `StandardSystemProperty.JAVA_EXT_DIRS`. We do not plan to remove the API, but note that, under recent versions of Java, that property always has a value of  `null`. ([38abf07](google/guava@38abf07))
-   `net`: Added  `HttpHeaders`  constants for  `Origin-Isolation`  and  `X-Request-ID`. ([a48fb4f](google/guava@a48fb4f),  [8319d20](google/guava@8319d20))
-   `reflect`: Added  `ClassInfo.isTopLevel()`. ([4106272](google/guava@4106272))
-   `util.concurrent`: Added  `ClosingFuture.submitAsync(AsyncClosingCallable)`. ([c5e2d8d](google/guava@c5e2d8d))
-   [Guava types can no longer be sent over GWT-RPC.](https://groups.google.com/d/msg/guava-announce/zHZTFg7YF3o/rQNnwdHeEwAJ)  Even the earlier, temporary way to reenable support (`guava.gwt.emergency_reenable_rpc`) no longer has an effect. ([0cb89dd](google/guava@0cb89dd))
-   `cache`: Fixed memory leak in  `LocalCache`  under  [j2objc](https://developers.google.com/j2objc). ([5e519d9](google/guava@5e519d9))
-   `collect`: Added two-element  `min`  and  `max`  methods to  `Comparators`. ([958186c](google/guava@958186c))
-   `collect`: Removed  `@Beta`  from  `Multimaps.toMultimap`. ([b6b4dc4](google/guava@b6b4dc4))
-   `collect`: Made the set returned by  `ImmutableMap<K, V>.keySet()`  serializable as long as  `K`  is serializable, even if  `V`  is not (and similarly for  `values()`). ([f5a69c3](google/guava@f5a69c3))
-   `collect`: Fixed bug in  `powerSet.equals(otherPowerSet)`  would erroneously return  `false`  if the two power sets' underlying sets were equal but had a different iteration order. ([215b1f0](google/guava@215b1f0))
-   `collect`: Eliminated  [j2objc](https://developers.google.com/j2objc)  retain-cycle in  `SingletonImmutableBiMap`. ([0ad38b8](google/guava@0ad38b8))
-   `eventbus`: Prevented  `@Subscribe`  from being applied to a method that takes a primitive, as that will never be called. ([554546c](google/guava@554546c))
-   `graph`: Made  `Traverser.breadthFirst()`  lazier, and optimized  `Traverser`  more generally. ([32f2d77](google/guava@32f2d77),  [b5210ca](google/guava@b5210ca))
-   `graph`: Added  `@DoNotMock`  to  `Traverser`. ([6410f18](google/guava@6410f18))
-   `io`: Deprecated  `Files.createTempDir()`. ([fec0dbc](google/guava@fec0dbc)) ([CVE-2020-8908](https://nvd.nist.gov/vuln/detail/CVE-2020-8908); continuing discussion in  [#4011](google/guava#4011))
-   `io`: Upgraded  `ByteStreams.copy(InputStream, OutputStream)`  to use the faster  `FileChannel`  if possible. ([a1e9a0b](google/guava@a1e9a0b))  [update: My mistake: This was  [rolled back](google/guava@e839f94), so it did not make 30.0.]
-   `math`: Added  `roundToDouble`  to  `BigDecimalMath`,  `BigIntegerMath`, and  `LongMath`. ([bee4f3c](google/guava@bee4f3c),  [2b5c096](google/guava@2b5c096),  [633abf2](google/guava@633abf2))
-   `net`: Added  `MediaType`  constants for several font/ types. ([571cf66](google/guava@571cf66))
-   `net`: Added  `HttpHeaders`  constants for  `Cross-Origin-Embedder-Policy(-Report-Only)?`. ([c3bf731](google/guava@c3bf731))
-   `testing`: Made  `EqualsTester`  test that non-`String`  objects are not equal to their  `String`  representations. ([c9570ea](google/guava@c9570ea))
-   `util.concurrent`: Added  `ClosingFuture`. ([52e048e](google/guava@52e048e))
-   `util.concurrent`: Removed the deprecated 1-arg  `ServiceManager.addListener(Listener)`. Use the 2-arg  `addListener(Listener, Executor)`  overload, setting the executor to  `directExecutor()`  for equivalent behavior. ([dfb0001](google/guava@dfb0001))
-   `util.concurrent`: Changed  `AbstractFuture.toString()`  to no longer include the  `toString()`  of the result. ([2ebf27f](google/guava@2ebf27f))
-   `util.concurrent`: Added  `awaitTerminationUninterruptibly`. ([f07b954](google/guava@f07b954))
@cpovirk cpovirk added this to the 31.0 milestone Sep 24, 2021
copybara-service bot pushed a commit that referenced this issue Sep 28, 2021
@cpovirk
Remove claims that guava-android works under Java 7.
fa9561e 
As of 31.0, it no longer does.

Relevant to #5269

RELNOTES=n/a
PiperOrigin-RevId: 399327005
@copybara-service copybara-service bot mentioned this issue Sep 28, 2021
Closed
copybara-service bot pushed a commit that referenced this issue Sep 28, 2021
@cpovirk
Remove claims that guava-android works under Java 7.
17e403c 
As of 31.0, it no longer does.

Relevant to #5269

RELNOTES=n/a
PiperOrigin-RevId: 399330047
@cpovirk cpovirk mentioned this issue Apr 14, 2022
Closed
@copybara-service copybara-service bot mentioned this issue Apr 17, 2023
Closed
copybara-service bot pushed a commit that referenced this issue Apr 17, 2023
@cpovirk
Use type-use @Nullable instead of declaration @CheckForNull in th…
7300252 
…e remaining guava-android sources (mostly testing utilities).

guava-android was using `@CheckForNull` because it couldn't use type-use annotations when it targeted Java 7. Now that [it's built with `-source 8 -target 8`](#5269), we can use type-use annotations.

RELNOTES=n/a
PiperOrigin-RevId: 522310996
copybara-service bot pushed a commit that referenced this issue Apr 17, 2023
@cpovirk
Use type-use @Nullable instead of declaration @CheckForNull in th…
8f0fda7 
…e remaining guava-android sources (mostly testing utilities).

guava-android was using `@CheckForNull` because it couldn't use type-use annotations when it targeted Java 7. Now that [it's built with `-source 8 -target 8`](#5269), we can use type-use annotations.

RELNOTES=n/a
PiperOrigin-RevId: 524912499
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
P3 no SLO package=general platform=java8 type=other Miscellaneous activities not covered by other type= labels
Projects
None yet
Milestone
31.0
Development

Successfully merging a pull request may close this issue.

Set -source 8 -target 8 in the open-source build, and remove Java8Usage. google/guava
4 participants
@ijuma @cpovirk @hiranya911 @ganeshbch