quickfix-developers — QuickFIX developers forum

[Quickfix-developers] Does QuickFix implement Message Queue mechanism to make sure all messages are delivered ?

[Frank <fr...@di...>]

Hi,

I'm trying to make use of quickfix to replace the old fix gateway(ugly =
code, not flexible, only support fix4.0) in my company.

I have many years of socket programming experience. Every time when I =
write a tcp socket program, I'll worry about buffering of read and write =
stream, queueing of messages ready to send, queueing of messages =
received and waiting for handling. If I add such kinds of handling in =
the socket program, the program will be really very stable and robust; =
if not, errors will occur after it runs for some time.

My question is: Does QuickFix have such kinds of handling ? especially =
message queue mechanism ?  May I borrow it to deal with other sockets =
(My program is a fix gateway, one side connects trading system (not fix =
protocol), one side connects ECN(fix protocol)) ?

QuickFix source code is so complicated that I can not figure it out by =
myself right now.

Thanks.

[Quickfix-developers] Connect and Disconnect Events

[Jon D. <jd...@wi...>]

Is there any way to determine when a SenderCompID has
Connected/Disconnected besides the onLogon() and onLogout() callbacks?

I need to perform some verification of the connection before 'letting it
in' and having callbacks for Connect and Disconnect would be helpful.

If there isn't, a pointer to the code for modification would be helpful.

Thanks,

jd

[Quickfix-developers] Windows build machine

[Oren M. <or...@qu...>]

We now have a automated build machine running Visual Studio 7.  Like 
the linux build, source and binary packages will be created off of the 
repository every hour.  You can get to the results of the latest builds 
from the development page: 
http://www.quickfixengine.org/developers.html

--oren

RE: [Quickfix-developers] QF read contents of .incoming / .outgoing?

[Brendan B. B. <br...@ka...>]

> On Aug 30, 2004, at 1:02 PM, Oren Miller wrote:
> It uses the MessageStore (such as the FileStore).  Not the log.

Thanks, that set me straight.

I expect that I'll be receiving 200MB+ of quotes and will be
connected Mon-Fri with breaks once a day.  Thinking incorrectly that
QF loaded <file>.incoming on each connect I was concerned that
connection times would increase each day.

Regards,
Brendan

Re: [Quickfix-developers] QF read contents of .incoming / .outgoing?

[Oren M. <or...@qu...>]

It uses the MessageStore (such as the FileStore).  Not the log.

--oren

On Aug 30, 2004, at 12:38 PM, Brendan B. Boerner wrote:

> QuickFIX Documentation: 
> http://www.quickfixengine.org/quickfix/doc/html/index.html
> QuickFIX FAQ: 
> http://www.quickfixengine.org/wikifix/index.php?QuickFixFAQ
> QuickFIX Support: http://www.quickfixengine.org/services.html
>
> Hi,
>
> I have a project which is using FIX42::QuoteRequest / FIX42::Quote to
> stream quotes.
>
> I have a recollection from debugging a problem last year that
> QuickFIX will read the contents of the .incoming / .outgoing FileLog
> when 'starting up' although looking in the source for v1.8.0 I don't
> see this.
>
> I figure that this is how its always been and my recollection is in
> error, can anyone confirm?
>
> Regards,
> Brendan
>
>
> -------------------------------------------------------
> This SF.Net email is sponsored by BEA Weblogic Workshop
> FREE Java Enterprise J2EE developer tools!
> Get your free copy of BEA WebLogic Workshop 8.1 today.
> http://ads.osdn.com/?ad_id=5047&alloc_id=10808&op=click
> _______________________________________________
> Quickfix-developers mailing list
> Qui...@li...
> https://lists.sourceforge.net/lists/listinfo/quickfix-developers
>

[Quickfix-developers] QF read contents of .incoming / .outgoing?

[Brendan B. B. <br...@ka...>]

Hi,

I have a project which is using FIX42::QuoteRequest / FIX42::Quote to
stream quotes.

I have a recollection from debugging a problem last year that
QuickFIX will read the contents of the .incoming / .outgoing FileLog
when 'starting up' although looking in the source for v1.8.0 I don't
see this.

I figure that this is how its always been and my recollection is in
error, can anyone confirm?

Regards,
Brendan

Re: [Quickfix-developers] Discrepancies in FIX44.xml?

