Release notes for portingset: 7.3a.06/SUN_SPARC/SOLARIS2.6/PA.2469Release notes for portingset: 7.3a.06 MaintReger: # 22150 (BDNT15135): Strange behaviour with BW 7.3.04.140 Date: Mon, 19 Apr 2004 10:21:17 +0200 Created on: MaintReger Type: bugfix Problem Description (Customer terms) Strange behaviour with BW 7.3.04.140 STEPS TO BE FOLLOWED : 1. Start session tsmsc1100m000 "Maintenance Sales Orders" 2. Find order GGS000010 3. Under Specific press 'All Lines' (tsmsc1110m000) 4. Mark line 10 5. Press enter --> session tsmsc1110m000 opens 6. close this session again 7. Mark line 10 8. Press enter --> nothing happens 9. Mark another (empty) line on the form OBSERVED RESULT : 1. Message 'Maintenance sales order line cannot be deleted for line status In Process' 2. It requires pressing OK several times to finish the message. EXPECTED RESULT : Also second time selecting and pressing enter the Maintenance Sales Order line . ADDITIONAL COMMENTS : Installing BW 7.5a.20 from Gemini does not reveal this issue Workaround n/a Motive source DF:145853 MaintReger: # 22010 (BDUX14370): QP: bind actual string length Date: Fri, 2 Apr 2004 13:40:56 +0200 Created on: MaintReger Type: bugfix Problem Description (Customer terms) Database drivers generate more network traffic than needed. Test Procedure Regression test Motive source DF:144257 MaintReger: # 21942 (BDUX14988): Removed DS-message when Extended Object Header is found. Date: Fri, 26 Mar 2004 11:44:48 +0100 Created on: MaintReger Type: bugfix Problem Description (Customer terms) When running an 3GL/4GL object on Reger that was compiled on Gemini, the following message can pop-up: Extended Object Header with xx bytes skipped. Workaround Don't deliver Gemini objects to a Reger environment that runs a portingset without this fix. Keep in mind though that the warning message is not nice, but that nothing goes wrong. Test Procedure Compile a small 3GL source with a Gemini bic-compiler and run it in a Reger environment. See that with the new portingset the warning message disappeared. The warning is still logged in the $BSE/log/log.bshell file though. MaintReger: # 21881 (BDUX14925): Bad log message(s) Date: Tue, 23 Mar 2004 11:05:29 +0100 Created on: MaintCorelli Type: bugfix Problem Description (Customer terms) Bad log messages: 1. from ipc_glob.c: Can't process group (message should not appear) 2. from pd_lpint.c: Error message should be informational message (permissions changed on lp) Test Procedure Run a bshell to check whether the 'can't change process group' message do not appear. The other log message appears when the license daemon is started. Motive source TCS:800-138046 MaintReger: # 21852 (BDUX11837): Add -dbgflow to Usage text Date: Mon, 22 Mar 2004 11:12:15 +0100 Created on: MaintCorelli Type: bugfix Problem Description (Customer terms) dbgflow option is not listed in Bshell Usage text Workaround not applicable Test Procedure run bshell6.2 -U to see the Usage, and check if -dbgflow is listed also Motive source DF:145821 MaintReger: # 21828 (BDNT14885): DF145598 Application toolbar is not complete Date: Thu, 18 Mar 2004 16:37:58 +0100 Created on: MaintReger Type: bugfix Problem Description (Customer terms) Run session tsbsc1110m000, mark line 10 and select Item details under the Specific menu. Asure that the Application Toolbar (Text) is alligned as: "Top" (Session menu: Tools/Options/Application Tools (Text) OBSERVED RESULT : In the Serialized Items session the Application Tool Bar is not complete/visable. Clicking the the Application Tools Bar with the mouse cursor makes it complete/viable again. Test Procedure See problem description Motive source DF:145598 MaintReger: # 21797 (BDNT14852): modified deliver.lst to copy additional files for cmfuiadap Date: Tue, 16 Mar 2004 16:23:35 +0100 Created on: MaintReger Type: bugfix Problem Description (Customer terms) Modified deliver.lst to copy the additonal files for cmfuiadap required by the bw installer. MaintReger: # 21764 (BDNT14821): Use new SSA names in console file Date: Mon, 15 Mar 2004 10:11:13 +0100 Created on: MaintCorelli Type: porting specific change Problem Description (Customer terms) Baan Management Snapin shows wrong name (no SSA). MaintReger: # 21721 (BDUX13938): Modifications to enable Teraspace from cf files Date: Wed, 10 Mar 2004 13:29:29 +0100 Created on: MaintCorelli Type: bugfix Motive source DF:135783 (Teraspace) MaintReger: # 21706 (BDUX14711): DF217307 Cannot use multiple trace values in debug bshell Date: Tue, 9 Mar 2004 13:54:39 +0100 Created on: MaintCorelli Type: bugfix Problem Description (Customer terms) BDB_DEBUG and TT_SQL_TRACE/BAAN_SQL_TRACE variable switching doesn't always work correctly using ottstpbshdebug. Workaround Large disk and set the flags on the command line Test Procedure the ottstpbshdebug session should show the correct values and not clear the toggles on the screen with incorrect setting of wrong BDB_DEBUG value Motive source DF:217307 MaintReger: # 21683 (BDNT14747): common ms_clients Date: Mon, 8 Mar 2004 08:36:04 +0100 Created on: MaintReger Type: bugfix Problem Description (Customer terms) ms_clients under coomon vob Motive source BRD:123 MaintReger: # 21635 (BDNT14697): Do not show paths of non-existent binaries in taskpad Date: Wed, 3 Mar 2004 13:15:30 +0100 Created on: MaintCorelli Type: porting specific change Problem Description (Customer terms) Do not show paths of services/binaries which do not exist in the baan management snapin. Workaround NA MaintReger: # 21626 (BDNT14687): Manual merge of BDNT14616 (Windows 2003 problems with Baan management snapin) Date: Tue, 2 Mar 2004 16:17:10 +0100 Created on: MaintReger Type: porting specific change Problem Description (Customer terms) Fixed several baan NT manager, alias SSA(R) Baan Enterprise Manager snapin. MaintReger: # 21588 (BDUX14643): rdbms_maint6.2 loops on crypted passwd starting with a dash Date: Thu, 26 Feb 2004 10:33:50 +0100 Created on: MaintCorelli Type: bugfix Problem Description (Customer terms) Adding an user from the DBA module hangs. Workaround Try the failing operation again. There is a big change that the new encryption generates another passwd without initial dash (random). Test Procedure Use a batch file like (Note that the encrypted passwd after -P starts with a dash): -F -O -F -absp -mbsp -pRnx|Y66[76.!!lbX5-vCd?Z6rmW8PX7| -csystem -iRsc~U8&V.6.!!lbX5-vCd?Z6rmW8PX7| -Gb50 -P-67%B.4z$:gZ@r} The current executable will loop forever on argument parsing. The fixed executable does its job (or return an appropriate error). MaintReger: # 21561 (BDNT14605): BaanMan: adapt column widths Date: Mon, 23 Feb 2004 11:58:17 +0100 Created on: MaintCorelli Type: bugfix Motive source TCS:800-134947 MaintReger: # 21551 (BDNT14505): Renamed Baan NT manager to SSA Baan Enterprise Manager Date: Mon, 23 Feb 2004 11:12:42 +0100 Created on: MaintCorelli Type: bugfix Motive source TCS:800-134947 MaintReger: # 21528 (BDUX14520): schedule not working correctly when using sockets and direct starting of the session Date: Thu, 19 Feb 2004 15:18:58 +0100 Created on: MaintCorelli Type: bugfix Problem Description (Customer terms) When starting a "server" program with the -server option of the bshell it seems that only one process gets ticks from the bshell. The server listens on a socket for connections, based on this info a slave is started. when a new connection is made the server is does not see this untill the first slave is finished. when starting the server from the user interface it is working fine. Workaround start the server from the User interface. Test Procedure create a server process like: *****SERVER PROCESS***** function main() { long fp, sock.lstnr, sock.id, child.pid, curpgrp string buffer(1024), port(10) port = "1248" sock.lstnr = sock.listen(port) set.pgrp(pid, pid) if sock.lstnr > 0 then while (1) |Never-Ending Loop sock.id = sock.accept(sock.lstnr) |Wait for Request curpgrp = get.pgrp(pid) child.pid = act.and.sleep("sockslaveo", port, sock.id, curpgrp) curpgrp = get.pgrp(child.pid) if child.pid then set.pgrp(child.pid, child.pid) reactivate(child.pid) endif endwhile endif } ******SERVER PROCESS****** Create a slave process like: ******SLAVE PROCESS******* function main() { long fp, sockid, orgpgrp, ret string buffer(1024), port(10) port = argv$(1) sockid = lval(argv$(2)) orgpgrp = lval(argv$(3)) e = sock.inherit(sockid) suspend(1) if sockid > 0 then |Start processing request e = sock.recv(sockid, buffer, 1024) select ttadv140.