Activity
From 09/19/2022 to 10/18/2022
10/18/2022
- 11:29 AM Bug #4149: BlockManager and TransactionManager handle TransactionType.NONE incorrectly
- @Finalizable.entry()@ is deprecated and needs to be removed, together with @TransactionManager.processEntry()@ - I ca...
10/10/2022
- 07:51 AM Feature #1970: improve ControlFlowOps method resolution by reflection
- 3821c/14283 fixed problems in 3821c/14268:
* for typed FUNCTION calls, convert the returned value to the expected ty... - 05:30 AM Feature #6820: reduce String.toLowerCase and toUpperCase usage and String instances
- New changes are in 3821c/14281:
* Reduced 'toLowerCase' usage in 'BufferImpl.attachDataSource'. - 05:29 AM Feature #6826: reduce the BDT usage in FWD runtime
- Ovidiu Maxiniuc wrote:
> @AnnotatedAst@:
> * @private String getPath(int max)@: it does not look like an optimizati...
10/08/2022
- 08:48 PM Feature #6826: reduce the BDT usage in FWD runtime
- Constantin Asofiei wrote:
> Some changes are in 3821c/14272. Ovidiu, please review.
Global notes:
* the replace...
10/06/2022
- 10:12 AM Feature #6826: reduce the BDT usage in FWD runtime
- Another idea is to replace @new unknown@ (conversion and runtime) with a singleton instance.
- 04:27 AM Feature #6826: reduce the BDT usage in FWD runtime
- Some changes are in 3821c/14272. Ovidiu, please review.
- 03:55 AM Feature #6826: reduce the BDT usage in FWD runtime
- The WrappedResource refactor will be done in #6827.
- 03:48 AM Feature #6826: reduce the BDT usage in FWD runtime
- There are resources like StaticTempTable where the state is kept as BDT, and converted APIs are being called to read ...
- 03:43 AM Feature #6826 (Review): reduce the BDT usage in FWD runtime
- 10:01 AM Feature #6416: explore and implement ideas to make OO 4GL more efficient
- Constantin Asofiei wrote:
> No, as the 4GL classes are loaded/initialized per each user context - they are *not* JVM... - 02:41 AM Feature #6416: explore and implement ideas to make OO 4GL more efficient
- Other notes:
* cache resolved object methods per type, arguments (see CFOps resolveLegacyEntry)
* cache 'method not... - 02:35 AM Feature #6416: explore and implement ideas to make OO 4GL more efficient
- Marian Edu wrote:
> Constantin Asofiei wrote:
> > A part of the problem is the instance initialization in 4GL; the... - 04:29 AM Feature #6821: java collection performance
- Eric: you may want to take a look at ScopedList, too.
- 04:26 AM Feature #6821: java collection performance
- Some changes are in 3821c/14272. Greg, please review.
- 03:14 AM Feature #6821: java collection performance
- For example, using Java for-each loop for an an ArrayList instead of a @for@ and getting the elements directly is mor...
- 03:09 AM Feature #6821 (WIP): java collection performance
- 04:28 AM Feature #6827: reduce the memory footprint of WrappedResource implementations: replace BDTs with ...
- Some changes are in 3821c/14272. Greg, please review.
- 03:54 AM Feature #6827: reduce the memory footprint of WrappedResource implementations: replace BDTs with ...
- This task is meant to look at WrappedResource implementations ('handles' in OE) and replace all BDT state with Java s...
- 03:53 AM Feature #6827 (WIP): reduce the memory footprint of WrappedResource implementations: replace BDTs...
- 04:27 AM Feature #6824: lazy initialization of collection class members
- Some changes are in 3821c/14272. Greg, please review.
- 03:28 AM Feature #6824: lazy initialization of collection class members
- For example, there are BlockDefinition collection fields which can be assigned only when that field is required. Som...
- 03:26 AM Feature #6824 (WIP): lazy initialization of collection class members
- 04:27 AM Feature #6822: performance of the FWD proxy implementation
- Some changes are in 3821c/14272. Greg, please review.
