------- ASISWG at WAdaS'93 On Tuesday evening, 29 June 1993, ASISWG held its second officially sanctioned SIGAda Birds-of-a-Feather session; this time, it was at WAdaS'93. The ASIS Working Group met again all day Friday, 2 July 1993, at MITRE's location in McLean, VA. Tom Shields chaired both of these meetings. After the general Birds-of-a-Feather session on Tuesday, ASISWG held a Technical Meeting. Tom Shields, chairman of ASISWG, welcomed everyone who had attended the ASIS BoF to stay for the ASISWG meeting. Tom distributed comments about ASIS from the Beijing Green Valley Software Company, Ltd. of mainland China. ISO/WG9 ASIS Work Currie Colket gave the current status of the ISO/WG9 work related to ASIS. There were a couple of enthusiastic responses about ASIS from ISO/WG9. However, they will wait until later this year (August) so that the standardization of Ada9X is not impacted. Five nations are needed to sponsor a standardization effort; ASIS currently has the support of Belgium, Canada, France, the United Kingdom, and the United States. We hope to have a Work Item established at the August meeting of ISO/WG9, which will be held in Boston during 16-20 August. Vendor Interest in ASIS When asked about the size of interest in ASIS with respect to vendors' customer bases, o Alsys/Telesoft indicated that its ASIS related customers number less than 20, but that they are "significant" customers o Rational's LRM (predecessor to ASIS) has a customer base of about half the entire Rational customer base; nearly all of these customers will move to ASIS. There was a discussion about collecting tests for ASIS implementations. Gary Barnes of Rational will put these tests, once they have been checked out, onto the AJPO host periodically. ASIS at WAdaS'93 ASISWG held a panel at WAdaS'93. The panel presenters were: o Tom Shields, ASISWG chair, of Paramax; Tom presented a brief history of ASIS and ASISWG and described ASIS Goals and Benefits. o Currie Colket, former Navy Deputy to the AJPO; Currie presented the AJPO's position on ASIS. o Steve Blake, ASISWG vice chair, of Alsys; Steve described the Alsys ASIS implementation. o Gary Barnes, ASIS archivist, of Rational; Gary discussed Rational's implementation of ASIS. o Experience reports were given by ASIS users: * Mike White of IBM Boulder; Mike discussed their use of ASIS in Phase I, a data retrieval and analysis program. * Alex Blakemore of Genoa Software Systems; Alex reported on an ASIS based style checker. * Cheryl Barbasch of ORA; Cheryl discussed the use of ASIS in the development of verification tools. Tom Shields recognized that the WAdaS'93 ASIS panel was successful and thanked those who participated. Many of the WAdaS'93 panelists mentioned that (the public-domain STARS version of) DIANA had been used in previous versions of their tools or applications, but that ASIS was better. ASIS had also been mentioned in the I-CASE RFP. ASIS at Tri-Ada'93 At Tri-Ada'93, ASISWG will be well represented. Tri-Ada'93 accepted Mickey White's paper, accepted ASISWG's "ASIS panel" proposal "ASIS: Tools with a View, or Exposing Ada", and accepted Steve Blake's half-day "ASIS Tutorial". Tom Shields requested an ASIS Birds-of-a-Feather session one evening during the Tri-Ada'93 conference. He will invite Oliver Cole and/or Tom Fleck of OC Systems to discuss an Interactive ASIS Browser and an interactive Ada cross-reference tool which uses ASIS, both currently in the development stage. Currie Colket, former Navy AJPO Deputy Director, gave an update on the ASIS-related ATIP (Ada Technology Insertion Project) contracts at the WAdaS'93 ASISWG meeting. There were four such proposals, but only three were funded; Tom will invite all three of these ATIP winners to speak at the ASIS session at Tri-Ada'93. o Captain David Cook of the Air Force Academy o Romel Rivera of Xinotech o Tom Strelich of GRC ASIS 1.1.0 We will try to have ASIS 1.1.0 (our official name of the ASIS 1.1 version with the changes we approved and will approve) by the time of the August ISO/WG9 meeting, in accordance with the following schedule: --------------------- ASIS CALENDAR FOR July/August 9 July All errata comments back to GARY BARNES. 