|
PureBytes Links
Trading Reference Links
|
"About this, I also speak not from
speculation, but from actual experience."
Does your letter assume that nobody else on this board has the same
experience? Maybe they have more. Honestly, 3 to 5 years to code a close
TS clone would be rediculous with a full time staff I outlined below. You
mentioned custom programming while you developed your app. That's got to
eat into your time. I'm talking about a dedicated staff working full time.
With test automation, good coders and people who understand software
development lifecycle (ie iterative not waterfall), you can deliver a good
piece of software like a close TS clone (charting rt, rt position monitoring
and a basic-like custom language with extensible C++) to market in less than
a year (minus the server which can be phased in) but no longer than 1 1/2 yr
max. But it will cost money to get good people. You've got to have good
people and that's hard to compare in letters like these.
-----Original Message-----
From: HDD95@xxxxxxx [mailto:HDD95@xxxxxxx]
Sent: Tuesday, October 09, 2001 2:49 AM
To: omega-list@xxxxxxxxxx
Subject: Re: BEST TRADING SOFTWARE
In a message dated 10/5/2001 1:54:30 PM Central Daylight Time,
stnahc@xxxxxxxx writes:
> --- Brian <blink64@xxxxxxxx> wrote:
>
> > <snip>
> >
> > Take a good (system) trader telling a good engineer what the trader
needs.
> > Give me 3 decent QA engineers, and 3 developers to implement and I
could
> > design and develop a product just as good as Tradestation (charting,
> > language, and if needed server). [snip]
>
> If you have only 3 QA + 3 developers ... it will take 3 to 5 years.
>
> [snip]
>
> not speculation, but from actual experience.
>
> Lawrence
Or here is another idea. Spend 12 years sitting in a corner and do it
yourself. Feed yourself by offering custom programming services to traders.
The traders you will meet along the way will be more than happy to give you
their input about your continually emerging software. Give your customers
what they are looking for. Fulfill their expectations and delivery to them
software with working features they are requesting, and they will be happy
customers.
I was recently publicly ridiculed in this mailing list for admitting that I
have spent 12 years doing exactly what I describe in the previous paragraph.
When Lawrence predicts 18 to 30 person years (according to my calculator),
it
makes me feel better. It is the difference between speculation and wishful
thinking, versus actual experience. I have listened to many speculations
over
the years, both on this mailing list and in private conversations. The scope
of development time for trading software, both commercial software and
proprietary software, is always underestimated. Lawrence and I both have
actual experience. Trading software is a difficult application.
But I have to be careful so I am not jumped on again. PowerST is not a
TradeStation feature for feature clone. Rather, it is a software product
which specializes in system testing. However, since it goes into much more
depth into system testing than TradeStation, there is some justification to
comparison of development times. Furthermore, it isn't only the development
time. The shake out period as real customers actually start using the
software also cannot be minimized. About this, I also speak not from
speculation, but from actual experience.
In a message dated 10/6/2001 2:33:49 PM Central Daylight Time,
simgenie@xxxxxxxxx writes:
> Business Strategy:
> Actually I don't think that trying to be just like product X is a viable
> business strategy when you have an upstart company attempting to take
market
> share away from an entrenched market leader. This was the basic strategy
of
> TraderWare and it wasn't successful. For a product to succeed it needs
merit
> above and beyond what other products on the market have. It needs that
> special something (differentiation) that sets it apart from the
competition.
> The easiest way for a startup product to succeed in a market with a well
> established product with more marketing muscle is to make their product a
> niche product (that doesn't compete directly) and then gradually move
into
> the other guy's market [snip]
This is a better summary of the PowerST approach than I have been able to
come up with myself. For example, I saw a void in software with programmable
portfolio level money management ability, so I went into that niche. At
first
this money management capability was available only with custom programming
services, which is very much of a niche because custom programming is
expensive. The next phase was the ability of end users to do their own
programming with a great deal of personalized hand holding help. That phase
is currently in transition to a more mass market product with more
comprehensive documentation which will allow users to come up to speed with
the software with less hand holding. PowerST also has other specialization's
in high end optimization, and in general a very powerful and highly
programmable testing environment.
But, yes, it is a niche product. It is for advanced traders who have already
used other trading software and realized the limitations. There is a
definite
emphasis on power and programmability, even when that leads to a tradeoff
that PowerST is a more complex and technical package.
As to what extent PowerST will move into "the other guy's market", who knows
what the future holds. I think it is difficult, if not impossible, to have a
single software product which both addresses the full extent of the
TradeStation mass market plus provides the level of high end testing that
PowerST provides. PowerST will always remain a more technical package with a
"power" accent. It was an original design goal, which led to the name "Power
System Tester", and will likely remain the permanent accent.
Bob Bolotin
Developer of "PowerST: The Power System Tester"
http://www.powertesting.com
|