- 03:20 AM Feature #6822: performance of the FWD proxy implementation
- There are two parts which need to be analyzed for performance impact (and object allocation):
* a new proxy creation... - 03:18 AM Feature #6822 (New): performance of the FWD proxy implementation
- 04:26 AM Feature #6820: reduce String.toLowerCase and toUpperCase usage and String instances
- Some changes are in 3821c/14272. Greg, please review.
- 03:08 AM Feature #6820: reduce String.toLowerCase and toUpperCase usage and String instances
- 4GL is by default case-insensitive. This proves tricky in FWD, as we need to i.e. keep registries with lowercased st...
- 02:51 AM Feature #6820 (Review): reduce String.toLowerCase and toUpperCase usage and String instances
- 04:26 AM Feature #6814: JMX instrumentation for FWD core, persistance and ORM layers
- Some changes are in 3821c/14272. Greg, please review.
- 02:23 AM Feature #6814: JMX instrumentation for FWD core, persistance and ORM layers
- The following are pending to be committed in 3821c:...
- 02:14 AM Feature #6814 (New): JMX instrumentation for FWD core, persistance and ORM layers
- 02:51 AM Feature #6819: refactor FWD proxy implementation to use ReflectASM instead of Java Method reflection
- In a large application tested with 6129a, @Utils.invoke(Method method, Object instance, Object... args)@ is being cal...
- 02:43 AM Feature #6819 (New): refactor FWD proxy implementation to use ReflectASM instead of Java Method r...
- 02:42 AM Feature #6818: analyze the performance of the FWD open client proxy client support
- In 6129a, there are some performance improvements in DataSetSDOHelper, to cache the meta fields for a temp-table, but...
- 02:37 AM Feature #6818 (WIP): analyze the performance of the FWD open client proxy client support
- 02:31 AM Feature #6817: replace INSTANTIATING-PROCEDURE and other handle<ExternalProgram> with the actual ...
- Some work related to this has already been done in 6129a.
- 02:30 AM Feature #6817: replace INSTANTIATING-PROCEDURE and other handle<ExternalProgram> with the actual ...
- The referent (which is held in an ExternalProgram) needs to be kept as a WeakReference.
The part which needs speci... - 02:29 AM Feature #6817 (Test): replace INSTANTIATING-PROCEDURE and other handle<ExternalProgram> with the ...
10/05/2022
- 12:10 PM Bug #6811: ERROR conditions converted to structured OO exceptions are not logged in OE
- When an an ERROR condition is thrown instead of raised, OE doesn't seem to log it in the appserver. For example, this...
- 12:08 PM Bug #6811 (New): ERROR conditions converted to structured OO exceptions are not logged in OE
- 05:01 AM Feature #6650: improve performance of Scopeable notification in TM.processScopeNotifications
- ca_upd_20221002b_3821c_14261.zip was committed to 3821c/14269
- 04:39 AM Feature #1970: improve ControlFlowOps method resolution by reflection
- ca_upd_20221002a_3821c_14261.zip was committed to 3821c/14268
10/03/2022
- 03:49 PM Feature #6650: improve performance of Scopeable notification in TM.processScopeNotifications
- Please merge ca_upd_20221002b_3821c_14261.zip into 3821c.
10/02/2022
- 12:18 PM Feature #6650: improve performance of Scopeable notification in TM.processScopeNotifications
- Constantin Asofiei wrote:
> And the update.
The latest update, previously had redmine errors. - 11:23 AM Feature #6650: improve performance of Scopeable notification in TM.processScopeNotifications
- And the update.
- 11:22 AM Feature #6650: improve performance of Scopeable notification in TM.processScopeNotifications
- There was a problem in an app related to dataset-handle parameters - BufferManager support is required for blocks def...
- 12:17 PM Feature #1970: improve ControlFlowOps method resolution by reflection
- Greg Shah wrote:
> @ca_upd_20220930a_3821c_14261.zip@ is the latest?
