FWD » Core Development » Base Language

09/25/2023

05:32 PM Bug #5538: fix readByte() for DirStream, ProcessStream, and others

Although the natural instinct is to create good code, which behaves as one would expect (in this case to fully popula... Ovidiu Maxiniuc

05:23 PM Bug #5538 (Review): fix readByte() for DirStream, ProcessStream, and others

Theodoros Theodorou

01:55 PM Bug #5538: fix readByte() for DirStream, ProcessStream, and others

Thank you for confirming the bug Constantin and Ovidiu. I have created a test under @testcases\oo\progress\io\input_s... Theodoros Theodorou

04:32 PM Feature #6649: improve the performance or SourceNameMapper runtime

> 4GL supports the following file types: @.p@, @.r@, @.w@, @.cls@, @.i@. I've also found @.pf@ (parameter file), @.cf... Greg Shah

04:28 PM Feature #6649: improve the performance or SourceNameMapper runtime

During conversion, we have a copy of the file-system of the 4GL project. That project might have some arbitrary path... Greg Shah

09:51 AM Feature #6649: improve the performance or SourceNameMapper runtime

Alexandru Lungu wrote:
> It is normal not to be found right? I mean, if you are running start.p and the only start.p... Dănuț Filimon

06:25 AM Feature #6649: improve the performance or SourceNameMapper runtime

Dănuț Filimon wrote:
> I've also found a problem with this idea where the possible paths from a @SinglePathLookup@ c... Alexandru Lungu

06:16 AM Feature #6649: improve the performance or SourceNameMapper runtime

> The one place I differ is that I was not thinking we would store the propath variants in the *PathLookup helpers. I... Alexandru Lungu

09/22/2023

03:29 PM Bug #5538: fix readByte() for DirStream, ProcessStream, and others

Constantin Asofiei wrote:
> This is an example where 'something' ends up in the memptr:
> [...]
>
> Only first b... Ovidiu Maxiniuc

03:08 PM Bug #5538: fix readByte() for DirStream, ProcessStream, and others

This is an example where 'something' ends up in the memptr:... Constantin Asofiei

02:38 PM Bug #5538: fix readByte() for DirStream, ProcessStream, and others

Indeed, in OE the content of @memptr@ variable is not filled with data from process stream. I do not know why. I trie... Ovidiu Maxiniuc

12:06 PM Bug #7460: implement Progress.Lang.Class:getMethods

For what we've been discussing about Progress.Reflect.Method related to caching (if is needed or not) and to the @=@ ... Constantin Asofiei

11:50 AM Bug #7460: implement Progress.Lang.Class:getMethods

Do we need a follow-up issue to resolve the remaining features/open items? Or should we clear them now, here? Greg Shah

11:12 AM Bug #7460: implement Progress.Lang.Class:getMethods

Branch 6410a was merged to trunk rev 14751 and archived. Constantin Asofiei

10:55 AM Bug #7460: implement Progress.Lang.Class:getMethods

Go ahead now. Greg Shah

10:52 AM Bug #7460: implement Progress.Lang.Class:getMethods

Greg, I think 6410a rev 14777 can be merged to trunk. Constantin Asofiei

11:12 AM Feature #6410: implement additional built-in OO classes/interfaces

Branch 6410a was merged to trunk rev 14751 and archived.
Marian: please let us know when we can create branch 6410b. Constantin Asofiei

06:33 AM Feature #6649: improve the performance or SourceNameMapper runtime

> # files ending in @.r@ or no extension will not be found if the simple name is @startproj.p@;
It's possible to sim... Dănuț Filimon

09/21/2023

10:32 AM Bug #5538: fix readByte() for DirStream, ProcessStream, and others

Greg Shah wrote:
> Try the @IMPORT@ statement with a @MEMPTR@ variable.
It worked for FWD, it triggered @ProcessS... Theodoros Theodorou

09:32 AM Feature #6410: implement additional built-in OO classes/interfaces

6410a was rebased from trunk rev 14747 - new rev 14777.
I'm putting this into conversion testing of some projects. Constantin Asofiei

05:06 AM Feature #6649: improve the performance or SourceNameMapper runtime

I had to discard my initial implementation since it was going in the wrong direction and started again from scratch b... Dănuț Filimon

09/20/2023

05:14 PM Feature #6649: improve the performance or SourceNameMapper runtime

I think we are largely on the same page.
The one place I differ is that I was not thinking we would store the prop... Greg Shah

02:10 PM Bug #7460: implement Progress.Lang.Class:getMethods

Constantin Asofiei wrote:
> On a side note, I'm trying to run the tests from @tests.oo.progress.reflect@ for @constr... Marian Edu

10:12 AM Bug #7460: implement Progress.Lang.Class:getMethods

Marian, I'm working on the review. I'll do any required changes directly on the branch and will prepare it for merge... Constantin Asofiei

09/19/2023

04:39 PM Bug #5538: fix readByte() for DirStream, ProcessStream, and others

> Can you please check if there is any way to check @ProcessStream.readByte()@?
Try the @IMPORT@ statement with a ... Greg Shah

