#CertDeveloper Betas?
121 messages in this thread
Could someone please explain the rational behind giving beta's (Yey!) to
certified developers only? Perhaps this is where this thread started
(sorry), but what's so special about developers? They would seem to be a
rather narrow audience, at best. I'm not a certdev, but I am definitely
not a "Joe User" (tho' I have been called that, on occasion). I have
worked on computers (mainframes mostly) for 12 years as a sysprog and have
lots of experience in trying to break things and in figuring out what broke
them. I AM NOT VOLUNTEERING, but, again, what's so special about certdevs?
Well, for one thing, we "certdevs" have signed a non-disclosure agreement.
"Joe User" hasn't. In other words, if we leak, we get plugged.
Several reasons:
1. They can test their software/hardware on it…in particular software
that used to work on 1.3 and now doesn't work on 1.4 to find out why.
2. They're generally more familiar with how things should work, and thus
why they don't work if there's a bug in the OS.
3. To give them an opportunity to insure their software will be ready when
the OS is released.
4. Beta releases have to be limited, general release would be disastrous.
5. Many users, blame problems on the OS, when the real problem is in what
the user is doing. Again, developers supposedly can differentiate between
the two types of problems.
6. Developers pay for the privledge.
Except for #6, and possibly #5, the reasons you note would appear to be
reasons for offering betas TO certified devlopers, but why to certdevs ONLY
(if I am correct that the betas are being offered to certdevs only)?
#6, that certdevs pay for the privledge, is pretty important if the fee is
high, say greater than $20, and with 1.4 there is the added complication of
the ECS.
As for number 5, that users would be more likely than developers to blame
problems on the OS than certdevs would be, perhaps. As I say, I, and
people like me, have enough background and experience to keep my mouth shut
unless I can back up a claim of an OS bug.
I'm sensitive to this issue, just now, because I'm working for a big
computer company which has raised secrecy to the level of religious dogma;
to the point where it is almost impossible for them to produce a product
which meets the needs and desires of its intended users.
CBM (actualy CATS) takes input from all on what features the users want.
This could be via letter, or communications via public networks that those
folks vist.
They also hold regular developers conferences and anyone willing to
pay the fee can attend. They also allow almost anyone to become a
certified developer for an initial fee of $75 plus a yearly fee of $50.
How do they judge "Joe Average User's" ability to properly test
software…especially OS software? Becoming a certified developer is one
of those ways.
There are very good reasons for secrecy. Number one, features in the beta
releases aren't firm. Things change. As with 1.2 and 1.3, copies of the
software got out very early (early betas) and suddenly there were lots of
folks who were very upset that some had the software,and they didn't. Not
only that, some features either changed, were dropped or added as the beta
levels went up. Again…that caused problems.
There is no firm date for the release of v1.4, it may be weeks off, or it
may be months off. If they allowed leaks, just like v1.3, magazine
articles would be full of stuff about v1.4 that may never appear or when it
does appear, it may work differently than the reviewer stated it would
work. And again, folks (some) would get upset because the feel that CBM may
have promissed some feature..and itdidn't appear. Finally, like all beta
software, it can cause you problems. It can crash your machine, it can
wipe out files, etc. Folks could give it a bad reputation..undeservedly
(is that a word?) because they wouldn't understand what they were dealing
with. If the release version gives you problems, fine then knock it…but
not a beta version.
v1.4 is a significant change from v1.3. Some software will break under
it…and thus CBM would like to know what software is going to break, and
to make sure that they aren't to blame (authors didn't follow the
rules..not that CBM has buggy code or broke their own rules). That's why
only developers should have it at this stage, and why secrecy is
necessary.
I can't say much since I'm under non-disclosure. My feeling for betas
going to commercial developers first then certified developers is so CATS
can control the info gained from their experience. That information will
more than likely be more reliable than if it came from the general user
population. A lot of things will break under 1.4. The vast majority of it
will be because the application program is written incorrectly, not because
of bugs in the OS. The developer community is more likely to be able to
determine the problem rather than report a general "XXX doesn't work, the
OS is broken". If something doesn't work the problem has to be narrowed
down to exactly what is broken, the app (most likely) or the OS. CATS
feels the developer to be best suited to track that down.
I'm going to make a very strange statement. The RELIABILITY of information
may not have anything to do with the USEFULNESS of that information.
Developers may be able to tell CATS what works and what doesn't, but are
developers, more than anyone else, able to tell CBM if their changes to the
product are useful, or just different? A very naive user would still be
able to answer questions like: What problems did you have in converting
your system to the new product? What new features did you find especially
useful? How much would you pay for this new release? Etc.
My point is that input from a selected, limited number of non-developer
types and even from naive users could be just as important. I will grant,
however, that servicing this type of user could be a pain.
Now that I've turned section Ramblings back on, I can hear what you guys
are saying. Reading over the other messages that weren't addressed
directly to me, (which explained why I only read some of them with 'rm',
and saw none of them with 'rtn',) I saw that some people were saying that
regular users should get betas of new OS software. I don't think that's a
good idea – obviously, a beta isn't open to changes. Theoretically. 🙂
But I had agreed when someone pointed out that it wasn't apparent that CBM
bothered listening to anyone but developers when it came to deciding what
goes into the next rev of the software. For example, I don't recall either
developer or user clamor for a version of uEmacs to be added, but one day,
there it was.
Would it do any good if I clamoured to get RID of Emacs? And ed? and edit?
Maybe they could stick TxEd on there – that's a good editor for starters…
once people get the hang of it, they could move up, but at least they'd
have a _real_ editor.
I've seen the "real" emacs… sucker has a command language thats common
lisp. More or less. It's so programmable, I couldn't even begin to tell
you.
Lucky you. ;^)
Anyway… those three editors, of which not ONE is worth the disk space
they take, would be well replaced by TxEd.