No, see attached. - 12:11 PM Feature #1970: improve ControlFlowOps method resolution by reflection
- @ca_upd_20220930a_3821c_14261.zip@ is the latest?
- 11:19 AM Feature #1970: improve ControlFlowOps method resolution by reflection
- Latest update is attached - for external programs, a separate call-site cache is being kept, as these are dependent o...
09/30/2022
- 01:18 PM Feature #1970 (WIP): improve ControlFlowOps method resolution by reflection
- Greg, please review these changes. The runtime for the caching is this:
* use the caching mechanism for event proc... - 05:29 AM Feature #6650: improve performance of Scopeable notification in TM.processScopeNotifications
- Greg Shah wrote:
> Code Review ca_upd_20220927a_3821c_14253.zip
>
> I have no objections. Has this been tested w...
09/29/2022
- 01:55 PM Feature #1970: improve ControlFlowOps method resolution by reflection
- > For external procedures, the only thing I can cache is the target Java classname - to avoid PROPATH resolution on e...
- 01:25 PM Feature #1970: improve ControlFlowOps method resolution by reflection
- For external procedures, the only thing I can cache is the target Java classname - to avoid PROPATH resolution on eac...
- 08:25 AM Feature #1970: improve ControlFlowOps method resolution by reflection
- > There is also the RUN SUPER or SUPER() case which is not supported in conversion yet...
Agreed. These seem impo... - 06:51 AM Feature #1970: improve ControlFlowOps method resolution by reflection
- @ControlFlowOps@ is being used in FWD internal classes, for example invoking a callback procedure for SAX parsing or ...
09/28/2022
- 08:56 AM Feature #1970: improve ControlFlowOps method resolution by reflection
- Greg, I'm leaving caching support only for internal procedures or function calls. Others, like ASYNC, ON SERVER, AS-...
09/27/2022
- 04:59 PM Feature #1970: improve ControlFlowOps method resolution by reflection
- I'd also go with an invocation syntax more like this:...
- 04:45 PM Feature #1970: improve ControlFlowOps method resolution by reflection
- > Unfortunately this approach of putting the static parts at the CALL_SITE makes debugging very hard
The resulting... - 02:57 PM Feature #1970: improve ControlFlowOps method resolution by reflection
- Unfortunately this approach of putting the static parts at the CALL_SITE makes debugging very hard - either way, I di...
- 04:30 PM Feature #6650: improve performance of Scopeable notification in TM.processScopeNotifications
- Code Review ca_upd_20220927a_3821c_14253.zip
I have no objections. Has this been tested with one of the large GUI... - 03:02 PM Feature #6650: improve performance of Scopeable notification in TM.processScopeNotifications
- Another update, includes some fixes after testing #6672.
- 12:07 AM Feature #6650: improve performance of Scopeable notification in TM.processScopeNotifications
- The persistence-related changes in @ca_upd_20220923a_3821c_14253.zip@ look good to me.
09/26/2022
- 02:10 PM Feature #1970: improve ControlFlowOps method resolution by reflection
- It looks good!
- 01:34 PM Feature #1970: improve ControlFlowOps method resolution by reflection
- The conversion is in testing and for this test:...
09/23/2022
- 10:12 AM Feature #6650: improve performance of Scopeable notification in TM.processScopeNotifications
- The latest update is attached. All projects look OK.
To fix the @p2j.oo@ code, I've analyzed it using a crude spo...
09/20/2022
- 11:30 AM Feature #6772: increase usage of ASSIGN bracketing
- Please see #6740 - we had an attempt to lock/unlock the frame for the duration of the widget ASSIGN, but this posed p...
- 09:33 AM Feature #6772: increase usage of ASSIGN bracketing
- In #6726, we found that a @CREATE <widget> ASSIGN ...@ was being handled with decreased performance because we are no...
- 09:30 AM Feature #6772 (New): increase usage of ASSIGN bracketing
Also available in: Atom