[Oren M. <or...@qu...>]

It's something that has been discussed, and for a short time there was 
even some discussion of using some variation of the QuickFIX xml 
format, but I think they wanted to go their own way.  Frankly I could 
care less if they put it together with an Access database, as long as 
they generate XML files.  In any case, I'd be more than happy to donate 
a process that will convert their database into a mysql database.  As 
long as the spec is distributed in a machine readable format, we can do 
with it what we will.

--oren

On Aug 30, 2004, at 8:35 AM, John Messmer wrote:

> QuickFIX Documentation: 
> http://www.quickfixengine.org/quickfix/doc/html/index.html
> QuickFIX FAQ: 
> http://www.quickfixengine.org/wikifix/index.php?QuickFixFAQ
> QuickFIX Support: http://www.quickfixengine.org/services.html
>
> You're right... that's very scary. Has anyone suggested that the FIX
> repository be maintained in a non-MS based d-base? I've noticed the
> strong MS base among the FIX folks. We non-MS folks need to be heard
> among the FIX community.
>
> IMHO, The very idea of the official FIX spec being distributed in a 
> Word
> doc versus an XML file format is insane! So what does that tell us?
>
> John
>
> On Mon, 2004-08-30 at 07:59, Caleb Epstein wrote:
>> QuickFIX Documentation: 
>> http://www.quickfixengine.org/quickfix/doc/html/index.html
>> QuickFIX FAQ: 
>> http://www.quickfixengine.org/wikifix/index.php?QuickFixFAQ
>> QuickFIX Support: http://www.quickfixengine.org/services.html
>>
>> On Sun, 29 Aug 2004 22:51:24 -0500, Oren Miller 
>> <or...@qu...> wrote:
>>
>>> I'm told that the spec is now being kept in an Access database
>>
>>      Wow.  Thats scary.
>
>
> -------------------------------------------------------
> This SF.Net email is sponsored by BEA Weblogic Workshop
> FREE Java Enterprise J2EE developer tools!
> Get your free copy of BEA WebLogic Workshop 8.1 today.
> http://ads.osdn.com/?ad_id=5047&alloc_id=10808&op=click
> _______________________________________________
> Quickfix-developers mailing list
> Qui...@li...
> https://lists.sourceforge.net/lists/listinfo/quickfix-developers
>

Re: [Quickfix-developers] QuickFIX Log Analysis

[Oren M. <or...@qu...>]

Actually we only currently have a Python interface (ruby is not yet=20
available), and that is only available on unix systems right now.

There is also this perl script in the contrib directory which can be=20
modified to do what you want:=20
http://cvs.sourceforge.net/viewcvs.py/quickfix/contrib/display_fixlog/

--oren

On Aug 30, 2004, at 4:41 AM, Joerg Thoennes wrote:

> QuickFIX Documentation:=20
> http://www.quickfixengine.org/quickfix/doc/html/index.html
> QuickFIX FAQ:=20
> http://www.quickfixengine.org/wikifix/index.php?QuickFixFAQ
> QuickFIX Support: http://www.quickfixengine.org/services.html
>
>> What you're looking for is really a perfect job for a Perl script
>> though.  Here's the skeleton of something that can read a quickfix
>> message store (or any file that just contains raw FIX messages one
>> per-line) and parses each message into a hash of field IDs and =
values.
>>  The rest it up to you:
>> #!/usr/bin/perl -w
>> use strict;
>> while (<>) {
>>      chomp;
>>      my %MSG =3D map { split /=3D/, $_, 2 } split /\001/;
>>      # Do something with %MSG
>> }
>
> Using Perl has the additional advantage, that it also runs on a lot of=20=

> non-UNIX systems.
>
> But also have a look at the Python and Ruby scripting languages.=20
> Actually for these languages there are native QF APIs so can avoid all=20=

> the parsing stuff but use QF directly from Python or Ruby. Since I am=20=

> neither a Python or Ruby programmers, others from the list may=20
> contribute the scripts.
>
> BTW, Caleb, did you consider adding your script snippets to the=20
> WikiFIX site?  It is actually meant as an repository for this kind of=20=

> stuff.
>
> Cheers, J=F6rg
>
> --=20
> Joerg Thoennes
>                                http://macd.com
> Tel.: +49 (0)241 44597-24      Macdonald Associates GmbH
> Fax : +49 (0)241 44597-10      Lothringer Str. 52, D-52070 Aachen
>
>
> -------------------------------------------------------
> This SF.Net email is sponsored by BEA Weblogic Workshop
> FREE Java Enterprise J2EE developer tools!
> Get your free copy of BEA WebLogic Workshop 8.1 today.
> http://ads.osdn.com/?ad_id=3D5047&alloc_id=3D10808&op=3Dclick
> _______________________________________________
> Quickfix-developers mailing list
> Qui...@li...
> https://lists.sourceforge.net/lists/listinfo/quickfix-developers
>

Re: [Quickfix-developers] Discrepancies in FIX44.xml?

