Re: ruby on rails
- From: Marco van de Voort <marcov@xxxxxxxx>
- Date: 11 May 2006 05:12:31 -0700
On 2006-05-10, Brian Moelk <bmoelk@xxxxxxxxxxxxxxxxxxxxxxxxxxxx> wrote:
"moreHow does "DevCo is the only provider of BDS languages" translate into
point.integration"? Regardless, I think we disagree on this fundamental
(see the previous mail, the part about how componentisation limits the
programmer, and increases legacy and makes it harder to stabilize)
More *integration* is more...not less.
This is more integration of external entities vs further integration of what
is already there.
I prefer the latter.
Simply because it is *harder* to do
doesn't mean that it is:
1) not worth doing (in terms of ROI)
2) limiting to the developer
Well. That's the point isn't it? You say it will give ROI because Microsoft
is doing the same. I say it isn't because of the same reason.
I mean _sustainable_. The investor will inject cash sure, but in timewant
to see money in return. So the customer base is left. But that moneyexists
for VS, NetBeans and Eclipse too. And they have additional revenues from
their other stuff.
Who knows if it will be *sustainable*.
Nobody. But that goes for this whole newsgroups. We work with the info we
have and intrapolate. Wildly usually :-)
How many years does it take to be called sustainable?
I wouldn't define it in years. I would say two product cycles after the
external sponsoring stops. If you survive that, and you verified you are not
in a downward spiral where every release sells less than the last one, you
are somewhat sustainable in business model.
Everything is temporary,
No. Long term profitability is quite necessary for survival :_)
the market has changed and DevCo must change in order to come up with a
"sustainable" strategy in the context of the market today.
No. It has an option to reposition for other markets.
Moreover, it's not even proven that Eclipse' model is sustainable
long term, if the main hype is over.
Please define how long "long term" is in specific measurements.
See above.
Typically public companies live quarter to quarter; as a private company
all that matters is that they are profitable enough to keep their
investors happy.
Should have sold to MS then before they became irrelevant.
Presumably, DevCo will be starting with profitable
products.
_but_ in a downwards cycle. It is about breaking that. Somewhere along the
line losses will set in, if one continues on this choice. Continuing the
rat race against MS will only require out of proportion investments that
will accelerate that problem IMHO.
We'll just have to wait and see what happens with them going forward.
Everything is IMHO here. However still I don't agree. If I choose a devel
tool, I often have to commit to maintainability for a couple of years.
Borland's future therefore is relevant ex ante.
Note that switching to MS was my boss' choice, not mine. It is also not
relevant for everybody (we are heavily into web apps, tradeoffs for
client-server apps may have different tradeoffs). Specially in this light I
can see reason in my bosses decision.
Moverover, if Apple releases its hold on XCode, another open competitorwith
already an ecosystem ready to expand to Wintel might emerge.
Perhaps...who knows.
It's a really wild speculation I know. But stranger stuff has happened, and
the motivation to bind developers is not unlike MS. (Though MS is a dev
tools corporation originally. Remember basic interpreters?)
add up. Unless the services get so big that the whole IDE business becomes
sideshow to attract people to the services.
Well, I think that *Borland* has chosen the exact strategy that you're
advocating.
Not really. I'm describing a strategy to keep the IDE business profitable.
It centers around the IDE business. So it is about DevCo, not Borland.
But where does that leave *DevCo*?
This _is_ about DevCo. The fact that Borland has services for its ALM tools,
doesn't mean that DevCo couldn't have any for the IDE branch.
IOW, if you don't believe that IDE's can stand alone as an independent
business, and must be a strategic asset in the context of a larger
business, then there's no point in DevCo existing at all.
And a good point it is. However that is a different question. I'm saying
that A (head on confrontation in the shrinkwrap dev tools market with MS and
the IBM sponsored Eclipse) is way, way out of their league.
I give some alternate courses (embedded, services), can think of some other
(align with some strategic partner that funds you, and bundles their tools,
but DevCo plans seem to rule such a construct out), but I can only hint at
that.
My point is more about A being a pretty certain dead course, simply based on
the magnitude of the competition.
But to declare defeat before the whole things starts is insane.
I'm not a soldier in the holy war of Delphi.
_Of course_ I leave if it hurts my own job/business, and so will everybody
that is not totally independant (iow without management or customers that
have a say in tools selection). Purely because the uncertainty is the
perfect weapon for those that want to abandon Delphi earlier.
Throwing all that you've got into direct competition with organisationsthat
can't be outcompeted, well, euh, that is pretty much the business'
_definition_ of suicide.
The assumption is that you're competing with organizations that can't be
"outcompeted", which I think is a false one. It's a defeatest attitude that
will never result in anything positive.
Note every defaitist attitude is wrong. It's like saying that boarding the
liveboat on the Titanic was defeatist. Guess what, they were running! And
most of those that stayed, stayed because they had no choice.
Borland/Delphians should realise that the golden age is over. Now it is
about saving what is left. Not sacrificing what is left in a desperate
attempt to win back the old grandeur. Because that will fail. For
certain.
I'm not advocating some "pie in the sky" attitude, what I'm saying is that
there are a whole bunch of *smart* people choosing to stay with DevCo,
choosing to take DevCo independent and they just might have some good ideas.