Even if Charlie _is_ a BIXer. ;^)
Ben
Amateur Radio Callsign is A A 7 A S
Does any kind of clamor, besides begging for obscure technical features,
appear to have much impact at CBM? 🙂 Yeah, the editor situation appalls
me. I agree, CBM would have been better off buying the rights to TxEd,
than wasting someone's time to "port and Amiga-ize" uEmacs, seeing that
it's already been done a dozen times, and in the Fish collection, no less.
(Sound of Charlie falling off his chair…)
I also think they should have bought Perry's RAM disk, instead of making
RAD. Think of all the disk space they could have saved, instead of doing
so many other things to make more room on Workbench.
yeah… <glowering, working up head of vapor-state H2O>
Ben
Amateur Radio Callsign is A A 7 A S
Yeah, the Amiga stock editors surely are terrible. Don't know why they
couldn't put in one like that really user friendly, powerful and flexible
editor that comes stock with that computer REAL serious computer folks use.
What's it called now?
Oh yeah — EDLIN ;^)
Ray Lewis
I think they dropped EDLIN in DOS 4.0.
EDLIN's still there in DOS 4.0, you just have to do a "full install" and
obey the prompts. Though Real Programmers Don't Use Editors; they use COPY
CON: <filename>
Great news!! If there is anything worse than ED, it has to be EDLIN.
……Betty
Well, that's certainly one good thing they did with 4. Everytime I get
upset about ed, I just think of edlin. On the bridge at work, when I want
to edit BAT and CONFIG files, I awrite them to vd0:, and use memacs!
Might delete memacs now, though. I got transscript, and it seems very
nice.
Ray Lewis
Hey, cut Edlin some slack! Well, OK, don't — it *is* pretty pitiful, but
it's *there*. Plus, some people are pretty awe-struck when they see you
type commands like "4,7,10c", "8<enter><F2>;:<F3><enter>" and such.
Being a computer consultant on the PC, sometimes impressiveness is
important. <grin> But seriously, I'd rather have Edlin than nothing
(editing is a bit tough using the COPY CON method).
Bryan
Slack? No way! That is without a doubt the most pitiful excuse for an
editor that I have ever had the misfortune to attempt to use. Ed beats it
hands down, six ways from Sunday, and probably 10 or 12 other cliches that
I can't think of right now, and Ed is pretty primitive.
-larry
Har!
We talked briefly with Commodore about putting a TxEd derivative on WB.
But at this point I think I agree with Perry K on this one, the best thing
Commodore can do with editors is just not to do anything at all for a while
C-:. Though I'd certainly hope they'd clean up NotePad's act.
Don't you think that the disk space saved would be worth the time? I mean,
if they save the cost of a disk in the upgrade, then (for instance) they
could give you 50 pesos per upgrade, and come out ahead. Heck, Charlie….
ed is 20k, edit is 18k, and emacs… well, I deleted emacs. but I'm sure
it's at least another 20k. Probably more. As for notepad – I think I
exposed that to a biological agent and watched it die screaming. Or was
that MaxiComm? hmmm.
Ben
Amateur Radio Callsign is A A 7 A S
I've often drooled at the possibility of making a $0.50 royalty on WB
disks, for saving disk space. But Commodore's bottom line is cost of
development, and time to market … I don't think they're too concerned
about a few million disks here and there. Too bad, it's what I do best. A
dinosaur in my own time.
Memacs is more like 50K, BTW. I'd have burned NotePad, except it
mysteriously keeps dissappearing before I can get out the torch C-:.
C
It still seems to me that the time it would take to completely delete all
vestiges of notepad, emacs, ed and edit off the disk set, added to the time
to install TxEd, wouldn't add up to 10 minutes – and _you_ could do it,
costing them nothing except when they sell a disk.
It's too bad CBM doesn't have to support the users – if they did, they
would have had all that trash fixed years ago.
Ben
Amateur Radio Callsign is A A 7 A S
There's also the issue of documentation, and supposed scripts which use
Edit, etc. Once anything is released, it's a lot easier to just keep it
there than to toss it, at least in C=land. So probably what would happen,
even if they did pick up TxEd or something like that for WBench, would be
that then there would be Five editors on WB, instead of Four.
Heck, while they're at it might as well go for an even half-dozen C-:
Memacs is 62940 bytes.
(Do the words "incredibly large" mean anything to you?)
Ray Lewis
They mean that TxEd oughta be there. Or something else, anyway, besides
that lame port.
Ben
Amateur Radio Callsign is A A 7 A S
Ray –
Seen on USENET in a discussion on possible acronym meanings:
EMACS – Eight Megabytes And Constantly Swapping
Jeff Schweiger
Notepad! I'd forgotten, there are _four_ editors supplied with the
machine, not just three.
Hmm… that's still 3… surely you can't call Notepad an _editor_? <grin>
Of course, you certainly can't call it a word-processor, either. May have
to give this one some thought 🙂 –Eet–
I've shown computerphobes (CLI-phobes?) how to edit their startup-sequences
with Notepad, so it's an editor.
I've written letters with it, using the various fonts (and I think
styles?), so it's a word processor <grin> –Eet–
There's one really great thing about Notepad; if you delete it you free up
lots of space!
Ron
Now there you have something! <grin> –Eet–
Now, now. Edit is the only editor I know that lets me edit a 5 MByte file
on a 3 MByte machine.
Are there any other Amiga editors that will do that?
I don't know – I've never seen a file that big… for that matter, I never
hope to. How would one transport it? Besides… the correct answer to THAT
problem is more memory – not editors that thrash the disk – I profoiundly
hope that the day of the limited memory machine is going before us, much
like the neanderthal, and for the reasons John uses – machines with small
memory capacity can't draw a file. ;^)
Ben
Amateur Radio Callsign is A A 7 A S
Just a few color separated super-bitmap pics on a postscript page, and you
get BIG text files.. 13megs is the largest I remember… 44meg removable
drives is how you get them to your service bureau
Sounds about as convenient as raytraced animations. And about as practical.
Ben
Amateur Radio Callsign is A A 7 A S
Gee, with virtual memory, would one _need_ and editor like edit?? –Eet–
no – good point.
