|
|||||||
REVMEDIA RevisitedTo ensure that your documentation is as up to date as possible we introduce an occasional series intended to bring previous REVMEDIA articles into line with release 2.1 of AREV. Note that we will normally emphasise information that has become outdated - new features will be left to later issues unless they directly impact on RTPSeries, VERBatim and @ATTACK. My thanks to all subscribers who have helped by notifying me of bugs or discrepancies. Thanks also to RevTech for fleshing out some of the previously "unsure" RTPs. Special thanks to Dave Mills of the Nature Conservancy Council. VOLUME 1 ISSUE 10001 Page 8 Reference is made to modifying $RTP1 to change the "Processing" 0002 message at logon. This is no longer necessary as all logon 0003 messages are stored in the DOS file REVBTMSG on REVBOOT and can be 0004 modified there. 0005 0006 Page 9 Codes S/L - the two codes are slightly different, not identical as 0007 stated. The L code is not able to access window registers. 0008 0009 Page 10 Alt-M no longer corrupts an active select list and the "locate by" 0010 anomaly has been fixed. VOLUME 1 ISSUE 2Page 2 FILE.LOC has been removed Page 3 The bug has been fixed. Page 3 The system has been modified so that if any of the prompts to be redisplayed or removed are multi-valued and/or required then all other necessary WINDOW_COMMON% variables are kept in step. Page 10 If the help level is anything other 0, screen corruption can occur whilst using GAS.BAR. Neil White of the Inland Revenue and Al Blake point out that the location parameter in the STATUP call cannot be a null value if the help level is not 0 (See 2.1 Addendum p8.69). It must contain a template name from @STATLIST. The template will not be affected by this operation. He suggests changing line 22 by inserting the following 0001 if @HELP.LEVEL # 0 then 0002 DUMMY = "LIST" 0003 end else 0004 DUMMY = "" 0005 end 0006 statup(PUSH$, DUMMY, SAVE.STATUS) Note further that if @REC.COUNT = 1, the bar will not clear down properly. This then effects subsequent calls. To avoid this problem test for @REC.COUNT > 1 rather than IF @REC.COUNT on line 17. See also the corrections in the editorial of Volume 1 Issue 3 for this code listing. VOLUME 1 ISSUE 3Page 4 Note that the "File" where the index is located is a dictionary file. VOLUME 1 ISSUE 4will break with a non- numeric message if alpha characters are summed. Page 7 RTP38 performs Filing System flushes, not complete frame flushes. Page 7 RTP42 is used to maintain system labelled common information. Page 10 This article has been superseded by the new index hook functionality. VOLUME 1 ISSUE 5support it. This tends to suggest that the functionality has been built into a lower level of RTP18 and RTP18.ENGLISH. Page 6 V6 now supports A B F and K. Page 6 V7 has been removed. VOLUME 1 ISSUE 6Page 3 Message display when using BTREE.EXTRACT can now be suppressed by using the optional flags provided. Parameters five and six in a BTREE.EXTRACT call are used to indicate options and error flags respectively. Passing an "E" in options will suppress all error messages, whilst passing an "S" will display all informational messages. (See 2.1 Addendum p 8.9 but bear in mind that it IS possible to suppress both kinds of message by passing SE or ES). Page 4 OSOPEN - the line "If the status is 2, the directory exists, if 2 a file exists with that name" should read "If the status is 2, the directory exists, if 4, the directory does not exist and if 0, a file exists with that name." Page 4 SETPTR - An additional error codes of 4 exists, meaning Not Enough File Handles. As a side point, users of products such as HIJAAK may get this message when trying to capture screens from within AREV due to the amount of files AREV holds open. To get round this, issue a FLUSH or PC at TCL - this will free up file handles. Page 6 INIT.VIEW has been amended to work better within called subroutines. However usage can be a little disconcerting unless properly understood. There are two ways to utilise report programs using INIT.VIEW, by calling subroutines that use it and by EXECUTE "RUN" programs / subroutines that use it. The latter is normally to be preferred. Unfortunately VIEWER behaves in an unusual manner which does not lend itself to a lucid explanation. The pseudo code following attempts to explain it. 0001 loop 0002 until <Esc> pressed 0003 begin case 0004 case key = PgDn 0005 if LAST_PAGE and FIRST_TIME 0006 run rest of subroutine then 0007 if called as subroutine then 0008 return to calling subroutine 0009 complete calling subroutine 0010 return to page logic 0011 else 0012 if RUN by EXECUTE/PERFORM then 0013 stay in page logic until <Esc> 0014 end 0015 end 0016 end 0017 else 0018 if not(LAST_PAGE) 0019 show next page 0020 end 0021 end 0022 case key = PgUp 0023 if not(FIRST_PAGE) then 0024 show previous page 0025 end 0026 end case 0027 repeat To clarify the above, try running the following sections of code and pressing PgDn at the bottom of the report 0001 * EXAMPLE 1 0002 perform "RUN BP TEST" 0003 * NB This only displays after <Esc> 0004 call msg("And back","","","") 0001 * EXAMPLE 2 0002 call TEST 0003 * NB this displays after 1st PgDn at bottom 0004 call msg("And back","","","") 0001 subroutine test 0002 call INIT.VIEW(40) 0003 @VIEW.MODE = 1 0004 heading "TEST REPORT" 0005 for X = 1 TO 150 0006 print "TESTING INIT VIEW " : X 0007 next 0008 call msg("Bottom of Report 1!","T2","","") 0009 for X = 1 to 150 0010 print "and more! " : X 0011 next 0012 call msg("Bottom of Report 2!","T2","","") 0013 return Note further that VIEWER now uses the video attributes set up in @AW so that the colour may be controlled by manipulating the following @AW fields :- < 1 > VIEW window border and title < 2 > HEADING/FOOTING and COLHEADING < 3 > Background to use in absence of any other (indicated by a CHAR(127) in the background byte position). < 8 > Main text colour < 9 > Border type Page 10 SECUREUSER - see editorial of issue 8 for correction to object code modification. (Volume 3, Issue 4, Pages 6-9) |
|||||||
| |||||||