03:41 PM Bug #5538: fix readByte() for DirStream, ProcessStream, and others

Greg Shah wrote:
> Anything using @INPUT THROUGH@, @INPUT-OUTPUT THROUGH@ or @OUTPUT THROUGH@ creates a child proces... Theodoros Theodorou

04:41 AM Feature #6649: improve the performance or SourceNameMapper runtime

Let me see if I understand your point correctly.
h3. Bootstrap
* We have a map with keys as unqualified procedu... Alexandru Lungu

04:00 AM Bug #7460 (Review): implement Progress.Lang.Class:getMethods

Please review. Marian Edu

03:59 AM Bug #7460 (WIP): implement Progress.Lang.Class:getMethods

Constantin Asofiei wrote:
> Marian, at this point, lets leave this behavior on the side for now. We need the implem... Marian Edu

09/18/2023

03:24 PM Bug #7460: implement Progress.Lang.Class:getMethods

Tried the obvious choice @BaseDataType@ as return type and this could work if we update the blocks to create a proxy ... Marian Edu

03:08 PM Bug #7460: implement Progress.Lang.Class:getMethods

Constantin Asofiei wrote:
> what I mentioned about @invoke@ to use @BlockManager.functio@).
Yeah, as said the @invo... Marian Edu

02:04 PM Bug #7460: implement Progress.Lang.Class:getMethods

Marian, at this point, lets leave this behavior on the side for now. We need the implementation of @progress.lang.me... Constantin Asofiei

02:02 PM Bug #7460: implement Progress.Lang.Class:getMethods

All this doesn't look like expected behaviour imho and since those are all internal objects anyway let me know if you... Marian Edu

02:00 PM Bug #7460: implement Progress.Lang.Class:getMethods

Constantin Asofiei wrote:
> Marian, what happens if you use @=@ operator, and not Equals/ToString methods? If they ... Marian Edu

01:51 PM Bug #7460: implement Progress.Lang.Class:getMethods

Marian, what happens if you use @=@ operator, and not Equals/ToString methods? If they are the same reference, this ... Constantin Asofiei

01:18 PM Bug #7460: implement Progress.Lang.Class:getMethods

Constantin Asofiei wrote:
> These @Parameter@, @Legacy4GLMethod@, etc, instances are immutable, right?
There seem... Marian Edu

03:11 PM Bug #7184: Incorrect conversion of concatenated string literals to a LONGCHAR

Constantin, can we close this? Vladimir Tsichevski

03:39 AM Bug #5538 (WIP): fix readByte() for DirStream, ProcessStream, and others

Theodoros Theodorou

09/15/2023

01:41 PM Bug #7460: implement Progress.Lang.Class:getMethods

Greg Shah wrote:
> Marian: Are you close to committing the next version of this?
I’ll try to push that by Monday,... Marian Edu

11:13 AM Bug #7460: implement Progress.Lang.Class:getMethods

Marian: Are you close to committing the next version of this? Greg Shah

10:26 AM Feature #6422: implement auditing support including AUDIT-POLICY and AUDIT-CONTROL

Eric: Please review. Greg Shah

10:18 AM Feature #6649: improve the performance or SourceNameMapper runtime

Another thing to consider with this "search inversion" idea is whether we can avoid the propath scanning process by r... Greg Shah

10:14 AM Feature #6649: improve the performance or SourceNameMapper runtime

I think the idea from #6649-17 has the greatest potential for improvement, though the caching from #6649-16 might als... Greg Shah

08:56 AM Feature #6649: improve the performance or SourceNameMapper runtime

From #6649-16, I understood that we should execute the lookup function with the unqualified filename, then resume sea... Dănuț Filimon

09/14/2023

08:17 AM Feature #6649: improve the performance or SourceNameMapper runtime

I did a test on a large customer application for #6649-16 where I got a total of *42846* calls to @SourceNameMapper.c... Dănuț Filimon

09/13/2023

08:16 AM Feature #6649 (WIP): improve the performance or SourceNameMapper runtime

Greg Shah

09/11/2023

11:24 AM Bug #7460: implement Progress.Lang.Class:getMethods

Marian Edu wrote:
> > ** @ErrorManager.recordOrThrowError@ must be followed by a @return@.
> Will fix that, as said... Constantin Asofiei

10:52 AM Bug #7460: implement Progress.Lang.Class:getMethods

Constantin Asofiei wrote:
> Review for 6410a rev 14571:
> * @LegacyClass.java@
> ** we need to cache the @Legacy4G... Marian Edu

08:11 AM Bug #7460: implement Progress.Lang.Class:getMethods

Review for 6410a rev 14571:
* @LegacyClass.java@
** we need to cache the @Legacy4GLMethod@ instances and resolve th... Constantin Asofiei

08:09 AM Bug #7604 (Closed): Switch does not honor case sensitivity of character

Greg Shah

07:49 AM Bug #7604: Switch does not honor case sensitivity of character

Greg, we can close this. Alexandru Lungu

07:37 AM Bug #7604: Switch does not honor case sensitivity of character

Yes, conversion seems good. Eduard Soltan

07:30 AM Bug #7604: Switch does not honor case sensitivity of character

Eduard, now that the change reached customer applications (newer binaries), please check that conversion is OK, so we... Alexandru Lungu