I make choices based on their strategy, not on heads.
OTOH I notice that all the old Borlanders, people that I remember as Borland
followers since the late eighties/early nineties all have droped out of
active development with Borland tools. (the few exceptions staying with
Borland are partners giving courses)
Do I believe that Delphi will capture significant marketshare? No. Do I
believe that Delphi can *increase* its current marketshare? Sure.
I'm not sure about that.
Borland is the _only_ tools vendor that gets its money by selling tools.
(should have been ... major ... tools)
Altova?
New to me. don't know how big or whatever they are. Seem to thrive on
shipping utilities, rather than tools.
ActiveState?
Lifts on Apache's hype. Can't see that happen for Delphi.
The only comparable organisation I can think of is Metrowerks. And they
first got out by focus on embedded, and they got out, totally.
Sure, but IIRC they haven't been independent for quite some time: Motorola
and now Freescale.
That's what I meant with "got out".
But I don't really know much about Metrowerks except that they did offer
development tools for Apple a while back.
Yes. And then Apple shipped XCode, and the Apple market was shrinking
already.
Markets without competition are not viable markets.
IMO, if a company isn't willing to compete, then they might as well upload
their code to sourceforge and call it a day.
The point is to compete "where". Or do you always pick the most difficult
task? It is not an evolutionary sensible thing to do :_)
Get out of where? .NET and Java?
While I'm tempted to say yes, I won't.
Why not? Cause that's where the *money* is? ;)
Agreed that there is money there. However that is also money that is most
risky to get, and requires huge investments.
IMHO they tried that starting with D8, and now another two versions futher,
maybe the conclusion should be that they missed the moneytrain, and get out?
But at least shun the main marketsfeature
where you are head on confront the others. IF you can find a niche in the
.NET or Java markets that's fine. Or have a non-duplicatable killer
that appeals to _all_ users, so that it actually adds to your revenue, and
allows you to keep prices high. Like they had in the past with easy
component GUI and DB, before everybody duplicated. (how many people really
use ECO?)
That's exactly right. If Borland is going to address the .NET and Java
markets they need to answer the tough questions directly:
Why should I choose BDS over VS for .NET development?
Why should I choose JBuilder over Eclipse for Java development?
And
Are these decisions made on purely IDE merit and/or in general factors
within my (read: DevCo's) control?
MS pushing VS or IBM pushing Eclipse simply makes a different impression than
Borland pushing Delphi.
IMO, they haven't been answering those questions well at all and they need
to start doing so. But that doesn't mean that I think it's *impossible* for
them to answer those questions.
The glass is not empty (because still wet) like theory?
For head on competition with MS? Of course! How can one not?
I haven't written them off. There are many things that DevCo can do that
MS won't do. For example, DevCo could revive their Kylix effort, or
pursue Mono aggressively, or look at supporting OSX, etc. There are ways
to compete against MS by attacking their Windows-only strategy.
I think all of them.
But IMHO abandon _at least_ the .NET part on Windows. (or better put it in
the fridge) You have to make
choices, and abandoning what you can't win is not a bad place to start.
Microsoft is
like thousand times bigger, and now seems to have its act together, and
pretty much already conquered the Windows based development market.
Which means that DevCo should consider expanding the platform support of
Delphi again or break away from Delphi and support something new like Ruby
or Python or PHP or whatever. But DevCo is larger than Delphi and we
shouldn't forget about that.
Lesson from Kylix: Most development still happens on windows, even if
deployed elsewhere.
Retain the IDE, target other platforms. Only support .NET as far as needed
to make it a nice "platform" in your setup, get out of the arms-race with
MS.
Borland is a sideshow.
You seem to like that word "sideshow". ;)
Problem of a non-native speaker I guess.
I would agree with that assesment, but they have to support .NET. What they
need to do is go beyond .NET.
ECO goes beyond. But the problem is there that it doesn't appeal to
everybody. They must have a real killer. Something that everything uses and
that is worth switching or rowing-against-the-stream for. If they don't have
it, stop trying, and get out.
Having a tools vendor that was primarily euh, your tools vendor, and not sb
who plugs those tools as part of a wider business based mostly on OSes and
Office might also be a good thing to emphasise.
Maybe they should take FPC's old attitude. We used to have the attitude, if
Borland doesn't do it (Linux before Kylix, Win64, WinCE), we do it. Maybe
they should do the same with MS. Live on the scraps and be agile.
I think that's exactly right.
It is dangerous path though. A few missers and you are ruined. Which is less
of a problem for a OSS project than for a corp. The worst we have to lose is
our spare time, and a bit of reputation.
.
- References:
- Re: ruby on rails
- From: Marco van de Voort
- Re: ruby on rails
- From: Brian Moelk
- Re: ruby on rails
- From: Marco van de Voort
- Re: ruby on rails
- From: Brian Moelk
- Re: ruby on rails
- From: Marco van de Voort
- Re: ruby on rails
- From: Brian Moelk
- Re: ruby on rails
- Prev by Date: Re: After a few days with VS.NET 2005
- Next by Date: Re: No wonder Borland Technical Documentation stinks
- Previous by thread: Re: ruby on rails
- Next by thread: Re: ruby on rails
- Index(es):
Relevant Pages
|
