Date:         Thu, 4 Jul 1996 03:21:25 GMT
Reply-To:     Jay Jaffe <j.j@IX.NETCOM.COM>
Sender:       "SAS(r) Discussion" <SAS-L@UGA.CC.UGA.EDU>
From:         Jay Jaffe <j.j@IX.NETCOM.COM>
Organization: Netcom
Subject:      Re: SAS AF Executable

Hi Don, how's yours selling?
Listen folks, don't you GET it?  SI totally knows in whose league they
play.  In fact, they get in TROUBLE when they try to extend themselves
where they shouldn't be, by overextension.  They need to stick to the
central wonderfulness that made them great:  The power of SAS reporting
procedures; statistical procedures; graphics procedures.  So what that
they have a user interface builder.  I mean that's nice and all, and I
might add a well-thought-out OO implementation (plus the best acronym
in the industry: SCL OOPS MOP), but it's no reason to use SAS software
over anything else.  So what that they have two decent programming
languages -- Data Step Language and Screen Control Language (as archaic
a misnomer nowadays as Statistical Analysis System (or San Antonio
Shoes); let's just say "SCL" forevermore).  (Actually, if you look at
it, the latter makes for a poor showing: how many times have you
desparately wanted to use SCL functions, and especially SCL lists, in
your DATA Step programs.  The best to hope for is a Rube Goldberg
mockup of DATA and PROC DISPLAY steps.  "They" -- this time it's a big
"they", the likes of Jeff P. & co., along with Doug C. in Host -- are
looking into what the heck might be done at Version 7 to give us more
compatible language features.
  So anyway, back to the point. The SAS System should be used when and
where its strengths lie and have lay historically: again, in their
astonishingly broad smorgasbord of analytic, statistical, reporting,
and graphics procedures.  Let Visual Basic or C++ try that: you'll wind
up spending $20K anyway, IF you can find it, and you will also find
that the heavy-hitter complex packages WILL charge royalties!
  How do you think SI makes money? From product splitting!  And what's
wrong with that if you're going for the big-iron, many-user accounts?
who love you because, as I've said (see below) it gets downright cheap
after a few score machines are listed to a site.
  Bottom line:  If you don't need the SAS tiger in your tank and are
just trying to work with databases, simple reporting, minimum
statistics, and a swell user interface, ... why the heck use SAS?
 
In <31DBE8AD.6517@ibm.net> Don Stanley <dstanle@ibm.net> writes:
>
>This is an interesting addition to this thread, in as much as its the
>first note I've seen that suggests a sense of happiness with SAS
current
>licensing, runtime general marketing policies. I'm pleased there are
>people out there who are happy with it.
>
>However, look at the replies from everyone else and a general thread
>emerges outside of the USA (I stil haven't seen much added to this
from
>USA). That thread is that we, as experienced consultants working with
SAS
>and SAS clients every day, are constantly fighting for the product and
 
>trying to change its image with people in the REAL WORLD. We often
spend
>more time defending the product, and trying to justify its cost
(because
>as soon as we can't we lose OUR credibility), than actually doing
>productive work for some clients. 3rd party SAS consultants are here
to
>use and promote the software by showing what it can do, and the
current
>policies of SI hinder that in a way that I suspect most SI staff
simply
>can't perceive.
>
>As for the suggestion by an experienced SAS user of many years
standing
>that we use another language, perhaps you haven't read much of this
>thread but its been about the need for SAS to re-establish its name in
 
>the market place so people think 'great product' instead of 'expensive
 
>product'. I frequently have people say to me that they never see any
>software written in SAS when we suggest the product to them, and that
I'm
>afraid is becoming a benchmark -- look at other products written using
 
>the product, then decide if the product looks right for us.
>
>If the considered opinion of our compatriots really is to go and
>use another language, then you will be kissing goodbye to some of the
>most experienced supporters of the software around. The day we start
>using another language will be the day we no longer have time to
commit
>to SAS as we do now.
>
>I'm still curious that the bulk of this thread has come from New
Zealand,
>Australia and Britain .... does this problem not exist in USA?
>
>
>Jay Jaffe wrote:
>>
>> Tim is correct, in part:  SAS programs will never compile to
>> stand-alone executables because of revenue considerations.  However,
>> this misses the point, the point being that the SAS System is
targeted
>> to corporate use and the mega-desktop license.  Economies of scale
>> rapidly kick in for a company that has, say, 100 machines ready for
>> deployment of SAS application(s); the product then becomes downright
>> cheap per machine.  BTW, I expect SI's happytimes VP of sales (let's
>> just call him "BJ") would agree: he said much the same thing at
SUGI.
>> Add to this the fact that you CAN compile SAS/AF code to run on
>> machines having only Base SAS (well, and SAS/Graph of course if you
>> want to use more than just text widgets), and you are not doing so
bad.
>>
>> If you want to sell software, write a provocative item you can
market
>> to those corporations in bulk, and not one for the individual or
small
>> business; or, make something so niche-specialized, thoroughly
valuable,
>> and cool that the price for a SAS license pales before your own
>> product's price. Or, develop in something other than the SAS System.
>>
>> - j
>> ==============================================
>>
>> In <vines.vq39+tegola@go01.comp.pge.com>
>> TWB2%Rates%FAR@GO50.COMP.PGE.COM writes:
>> >
>> >There are a couple of reasons why I am not holding my breath for a
SAS
>> >compiler.
>> >
>> >First, the executable (with associated .DLL's) would be about 120
>> >megabytes judging by the amount of space SAS is taking right now.
>> Maybe
>> >half of that is maps and other stuff I would not distribute, so my
>> >package would only be 60 megabytes.  Each of my customers would
have
>> to
>> >have a ZIP drive.
>> >
>> >Second, since each customer who uses my product no longer licenses
>> SAS,
>> >SI would have to charge me enough for the compiler to make up for
that
>> >lost revenue.  That means your compiler would cost, say, $1,000 X
100
>> >sales = $100,000 the first year.  The second year you would pay
>> $50,000
>> >to cover the lost maint. fees, and another $100,000 for your new
>> sales.
>> >After 5 years, you are paying $250,000 per year even if you sell
>> >nothing.  To me, that is a little too dear.  Oh, and your users
could
>> >not call SI with problems nor for upgrades.
>> >
>> >I think most people who think they want a SAS compiler really want
to
>> >deliver SAS applications to users without convincing the users to
>> >deliver $1,000 to SI.
>> >
>> >Tim Berryhill - Contract Programmer and General Wizard
>> >TWB2@PGE.COM
>> >Frequently at Pacific Gas & Electric Co., San Francisco
>> >The correlation coefficient between their views and
>> >my postings is slightly less than 0
>> >
>> >PS: Is someone aware of an Access compiler?  We license Access for
>> each
>> >user (not just for each programmer).
>> >----------------------[Reply - Original
Message]----------------------
>> >
>> >
>> >> I would gladly pay dearly for a one-time license of SAS if I
could
>> >then
>> >> distribute what I've created.
>> >
>> >More than anything, I want to be able to write and market software
>> >written in FRAME. After viewing MSACCESS and DELPHI and seeing what
a
>> >large market share SAS loses to them through pricing, I wonder if
>> >dedicated and honest committement to the product and company is
really
>> >in my best interests.
>
>--
>==========================+++++=======================================
====
>Don Stanley
>Don Stanley Consulting Services
>NEW ZEALAND            DSTANLE@IBM.NET