09/07/2023

10:59 AM Feature #7045: re-implement "normal" (non-abend) stack unwinding to avoid throwing an exception

I've continued working on 7045b. I mostly covered some issues with @CompoundQuery@. Currently, I am still facing some... Alexandru Lungu

09/06/2023

05:24 AM Feature #6410: implement additional built-in OO classes/interfaces

Constantin: Please review.
We need to get our OO support to the 100% mark sooner rather than later. Otherwise we ... Greg Shah

09/05/2023

08:09 AM Feature #6410 (Review): implement additional built-in OO classes/interfaces

Completed most of getMethod/getMethods implementation, please review.
There are a few things remaining:
* package... Marian Edu

08/31/2023

10:56 AM Bug #7656: investigate performance of context-local helpers (e.g. TransactionManager.TransactionH...

Agreed. Greg Shah

10:49 AM Bug #7656: investigate performance of context-local helpers (e.g. TransactionManager.TransactionH...

Greg Shah wrote:
> Full changes.
OK, then I think we need this:
* for all inner classes which have fields accessed... Constantin Asofiei

10:43 AM Bug #7656: investigate performance of context-local helpers (e.g. TransactionManager.TransactionH...

Full changes. Greg Shah

10:30 AM Bug #7656: investigate performance of context-local helpers (e.g. TransactionManager.TransactionH...

Thank you, Alexandru.
Greg: how do you want to proceed? Get the full changes or just the critical classes?
Constantin Asofiei

10:22 AM Bug #7656: investigate performance of context-local helpers (e.g. TransactionManager.TransactionH...

I tested with 7156b/rev. 14732 (which was on 8.193s) patched with 7656a / rev. 14722 + 14723 and got 8.158ms. This me... Alexandru Lungu

04:51 AM Bug #7656: investigate performance of context-local helpers (e.g. TransactionManager.TransactionH...

Alexandru, in 7656a rev 14723 I have the automated changes to modify 'private' to 'package private' fields/methods wh... Constantin Asofiei

06:34 AM Support #4032: block/transaction processing context-local and other performance improvements

4032b was rebased from trunk rev 14721.
4032b rev 14722 contains a small batch of performance improvements related... Constantin Asofiei

08/29/2023

05:01 PM Bug #7759: improve CASE statement for integer type

Constantin,
I was not aware of the 64 bit limitation of Java @switch@ statement. Probably we will need to use the... Ovidiu Maxiniuc

06:59 AM Bug #7759: improve CASE statement for integer type

What I had in mind (for integer type, as Java switch AFAIK works with 32 bit values, not 64, at least for Java 8), is... Constantin Asofiei

06:56 AM Bug #7759: improve CASE statement for integer type

From an email discussion:
Constantin Asofiei wrote:
> Quick question, I don't recall if this is how it is suppose... Constantin Asofiei

06:54 AM Bug #7759 (New): improve CASE statement for integer type

Constantin Asofiei

07:46 AM Feature #6410: implement additional built-in OO classes/interfaces

Marian Edu wrote:
> Constantin Asofiei wrote:
>
> > Thanks, so we need what I mentioned before, we need to fix @B... Constantin Asofiei

07:45 AM Feature #6410: implement additional built-in OO classes/interfaces

Constantin Asofiei wrote:
> Thanks, so we need what I mentioned before, we need to fix @BaseObject.getNextSibling@... Marian Edu

07:45 AM Feature #6410: implement additional built-in OO classes/interfaces

Marian Edu wrote:
> Constantin Asofiei wrote:
> > Please check also what @Progress.Reflect.Method:next-sibling@ ret... Marian Edu

07:44 AM Feature #6410: implement additional built-in OO classes/interfaces

Marian Edu wrote:
> Constantin Asofiei wrote:
> > Please check also what @Progress.Reflect.Method:next-sibling@ ret... Constantin Asofiei

07:43 AM Feature #6410: implement additional built-in OO classes/interfaces

Constantin Asofiei wrote:
> Please check also what @Progress.Reflect.Method:next-sibling@ returns - is it always unk... Marian Edu

07:39 AM Feature #6410: implement additional built-in OO classes/interfaces

Please check also what @Progress.Reflect.Method:next-sibling@ returns - is it always unknown? If yes, maybe this (and... Constantin Asofiei

07:37 AM Feature #6410: implement additional built-in OO classes/interfaces

Constantin Asofiei wrote:
> Marian, can these be created outside of FWD runtime (i.e. by the application?) If these... Marian Edu

07:25 AM Feature #6410: implement additional built-in OO classes/interfaces

Marian, can these be created outside of FWD runtime (i.e. by the application?) If these can only be created by FWD r... Constantin Asofiei

07:19 AM Feature #6410: implement additional built-in OO classes/interfaces

From Marian Edu:
> think I've stumbled upon the same problem before but can't remember exactly the use case nor the ... Constantin Asofiei