Isaac_K_Rabinovitch@cup.portal.com (12/11/87)
After my recent comments on Fastback's incompatability with certain
hardware configurations, I got an interesting letter from a certain
cmaag@csd4.milw.wisc.edu, disputing my moral assumptions. His return
address doesn't work for me, and besides it occurs to me that the issues
we're arguing are of general interest: what are a software publisher's
responsibilities? And what does the word "standard" mean? If you think
this sort of argument is out of place in this newsgroup, by all means
flame me.
To: cmaag@csd4.milw.wisc.edu
Subject: fastback
Lines: 55
Date: Wed Dec 9 14:03:31 1987
Message-Id: <8712091403.1.29668@cup.portal.com>
X-Origin: The Portal System (TM)
X-Possible-Reply-Path: Isaac_K_Rabinovitch@cup.portal.com
X-Possible-Reply-Path: sun!portal!cup.portal.com!Isaac_K_Rabinovitch
1>
->In article <1802@cup.portal.com> you write:
->[...]
->>
->>That's unnecessarily intemperate. The fact is, there's little in the way
->>of hardware standards for PCs, so given the large number of RLL drives out
->>there, the onus is on Fastback's publishers to either make the product
->>work safely on RLL drives or make it clear that it shouldn't be used with
->>them.
->>
->[...]
->If software publisher's spent their time testing every possible combination
->of controller/drive/clone/DOS version/resident software/etc... they would
->never bring a product to market. It is the onus of the *user* of
->somewhat non-standard equipment to assure that software designed for
->*other* (standard) hardware will work reliably on their machine.
You're not a very careful reader. Look again! Where do I say Fifth Generation
should test *all* combinations? I simply point out that RLL is a *very*
commonly used technology, sufficient to justify documenting Fastback's
incompatibility with it. You also seem to be assuming some nonexistant
disk technology standard for PCs (and simply ignoring my statement that
there isn't).
BTW, Fifth Generation *did* check out DOS version and resident software
compatibilities: read the documentation.
->>I myself bought Fastback after hearing it praised to the skies by nearly
->>every user I talked to. It turns out to be incompatible with my floppy
->>controller! If there's any blame to be laid here, it's in the fact that
->>the PC marketplace still resemble a hobby shop.
->>
->[...]
->And why is your floppy controller incompatible? I would guess it is because
->it is non-standard. Again, it is impossible to check for this type of
->thing. I hope that you let the company know about your situation so
->they can program a fix.
My floppy/hard controller is the one that comes with all Heath/Zenith AT clones.
Zenith has a very good compatibility reputation, and Fastback is the
only software I've had trouble with. The weirdness seems to happen when
Fastback tries to format the floppies (which it can supposedly do in
one third the time of DOS). Judging from what I've been told (there was
an interesting discussion on the net about a month ago), this special
formatting procedure can trip over hardware differences that are indetectable
by any other procedure.
Fifth Generation seems to be aware of this problem: the new version of
FB allows you to use disks formatted with DOS. (Unfortunately the upgrade
fee is hopeless.)
-> Does the box specifically list your floppy controller
->as being compatible? If not, you don't have a gripe.
The only hardware requirement mentioned on the box is "128K memory and one
floppy drive". The box does not mention or exclude *any* specific brand
of compatibile; there's a clear implication that it works on them all.
Isaac Rabinovitch
Disclaimer: Just because I think you're wrong, doesn't
mean I don't think you're a fun person!
:-)brad@looking.UUCP (Brad Templeton) (12/13/87)
In article <1907@cup.portal.com> Isaac_K_Rabinovitch@cup.portal.com writes: >what are a software publisher's >responsibilities? And what does the word "standard" mean? If you think >this sort of argument is out of place in this newsgroup, by all means >flame me. Not out of place yet. If comp.industry or comp.industry.pc or whatever is created to provide a forum for non-technical computer related discussions, then that would be the place. Now, on to compatability rules: If you're a non-major software vendor (I am), then this is a very tough issue. There's no way in the world you can test on even a fraction of the configurations out there. Even companies as large as Microsoft have trouble doing this. You try as many machines as are available to you, and then you count on beta testers. But beware, for beta testing is harder than people think. You will really only have a small number of good beta testers. Particularly if you're making a specialty product in a niche market. If you're making hardware configuration dependent software like a special device driver, you do what you can. But in the end the best a small company can do is try it on the major computers in their standard configurations and say, "For use with IBM-compatible computers." Unfortunately, tons of things are not in the realm called IBM-compatible. If your dealer sells you a machine, saying it's compatible, and a program doesn't work on it, your dealer was wrong. Which is not to say that software vendors don't try to run on as many machines as they can. If they don't run on your clone, they lose a sale. But you must understand that there *is* a reason why IBM gets to sell for more, and if you decide to buy something cheaper, you should fully expect to not get some things. > >I simply point out that RLL is a *very* >commonly used technology, sufficient to justify documenting Fastback's >incompatibility with it. Was it that common at the time your copy of FastBack was written? Here's where larger companies get in trouble -- they can't make a new release with every new hardware change. They release once or twice a year, if they're very efficient. Lotus 123, the most successful software ever, releases once every 2 years. Believe me, it's not a matter of "edit the hardware page to say don't use RLL". These manuals are printed in 10,000 copy lots, and new releases and print runs must be coordinated to avoid inventory wastage. RLL was not that common a year ago. Don't be so surprised. -- Brad Templeton, Looking Glass Software Ltd. - Waterloo, Ontario 519/884-7473
Isaac_K_Rabinovitch@cup.portal.com (12/16/87)
brad@looking.UUCP (Brad Templeton) writes:
->>I simply point out that RLL is a *very*
->>commonly used technology, sufficient to justify documenting Fastback's
->>incompatibility with it.
->
->Was it that common at the time your copy of FastBack was written? Here's
->where larger companies get in trouble -- they can't make a new release with
->every new hardware change. They release once or twice a year, if they're
->very efficient. Lotus 123, the most successful software ever, releases once
->every 2 years.
A good point, though I hope most software developers don't set out to operate
like Lotus! Anyway, the files on my Fastback diskette were created last
June 1. I *think* RLL drives were common then, but since I don't use them
I don't really know.
->Believe me, it's not a matter of "edit the hardware page to say don't use
->RLL". These manuals are printed in 10,000 copy lots, and new releases
->and print runs must be coordinated to avoid inventory wastage. RLL was
->not that common a year ago. Don't be so surprised.
Which is why The Lord invented Update Notices. There's a one-page update
stuffed in the Fastback manual, but it's a year old!
I don't suppose Fifth Generation has done that bad a job, considering the
kind of technology they have to work with. But neither are all the people
complaining of Fastback problems demented crybabies.
Addressed to the public at large:
Probably we've worn this topic out, but here's a suggested idea for the
next Big Software Controversy: when you see a posting from someone who
complains that your favorite program blew up on him, don't assume the
person is a brain-damaged wimp. Even if the fault is his and not the
software's (and you can *never* know that a piece of software doesn't
have some tragic bug you've never run accross) your responsibility to
those less expert than you is to help them raise their expertise, not
cuss them out for having been born a twit. Besides, you may have something
to learn yourself!
Isaac Rabinovitch
Disclaimer: Just because I think you're wrong, doesn't
mean I don't think you're a fun person!
:-)