Ben
Amateur Radio Callsign is A A 7 A S
Good point only for the issue of being able to edit a file larger than
physical memory available. Not a good point for other uses of Edit, like
script driven editing.
-larry
Jim,
Certainly …. even with virtual memory, one should use as little
of it as possible. Loading a huge file into memory to edit it (virtual or
real memory) doesn't meet that objective.
Don
Hmm.. I always figured that with virtual memory, the file would be in
memory always, since it was on the disk… guess I thought wrong.
–Eet–
Don and Eet –
Working for a living, as I do, in a virtual world I can bring some
experience about whether or not it is a good idea to load an entire large
file into virtual memory. I work with VAX/VMS. In VMS there are two major
editors, EDT and EVE. EDT, the elder editor, only reads in a part of the
file. EVE, the younger and more vigorous, reads the entire file into
virtual memory.
Keeping functions and features completely out of the picture, EDT opens and
is ready for editing much sooner on larger files than is EVE. On the
otherhand, once EVE has read in the file, it moves around it much more
quickly. My least favorite thing to do in EDT is read in a large file and
then goto the end of the buffer. Personally I prefer EVE. I know both
_very_ well. While it may appear to be a waste of memory, keep in mind
that a virtual memory, multitasking machine, need only keep some of the
file actually in memory. The remainder, as necessary, can be paged to
disk.
My 2-cents.
– Nelson …
Thomas,
WordPerfect.
Falls into the wordprocessor category rather than "editor" category,
but close enough?
Lloyd
What keeps WordPerfect in the "wordprocessor" rather than "editor" category
is is, of all things, its line length limit! You are correct that it can
operate upon a file which is much, much bigger than memory (but watch out
for "disk full" messages, because it can require a disk scratch area twice
as big as the file being edited). However, version 4.1 has a hard-coded
line length limit of 250 characters. That means any verbose stuff gets a
soft-return inserted.
Also, WordPerfect can get grumpy about non-alphanumeric characters in the
file. Though one can learn to read [HRt][TAB][BOLD] as easily as hex, the
lookup table is not standard. 🙂
Thomas,
Hard drive users hopefully wouldn't have to worry about the "disk
full" message. Floppy users would be masochists to be using WordPerfect as
an editor anyway ;^).
GAWD, if I had programs that made use of the full 250 character line
length "limit", I think I'd go nuts trying to read/edit/debug that thing.
Line lengths that fit within the SCREEN – or at least a PRINTED PAGE – are
more to my liking. That may just be a personal preferences though.
Most of the source code I work with is alphanumeric anyway, so there
is no big problem there. I'd want to save the file as a TEXT FILE, rather
than in WordPerfect's file format. I suspect it would drive most compilers
pretty bonkers trying to figure THAT out!
P.S. WP has an "editor" on their LIBRARY disk, I believe. I don't
have it, so I can't say if it will edit "unlimited" length files – or be
any better handling the "problem" areas you mentioned.
And I *don't* use WordPerfect as a programming editor, in case you
were likely to ask.
Lloyd
The MS-DOS WordPerfect Library will handle the unlimited line length and
character set, but does not appear to buffer to disk.
I think the whole problem is that Commodore would have a severe
signal-to-noise ratio on bug reports, and would most likely be swamped with
Joe Users who couldn't figure out how to get out of the CLI. Allowing
people to beta test software who really aren't qualified to beta test it is
just asking for alot of headaches.
-Mike
I guess I must have bought the wrong machine. I thought that any company
selling ANY product would have a keen interest in what it's users/consumers
thought of the product. A "lets make it for deveolpers and see if anyone
else will buy the beast" attitude is hardly likely to make an impact on the
buying public.
Who is qualified to beta test, then? Software writers whose main interest
is putting in "hooks" so it will interface with their own software?
Hardware engineers whose interest is to see that their ideas of what should
be attached after market could be included? If the hardware/software is
broken, even those common users you seem to be so down upon should be able
to pick up on it. OK, so maybe they won't pick up on the AREXX hooks being
faulty, but they should be able to know when the machine GURUs on power up,
or the mouse pointer locks in place whenever a certain icons is chosen.
AND… those things are just the types of things that will drive the buying
public away from a machine, not the lack of AREXX or whatever.
I am now going out to get my brain washed, so maybe I can forget what
ENDCLI means.
Bill, K3OZM
The fact is, developers are going to find far more bugs in the system then
the average user will due to the fact that developers are the ones that
usually push the system farthest to its limits, as well as the fact that
developers are more conscientious (in general) about isolating exactly
where a problem might lie.
Of course you're going to have some developers who may be the world's worst
beta tester, and an 'average user' who is a fantastic beta tester. But on
the average the developers are going to do a better job. Everything you
mentioned above – seeing GURUS, etc., are going to be caught by developers
as well as end-users. But its the developers are going to catch the more
advanced things in addition to the simple ones. And since beta
distribution is by virture limited in nature, I believe the majority of
people who are involved in it should be developers.
-Mike
Bill:
Beta test is AFTER all the decisions have been made. It is for bug
fixing. User input is useful at ALPHA and before. It is too late once it is
in beta to make design changes for that release. I am all for users to have
all the input they want while decisions are being made. When it's in beta,
I want someone testing it that will recognise a bug when they see one and
will have an idea as to the likely solution.
Clint <VRLI>
I think the issue here is one of beta versus gamma or perhaps focus group
testing. That Commodore has never really listened much to users is a
given; that they ought to is a fact. Look at how far Apple was able to get
when they did so.
In any case, I quite agree that C= should BEFORE WRITING A STITCH OF CODE,
have gotten impressions from users as to what would be needed/desirsable in
a new release. Personally, I would agree to be just such a Gamma tester,
being an experienced computer professional, but not a registered developer.
I would be only too glad to take part in focus groups concerning OS
contents. (Hear that, Dave Minnich?)
–Gabe
Gabe,
Commodore DOES have places where they ask for suggestions from users – and
not from just developers. It just happens that they have not asked on CIS.
