| From: | Jean-Pierre |
| Date: | 5 Aug 2000 at 22:38:01 |
| Subject: | AMIOPEN: To Windoze or Not to Windoze (was: reply to Gary) |
Aaron seems to be hitting it spot on -
I must admit maybe initially I thought what's all the fuss about,
BUT: once you begin reminiscing about :
-all the positive coverage Ami got because of running on Linux (and no doubt
got a lot
of Linux programmers interested in doing something with/for Ami)
-The fact that a basic Linux installation is tres simple to do in parallel
with a Windoze
installation for those bored Windoze developers that may want to try it
(a linux magazine plus an hours work gives you an installation capable of
running the SDK)
-All the bad attention Gateway got by swapping their 'chosen' OS out after a
few months in
(ok, this is a totally diff things as its just the underlying OS for the
Dev Sys, but still...)
-The fact that the Amiga Active editor praised Amiga for choosing Linux as
the underlying OS
for the devbox, more or less saying 'god knows what would have happened if
they chose Windoze'
-The fact that like Aaron pointed out, anyone really committed to doing
truly great innovative
things would have no problem playing with an alternative OS....and that
having the SDK running
on an unreliable OS (as everyone here can testify, Linux is rock-stable in
comparison) is
going to negatively influence ppl's perception of Ami's NG products....
-The support for developing under Linux is only know getting properly
started - time spent
on the SDK for Windoze will destract from valuable time that could be spent
on getting the
real hardcore/early adopting developers that will make the real difference
going strong...
Please bare in mind that this is comment comming from someone that runs
windows
alongside Linux - I'm no huge Linux fan by a long way
Perhaps the compromise would be to move the Windoze SDK back and just give
dedicated to the few key partners that have indicated that they will only
develop for
the new AMI anytime soon if a Windows SDK is available -are there any of
those?
Perhaps a high level of support for those few essential to Amiga plans would
help.
Besides which, (embedded) Linux is rapidly becoming a standard for ppl that
target
Amiga's market - like the consortium of Japanese companies that are gunning
to develop
Embedded Linux as the standard for IA's, PDA's etc.
Perhaps it would be prudent to kick up the priorty for Native SDK and
Superbly supported
Linux development before catering to the laggards that are Windoze
developers?
No offence to the open-source ppl here that have stated here they would like
a Windoze
based SDK, but then there are those that want Linux-PPC and that's obviously
not
seen as a priority here either. If a company like Hyperion can be
professional enough to
run Linux, if ID's John Carmak loves Linux, I'd like to know exactly who the
key partners
are that can't live without a Windoze SDK....
Don't know if these ramblings make sense, no need to reply either, but I did
feel there
was a need to underline Aaron's comments.
Regards,
JP
>
> On Tue, 1 Aug 2000, Gary Peake wrote:
>
> > Hope no one minds my follow-up here. Aaron raises some good points
> > that need to be discussed even if this is a developer list. I know you
> > guys are busy, so please bear with me while I answer.
>
> Sincere thanks for responding, Gary, (to both this and my last
> questions on a similar lines) but I think we can drop the implication
> that any of this is "off-topic" or out of place. Concerns relevant to
> Developers rather inseperably include matters of policy and practice
> announced by Amiga Inc. and it's developer support organizations.
>
> > >> A version of the SDK for Windows will be made available next month.
> > >
> > >I find this a most untimely and disturbing announcement.
> > >... Any concession to the monopoly
> > >corruptions at this stage of Amiga development seems hugely premature
> > >and remains, despite all fronts and assurances, a drain on resources
> > >and a serious contradiction to the alternative community commitments
> > >that Amiga Inc. has so successfully stated, solicited and exploited to
> > >date.
> >
> > There is another aspect to this that needs to be looked at. In our
> > better years, some of the most prolific and ingenious developers were
> > programming for the Amiga. Many no longer keep up with what is
> > happening and there is no way to find them all.
> >
> > We decided back a few months ago to engage as many developers as
> > possible in the future of the Amiga. To do so, it is necessary to go
> > where they are. Once we have their interest, then we feel they will
> > stay just as you have.
> >
> > We are not looking to develop FOR Windows, but to be able to develop
> > ON Windows. There is a marked difference between the two.
>
> The *intentions* of a windoze SDK release are not being questioned
> here. The probability of those intentions ever being realized and the
> likelyhood of serious damaging consequences are the issues of concern.