[John M. <jo...@tr...>]

You're right... that's very scary. Has anyone suggested that the FIX
repository be maintained in a non-MS based d-base? I've noticed the
strong MS base among the FIX folks. We non-MS folks need to be heard
among the FIX community.

IMHO, The very idea of the official FIX spec being distributed in a Word
doc versus an XML file format is insane! So what does that tell us?

John

On Mon, 2004-08-30 at 07:59, Caleb Epstein wrote:
> QuickFIX Documentation: http://www.quickfixengine.org/quickfix/doc/html/index.html
> QuickFIX FAQ: http://www.quickfixengine.org/wikifix/index.php?QuickFixFAQ
> QuickFIX Support: http://www.quickfixengine.org/services.html
> 
> On Sun, 29 Aug 2004 22:51:24 -0500, Oren Miller <or...@qu...> wrote:
> 
> > I'm told that the spec is now being kept in an Access database
> 
>      Wow.  Thats scary.

Re: [Quickfix-developers] Discrepancies in FIX44.xml?

[Caleb E. <cal...@gm...>]

On Sun, 29 Aug 2004 22:51:24 -0500, Oren Miller <or...@qu...> wrote:

> I'm told that the spec is now being kept in an Access database

     Wow.  Thats scary.

-- 
Caleb Epstein
cal...@gm...

Re: [Quickfix-developers] QuickFIX Log Analysis

[Joerg T. <Joe...@ma...>]

> What you're looking for is really a perfect job for a Perl script
> though.  Here's the skeleton of something that can read a quickfix
> message store (or any file that just contains raw FIX messages one
> per-line) and parses each message into a hash of field IDs and values.
>  The rest it up to you:
> 
> #!/usr/bin/perl -w
> use strict;
> while (<>) {
>      chomp;
>      my %MSG = map { split /=/, $_, 2 } split /\001/;
>      # Do something with %MSG
> }
> 

Using Perl has the additional advantage, that it also runs on a lot of 
non-UNIX systems.

But also have a look at the Python and Ruby scripting languages. 
Actually for these languages there are native QF APIs so can avoid all 
the parsing stuff but use QF directly from Python or Ruby. Since I am 
neither a Python or Ruby programmers, others from the list may 
contribute the scripts.

BTW, Caleb, did you consider adding your script snippets to the WikiFIX 
site?  It is actually meant as an repository for this kind of stuff.

Cheers, Jörg

-- 
Joerg Thoennes
                                http://macd.com
Tel.: +49 (0)241 44597-24      Macdonald Associates GmbH
Fax : +49 (0)241 44597-10      Lothringer Str. 52, D-52070 Aachen

Re: [Quickfix-developers] Discrepancies in FIX44.xml?

[Oren M. <or...@qu...>]

The first two are probably issues with the parsing.  Instead of making 
it easier, the FPL has made it much more difficult to create a data 
dictionary.  Back in FIX.4.2 the entire spec was available in one html 
file.  Not ideal but relatively easy to parse.  In the past two 
releases however they have only released the spec as 7 Word documents.  
This makes parsing difficult, particularly since these documents are 
seemingly created by hand and there isn't perfect consistency in how 
things are laid out.  We tried to account for all of the exceptions, 
but some slip by.  So we just try to fix them as they are reported.

As for the LinesOfText, well, Fiximate lists it as NoLinesOfText, while 
the official FIX 4.4 specification document lists it as LinesOfText.  
Since the document is supposed to define the standard, that's what we 
go with (even through NoLinesOfText is what it really should be if you 
go by the naming convention).  A bit worrisome since it appears someone 
is keeping two versions of the spec.  I think it is likely due to a 
transition the FPL is going through.  I'm told that the spec is now 
being kept in an Access database, so hopefully future versions will be 
generated off of this database.

