🚀 go-pugleaf

"JLoz" <jfsierra@excite.com> wrote in message
news:4679c15b$2@newsgroups.borland.com...

> The error message I get is
>
> EAssertion Failed with 'Assertion Failure IdGloba.pas Line 2528'

It would be helpful if you would show what is actually on that line.
There have been so many releases and snapshots of Indy, there is no
way to know which one you are actually using.  For instance, there is
no assertion on that line in my copy of IdGlobal.pas.  Line 2528 is
inside the TIdLocalEvent constructor.

> The way the attachments are sent is different from what I was
> accostumed to do with Indy 9

Attachments work the same as they did in Indy 9.  The only difference
is that Indy 9 supported only file attachments whereas Indy 10
supports a customizable attachment model.  So the TIdAttachment class
is now an abstract base class in Indy 10 and can't be instantiated
directly anymore.  TIdAttachmentFile in Indy 10 is the direct
replacement for TIdAttachment in Indy 9.  Most of your code that used
TIdAttachment in Indy 9 does not need to change in Indy 10, as the
methods/properties are the same.  The only real change you need to
when you instantiate a new attachment, which you already updated that
code.

> I had problems trying to use IdAttachementFile as it was not being
recognized.

Then you did not read the documentation, as it tells you which unit
each class/component is declared in.


Gambit

Here are the lines:

2526:   function IndyLength(const ABuffer: String; const ALength: Integer 
= -1; const AIndex: Integer = 1): Integer;
2527:   begin
2528:     Assert(AIndex >= 1);
2529:     if ALength < 0 then begin
2530:       Result := Length(ABuffer)-AIndex+1;
2531:     end else begin
2532:       Result := Min(Length(ABuffer)-AIndex+1, ALength);
2533:     end;
2534:   end;


Thanks again,


"Remy Lebeau (TeamB)" <no.spam@no.spam.com> wrote in message 
news:4679fa58$1@newsgroups.borland.com...
>
> "JLoz" <jfsierra@excite.com> wrote in message
> news:4679c15b$2@newsgroups.borland.com...
>
>> The error message I get is
>>
>> EAssertion Failed with 'Assertion Failure IdGloba.pas Line 2528'
>
> It would be helpful if you would show what is actually on that line.
> There have been so many releases and snapshots of Indy, there is no
> way to know which one you are actually using.  For instance, there is
> no assertion on that line in my copy of IdGlobal.pas.  Line 2528 is
> inside the TIdLocalEvent constructor.
>
>> The way the attachments are sent is different from what I was
>> accostumed to do with Indy 9
>
> Attachments work the same as they did in Indy 9.  The only difference
> is that Indy 9 supported only file attachments whereas Indy 10
> supports a customizable attachment model.  So the TIdAttachment class
> is now an abstract base class in Indy 10 and can't be instantiated
> directly anymore.  TIdAttachmentFile in Indy 10 is the direct
> replacement for TIdAttachment in Indy 9.  Most of your code that used
> TIdAttachment in Indy 9 does not need to change in Indy 10, as the
> methods/properties are the same.  The only real change you need to
> when you instantiate a new attachment, which you already updated that
> code.
>
>> I had problems trying to use IdAttachementFile as it was not being
> recognized.
>
> Then you did not read the documentation, as it tells you which unit
> each class/component is declared in.
>
>
> Gambit
>
> 

Gambit,

According to your answer I already changed the code in in theory it should 
work, right?

I tried entering a wrong path and I get an EFOpenError with message 'Cannot 
open file xxx'.  That tells me its trying to open the file.  When the path 
is correct the assertion error occurs in the line

IdSMTP1->Sent(IdMessage1)

Thanks in advance for your help.