* |this select to let the slave run for some time te start the second from ttadv140 where ttadv140._compnr = 000 order by ttadv140._index1 selectdo endselect buffer = sprintf$("Slave Process-%u002 %U001", utc.num(), utc.num() ) e = sock.send(sockid, buffer, 1024) sock.close(sockid) endif } ******SLAVE PROCESS****** Create a client process like: ******CLIENT PROCESS****** function main() { long fp string buffer(1024), w.request(1024), w.reply(1024) long sock_open, sock_write, sock_flush, sock_read sock_open = sock.connect(hostname$(), "1248", 5) w.request = "Calling Program PID -> " & str$(pid) & " Date/Time - " & sprintf$("%u002 %U001", utc.num(), utc.num() ) if sock_open > 0 then sock_write = sock.send(sock_open, w.request, 1024) sock_flush = sock.flush(sock_open) sock_read = sock.recv(sock_open, w.reply, 1024) sock.close(sock_open) endif } *******CLIENT PROCESS******* compile the server to socksrvo, the client to sockclio and the slave to sockslaveo start the server like: bshell6.x -server ./socksrvo then start from a other connection the client 2 times, when doing this from unix it can be like: ba6.x ./sockclio then check with bshcmd6.x -p to see if there are more slave processes running at the same time. Motive source DF:218263 MaintReger: # 21503 (BDUX14555): Problem with compressing/decompressing empty files Date: Wed, 18 Feb 2004 17:29:15 +0100 Created on: MaintBaanIVc Type: bugfix Problem Description (Customer terms) On DEC, compression of zero-byte file results in a wrong .Z archive. Decompressing that archive doesn't result in (again) a zero byte file. MaintReger: # 21495 (BDUX14568): Oracle 10G support Date: Wed, 18 Feb 2004 16:05:43 +0100 Created on: MaintReger Type: porting specific change Motive source BRD:BRQ000293 MaintReger: # 21483 (BDNT14554): DF144721 TAB formatting problem with multiline editor Date: Tue, 17 Feb 2004 12:43:19 +0100 Created on: MaintReger Type: bugfix Problem Description (Customer terms) The customer has a problem with the functionality that the multiline editor (tttxt2230) stores a as a tab sign in the database and a tab sign is represented with 8 spaces while it is printed. Inside the editor you see the cursor jumping to the next position that can be divided by 8. So the formatting you enter in the editor will be destroyed in the printout. This should be consistent. Workaround Use spaces in stead of TAB's Test Procedure Use the Baan Text editor and enter tab's Motive source DF:144721 MaintReger: # 21466 (BDNT13559): BW: set reason TsTerminationProcessHandle only when process could be started Date: Mon, 16 Feb 2004 11:52:47 +0100 Created on: MaintCorelli Type: bugfix Problem Description (Customer terms) The process was not started, but the reason given by BW was TerminationProcessHandle (7). This is not correct. The reason should be TerminationCreateProcess (4) Workaround no workaround Test Procedure Start a .bat file using app_start. Take care that the .bat file cannot be started Check the reason in returned. Refer also to defect for reproduction from TG package. Motive source DF:144386 MaintReger: # 21452 (BDNT14451): Thai spacing problem in bwprint Date: Fri, 13 Feb 2004 12:28:06 +0100 Created on: MaintBaanIVc Type: bugfix Problem Description (Customer terms) Spacing between Thai characters is going wrong when some of the combination characters are printed. Workaround n/a MaintReger: # 21448 (BDUX14464): Informix: Enabling alias in Delete statement with subquery Date: Fri, 13 Feb 2004 12:08:56 +0100 Created on: MaintReger Type: bugfix Problem Description (Customer terms) NQP regression test timeout_out/001c fails with core dump in informix Test Procedure Executing the below query should not dump core on informix. delete from dbtst190 where empno in ( select a.empno from dbtst120 a, dbtst120 b, dbtst120 c where a.empno > b.empno and b.empno > c.empno and a.empno in ( select a.empno from dbtst120 a, dbtst120 b, dbtst120 c where a.empno in ( select a.empno from dbtst120 a, dbtst120 b, dbtst120 c where a.empno > b.empno and b.empno > c.empno and a.empno in ( select a.empno from dbtst120 a, dbtst120 b, dbtst120 c where a.empno > b.empno and b.empno > c.empno and a.empno in ( select a.empno from dbtst120 a, dbtst120 b, dbtst120 c where a.empno > b.empno and b.empno > c.empno ) ) ) ) and a.empno > b.empno and b.empno > c.empno ) MaintReger: # 21432 (BDNT13771): When startup prog is DEM Browser problems when logging in Baan using Webtop Date: Thu, 12 Feb 2004 11:40:22 +0100 Created on: MaintCorelli Type: bugfix Problem Description (Customer terms) Error for users that have only a dem browser configured in Webtop2.0 If a user has only assigned the dem browser in the maintain user data,it will give errors in Webtop 2.0 and the process browser