--oren

On Aug 29, 2004, at 3:36 PM, James Wiggs wrote:

> QuickFIX Documentation: 
> http://www.quickfixengine.org/quickfix/doc/html/index.html
> QuickFIX FAQ: 
> http://www.quickfixengine.org/wikifix/index.php?QuickFixFAQ
> QuickFIX Support: http://www.quickfixengine.org/services.html
>
>
>  Folks,
>
>    I'm working with the FIX44.xml specification document off the
> QuickFIX site and I've spotted what seem like discrepancies between
> it and the FIX4.4 spec at fixprotocol.org.  The MDEntryRefID "group"
> under MarketDataIncrementalRefresh doesn't appear to jibe with what
> is defined in the fixprotocol.org FIXIimate documentation.  Is this
> a "group" that isn't a repeating group, but just an organizational
> convenience?  There is no "NoMDEntryRefID" defined in FIXimate to
> indicate that MDEntryRefID is a repeating group.  FIXimate defines
> it as a string field.  What's up, here?
>
>    Secondly, the FIX header clearly specifies a field "NoHops" to
> indicate the number of "HopCompID" groups, but FIX44.xml has:
> ...
>     <field name="LastMsgSeqNumProcessed" required="N" />
>     <group name="HopCompID" required="N">
>       <field name="HopCompID" required="N" />
>       <field name="HopSendingTime" required="N" />
>       <field name="HopRefID" required="N" />
>     </group>
> </header>
> ...
>    A typo?  Or is this deliberate?  Shouldn't it be instead:
> ...
>     <field name="LastMsgSeqNumProcessed" required="N" />
>     <group name="NoHops" required="N">
>       <field name="HopCompID" required="N" />
>       <field name="HopSendingTime" required="N" />
>       <field name="HopRefID" required="N" />
>     </group>
> </header>
> ...
>
>    Finally, the "News" message type defines the group number field
> for "Text" as "NoLinesOfText" but FIX44.xml contains:
> ...
>       <group name="LinesOfText" required="Y">
>         <field name="Text" required="Y" />
>         <field name="EncodedTextLen" required="N" />
>         <field name="EncodedText" required="N" />
>       </group>
> ...
>    So, shouldn't it be:
> ...
>       <group name="NoLinesOfText" required="Y">
>         <field name="Text" required="Y" />
>         <field name="EncodedTextLen" required="N" />
>         <field name="EncodedText" required="N" />
>       </group>
> ...
>
>    That said, I've gotten an enormous amount of benefit from these
> spec files and want to send profuse thanks to whomever built them
> and then made them available!
>
> regards,
> Jim Wiggs
>
>
>
> -------------------------------------------------------
> This SF.Net email is sponsored by BEA Weblogic Workshop
> FREE Java Enterprise J2EE developer tools!
> Get your free copy of BEA WebLogic Workshop 8.1 today.
> http://ads.osdn.com/?ad_id=5047&alloc_id=10808&op=click
> _______________________________________________
> Quickfix-developers mailing list
> Qui...@li...
> https://lists.sourceforge.net/lists/listinfo/quickfix-developers
>

[Quickfix-developers] Discrepancies in FIX44.xml?

[James W. <wi...@wi...>]

 Folks,

   I'm working with the FIX44.xml specification document off the
QuickFIX site and I've spotted what seem like discrepancies between
it and the FIX4.4 spec at fixprotocol.org.  The MDEntryRefID "group"
under MarketDataIncrementalRefresh doesn't appear to jibe with what
is defined in the fixprotocol.org FIXIimate documentation.  Is this
a "group" that isn't a repeating group, but just an organizational
convenience?  There is no "NoMDEntryRefID" defined in FIXimate to
indicate that MDEntryRefID is a repeating group.  FIXimate defines
it as a string field.  What's up, here?

   Secondly, the FIX header clearly specifies a field "NoHops" to
indicate the number of "HopCompID" groups, but FIX44.xml has:
...
    <field name="LastMsgSeqNumProcessed" required="N" />
    <group name="HopCompID" required="N">
      <field name="HopCompID" required="N" />
      <field name="HopSendingTime" required="N" />
      <field name="HopRefID" required="N" />
    </group>
</header>
...
   A typo?  Or is this deliberate?  Shouldn't it be instead:
...
    <field name="LastMsgSeqNumProcessed" required="N" />
    <group name="NoHops" required="N">
      <field name="HopCompID" required="N" />
      <field name="HopSendingTime" required="N" />
      <field name="HopRefID" required="N" />
    </group>
</header>
...

