Menu
▾
▴
[GT] Re: Future of GeniusTrader
|
From: Thomas W. <we...@ms...> - 2009-10-16 05:31:10
|
<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
<meta content="text/html;charset=ISO-8859-1" http-equiv="Content-Type">
</head>
<body bgcolor="#ffffff" text="#000000">
João,<br>
<br>
thanks for sharing your story and your feeling about the project.<br>
<br>
I'll try to give my 2c....<br>
<br>
I disagree with your assessment because I have a totally different
understanding of the goals of GT.... hear me out..... <br>
<br>
I personally use GT because it is the way it is, warts and
complications and all.... There are indeed much easier to use free or
commercial charting tools, even tools that evaluate trading systems.
However, that is not what I think GT is. GT is a platform to build your
own trading software. The scripts that exist are just examples, I
think, of what one can do.<br>
<br>
I have used several other such software, most extensively some
commercial products. But these products do not really help me. They let
me write a little function that I can use to evaluate whether to trade
or not, or to backtest. But they do not allow me to write complex
systems that involve multiple markets, do cross market analysis, etc.<br>
<br>
GT supports all that beautifully. Yes, you need to understand how to
write the application that does that for you, but there is a lot of
infrastructure there to do so. I have implemented quite an extensive
analysis system on top of GT in that manner. It is very unique to the
way I invest (I run my own quantitative hedge fund, so to speak). I
could not have implemented that analysis on any other of the freely
available platforms as easy, and certainly not on any of the commercial
products that I can afford.<br>
<br>
So my take is different. I think that most GT users (those that are
left over) are like me. They have built an investing platform on top of
GT. They don't discuss on the system traders list, as these
applications are unique and cannot easily be compared. But they discuss
on the developers list, as there is still plenty to fix.<br>
<br>
I believe you are correct that those who look at GT as a Metastock
equivalent will be disappointed. But to be honest, it is actually not
that bad as is compared to Metastock, except that Metastock is better
in dealing with the graphics, of course. <br>
<br>
OK. All that said, I think that the goals that you describe can be
satisfied by the GT platform also. I think that GT will easily support
the web-based application I understand you envision. In fact, I run my
analysis every night on a server and generate the results so that I can
review them on the web. It would be just a small step to an interactive
mode.<br>
<br>
Chia Liang's ideas of interactive charting are great as well, and
should be feasible to implement.<br>
<br>
The CPAN branch will go a great way in making it easier to get started.
Maybe the features that you and Chia Liang are talking about can help
bring the "develop a system" people in as well...<br>
<br>
I would love to see the web-based trading system that you are talking
about. But I do hope that it will be on the existing infrastructure, so
that those who want to use GT as analysis engine can also benefit from
that innovation....<br>
<br>
Best regards, Th.<br>
<br>
<br>
<br>
<br>
João Costa wrote:
<blockquote
cite="mid:d8e...@ma..."
type="cite">
<div>OK, this is a bit of a long one, so, grab a cup of whatever your
favourite beverage, and stay with me here for a minute :)</div>
<div><br>
</div>
<div>I've been quietly following on the background the latest
developments in the GT world with interest.
<div><br>
</div>
<div>While I stopped contributing since about one year ago, GT is in
a way a very special project to me. Before my first contribution to
the project, I had read maybe one online perl tutorial. Today I hold a
full time job as a perl developer for an organization offering hosted
application services, and my experience in contributing to GT was
essential for that.</div>
<div><br>
</div>
<div>I suppose my reasons for having stopped contributing might have
been because I felt the project was going nowhere and I didn't feel
capable of changing that. We got stuck with old processes, innovation
didn't seem to happen, and in general, there was simply no direction or
coordination of efforts in where to go.</div>
<div><br>
</div>
<div>But if I drill further down, my real reason for quitting was
that I just don't believe in the concept. GT was always a hard
application to use. First you have to install it which often proved
difficult for many people. Then you need to setup a data source and
keep it up to date. If you're doing stocks, this means dealing with
things like splits as well.</div>
<div><br>
</div>
<div>Quite a few people will have given up by now. Only if you're
one of the minority who made it past these hurdles can you start
thinking back on why did you want to install GT in the first place ?
Right, to build trading systems. So now you start mucking around with
the scripts, except often times they don't work in an obvious way and
you find yourself mucking through perl code instead of writing your
trading system.</div>
<div><br>
</div>
<div>Which in part I think explains why we have some traffic in the
devel list, but none in the system traders list. They've all given up
by now! All that's left is us, the developers :) And in my view that is
what's fundamentally wrong with GT. It's an application with a target
audience completely different than the people who actually use it.</div>
<div><br>
</div>
<div>Now, I realize there seem to be some ideas out there which I
believe have great potential. <span class="Apple-style-span"
style="font-family: 'Times New Roman'; font-size: medium;">Chia-liang
mentioned adapting GT for real time trading, I think this is superb.