JLoz

"Remy Lebeau (TeamB)" <no.spam@no.spam.com> wrote in message 
news:4679fa58$1@newsgroups.borland.com...
>
> "JLoz" <jfsierra@excite.com> wrote in message
> news:4679c15b$2@newsgroups.borland.com...
>
>> The error message I get is
>>
>> EAssertion Failed with 'Assertion Failure IdGloba.pas Line 2528'
>
> It would be helpful if you would show what is actually on that line.
> There have been so many releases and snapshots of Indy, there is no
> way to know which one you are actually using.  For instance, there is
> no assertion on that line in my copy of IdGlobal.pas.  Line 2528 is
> inside the TIdLocalEvent constructor.
>
>> The way the attachments are sent is different from what I was
>> accostumed to do with Indy 9
>
> Attachments work the same as they did in Indy 9.  The only difference
> is that Indy 9 supported only file attachments whereas Indy 10
> supports a customizable attachment model.  So the TIdAttachment class
> is now an abstract base class in Indy 10 and can't be instantiated
> directly anymore.  TIdAttachmentFile in Indy 10 is the direct
> replacement for TIdAttachment in Indy 9.  Most of your code that used
> TIdAttachment in Indy 9 does not need to change in Indy 10, as the
> methods/properties are the same.  The only real change you need to
> when you instantiate a new attachment, which you already updated that
> code.
>
>> I had problems trying to use IdAttachementFile as it was not being
> recognized.
>
> Then you did not read the documentation, as it tells you which unit
> each class/component is declared in.
>
>
> Gambit
>
> 

"JLoz" <jfsierra@excite.com> wrote in message
news:467a6b66$1@newsgroups.borland.com...

> Here are the lines:

According to that, an AIndex of 0 or lower is being passed to the
function.  That should never happen.  Then again, IndyLength() is used
in a lot of places, so it would be useful if you could track down
further which specific call is the faulty one.


Gambit

"JLoz" <jfsierra@excite.com> wrote in message
news:467a8244@newsgroups.borland.com...

> I tried entering a wrong path and I get an EFOpenError with
> message 'Cannot open file xxx'.  That tells me its trying to open
the file.

That is what it is supposed to do.  You are trying to send a message
with a file attachment, so the file has to be opened so its data can
be encoded for transmission.


Gambit

Thanks for the suggestion.  However, I am not exactly sure how I could track 
down the faulty call since I would not be able to step into any pascal unit. 
I tried going into the IdAttachmentFile.pas file and see the constructor but 
it is not clear to me where it goes.

I would be glad to do the work, but could you please give me some tips?

Thanks again,

JLoz

"Remy Lebeau (TeamB)" <no.spam@no.spam.com> wrote in message 
news:467ab2f7$1@newsgroups.borland.com...
>
> "JLoz" <jfsierra@excite.com> wrote in message
> news:467a6b66$1@newsgroups.borland.com...
>
>> Here are the lines:
>
> According to that, an AIndex of 0 or lower is being passed to the
> function.  That should never happen.  Then again, IndyLength() is used
> in a lot of places, so it would be useful if you could track down
> further which specific call is the faulty one.
>
>
> Gambit
>
> 

"JLoz" <jfsierra@excite.com> wrote in message
news:467ab52b@newsgroups.borland.com...

> Thanks for the suggestion.  However, I am not exactly sure how
> I could track down the faulty call since I would not be able to
> step into any pascal unit.

You could recompile Indy with debug enabled.  Then you could step
through it.

> I tried going into the IdAttachmentFile.pas file and see the
> constructor but it is not clear to me where it goes.

This is not a constructor issue.


Gambit

Thanks again.

The problem is that I never got it to compile with the debug as the Zlib.pas 
file makes the compiler crash and the only way I got it to work was to 
disable debug (this by creating the bpk files manually using the IDE).

If you could help me determine why the following lines fail in the 
FULLC5.bat file that I got by modifying FULLC6.bat, I would probably be able 
to make it work and probably send the new FULLC5.bat to share with other 
people.  I am using the latest snapshot by the way (Version 10.1.6 I think)

dcc32.exe IndyProtocols50.dpk /O..\Lib\Protocols\objs /DBCB /M /H /W /JPHN 
/N. /U. -$d-l-n+p+r-s-t-w-y- %2 %3 %4

dcc32.exe IndyProtocols50.dpk /O..\Lib\Protocols\objs /DBCB /M /H /W /N. 
/U. -$d-l-n+p+r-s-t-w-y- %2 %3 %4

The first line fails and then by ignoring the error I get the other one to 
run and it seems to work.  However, since I am not familiar with the command 
line parameters for dcc32.exe I am not exactly sure the differences.