   Finally, the "News" message type defines the group number field
for "Text" as "NoLinesOfText" but FIX44.xml contains:
...
      <group name="LinesOfText" required="Y">
        <field name="Text" required="Y" />
        <field name="EncodedTextLen" required="N" />
        <field name="EncodedText" required="N" />
      </group>
...
   So, shouldn't it be:
...
      <group name="NoLinesOfText" required="Y">
        <field name="Text" required="Y" />
        <field name="EncodedTextLen" required="N" />
        <field name="EncodedText" required="N" />
      </group>
...

   That said, I've gotten an enormous amount of benefit from these
spec files and want to send profuse thanks to whomever built them
and then made them available!

regards,
Jim Wiggs

[Quickfix-developers] enum values for Product field 460

[John M. <jo...@tr...>]

QF's FIX44.xml file lists only 9 enumerated values for the Product field
(tag 460). However, the 4.4 spec and FIXimate lists 13. The differences
are the following values:
10 = MORTGAGE
11 = MUNICIPAL
12 = OTHER
13 = FINANCING

Have these values been deprecated in the 4.4 spec or is QF's FIX44.xml
file not up-to-date?

Thanks in advance, John Messmer

Re: [Quickfix-developers] QuickFIX Log Analysis

[Caleb E. <cal...@gm...>]

On Fri, 27 Aug 2004 15:48:20 +0200, Joerg Thoennes
<joe...@ma...> wrote:
> >>
> >>      grep -c '35=8' logfile
> >>
> >> counts the number of ExecutionReports.
> 
> If it does, you could easily add a grep for the date.  The same applies
> for NewOrderSingle ('35=D').

If you have GNU grep I'd suggest making the regexp a little stricter
so you don't get any false positives (e.g. something like 135=8):

     grep -c '[[:cntrl:]]35=8' logfile

What you're looking for is really a perfect job for a Perl script
though.  Here's the skeleton of something that can read a quickfix
message store (or any file that just contains raw FIX messages one
per-line) and parses each message into a hash of field IDs and values.
 The rest it up to you:

#!/usr/bin/perl -w
use strict;
while (<>) {
     chomp;
     my %MSG = map { split /=/, $_, 2 } split /\001/;
     # Do something with %MSG
}

-- 
Caleb Epstein
cal...@gm...

RE: [Quickfix-developers] RE: FIX42::QuoteRequest operator=()

[Brendan B. B. <br...@ka...>]

On August 27, 2004 at 10:52 AM, Oren Miller wrote:

> I would venture to guess that it may be a bug in the STL that comes 
> with VC 6.0 (see earlier threads regarding its shaky implementation).  
> Have you tried running this test with VC 6.0 and STLPort?

Just tried it.  Nope, that doesn't seem to be it (debugged into the
A::foo::operator=() and confirmed I'm using STLport's and not VC++'s
map/string).

Regards,
Brendan

Re: [Quickfix-developers] RE: FIX42::QuoteRequest operator=()

[Oren M. <or...@qu...>]

I would venture to guess that it may be a bug in the STL that comes 
with VC 6.0 (see earlier threads regarding its shaky implementation).  
Have you tried running this test with VC 6.0 and STLPort?

--oren

On Aug 27, 2004, at 10:40 AM, Brendan B. Boerner wrote:

> QuickFIX Documentation: 
> http://www.quickfixengine.org/quickfix/doc/html/index.html
> QuickFIX FAQ: 
> http://www.quickfixengine.org/wikifix/index.php?QuickFixFAQ
> QuickFIX Support: http://www.quickfixengine.org/services.html
>
>> Using VC++ 6.0, the following copy of a FIX42::QuoteRequest causes a
>> stack overflow as FIX42::Message::operator=() recursively calls
>> FIX42::Message::operator=() instead of calling up to
>> FIX::Message::operator=() e.g:
>
> ...
>
>> This is probably a VC++ 6.0 bug - I've also seen it when copying
>> FIX41::NewOrderSingle as well.
>
> This does appear to be a bug in VC++ 6.0 as the following will
> reproduce it:
>
> namespace A {
>     class foo
>     {
>         public:
>             typedef std::map <int, std::string> t_mapStr;
>
>             t_mapStr m_mapStr;
>     }; /* end class foo */
>
> } /* end namespace A */
>
> namespace B {
>     class foo : public A::foo
>     {
>         public:
>             int m_y;
>     }; /* end class foo */
>
> } /* end namespace B */
>
> A::foo l_fooA_1;
> A::foo l_fooA_2;
> B::foo l_fooB_1;
> B::foo l_fooB_2;
>
> // ok
> l_fooA_1 = l_fooA_2;
>
> // stack overflow
> l_fooB_1 = l_fooB_2;
>
>
> Regards,
> Brendan
>
>
> -------------------------------------------------------
> This SF.Net email is sponsored by BEA Weblogic Workshop
> FREE Java Enterprise J2EE developer tools!
> Get your free copy of BEA WebLogic Workshop 8.1 today.
> http://ads.osdn.com/?ad_id=5047&alloc_id=10808&op=click
> _______________________________________________
> Quickfix-developers mailing list
> Qui...@li...
> https://lists.sourceforge.net/lists/listinfo/quickfix-developers
>