Several months ago, they were discussing all kinds of wish lists for 1.4
with people on Usenet. That network includes some developers, and quite a
few highly technical types – but it is also available to students of all
varieties, and to lots of non-students as well. Quite a mix of people ,
really. There were long discussions there about what ought to be in 1.4,
with all kinds of people getting involved. Commodore folks were obviously
listening. I suspect that they have limited amounts of time for reading
networks, so they have chosen one or two. Maybe our turn will come for
1.5?? We can hope and wish.
……Betty
Well, Betty, for those of us who consider CI$ to be the be-all and end-all
(for whatever reasons, so no discussions on that topic, please) the
inclusion of Compuserve would be a breath of fresh air.
Until then, <*sigh*> I guess we wait.
–Gabe
Developers will notice those things too. I think you're accusing them of
being too myopic. They have an interest in the end-user being happy, as
much as CBM. And some of them will have conflicting interests, which should
either cancel out, or make doubly sure that something is wanted and works.
–Eet–
Some developers will notice those things, and they were just examples –
probably poor ones since it would be difficult to FAIL to notice them. The
point is that many developers have a direct monitary interest in software
and hardware that will use the system. If a conflict between the system
under test and their own product should occur, it is only natural that they
would choose to view such a conflict as a problem area – whether or not it
is for the good of the whole user community. The product puts the dollars
in their pockets, not necessarily the entire community.
As for developers having an interest in the end user being happy, well…
just look back a few messages. A representative of Manx, and a developer,
brought on all of my comments by his snide reference to average users being
"unable to find their way out of CLI." With that opinion of the end user,
how can ANYONE expect him to have "an interest in the end-user being
happy.?" I know that if I should ever decide I want to program on the
Amiga, it sure won't be by using Manx.
Developers get that title by BUYING it. It does not mean anything more
than they paid a fee to Commodore for the privilege of calling themselves
Certified Developers. (There may be a minor screening procedure – the
check didn't bounce.) There are others in the user community who may be
many times more capable of being developers, but who, for whatever reason,
chose not to pay Commodore for the privilege of being "certified." Buying
the title does not necessarily "qualify" an individual to be a good tester,
developer, or janitor.
Bill, K3OZM
Just want you to know that not all developers blame problems between the
new OS and their programs on Commodore. We have a number of programs
which, though not completely broken, do no work properly with the 1.4 as it
now stands. We plan on addressing these problems. If we told Commodore to
change the OS to avoid the problems, they would have to do so at a
sacrifice of new features. We're not willing to make that sacrifice, even
though the problems will end up showing in our product. (We are already
addressing these problems for our next patch).
— Mike Roth / Lattice
P.S.: I think I stated things vaguely enough so as to not violate
any nondisclosure agreements. 🙂
I congratulate you and your company on such an attitude. Unfortunately, it
is almost directly opposite of what is perceived as normal human behavior.
Most humans, seeing that the OS conflicts with their product, would see the
new OS as a threat to their livlihood (the product) and react in a manner
so as to defend that livlihood. There is nothing wrong with this attitude,
as it is perfectly normal, but it is the reason that conflict of interest
laws and policies are in effect throughout industry and government. By
using humans who have a vested interest in seeing that the new OS is
compatable with THEIR product (read livlihood), the probability of having
the human reaction to conflicts turn against changes, good or bad,
increases dramatically. That is my objection to exclusively using
developers as testers.
Bill, K3OZM
Most developers are using the 1.4 stuff to make sure that they have support
in place when it is released. CBM is trying to make sure that they can do
that effectively.
Developers or any stripe would have a hard go of it if they just sit around
and whine about 1.4 being "broke" when other developers have made the
compatability a fact – it becomes obvious who is wearing the shoe, neh?
Ben
Amateur Radio Callsign is A A 7 A S
First, you say developers generically, then say they paid for the
privelidge of calling themselves 'Certified Developers', whereas I referred
_specifically_ to 'Commercial Developers', where the screening _is_ much
higher. I'm sure there are many construction workers that would make
outstanding physicists, but, they didn't 'pay' (the $$ to go to a college
to learn the trade) to become physicists, and aren't qualified to act as
such. You seem to assume that commodore cannot give you something in return
for registering with them, which is not the case. I know for a fact that
that specific Manx representative went out of his way to make an end-user
happy. And while that comment might have been phrased snidely (is that a
word? :), there is a grain of truth in it. One of the complaints some
end-users have is the fact that they _have_ to find their way into and out
of the CLI. Smart people, too. Would these folks make good beta testers?
Probably not, because they don't know the system well enough (yet) to know
_why_ what went wrong, went wrong. If you resent Mike Spille's 'snide
remark' about end-users, consider how I feel following this thread. You've
made assumptions as to the motives of Amiga developers in a generic sweep,
while I'm attempting to become a 'good' developer (following the rules,
getting registered, etc, etc..). And I personally _do_ have an interest in
the end-user and what s/he thinks, because s/he won't buy what s/he doesn't
like to use. There's no point in creating a product if the end-user doesn't
like it, so when a developer is developing a product, they're going to be
more concerned with what the end user thinks than what commodore thinks. If
commodore does something to the OS that violates their software (ruins the
gui, whatever), they'll scream and holler. If enough scream and holler, CBM
_better_ listen. Because of all the developers I know (I know 1 or 2 :), at
one time or another (almost) all of them were end-users.
–Eet–
I am having trouble following the logic of your message.
Suppose, for example, that Lotus (of 1, 2,3 fame) decided to become active
in developing (or even porting existing product to) for the Amiga.
Certainly, Lotus can afford whatever fee is required to become a
"Commercial Developer," and Commodore would certainly welcome them with
open arms, even to the point of waiving some of the screening which you
seem to be caught up in. Now, Lotus would have bought it's status as a
Commercial Developer both with cash on the barrelhead, and use of it's
corporate reputation. The minor detail of Lotus' reputation and experience
being on a completely different platform would have nothing to do with
anything. As a developer, Lotus would be "qualified" to test the operating
system of the Amiga. My question is why should Lotus be considered more
capable of testing a new operating system than John Smith, who has written
dozens of PD and shareware programs for the Amiga but hasn't received
enough in return to afford the fees to become a "Commercial Developer?"