He also mentioned making GT development more agile and test driven, in
my opinion, spot on. There's the CPAN effort which seems to be driven
mostly by Erik with some great help from Thomas and Robert, who on top
of that have been tireless in giving a helping hand to newcomers, and
these among contributions by other people whom I'll have to fail to
mention.</span></div>
<div><font class="Apple-style-span" face="'Times New Roman'"><span
class="Apple-style-span" style="font-size: medium;"><br>
</span></font></div>
<div><span class="Apple-style-span"
style="font-family: 'Times New Roman'; font-size: medium;">But to me
the fundamental flaw remains. The entry barriers are just too high
from the IT side. Even if you are the super guru of trading systems,
you're unlikely to ever use GT unless you hire an IT team to run and
manage it.</span></div>
<div><br>
</div>
<div>And this is where my idea on what the future of GT should be
comes in. I think it should be an online application running in a
hosted environment. All the setting up would only have to be done once
by us, IT geeks. The people out there who actually want to setup, test
and run trading system could focus on their area of expertise. And in
the end I believe we could have an application with a defined target
audience and some great potential going forward.</div>
<div><br>
</div>
<div>I know there are some technical and logistical challenges in
making something like this work, but to be honest, I deal with this
kind of thing everyday. It can be done, and it can be successful. But
it does require a concentrated effort by a willing team. If I could
find a few people willing to work on this I'd love to put my best
effort into it, so, anyone interested ?</div>
<div><br>
</div>
<div>If on the other hand people generally decide to go on working in
a desktop application, I'll just keep watching from the sidelines, but
let me give you my sincere best wishes for the future of the project.</div>
<div><br>
</div>
<div>João</div>
</div>
<br>
<br>
<div class="gmail_quote">On Thu, Oct 15, 2009 at 2:35 PM, Raphael
Hertzog <span dir="ltr"><<a moz-do-not-send="true"
href="mailto:ra...@ge...">ra...@ge...</a>></span>
wrote:<br>
<blockquote class="gmail_quote"
style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">(Please
keep me in CC on replies as I don't really read the list)<br>
<br>
Hello,<br>
<br>
Chia-liang Kao mailed me recently to be granted commit access to<br>
GeniusTrader. I'm glad that the project seems to take off after<br>
having languished so many years.<br>
<br>
However, even if I initiated the project, I'm not really leading it any<br>
more. Except me 4 persons have commit access nowadays:<br>
- joao: João Costa <<a moz-do-not-send="true"
href="mailto:joa...@zo...">joa...@zo...</a>>
(last seen in 2008)<br>
- olf: Oliver Bossert <<a moz-do-not-send="true"
href="mailto:ol...@ol...">ol...@ol...</a>> (last seen in
2006)<br>
- ras: "Robert A. Schmied" <<a moz-do-not-send="true"
href="mailto:ra...@ac...">ra...@ac...</a>> (active)<br>
- thomas: Thomas Weigert <<a moz-do-not-send="true"
href="mailto:we...@ms...">we...@ms...</a>> (active)<br>
<br>
I don't want to be a blocker for the project so I would suggest to move<br>
the project on a forge so that I'm no longer the sole project admin that<br>
can grant commit access.<br>
<br>
As part of my Debian work, I happen to be administrator of the<br>
<a moz-do-not-send="true" href="http://alioth.debian.org"
target="_blank">http://alioth.debian.org</a> forge so if that forge
could suit you, I'd be<br>
glad to move the project there. It supports all modern VCS that could<br>
be of interest for the project (SVN, Git, bzr, hg, ...).<br>
<br>
What do you think ? (Input from the new contributors who have no commit<br>
access is welcome too)<br>
<br>
Thomas and Robert, if you think that I should immediately grant commit<br>
access to some new contributors, please tell me and I'll create them<br>
an SVN account.<br>
<br>
Cheers,<br>
<font color="#888888">--<br>
Raphaël Hertzog -+- <a moz-do-not-send="true"
href="http://www.ouaza.com" target="_blank">http://www.ouaza.com</a><br>
<br>
Freexian : des développeurs Debian au service des entreprises<br>
<a moz-do-not-send="true" href="http://www.freexian.com"
target="_blank">http://www.freexian.com</a><br>
</font></blockquote>
</div>
<br>
</blockquote>
</body>
</html>
|