[Quickfix-developers] RE: FIX42::QuoteRequest operator=()

[Brendan B. B. <br...@ka...>]

> Using VC++ 6.0, the following copy of a FIX42::QuoteRequest causes a
> stack overflow as FIX42::Message::operator=() recursively calls
> FIX42::Message::operator=() instead of calling up to
> FIX::Message::operator=() e.g:

...

> This is probably a VC++ 6.0 bug - I've also seen it when copying
> FIX41::NewOrderSingle as well.

This does appear to be a bug in VC++ 6.0 as the following will
reproduce it:

namespace A {
    class foo
    {
        public:
            typedef std::map <int, std::string> t_mapStr;

            t_mapStr m_mapStr;
    }; /* end class foo */

} /* end namespace A */

namespace B {
    class foo : public A::foo
    {
        public:
            int m_y;
    }; /* end class foo */
 
} /* end namespace B */

A::foo l_fooA_1;
A::foo l_fooA_2;
B::foo l_fooB_1;
B::foo l_fooB_2;

// ok
l_fooA_1 = l_fooA_2;

// stack overflow
l_fooB_1 = l_fooB_2;


Regards,
Brendan

Re: [Quickfix-developers] QuickFIX Log Analysis

[Joerg T. <Joe...@ma...>]

> We would like to implement features 
> Like 	1) sent newordersingle Filter by date
> 	2) executionreport received Filtered by date ...
> and provide a facility to geek the messages which are sent already.

Does this UNIX grep work for you?

>> If you just want to do some statistics, this can be easily achieved 
>> using UNIX tools, e.g.
>> 
>> 	grep -c '35=8' logfile
>> 
>> counts the number of ExecutionReports.

If it does, you could easily add a grep for the date.  The same applies 
for NewOrderSingle ('35=D').

Cheers, Jörg

-- 
Joerg Thoennes
                                http://macd.com
Tel.: +49 (0)241 44597-24      Macdonald Associates GmbH
Fax : +49 (0)241 44597-10      Lothringer Str. 52, D-52070 Aachen

RE: [Quickfix-developers] QuickFIX Log Analysis

[Ramprakash U. <ram...@ca...>]

We would like to implement features=20
Like 	1) sent newordersingle Filter by date
	2) executionreport received Filtered by date ...
and provide a facility to geek the messages which are sent already.

Does any tool available for this?=20

TIA,
Ramprakash


-----Original Message-----
From: Joerg Thoennes [mailto:Joe...@ma...]=20
Sent: venerd=EC 27 agosto 2004 13:43
To: Ramprakash Umapathy
Cc: 'QuickFIX-Developers'
Subject: Re: [Quickfix-developers] QuickFIX Log Analysis


> Is there any free (Open Source) FIX Log Analysis Program available? We =

> are developing an QUICKFIX Initiator application which sends orders,=20
> receive execution reports etc., At the end of the day, we would like=20
> to do analysis of the number of messages(orders) sent, execution=20
> reports received etc., Is there any free tool available?

If you just want to do some statistics, this can be easily achieved=20
using UNIX tools, e.g.

	grep -c '35=3D8' logfile

counts the number of ExecutionReports.

What else do you want to do?