What has paying to go to college to do with being qualified to act as a
physicist? Many technological breakthroughs throughout history have been
made by folks who have not been to college. In fact, there are many people
throughout industry functioning in, and considered qualified by their
employers, professional areas who have not paid to go to college. They may
have amassed knowledge through years of hands-on experience, self teaching,
or even be childhood prodigies, but they are recognized as qualified to
perform in their fields both by their employers and others in those fields.
<continued>
<continuation>
On the other hand, there are those in the same professions holding degrees
from institutions of higher learning who are unqualified to hold those
positions. College is nice, and often a shortcut to a professional
position, but it does not NECESSARILY qualify one to fill such a position.
What did I ever say to make you think that I "assume that Commodore cannot
give you something in return for registering with them, which is not the
case?" All along, I've been saying that those who register (by paying) are
'blessed' with the title "Certified Developer." I also know that they are
eligible for pre-releases of software and hardware, and going to DevCons,
and maybe getting drunk in hospitality suites at other conventions (or
whatever.) What I did say is that paying the bucks up front does not
NECESSARILY qualify anyone to be a good tester.
I'm not at all sure what assumptions I've made as to the motives of Amiga
developers in a generic sweep. I think you may be referring to my concerns
about a possible conflict of interest. If in testing the new operating
system, a "commerical developer" finds something that will cause his
hardware or software product to malfunction, it is normal for him to tend
to favor his own product. It is that product that puts food on his table,
after all, and having it broken by a new operating system can be perceived
as a threat to his wellbeing. For that reason, the potential of a conflict
of interest is very real, and even likely, when the entire pool of testers
is taken from folks in this category. Some developers will run counter to
this trend, others will overcome the perceived threat, but others will
<continued>
<continuation>
succumb to it. Diluting the pool of testers with "common users" could well
balance the results of the testing in favor of the population in general,
thus mitigating the natural tendency of those with a vested interest in
maintaining their own products unchanged from biasing the test.
There are certainly some users who cannot find their way in or out of the
CLI. There are others who are uncomfortable in that environment. By the
same token, there are some developers who are uncomfortable in the
Workbench environment. Yet, virtually every product for the Amiga can be
used in either environment… yes – there are exceptions, but they are
relatively few in number. It is unlikely that any of the foregoing would
be good beta testers, developer status or not. At the same time, there are
quite a few people who are comfortable in either environment, with or
without developer status. Some of these later people would be good
testers, others wouldn't. Having an official title does not NECESSARILY
cause anyone to be a good tester, nor does lack of one NECESSARILY preclude
anyone from being a good tester.
The wider the testing populace, the more likely the test is to be unbiased.
Of course, there has to be some limit – nobody could keep track of an
extremely large number of testers. Neither would selecting 1000 names at
random from owner registration cards and shipping them software to play
with be particularly effective. (Though it might or might not work once or
twice.) As I mentioned earlier, developers may have a vested interest in
seeing a new operating system, if for no other reason than to adjust their
<continued>
<continuation>
own products to become compatable with it. Users also have an interest in
a new operating system, since they are going to have to use the silly thing
once it becomes public. The testing pool should be taken from both groups.
Bill, K3OZM
Bill – also, there are those (talented individuals) who get to go to
college for 0 $… I think CBM should have a catagory for them, too. For
instance, lets say (I do NOT know this, btw) that Leo Schwab wants to join
in the "fun". He's already proved his ability to handle the system in
marvelous ways, which is _exactly_ what CBM needs. Perhaps (again, I have
NO idea) he's not in a financial position to become one. He could make
application to CBM, they could say, "Hey! This guy is Good!" and take him
under the ol feline wing.
Wouldn't hurt a bit…
CBM is pretty silly for putting up a cash barrier to developers; of all
people, they oughta know that the Amiga hasn't exactly been a gold mine –
and a lot of that is _directly_ CBMs fault. it certainly is the best
machine out there.
Ben
Amateur Radio Callsign is A A 7 A S
Bem, I don't argue with that idea. But, Leo has proven his competence, you
said that yourself. –Eet–
Ben,
They do not always put up a cash barrier. They do recognize certain
qualities in people and invite them to participate in one or another aspect
of development. Further than that I will not say. Suffice it to say that
CBM does not tell the world at large everything they do.
-larry
What I meant about the construction workers/physicists was that the
construction workers were working construction, NOT working as physicists.
Sure, there are those working in any field that are very competent but
don't have the 'degree'. There are also those with the degree that aren't
competent. Both cases are in the minority. And, if we move it back to the
subject of developers, yes, there are competent developers that aren't
registered with commodore, and there are no err, doubt incompetent
developers who are registered… but in _both_ instances they're
developers. While a company doesn't need registration to be competent at
development, by the same token someone with _no_ experience at development
can't be considered competent.While the end-users input might be worthwhile
_before_ any work is done on the next revision of the OS, they are _not_
necessarily going to be worth the investment in letting them _test_ the
next revision for bugs and flaws. They're input would be exactly what is
needed for 'ergonomics' of the next revision, and the time for that is in
the design phase, not the implementation/testing phase. When I get a guru
81000009, I know immediately what _kind_ of thing is wrong, not "some weird
numbers", and what to check to see if it's the program I'm working with or
actually something in the OS. (an example, I'd been getting that particular
guru on exit from my program for some time, worrying because it appeared to
happen _after_ my program exited… I knew _what_ to look for and _where_
to look for it, and it did turn out to be my code (still using 1.3, btw)).
A program I'm beta testing right now (a term program) appears to lock-up at
bizarre points. Through experimentation, I've discovered that it is _not_
in fact locking up, but the input/output is getting 'lost'. A 'normal'
end-user might possibly have caught this distinction (more likely, not),
but, could s/he take it any further? Would s/he know what things to try
next time to help narrow down the cause of the bug? again, probably not.