You might remember my previous post where I had compiler errors due to some 
structures in the API header that are not available for BCB5.  Well, I 
commented out those lines in Winsock2.hpp and gave it a try since I do not 
have any choice being stated that Indy10 will probably not be tested in BCB5 
in a long time, if at all.

If at least I could compile it without issues it would be great.

Thanks for your patience,

JLoz

"Remy Lebeau (TeamB)" <no.spam@no.spam.com> wrote in message 
news:467abdaf@newsgroups.borland.com...
>
> "JLoz" <jfsierra@excite.com> wrote in message
> news:467ab52b@newsgroups.borland.com...
>
>> Thanks for the suggestion.  However, I am not exactly sure how
>> I could track down the faulty call since I would not be able to
>> step into any pascal unit.
>
> You could recompile Indy with debug enabled.  Then you could step
> through it.
>
>> I tried going into the IdAttachmentFile.pas file and see the
>> constructor but it is not clear to me where it goes.
>
> This is not a constructor issue.
>
>
> Gambit
>
> 

JLoz wrote:

>You might remember my previous post where I had compiler errors due to some 
>structures in the API header that are not available for BCB5.  Well, I 
>commented out those lines in Winsock2.hpp and gave it a try since I do not 
>have any choice being stated that Indy10 will probably not be tested in BCB5 
>in a long time, if at all.

That is not the usual method to use in the case of missing
definitions.  The proper thing to do is research the missing
definitions and add them.
Here's one one the missing structures.
http://msdn2.microsoft.com/en-us/library/ms740504.aspx

Thanks!

In BCB5, which header file should I add this to?

JLoz
"Bob Gonder" <notbg@notmindspring.invalid> wrote in message 
news:t1un73t3iuqf0l49rnj9075f4i0f7cc199@4ax.com...
> JLoz wrote:
>
>>You might remember my previous post where I had compiler errors due to 
>>some
>>structures in the API header that are not available for BCB5.  Well, I
>>commented out those lines in Winsock2.hpp and gave it a try since I do not
>>have any choice being stated that Indy10 will probably not be tested in 
>>BCB5
>>in a long time, if at all.
>
> That is not the usual method to use in the case of missing
> definitions.  The proper thing to do is research the missing
> definitions and add them.
> Here's one one the missing structures.
> http://msdn2.microsoft.com/en-us/library/ms740504.aspx
>
> 

JLoz wrote:

>In BCB5, which header file should I add this to?

At the bottom of the page it tells you where it is normaly found.

Or you could just put them in idmissing.h and add that to the Indy
modules that complain.

// regular headers first, then
#ifndef SOCKADDR_STORAGE
#include	<idmissing.h>
#endif

Thanks a lot!

I will take care of this.

JLoz
"JLoz" <jfsierra@excite.com> wrote in message 
news:467c0073$1@newsgroups.borland.com...
> Thanks!
>
> In BCB5, which header file should I add this to?
>
> JLoz
> "Bob Gonder" <notbg@notmindspring.invalid> wrote in message 
> news:t1un73t3iuqf0l49rnj9075f4i0f7cc199@4ax.com...
>> JLoz wrote:
>>
>>>You might remember my previous post where I had compiler errors due to 
>>>some
>>>structures in the API header that are not available for BCB5.  Well, I
>>>commented out those lines in Winsock2.hpp and gave it a try since I do 
>>>not
>>>have any choice being stated that Indy10 will probably not be tested in 
>>>BCB5
>>>in a long time, if at all.
>>
>> That is not the usual method to use in the case of missing
>> definitions.  The proper thing to do is research the missing
>> definitions and add them.
>> Here's one one the missing structures.
>> http://msdn2.microsoft.com/en-us/library/ms740504.aspx
>>
>>
>
>