Cheers, J=F6rg

--=20
Joerg Thoennes
                                http://macd.com
Tel.: +49 (0)241 44597-24      Macdonald Associates GmbH
Fax : +49 (0)241 44597-10      Lothringer Str. 52, D-52070 Aachen

Re: [Quickfix-developers] QuickFIX Log Analysis

[Joerg T. <Joe...@ma...>]

> Is there any free (Open Source) FIX Log Analysis Program available? We are
> developing an QUICKFIX Initiator application which sends orders, receive
> execution reports etc., At the end of the day, we would like to do analysis
> of the number of messages(orders) sent, execution reports received etc., Is
> there any free tool available?

If you just want to do some statistics, this can be easily achieved 
using UNIX tools, e.g.

	grep -c '35=8' logfile

counts the number of ExecutionReports.

What else do you want to do?

Cheers, Jörg

-- 
Joerg Thoennes
                                http://macd.com
Tel.: +49 (0)241 44597-24      Macdonald Associates GmbH
Fax : +49 (0)241 44597-10      Lothringer Str. 52, D-52070 Aachen

[Quickfix-developers] QuickFIX Log Analysis

[Ramprakash U. <ram...@ca...>]

Hi all,

Is there any free (Open Source) FIX Log Analysis Program available? We are
developing an QUICKFIX Initiator application which sends orders, receive
execution reports etc., At the end of the day, we would like to do analysis
of the number of messages(orders) sent, execution reports received etc., Is
there any free tool available?

Many Thanks,
Ramprakash

[Quickfix-developers] FIX42::QuoteRequest operator=()

[Brendan B. B. <br...@ka...>]

Hi,

Using VC++ 6.0, the following copy of a FIX42::QuoteRequest causes a
stack overflow as FIX42::Message::operator=() recursively calls
FIX42::Message::operator=() instead of calling up to
FIX::Message::operator=() e.g:

FIX42::QuoteRequest::operator=()
    --> FIX42::Message::operator=()
        --> FIX::Message::operator=()
            --> FIX::FieldMap::operator=()

as is the case on Linux (Fedora C1, gcc 3.4.0).

    FIX::Message msg_1;
    FIX::Message msg_2;

    FIX42::QuoteRequest quoteRequest42_1;
    FIX42::QuoteRequest quoteRequest42_2;

    msg_1 = msg_2;                          // ok
    quoteRequest42_1 = quoteRequest42_2;    // stack overflow

I tried implementing a FIX42::Message &FIX42::Message::operator =(const
FIX42::Message &rhs):

    Message &operator =(const Message &p_message)
    {
        if (this == &p_message)
        {
            return(*this);
        } /* end if */

        *((FIX::Message *) this) = p_message;
        return(*this);
    }


This seems to address the copy problem and work on Win & Lnx.

This is probably a VC++ 6.0 bug - I've also seen it when copying
FIX41::NewOrderSingle as well.

TIA for any insights / comments.

Regards,
Brendan

RE: [Quickfix-developers] Obtaining original request using RefMsgType

[Ramprakash U. <ram...@ca...>]

Hi,
=20
I think you are referring RefSeqNum (tag 45) in the FIX Reject Message.
=20
See this post if it related to your question
http://sourceforge.net/mailarchive/forum.php?thread_id=3D5258597
<http://sourceforge.net/mailarchive/forum.php?thread_id=3D5258597&forum_i=
d=3D103
> &forum_id=3D103
=20
Ramprakash
=20
-----Original Message-----
From: qui...@li...
[mailto:qui...@li...] On Behalf Of =
Matt
Proud
Sent: marted=EC 24 agosto 2004 20:46
To: qui...@li...
Subject: [Quickfix-developers] Obtaining original request using =
RefMsgType


I had a few related questions:
=20
1. Is there a way to obtain the original fix message using the =
RefMsgType
field from a FIX reject message?=20
=20
Example: I am sending a FIX New Order single message but am receiving a =
FIX
Session-level reject message from the exchange. I need access to the FIX =
New
Order single message to access the ClOrdId.
=20
2.  From the onLogout method, is there a way to obtain the Logout =
message?
=20
Matt P.

[Quickfix-developers] Obtaining original request using RefMsgType

[Matt P. <mp...@ca...>]

I had a few related questions:
=20
1. Is there a way to obtain the original fix message using the
RefMsgType field from a FIX reject message?=20
=20
Example: I am sending a FIX New Order single message but am receiving a
FIX Session-level reject message from the exchange. I need access to the
FIX New Order single message to access the ClOrdId.
=20
2.  From the onLogout method, is there a way to obtain the Logout
message?
=20
Matt P.