<continuation> As for Lotus, look at the _success_ they've had on that
other platform. Nothing succeeds like success. They could make _major_
blunders in moving to the new platform, but because of the resources they
must have at their disposal, the difficulty fixing them would be minimal
(certainly less than someone with experience on NO platform).
I notice in another message of yours that 'generic sweep' you make of
developers. The majority of developers realise that they are in business to
move a product out the door. It must be what the end-user wants, it must be
what the end user will buy. Opinions on that may vary, but, that's still
the bottom line. If there were only a few developers developing for the
Amiga, I might agree that they might become isolated from the end-users,
and therefore do more harm than good as beta testers. The same is true of
an end-user whose only use for the machine is counting peanut shells after
the circus leaves town.But there are a lot of developers out there. And, as
I said before, most (h*ll, all) of them are end-users as well. I don't know
any developers that don't use software themselves, such as Deluxe Paint,
Digiview, Power Windows, SuperBase, Maxiplan, and so on. To quote Bob
Heinlein (one of my favorite pasttimes :): "The greatest productive human
force is greed". Those that put ego above greed don't stay in the business
long. –Eet–
Bill: Commercial Developers have to have successful products out in the
Amiga marketplace, as well as pay $450/year for the privledge of being a
commercial developer. Having a successful product in the marketplace is as
good of a criteria as I can think of to qualify as a good beta tester.
Clint <VRLI>
OK. Redmond Cable has several successful cables out for the Amiga,
particularly since Flicker Fixer came along. If they wanted to pay the
$450, they would technically be qualified to be a Commercial Developer.
Now, I don't see any particular qualification on the part of Redmond to
test an OS. On the other hand, the author of the majority of the Operating
System would, on the face of it, seem to be the best of all possible
testers. However, the author is probably the WORST of all potential
candidates to be a tester. He's just too close to the project, knows all
the ways to work around problems (and probably does unconsciously), and
would likely write off more than a few problems as hardware glitches. A
successful software author, having found a way to do something in software,
may have his technique "locked" in his own mind, and possibly would not be
receptive to another technique, particularly when the other technique may
cause his own program to function differently than it was designed. I
really don't know of any particular criterion which could be applied across
the board to ensure that any candidate would be a good tester.
Bill, K3OZM
Bill, actually, the author makes a bad tester because he's unconciously
learned what not to do ("doc it hurts when I do this." "don't do that.").
However, Redmond might well be a good and qualified beta tester, after all,
the OS does include device handlers/drivers.
–Eet–
Bill,
You have a warped view of what happens when a bug is reported. I can
assure you that a developer who is insistent on putting his own program and
his own method above a bug fix, while other developers have successfully
done the same operation in acceptable ways that do not break, will soon
find himself inundated with folks telling him not only to bug off, but
telling him how to bug off and still get the job done.
You know, a while ago, someone from CBM made a statement about a certain
field in the Gadget structures, and claimed that it would completely change
meaning in the next release. I objected, strongly, not because I had used
that field in any way, but because I knew that some developers had used it
in ways that were going to break a lot of code, and that their usage was,
if not entirely kosher, at least reasonable, given the way it was
documented.
I got flamed royally for that criticism, from a number of developers,
both in public and in private, and virtually noone would admit that they
had used that field in public. The end result? A decision was made to go
ahead and use that field for its new purpose, but only in conjunction with
the assignment of a meaning to a previously reserved bit in another field,
that had been very specifically documented as reserved.
What does it all mean? I had made my comments, not as a developer, and
not as an authority on gadgets, but as a person who had noticed a possible
problem that did not have to make it into the next release. Everyone,
including those who had flamed me and those who suported me, were happy
with the result. You do not have to be a certified developer in order to
comment on things. Get involved if you feel strongly about it. Pay the
money if you want to test the new stuff, or show by your actions that you
are very good at testing and perhaps they'll waive the fees.
-larry
Bill:
True, and I expect that C= is aware of this. Having a product out on the
Amiga is the minimum requirement for Commercial status, it is not a
guarantee that C= will accept your request. If a company's product does
not require detailed technical info from C= to be maintained, and the
company is not working on a product that does, I expect such a request
might be rejected. But in that case, I expect that the company in question
would not fork over $450-500 just for the fun of it.
Clint <VRLI> P.S. Most hardware
vendors DO need to know if their hardware will play with the latest version
of the OS.
450 a year – but 500 to start. :^)
Ben
Amateur Radio Callsign is A A 7 A S
Bill,
That was NOT a snide comment. It did accurately reflect a whole class of
users; those being users that only operate through the Intuition interface.
It is not a putdown. It is not indicative of an attitude that the user who
can't find his way out of a CLI is to be scorned. It is, in fact, an
acknowledgement that there are all manner of folks out there for whom the
CLI is a place unvisited. It is because there are users that operate in
widely differing ways, that the developers are providing ways to operate
that do not require trips to the CLI. If you want to get your back up about
something, please do it over real issues, and not those you read into the
postings of others.
-larry
Yeah, those users are a real pain. And you can give lessons in ignoring
*them*.
You seem to have the term 'beta release' confused with 'release'. A beta
is a release with lots of bugs, that's still in testing phase. It is not a
release where alot of things are going to change, or where suggestions are
going to be too helpful. The company that puts it out just wants to know
what bugs are there, and how to fix 'em. Would you feel comfortable giving
a beta release to someone doing something sensitive who didn't quite
understand that is was really a beta and not supposed be used w/ senstive
things?
BTW, would you care to elaborate on 'and you can give lessons in ignoring
*them*'??
-Mike
You don't answer Manx V5.0 questions, as far as I can tell from reading the
messages.
Nobody _asks_ Manx V5.0 questions, as far as _I_ can tell from reading the
messages…. but of course, anyone that knows me can tell you that that is
hardly surprising. –Eet–
Well, I asked, Dan Hankins (?) asked, and a couple others as well since
Jan. have asked Manx V5.0 questions. Dan got the closest thing to an answer
but it wasn't satisfactory. Then Mike let drop the existence of a RSN bug
fixer for nominall $$. Perhaps there _was_ no answer for the questions at
the time. Just wanted to rag him for accusing CBM of the approach he is
using.