can't be seen. If the same user activates the menu browser he will be able to logon to webtop and can change to the process browser in the navigation pane. Workaround Specify the Menu Browser Test Procedure Pre-requisites: - Have a iBaanERP5.0c installation running - Have the "tm" package version 2.4 installed, so you can run otmboaasmserver - Have a correctly configured busproperties file (in %SystemRoot%) - Have ASM (I used version 1.1_PA.2361_upd1) installed and correctly configured for OpenWorld. (refer to "iBaan OpenWorld Adapter Suite 2.3 - Installation and Configuration Guide for Adapter for iBaanERP", U7978A US) - Have your User Data Startup Program set to "DEM Browser" - Run regression test, by executing: jview com.eibus.test.regression.B3test "cn=koen,o=baan.com" 21 (fill in correct data) (be sure that your CLASSPATH variable contains the path to the b3.jar file) - If it works fine the test will continue to run, and if it works not fine the test directly stops after "Initializing manager 1. . ." You can also test by logging in Baan using Webtop 2.0. But in that case you need also to setup Webtop 2.0. Motive source DF:143856 MaintReger: # 21390 (BDUX14350): Informix: DF143671 Informix error 710 Date: Mon, 9 Feb 2004 11:34:19 +0100 Created on: MaintReger Type: bugfix Problem Description (Customer terms) If an existing table is dropped twice, the second drop table give Informix error 710 "Table has been dropped, altered or renamed.". Workaround Set inf_use_prepare_cache:0 in db_resource file. Test Procedure Drop an existing table twice. #1 ddrgt500 TDRTBL 0 #1 ddrgt500 TDRTBL 0 You should not get error 710. Motive source DF:143671 MaintReger: # 21386 (BDUX14465): Lazy audit start gives problem in bdbreconfig Date: Mon, 9 Feb 2004 11:33:09 +0100 Created on: MaintCorelli Type: bugfix Motive source BRD:BRQ000322 MaintReger: # 21371 (BDUX14455): Lazy start of audit server Date: Fri, 6 Feb 2004 15:09:33 +0100 Created on: MaintReger Type: bugfix Problem Description (Customer terms) Curtomer enhancement request. A lot of bshells never update audited tables. This means that a lot of audit_srv processes are started but never used. Test Procedure regression test Motive source BRD:BRQ000322 MaintReger: # 21362 (BDNT14356): Minor changes regarding SSPI. Date: Thu, 5 Feb 2004 15:45:32 +0100 Created on: MaintReger Type: porting specific change Motive source DF:141172 MaintReger: # 21341 (BDUX14422): OS400 Portingset cannot handle memory allocation more than 16 MB Date: Wed, 4 Feb 2004 13:26:32 +0100 Created on: MaintReger Type: bugfix Problem Description (Customer terms) customer runs specific standard Baan-sessions, where this is so programmed that an allocation of variables >16 MB can occur. If such - mainly long running sessions - were started by job, errors were not logged and the whole job-process aborts Test Procedure Create a test script: function main() { double ret(1) based alloc.mem (ret, 2100000) message("done") } Run this object. Motive source DF:135783 MaintReger: # 21254 (BDNT14345): DF218099 Disappearing secondary toolbar options Date: Wed, 28 Jan 2004 15:52:00 +0100 Created on: MaintReger Type: bugfix Problem Description (Customer terms) In some sessions with text-buttons in the toolbar, not all toolbar buttons are visible when the session starts. For example session: tirpt1201m000 on gamma-reger. Or session: ttzzzmjg on system ping. Workaround Undoc/redoc the toolbar. Test Procedure See problem description Motive source DF:218099 MaintReger: # 21213 (BDNT13843): CR/LF character in single line field on grid Date: Mon, 26 Jan 2004 10:58:40 +0100 Created on: MaintCorelli Type: bugfix Problem Description (Customer terms) When a using copy/paste action of text from e.g. Word to a field on a grid in the Baan software characters like CR/LF (carriage return/line feed) are copied as well. Records containing these characters can result in erroneous sort action later. Workaround Find the wrong record manually and remove the CR/LF characters. This workaround is hardly manageable. Test Procedure Start session tcmcs0145m000 (Areas). Start MSWord and create a new document. Enter some lines of text. Select the text in MSWord, including one or more Line Feeds. Paste the selection in the description in session tcmcs0145m000. The CR/LF characters should be filtered and changed to spaces. Motive source DF:217408 Total number of changIDs displayed: 31