16 July GARY BARNES to have ASIS 1.1.0 to everyone on net. 30 July Comments back to GARY BARNES. 6 August Final draft version of ASIS 1.1.0 on AJPO host. 13 August Functionality of ASIS 1.1.0 finished. 16-20 August ISO/WG9 meeting. TOM SHIELDS to make sure that ASIS is on the agenda for this meeting. 27 August TOM SHIELDS to email short synopsis of ISO/WG9 meeting to everyone. --------------------- Special Issue of AdaLetters We discussed the idea of publishing a special issue of AdaLetters which contains only ASIS materials. The special issue should contain the following three items: 1. Introductory material; this might include some slides from ASIS Tutorial(s). 2. Updated Implementors Guide 3. ASIS 1.1.0. ASIS Artifacts on the AJPO Host Eventually, we would like to have the following ASIS artifacts uploaded to the AJPO host machine (ajpo.sei.cmu.edu). o ASIS 1.1.0 documentation and Ada files o Structured Discussion and ASIS LRM Cross-reference o Implementors Guide (currently this is an overview document) o ASISWG Charter o ASISWG Mailing List o ASISWG Calendar of Events o ASISWG Schedule of Meetings o Resources available, i.e., hardcopy slides, papers, etc. and who to get them from. o AJPO Position Paper o ASIS Users List o ASIS Vendors and Implementors List o Planned Products o Bibliography One idea discussed was to formalize the Tutorial(s) that have been given in order to develop some Introductory ASIS material (similar to the John Barnes Ada9X Tutorial available on the AJPO host). The ASIS Overview document should be for ASIS application developers. ISO Standardization of ASIS (Version 2.0) We reviewed the schedule for ISO Standardization. The assumed dates are so that we can easily fit in with the Ada9X standardization schedule. In the following, '*' indicates important milestones for the Ada9X standardization process to meet; '+' indicates that the major emphasis for ASISWG will turn to ASIS 2.0 (for Ada9X) and that ASIS 1.1.0 will be relatively stable. ---------------------------------------------------------------- Aug 93 + "Start of Schedule"; WG9 New Work Item request for ASIS 1.1.0 as a proposed Committee Draft Sep 93 ASISWG meets two nights during Tri-Ada'93 Nov 93 * Ada9X complete; submitted to ISO as DIS Feb 94 At end of ASISWG meeting, ASIS 2.0 requirements defined March 94 ANCOST'94 in Williamsburg, VA April 94 ASIS "2.0 draft 0" -- includes Ada9X syntax, Ada9X semantics (for Ada83 subset); consider candidate architectures and come up with a trial list. June 94 WAdaS'94; select finalist architecture, put on AJPO host for internal review and discussion. Ada9X compiler vendors could start prototyping the selected ASIS architecture (even GNAT). Aug 94 ASIS "2.1 draft 1" -- all Ada9X core semantics Sep-Nov 94 Tri-Ada'94; develop document with BNF Cross-references for Ada9X Dec 94 * Expected ANSI and ISO approval of the Ada9X standard Feb 95 All Core Ada9X syntax and semantics "complete" (this is the real ASIS 2.0); first public review? April 95 ASIS "2.2 draft 2" -- add Annexes for Real-Time, MIS, Security, Distribution June 95 WAdaS'95; the real ASIS 2.1 -- add necessary information for the Ada9X Annexes ---------------------------------------------------------------- ASISWG will meet again at Tri-Ada'93 in September. By that time, ASIS 1.1.0 would have been delivered to ISO. At this meeting, perhaps we can hear presentations on different approaches to the development of ASIS 2.0 with rationale. Then we can debate those approaches. The approaches identified thus far are: 1. Do Nothing, that is, just add stuff for Ada9X to the then-current standard. 2. Let the LRM syntax drive the creation of ASIS 2.0. 3. Let ASIS 2.0 be basically one procedure that is context driven; this is effectively the schema approach (similar to DIANA); it is also known as the cursor or schema model (from John Nissen's email discussion). Other issues to be resolved about ASIS 2.0 include: o Dynamic semantic capability. o Addition of a write capability (if so, at what level?). o About 200 proposals deferred from version 1. -------------------------------------------------