Best wishes to your brother. Any good news? Rob.
Now, about Manx… my comment meant that I couldn't know if anyone had even
asked any questions, since I spend 95% of my time in co, rather than in the
message base. <grin> –Eet–
"Here, I know you just learned how to drive that Chevy automatic, but I
want you to test this here new V12 5-speed. Just take it out on the track
and do a few quick 200 MPH laps, and let me know where the suspension needs
work."
Sheesh!
I don't care who beta-tests. I just think the Manx rep has a lot to say on
every subject except Aztec C.
Touche!!
……Betty
🙂
-larry
Probably they chose the developers because they are in contact with them;
the developers have something to lose if they break their non-disclosure
agreements (loss of developer status, with no right to be reinstated, could
happen), and people from a wide range of interests are included in the
certified developer group. They are not all heavy program developers at
all. The commercial developers are those who have products on the market.
Certified developers can be almost anyone who can show some kind of need
and is willing to pay the fee. This includes lots of users – such as me.
I need the advance information and the opportunity to try things in order
to give help and information here, and in order to write articles for
magazines. That was good enough. Some of the developers are teachers who
need the information that this status gives them. There are LOTS of
reasons why someone who was not primarily a programmer would want to be a
certified developer. Seems to me that Commodore made a good choice there.
This discussion started off with my complaint that Amiga users–that is,
non-developers–seemed to be shut out of the 1.4 beta-testing process. I
had in mind people who use Amigas in serious ways in professional
environments–teachers, scientists, artists, businessfolk, etc.–altho I
suppose "Joe & Jane User" (used in subsequent messages) connotes people
less advanced than that. My mistake. It turns out, tho, that certified
developers (as opposed to _commercial_ developers) can also beta test, so
my concern is partly alleviated. I also mentioned retargetable
graphics…this is something I hope we get as soon as possible.
I guess I agree with everything Don has said, but I do think CATS'
user/developer feedback apparatus needs some restructuring. First of all,
the title "Certified Developer" is something of a misnomer. Certified
developers aren't necessarily developers at all. Another category is
needed, something like "User Consultant," to encompass experienced, expert
users who are not developers, who feel put off by the word "developer," but
who still want to contribute to the maturation of the machine (via
alpha/beta-testing OS upgrades, offering suggestions with the knowledge
that those suggestions will be evaluated, or whatever). Then, if CATS were
to add such a category, it would be crucial for them to *advertise* it. An
ad in the major Amiga rags–say 3 or 4 months after 1.4 is released–to the
effect that: "We want experienced Amiga users to be involved in the Amiga's
future! Do you want to directly contribute to the Amiga's future
direction? Then join our new User Consultant program. Call 1-800-XXX-XXXX
for more information on how you can let your voice be heard." would do the
job. (Choose your own prefered wording, of course <grin>.)
Callers, after screening, might be offered the same benefits (at the same
rates) as current certified developers, minus the current terminology, or
they might be offered something different if CATS sees fit to do so. The
input of such folks would likely be of most use to CATS before and during
the _alpha_ stage of OS upgrade development; even so, they shouldn't be
excluded from beta testing. At any rate CATS would benefit not only from
the additional input but also from the good PR, and we users would benefit
from a machine more tailored to our wants and needs. IMHO, anyway. 8^)
Actually, the biggest thing about Certified Developers is that _THEY_ have
taken the first step. THEY told Commodore that they were intersted in
doing thing for the Amiga.
Victor A. Wagner, Jr.
Just an observation here. Almost everyone who's asked for C= to have
non-developers as Beta testers have added a remark along the lines of "I AM
NOT VOLUNTEERING". Wonder why that is?
-Mike
I'd volunteer in a heartbeat. I think it is important for a company like
CBM to have actual users with no commercial interest in the product as
testers. There would be no conflicts of interest and the company would
benefit from a representative of the total user population they are
targetting.
Your comment in the previous message about using only those who take the
time and trouble to become certified developers as testers is rather narrow
minded. For example, I have absolutely no interest in programming the
Amiga, nor have I any interest in developing hardware for it. What benefit
could I derive from becoming a certified developer, knowing full well that
I have no intention of developing anything, other than to enrich Chairman
Gould? I do, however, have an interest in USING the Amiga in a number of
areas. —- Unless, of course, it is really true that the Amiga is only a
hacker's machine, and CBM has no interest in serving anyone not in that
niche.
Bill, K3OZM
True, Certified/Commercial developer doesn't fit for most people. However,
CBM _does_ need some type of criteria for picking beta testers. I'm afraid
just any old person walking up and saying "Hi, I'd like to beta test your
software" just doesn't do it. CBM needs a way to insure that 1) the person
isn't gonna leak sensitive information when CBM doesn't want it known, and
2) CBM needs to know the person is actually going to test the stuff
reasonably well. Perhaps the suggestion some one made of "User Consultant"
would be a step towards assuring this type of thing. How one would become
a "User Consultant" would have to be restricted in some way however, or
else we'd be back to square one.
-Mike
I don't see "just any old user" sending a letter to CBM and volunteering to
beta test. I'm sure there would be a few in that category, but anyone
taking the time to write the letter would likely be reasonably competent.
Those volunteers could be allowed an interim upgrade or two, and, if they
worked out as testers, could be permitted full tester status subsequently.
In fact, that probably wouldn't be a bad idea for anyone being considered
for testing.
As far as I know, NDAs are pretty standard for use with testers anywhere.
The mere fact that a user/tester didn't pay the fee for the privilege of
signing such an agreement isn't likely to make him any more "leaky" than
someone who did pay the fee.
I like the idea of "user consultant." If only CBM would listen to
suggestions and reports from them, we'd probably all have a better machine.
Bill, K3OZM
Bill,
If you think that anyone who writes a letter would be reasonably
competent…read some of the letters to the editor of your favorite
magazine and/or newspaper, etc.
