NetNews Usenet Archive 1992 #31/NN_1992_31.iso/spool/comp/lang/ada/3824

Text File | 1992-12-28 | 3.7 KB | 76 lines

Newsgroups: comp.lang.ada
Path: sparky!uunet!spool.mu.edu!agate!usenet.ins.cwru.edu!eagle!lims01.lerc.nasa.gov!withrow
From: [e-mail] (Jim Withrow)
Subject: Re: C++ vs. Ada -- Is Ada loosing?
Message-ID: <[e-mail]>
Lines: 62
Sender: [e-mail]
Nntp-Posting-Host: 139.88.33.2
Organization: NASA Lewis Research Center, Power Systems Facility
References: <[e-mail]> <[e-mail]>
Date: Mon, 28 Dec 1992 15:23:49 GMT
Lines: 62

In article <[e-mail]> [e-mail] (Michael Feldman) writes:
>From: [e-mail] (Michael Feldman)
>Subject: Re: C++ vs. Ada -- Is Ada loosing?
>Date: Wed, 23 Dec 1992 22:18:17 GMT
>In article <[e-mail]> [e-mail] (Mark Bayern) writes:
>>
>>I'm afraid you're absolutely correct! But you don't have to look
>>into the update cycle to see Ada's commercial problems.  Look at the
>>original cost of getting the language.  Want to use C in your
>>embedded 80x86? Both MSC and TurboC list for (much) less than
>>$1000.  That and a little work and you're up and running.  Ada? How
>>about $25,000 for a license, and $5,000 for the _required_ first
>>year 'support'!
>>
>Huh? I'm curious. Whose 80x86 target compiler is $25,000? RR, AETech,
>Meridian and Alsys' DOS extended-mode compilers are all in the $1-2k range,
>commercial prices (maybe less). Who else is there?
>
>Are you thinking of the commercial prices for Unix-server-host compilers?
>I'll admit that those are too high, but not the DOS-host ones. Is your
>information up to date? Meridian? Alsys? You awake out there?
>
>If this post's information is correct, we are all in BIG trouble. If not,
>this is a great example of the lousy job we are doing in the Ada
>community in getting the facts out into the street, guys.
>
>Mike Feldman


When I first read his post I though the same as you did.  But then I read it
again and realized that he was talking about the cross compilers which run
on a dos machine and target to a bare board 80x86.  I have the Alsys 386 DOS
compiler and it will not target to a bare board (embedded) 80x86.  I also
have an Alsys 386 DOS to 80386 cross compiler.  That compiler is really the
same as the all DOS compiler, but has support for various bare board
computers, including a PC with no disks (hard or floppy).  You can ROM the
Ada executable or transfer it through the serial port.  The prices he is
talking about do apply to that very same cross compiler from Alsys.  They
gorge you for 25K + 5K for the first compiler.  Each additional compiler
costs 5K + 1K.

He does admit that he did have to do some work to get the equivalent to the
RTE moved over.

If I did not work on a project which REQUIRED the use of embedded Ada, I
don't think I would have purchased that 30K cross compiler from Alsys.  I
see no justification for the high cost of that compiler.  (That cost 6x more
than the cost of the 486/DX2-66 computer you use to work on the project!)
Imagine if you had 4 people working on the project, you would need 48K for
the four compilers, for just the first year!

Although I am talking about Alsys pricing, I have had experience with other
Ada compiler vendors who have prices which are even higher than Alsys.

Mike is correct regarding the need to get the facts about pricing out.
There might be a way to shame the Ada vendors into lowering their prices.
Although I hate to see the childish "C can do this in 2 statements and it
takes 10 statements to do that in Ada", I would like to see comparisons of
cost to do a project in Ada vs. C.  (When refering to cost, I mean cost of
tools, compilers, editors..., not hours, which is too subjective.)

Jim Withrow