>
> First major problem on the list is that the announcement itself is a
> contradiction to previously stated and implied directions of Amiga
> Inc., starting with the direct and repeated McEwen quote:
>
> "Our business plan is scaled to the existing Amiga community."
>
> Prior to the release of the Linux SDK, I can find no mention anywhere
> of windoze support in any of the development and business plans.
> All observable actions of AI through the Linux SDK release involved
> soliciting and securing alternative partners and developers
> *exclusively*. The real world effect of any windoze support now is
> that it paints AI as being highly deceptive and manipulative with the
> very communities that are the most central and essential to its
> success.
>
> The downsides of the proposed windoze SDK release are hardly limited
> to community relations, either. The real world effect is that all
> efforts directed to windoze support will directly and insidiously
> usurp the resources of alternative efforts on several fronts.
>
> (-: ...and your previous reply that doing a windoze release won't
> require any of AI's limited resources is a nice thought, but I'm
> afraid the Keebler Elves are already under contract at the cookie
> factory. :-)
>
> Just from the posts to this thread, the "Black Hole" syndrome that
> history and I suggest is clearly exemplified. Your own short "target
> markets" list shows that legacy Amiga support is totally absent from
> the development plans. Support that has been promised (yet again) to
> the dedicated and patient, the ones who stayed focused, the ones who
> are already on board for making the Next Generation a possibility,
> that support is being handed instead to those who abandoned
> alternative thinking long ago, the ones whose efforts supporting new
> choices like Amiga NG are the *least* likely to be enlisted.
>
> Even among the persistent developers here, the ones who took AI at
> their word and example by investing in Amiga alternatives with the
> Linux SDK, the cries for side-grades prove the guaranteed exodus
> away from the goals once windoze enters the mix.
>
> Perhpas the most overlooked aspect, the well known and well publicized
> element that everyone clammoring for a windows SDK is perpetually
> ignoring in their arguments, is that the Tao/Amiga SDK runs MISERABLY
> when it has to wade through windoze. The generalized comparisons
> suggest that under Linux, the Amiga OE system runs twice as fast and
> is twice as capable as it ever will be on a windoze box.
>
> So let me get this straight: We *want* to introduce the new Amiga to
> the windoze world in the worst possible light?? We *want* to encourage
> development for it in the most inhibiting and crippled environment??
> We believe that SDK install complications and operational bugs are
> going to be magically cured by running it on top of a bloated and
> unreliable OS??
>
> (-: We think advertising sushi as cold dead fish is a good idea?? We
> think Commodore was really adept at marketing??? :-)
>
>
> > >The first obvious and unavoidable result of this action will be the
> > >destruction of the market for the alternative Developer Systems.
> >
> > Our intended target market for the original SDK, as we did state, was
> > Linux developers who believe in Open Source. We have a commitment to
> > open source and want to open as much as we can to the open source
> > developers.
>
> We seem to have some divergent definitions for the concept of
> commitment then...
>
> > The Dev Box is targeted at Amiga developers who can not or do not want
> > the problem of installing Linux on a computer. Nothing we have done
> > has hurt either targeted market.
>
> I do agree that the Linux package handling has been consistent, at
> least up to the introduction of the windoze problems.
>
> (-: Though you might be doing better with the current SDK if you
> started refering to Linux installation as an "opportunity" instead of
> a "problem". ;-) I saw it as an opportunity, and am therefore enjoying
> exploring and learning a huge amount of relevant information that will
> not only make me a much better AmigaNG developer, but is significantly
> adding to my store of marketable I.T. skills.
>
> > The third leg is to involve some ex_Amiga (now Windows) developers in
> > the process by offering an SDK that works on top of Windows...
>
> ...with the worst possible performance. That will grab 'em! ;-)
>
> > The Fourth stage is a free standing SDK.
>
> ...which will show the New Amiga product at its best. Until then, of
> couse, it's the Linux environment that does this.
>
> > Four targets, four separate markets, four separate products.
>
> All these plans would be sensible and the various contradictions
> defensible if, indeed, the markets were so clearly delineated.
>
> The reality is that there are a limited number of trees in the forest,
> and the ones that have succumbed to the fire aren't likely to be any
> use for builiding. Until you can put out the flames and make a
> clearing for new growth, there is absolutely no point in watering
> charcoal, let alone burning any of the trees behind the fire
> break.
>
> > As Bill stated at AmiWest, we are NOT going to be charging for SDK
> > upgrades that come through this process. We are going to make them
> > available to everyone who has purchased any of the target SDK's.
>
> A very fair and constructive policy. I'm ready for my Linux based
> Amiga emulator any time. I would love to be integrating all my
> familiar Amiga development tools and showing the emulator off to my
> many friends in the legacy community that have been waiting for
> something/anything exciting for so very many years.
>
> > >Again in the Amiga marketing tradition, the first ones left slashed
> > >and bleeding from a premature windows releases will be that tenacious
> > >but waivering handful of remaining Amiga dealers;
> >
> > Why would they be slashed and burned?
>
> Because there is a tremendous amount of overlap in the markets you are
> you have outlined. The proposed windoze SDK, thanks to monopoly
> pervasiveness, will appear to most as the easiest solution and all but
> eliminate incentive to obtain one of the Linux DevBox Systems. The
> (*late*) addition of including Trailblazer developer registration in
> the deal may salvage some sales, but the majority will be lost in the
> black hole.
>
> > We are working closely with the
> > few remaining dealers and developmental companies. I talk to them as
> > does Randy and Bill on a weekly basis. They are going to have a
> > chance to finally expand their markets rather than watching as their
> > markets shrink to nothing.
>
> The continued vacuum in legacy support is up to AI to address now.
>
> If these dealers have invested in stocking Developer systems, the
> evidence indicates that they will be badly burned by a windoze based
> SDK release. To salvage what they can from the loss, they will
> probably feel forced to re-package the systems as windoze boxes and
> put still more money and users in Black Billy's pocket.
>
> > >Of course, once you point the sword the wrong direction, you have to
> > >start applying bandages, with the common result of adding more gashes
> > >in the process.
>
> > >Perhaps an attempt to patch the critical wound from a windows
> > >SDK release would exlain why such a nasty cut was delivered to all the
> > >committed developers who had already invested in the Linux SDK release
> > >and/or compatible systems to run it on.
> >
> > Why not offer a value add so that Dealers CAN improve their market
> > positions? You can't argue on one point that we are killing them and
> > on another that we are favoring them?
>
> I was arguing that the Trailblazer addition can be readily perceived
> as a "fix" attempt, anticipating the market erosion that would come
> with windoze support. The fix may help DevBox Dealers, but present
> early adopters may feel a little slighted that this $1000 value was
> not on the table when they made decisions for when/how to get on
> board.
>
> > >The only thin ray of light that has escaped the black hole so far is
> > >Gary Peake's follow up announcement that a Stand Alone SDK is close to
> > >release as well. Of course, any windows release within 6 months of
> > >this will overshadow that pretty completely. Why the stand alone SDK
> > >and OE releases aren't the ONLY focus of AI at present is beyond me.
> >
> > Because some developers want to be a part of the beginning, want the
> > fame or glory of having produced a part of what we are doing.
>
> Good. Bring them on. With the Linux DevBox and the inclusion of
> Trailblazer status in the price tag, AI has gone to great lengths to
> make being part of the beginning as simple, practical, constructive
> and affordable as possible.
>
> If these developers are at all consistent and sincere to the goal of
> building a New Amiga, then they will want to develop for it in its
> most efficient run time environment and introduce the New Amiga to
> others in its best possible form. Until the stand-alone OE, that "best
> possible" scenario can only be achieved via Linux, so these commited
> developers will be inviting the opportunity of exploring, learning and
> using that well supported, marketable and extremely relevant
> alternative.
>
> By my dictionary, THAT is commitment. If AI was living up to the
> definition, then the current Linux SDK offerings, Legacy Emulation and
> Stand-alone SDK / OE releases would be the exclusive priorities and
> directions that they would be pursuing. Anything else is a distraction
> and a detraction from their commitment.
>
>
> /* AArexx AAron */
> /* (comment required) */
> options RESULTS
> ADDRESS 'aaron@pd.org'
>
>
>
> Subscribe/Unsubscribe: open-request@amiga.com
> Amiga FAQ: http://www.amiga.com/faq.html
>
Subscribe/Unsubscribe: open-request@amiga.com
Amiga FAQ: http://www.amiga.com/faq.html