Then imagine those that don't get printed!
Don
The user doesn't have anything to loose if he's leaky, where as a developer
has lots to loose, because it's unlikely that CBM would actually prosecute
someone who bent the non-disclosure agreement, unless it was unbelievably
blatant.
-Mike
Well…… I haven't heard on any prosecutions, but 1.3 Beta (various
versions) and Gamma were all over the place months before 1.3 was
officially released. Portions of 1.3 were even in commercial packages long
before 1.3 was released. Since these versions were'nt available to the
"common user," they had to come from either a CBM employee or a certified
developer, or more than one. Seems to me that certified developer status
doesn't preclude leaks either.
Bill, K3OZM
Bill,
Things have changed greatly since 1.3 came out. I agree with you – it was
all over the place. That is not true with 1.4, though. At least, not yet.
I assure you that I would not risk my right to be have developer status by
spreading it around, and I'm not even developing commercial products.
……Betty
You won't even talk to ME about it! <grin> –Eet–
Not until you have your own!
……Betty
<snif> I'm waitin' as fast as I can, Betty, ma'am! <snif, snif> –Eet–
And security seems much tighter now with 1.4, IMAO. –Eet–
How would you be differentiated from the other end-users? How would CBM be
able to keep you from yakking? How could CBM be assured of bug-reports from
you (on a fairly regular basis)? All of these problems can be covered by
developer certification. –Eet–
For one of two reasons: a)they don't want to toot their own horn, b)they
don't think they themselves have the knowledge of AmyDOS to beta test.
However, there are users out there who would make great beta testers, and
should be invited in.
John
John,
And who "separates the wheat from the chaff?" Does CBM (or any
other company) ask for testers and then draw straws to see who "wins", do
they just make beta products available to anyone…and those that give good
test reports are permitted to "advance" to bigger and better beta products?
Or…do you make sure someone is serious, by asking for an initial
fee of $75 for the first year, and $50 after that. And then provide those
people with the technical information necessary to allow them to properly
test things? Along with some other benefits?
Don
Somehow, I don't see that contributing $75, or $50, or however much you
want to say, to Commodore makes a person either serious or competent enough
to test anything. As nearly as I can figure, it means that the person had
the money to spare at the time. I've got no idea how many certified
developers there are, but I'm sure that quite a number of them pay the fees
for things like pre-release software/hardware, seeing their name listed
with an impressive title, etc.
Bill, K3OZM
Bill,
That's true…just paying the fee doesn't make a person competent
either. However, I'd bet if you grabbed 50 names at random from the
developer's list, 50 from our membership base, and 50 from Amiga owners in
general…you'd find the average level of competence to also be in about
the same order (developers, folks here, owner in general).
Some system must be used. There's over a million Amiga owners out
there. Presuming only 1% decided to respond to a plea for beta testers,
that'd be 10,000 users to "sift" thru trying to find those that would be
competent enought to really give the software a workout.
Remember…this is the Operating System…this isn't just some
application program. Unless you're writing programs, you can't very well
test many of the functions. All "Joe Average User" could do is report that
Program X crashes under 1.4 when selecting menu item Y…whereas it worked
under 1.3. Joe Average Developer could not only state that…be they could
also send in a sample piece of code. Joe Average Developer would also kow
if they pulled a trick that used to work on 1.3, but now doesn't on 1.4 and
thus…the "fault" is theirs…not the new OS. Thus, they'd know they had
to change their code and not make an erroneous bug report.
Also…there is nothing stopping Joe Average User from sending in
bug reports when they come across them. It's not at all uncommon for me to
see reports of problems…or suggestions for changes on the various
networks I "visit".
Don
Give a written exam to users . . . They pass, they beta test. Simple as
that.
John
John,
And who is going to administer these written tests? Who is going
to pay the costs, grade them, decide of the 10,000 entrants how many of
those that pass really will get beta copies, etc.
Don
That doesn't necessarily make them _responsible_ enough, tho. –Eet–
Don… not to be picky or anything, but CBM america does not have "over a
million Amiga owners" to pick from. The Amiga market in the US (even
including canada) isn't even half of that.
Ben
Amateur Radio Callsign is A A 7 A S
Ben,
That's true…but CATS covers the whole Amiga community, not just
the North American community don't they?
But you're right…I did use "CBM" when I should have just said
"Commodore-Amiga"
Don
Hmm…. how a bout $450 for commercial status (and stricter requirements to
even qualify)? Someone willing to pay almost the price of an A500 for
commercial status, plus having a track record, plus having a product to
develop all indicate to me a more responsible attitude than sending in a
letter ("gee, I wanna beta test") would. –Eet–
Mike,
The answer to that, in my case, is simple. Too lazy. And I'm suspicious
of Dave Minnich, who seems to always be there, waiting for an oppourtunity
to get people INVOLVED. I guess I'd have to participate if CBM started
somekind of program where they wanted people like me to help out… Maybe.
Meanwhile, it's fun to b*tch.
Norvel
Dave does look like he wants to get people involved – and like he cares
about the Amiga.
Hey! I know! He works for _Atari_ !!!
:^)
Ben
Amateur Radio Callsign is A A 7 A S
Ben,
How dare you mention that obscenity in connection with me!!! 😀 (just
kidding, folks…Atari has some fine machines (if you can find them)) ;>
Actually, I work in the automotive department of a KMart and have nothing
whatsoever to do with computers. 😀
Dave
Commodore Online Relations
SO the truth is out. The new rumoured Amiga 1423.5 unit is a car portable
that doubles as a trip computer. I knew I could weasle it out of ya!
Ben
Amateur Radio Callsign is A A 7 A S
Ben,
You have that backwards….
It's a trip computer that doubles as a home portable door bell. ;D
Dave
Commodore Online Relations
I never was any good at mind-reading. :*)
Ben
Amateur Radio Callsign is A A 7 A S
Ben,
Mind??? I'm beginning to wonder if I still HAVE one! 😀
Dave
Commodore Online Relations