comp.lang.ada
 help / color / mirror / Atom feed
From: Roy Grimm <ragrimm@collins.rockwell.com>
Subject: Re: Rational Apex
Date: 1998/08/05
Date: 1998-08-05T00:00:00+00:00	[thread overview]
Message-ID: <35C8AA20.794B@collins.rockwell.com> (raw)
In-Reply-To: 6qa2n2$4d5@newshub.atmnet.net

Christopher Green wrote:
> 
> Apex is not just a compiler; it's a way of life.

So true.

> Apex has its own highly structured way of doing things.  It is insistent
> about using its own directory structure and file naming conventions, at
> a minimum.  It will reformat your code according to its defaults, unless
> you force it not to.  It has its own CM subsystem (Summit) built in.

I'm assuming you mean Change Management instead of Configuration
Management.  Summit is available as one of the many layered products. 
However, you don't have to use it to track changes  You're free to use
your old tools.  In many of our legacy projects, we still use our old
Change Reporting tools instead of Summit.

> Apex can be run from the command line, but this does not reduce the need
> to organize your code into Apex subsystems and views, and at least to
> observe Apex file naming conventions.

But this isn't necessarily a bad thing.  Getting used to their structure
of subsystem/views with multiple history CMVC takes a while, but adds a
layer of organization that makes managing large projects significantly
easier.  Our department makes a reusable bootstrap loader and software
loader that is compiled to several different processors.  With Apex, we
assign different version history trees to critical configuration files,
make the rest of the code reference those specific packages.  When we
make a change to one high level package, it's a few mouse clicks and
we've updated every other platform we want to.  Very slick, very easy. 
I'll never miss the old Unix/SCCS days, or that brief stint with
WindozeNT and Micro$oft Visiual Source Save...  *shudder*

> Basically, Apex is a heavyweight solution to many of the problems that
> arise in developing software on a very large scale.  But you need to use
> Apex the way Rational designed it; if you try to impose your own way of
> doing things on Apex, you will spend all your time fighting it rather
> than getting work done.

That's half true.  Apex has it's way of doing things.  One of those ways
is providing the flexibility to manage your project with their tools or
someone else's.  The basic file and directory structure is provided for
you.  Everything else is configurable.  Don't want to use Rational Rose
for documentation?  Don't.  Want to use another compiler?  The Apex IDE
has the Remote Compilation Interface that lets you run any third party
compiler you want.  Hell, I've worked on Apex running on a Sun
workstation that remotely compiles on a VAX/VMS workstation.  A little
set up for the 'rsh' and NFS interface and you're good to go.  It's all
automatic from within the IDE.  Just hit "link" and away it goes.

If you've got just one line of products on one target architecture, Apex
might be overkill.  If you've got dozens of projects with several
different projects and requirements to support legacy compilers and what
not, Apex is the way to go.

-- 
Roy A. Grimm             |  Tel:   (319)295-8099
Rockwell Collins, Inc.   |  Fax:   (319)295-8940
Cedar Rapids, Iowa       |  email: ragrimm@collins.rockwell.com

When you think how well basic appliances work, it's
hard to believe anyone ever gets on an airplane.     --Calvin

My opinions don't necessarily match those of my employer.




  reply	other threads:[~1998-08-05  0:00 UTC|newest]

Thread overview: 12+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
1998-08-05  0:00 Rational Apex Glenn
1998-08-05  0:00 ` David  Weller
1998-08-05  0:00   ` Christopher Green
1998-08-05  0:00     ` Roy Grimm [this message]
1998-08-07  0:00       ` Lowe Anthony A
1998-08-11  0:00         ` Gene Ouye
1998-08-06  0:00     ` Matthew Heaney
1998-08-14  0:00   ` Samuel T. Harris
1998-08-08  0:00 ` Chris Warwick
1998-08-09  0:00   ` Corey Ashford
  -- strict thread matches above, loose matches on Subject: below --
1998-08-20  0:00 James Amendolagine
1998-08-27  0:00 ` Samuel T. Harris
replies disabled

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox