SQLRDD request for advice

[Andrzej Morgiewicz]

Hi
I'm going to port my CDX based applications to SQL. Now it's time to choose SQL platform. I'm going to use SQLRDD to do fast conversion and then lately optimize, so I've a question to all SQLRDD users. Which SQL serwer, in yours opinion and experience, is the best choice ? I mean especially whitch is the best in terms of speed with no optimatization, just simply porting. I suppose that some platforms are "more friendly" and others less to SQLRDD methodology. Both free and commercial servers can be considered. Application is same kind of accounting and real estate rent meangment bussines suport software. Same files up to 1 000 000 recs or more but mostly max several thousands.
Andrzej Morgiewicz

[Patrick Mast]

Hello Andrzej,

> I'm going to port my CDX based applications to SQL. Now it's time to choose SQL platform. I'm going to use SQLRDD to do fast conversion and then lately optimize, so I've a question to all SQLRDD users. Which SQL serwer, in yours opinion and experience, is the best choice ? I mean especially whitch is the best in terms of speed with no optimatization, just simply porting. I suppose that some platforms are "more friendly" and others less to SQLRDD methodology. Both free and commercial servers can be considered. Application is same kind of accounting and real estate rent meangment bussines suport software. Same files up to 1 000 000 recs or more but mostly max several thousands.

I did the same with my WinFakt! application. We converted from DBFSIX to SQLRDD. We use PostgreSQL as our SQL server. We tested a few and this was the tastes and best compatible with SQLRDD.

Patrick

[cul...@gmail.com]

Hi

Same opinion as Patrick, I suggest to use potgresql

Regards
Luiz

[elviraa...@gmail.com]

Hi Andrzej,

SQLRDD is not fully debugged yet.

It is used only by very very few users. Many bugs are pending to solve.

To see yourself, try to download a demo for xHarbour and Borland...

Also, check news.xharbour.com. See how few messages there are, and the pending bugs reported.

We purchased it last year and support was inexistent. Also, I don´t know why they do not pay attention to that issues. The hard work at SQLRDD is done, but some flecks are remaining and no one at xHarbour.com pays attention. As a businesswoman I do not understand.

We moved to ADO instead.

Good luck!!.

[Patrick Mast]

Hello Elvira,

> SQLRDD is not fully debugged yet.

We use SQLRDD on a daily bases here. Works great for me! I use xHarbour Builder + SQLRDD

> It is used only by very very few users. Many bugs are pending to solve.

Please let me know what bugs are pending so we can try to fix them asap, Thank you!

> To see yourself, try to download a demo for xHarbour and Borland…
Ok. ;-)

> Also, check news.xharbour.com. See how few messages there are, and the pending
> bugs reported.

Few messages does not mean open bugs ;)

> We purchased it last year and support was inexistent. Also, I don´t know why they do not pay
> attention to that issues. The hard work at SQLRDD is done, but some flecks are remaining and
> no one at xHarbour.com pays attention. As a businesswoman I do not understand.

Sorry you feel that way.

> We moved to ADO instead.
> Good luck!!.

Also good luck to you and than you for your thoughts. It can only help us to make our software better.

Patrick

[Elvira]

Hello Patrick,

Very surprised by fast answer here and not in support ;).

> We use SQLRDD on a daily bases here. Works great for me! I use xHarbour Builder + SQLRDD

Of course, you won´t say opposite.


>
> > It is used only by very very few users. Many bugs are pending to solve.

> Please let me know what bugs are pending so we can try to fix them asap, Thank you!
>

You already know them.

At your forum: Lucas (opening inexistent tables produces GPF), Cassiano (index unique), Kokoo (recno(), SR_Recno(), OrdKey()), Oscar (Interbase), Ariel Paredes (can´t work with latest Postgres), ...


Hope you are doing good with Winfakt.

Best regards,

[Patrick Mast]

Hey Elvira,

> Very surprised by fast answer here and not in support ;).

I try to do so in Support also!! ;-)

> > We use SQLRDD on a daily bases here. Works great for me! I use xHarbour Builder + SQLRDD

> Of course, you won´t say opposite.

Yes, if I need SQL i SHLOUD use SQLRDD. The good thing is that it works great for me.

> > > It is used only by very very few users. Many bugs are pending to solve.
> > Please let me know what bugs are pending so we can try to fix them asap, Thank you!
> >

> You already know them.
> At your forum: Lucas (opening inexistent tables produces GPF)

Problem is XBrowse uses OrdkeyNo() and SQLRDD does not support this function.

>Cassiano (index unique), Kokoo (recno(), SR_Recno(), OrdKey()), Oscar (Interbase),
> Ariel Paredes (can´t work with latest Postgres), ...

I will ask Luiz to have a closer look at these. Thanks.

> Hope you are doing good with Winfakt.

Thank you Elvira ;-)

Patrick

[Elvira]

Hello Patrick,

> I try to do so in Support also!! ;-)

Please, look at the dates of the messages.

Maybe the news system is not as better as a forum, as Fivetech for example has.

For out customers, we use http://www.kayako.com/


>
> > > We use SQLRDD on a daily bases here. Works great for me! I use xHarbour Builder + SQLRDD
> > Of course, you won´t say opposite.

> Yes, if I need SQL i SHLOUD use SQLRDD. The good thing is that it works great for me.
>

Plase, try to use in xHarbour.org and BCC 5.82 ;)


> > At your forum: Lucas (opening inexistent tables produces GPF)

> Problem is XBrowse uses OrdkeyNo() and SQLRDD does not support this function.

Plase, have you looked at that?. He says "opening inexistent tables produces GPF", nothing to do with OrdKeyNo()

>
> >Cassiano (index unique), Kokoo (recno(), SR_Recno(), OrdKey()), Oscar (Interbase),
> > Ariel Paredes (can´t work with latest Postgres), ...

> I will ask Luiz to have a closer look at these. Thanks.
>

They are so importat ;). For example, latest Postgres SQL

Willing you fix the problems ;). In fact, as I say before, only some flecks are missing to have a great tool. I do not understand why xHarbour.com does not focus on this.

Also, the architecture of the product is very good. Who did it?. The parser was a great idea.

[cul...@gmail.com]

Elvira
As far i know, their is no bugs on sqlrdd,
all bugs reported is fixed.

And many of the problems, is resolved direct by direct email to me

Regards

Regards
Luiz

On 15 jul, 10:08, Elvira <elviraadorac...@gmail.com> wrote:
> Hello Patrick,
>
> > I try to do so in Support also!! ;-)
>
> Please, look at the dates of the messages.
>
> Maybe the news system is not as better as a forum, as Fivetech for example has.
>

> For out customers, we usehttp://www.kayako.com/

[Elvira]

Luiz,

First of all, thank you for your dedication to xHarbur development.

If you read them, they are bugs I am afraid

Please, try to use in xHarbour.org and BCC 5.82 ;)


And also, please look at the posts.


> &gt;
> &gt;
> &gt; &gt; &amp;gt;Cassiano (index unique), Kokoo (recno(), SR_Recno(), OrdKey()), Oscar (Interbase),
> &gt; &gt; &amp;gt; Ariel Paredes (can´t work with latest Postgres), ...
> &gt; &gt; I will ask Luiz to have a closer look at these. Thanks.
> &gt;
> &gt; They are so importat ;). For example, latest Postgres SQL
> &gt;
> &gt; Willing you fix the problems ;). In fact, as I say before, only some flecks are missing to have a great tool. I do not understand why xHarbour.com does not focus on this.
> &gt;
> &gt; Also, the architecture of the product is very good. Who did it?. The parser was a great idea.

[Andrzej Morgiewicz]

Thanks Elvira for your opinion.
I bought SQLRDD in 2009 and made a quick conversion of my simplest module and it seemed to work properly except of same issues ( for exemle memo fields were reported as character etc ) but at that time i had to leave further conversions and tests due to other works that had have been to do. You made me worried because i was convinced that SQLRDD is (almost) finished and now it will be quict an painless road to SQL :(. I hope that my question and yours answer will push same things ahead as could be seen in discussions on this forum :).
Andrzej Morgiewicz

[Andrzej Morgiewicz]

> &gt; You already know them.

> &gt; At your forum: Lucas (opening inexistent tables produces GPF)

> Problem is XBrowse uses OrdkeyNo() and SQLRDD does not support this function.
>

Hi Patrick,
It could be a problemm for me too because i'm using hwgui & hbrowse which is based on ordeyno()/ordkeycount() cdx functions, maybe there are same 'replacemnts' in SQLRDD, recently i haven't study any SQLRDD docs so i don't remember all features.
Andrzej Morgiewicz
PS Anybody uses SQLRDD with HWGUI or HWGUI with SQL oriented database browser?
( my apps are based on browsers and data forms dynamicaly built and configured in same kind of data dict)?

W dniu niedziela, 15 lipca 2012 14:37:17 UTC+2 użytkownik Patrick Mast napisał:
> Hey Elvira,
>
> &gt; Very surprised by fast answer here and not in support ;).

> I try to do so in Support also!! ;-)
>

> &gt; &amp;gt; We use SQLRDD on a daily bases here. Works great for me! I use xHarbour Builder + SQLRDD
> &gt; Of course, you won´t say opposite.

> Yes, if I need SQL i SHLOUD use SQLRDD. The good thing is that it works great for me.
>

> &gt; &amp;gt; &amp;amp;gt; It is used only by very very few users. Many bugs are pending to solve.
> &gt; &amp;gt; Please let me know what bugs are pending so we can try to fix them asap, Thank you!
> &gt; &amp;gt;

> &gt; You already know them.

> &gt; At your forum: Lucas (opening inexistent tables produces GPF)

> Problem is XBrowse uses OrdkeyNo() and SQLRDD does not support this function.
>

> &gt;Cassiano (index unique), Kokoo (recno(), SR_Recno(), OrdKey()), Oscar (Interbase),
> &gt; Ariel Paredes (can´t work with latest Postgres), ...

> I will ask Luiz to have a closer look at these. Thanks.
>

> &gt; Hope you are doing good with Winfakt.

[Patrick Mast]

Hey Andrzej,

> Hi Patrick,
> It could be a problemm for me too because i&#39;m using hwgui &amp; hbrowse which is based on ordeyno()/ordkeycount() cdx functions, maybe there are same &#39;replacemnts&#39; in SQLRDD, recently i haven&#39;t study any SQLRDD docs so i don&#39;t remember all features.

> Andrzej Morgiewicz
> PS Anybody uses SQLRDD with HWGUI or HWGUI with SQL oriented database browser?
> ( my apps are based on browsers and data forms dynamicaly built and configured in same kind of data dict)?

Thank you for your thoughts. The best way you guys can help us is to provide reduced samples that shows bugs. We will be GLAD to fix them. REALYY! ;-)

I myself use SQLRDD everyday, so its even in MY personal interest to fix SQLRDD bugs.

So, please, go ahead and send us your reduced samples. Lets make SQLRDD the best choice for our users! Thank you!

Patrick

[Elvira]

Hi Andrzej,

Glad to see I am not alone with the problems

This forum seems to be very efective, as Patrick answers very very quick. I can´t say the same to the multiple emails sent to him or the activity at the news server.

It seems that Patrick only uses the product is Postgres but they do not test on other platforms, particulary on Mysql.

In fact, if you check to the libs and dlls for MySQL, Postgres, MS SQL, etc, thet are very old, from 2006. They do not use latest versions of servers and client libs. Here, more flecks pending ;).

[Patrick Mast]

Hey Elivira.

Thanks again for your thoughts ;-)

You can also help SQLRDD by sending us reduced samples that shows the bugs so we can try to fix them. Please do not use any 3rd party GUI for the samples.

Thank you for your support!

Patrick

On Monday, July 16, 2012 11:35:31 AM UTC+2, Elvira wrote:
> Hi Andrzej,
>
> Glad to see I am not alone with the problems
>
> This forum seems to be very efective, as Patrick answers very very quick. I can´t say the same to the multiple emails sent to him or the activity at the news server.
>
> It seems that Patrick only uses the product is Postgres but they do not test on other platforms, particulary on Mysql.
>
> In fact, if you check to the libs and dlls for MySQL, Postgres, MS SQL, etc, thet are very old, from 2006. They do not use latest versions of servers and client libs. Here, more flecks pending ;).
>
>
> El lunes, 16 de julio de 2012 09:34:07 UTC+2, Andrzej Morgiewicz escribió:

> &gt; Thanks Elvira for your opinion.
> &gt; I bought SQLRDD in 2009 and made a quick conversion of my simplest module and it seemed to work properly except of same issues ( for exemle memo fields were reported as character etc ) but at that time i had to leave further conversions and tests due to other works that had have been to do. You made me worried because i was convinced that SQLRDD is (almost) finished and now it will be quict an painless road to SQL :(. I hope that my question and yours answer will push same things ahead as could be seen in discussions on this forum :).
> &gt; Andrzej Morgiewicz

[Gerard de Vries]

We have been using SQLRDD with MSSQL Server for about 4 years now. In
general, in my experience, SQLRDD is pretty stable, however, there are a few
bugs. When I reported the bugs way back when, I was told everything worked
fine, and that there were no bugs. So, I had to program around the bugs,
which worked out okay. SQLRDD has definitely made the switch to working with
SQL instead of DBF pretty easy though. All in all a very good product in my
opinion, but I wish the bugs I reported had been handled differently.

Gerard

"Patrick Mast" schreef in bericht
news:673248b3-6884-489d...@googlegroups.com...

Hey Elivira.

Thanks again for your thoughts ;-)

You can also help SQLRDD by sending us reduced samples that shows the bugs
so we can try to fix them. Please do not use any 3rd party GUI for the
samples.

Thank you for your support!

Patrick

On Monday, July 16, 2012 11:35:31 AM UTC+2, Elvira wrote:
> Hi Andrzej,
>
> Glad to see I am not alone with the problems
>
> This forum seems to be very efective, as Patrick answers very very quick.
> I can�t say the same to the multiple emails sent to him or the activity at
> the news server.
>
> It seems that Patrick only uses the product is Postgres but they do not
> test on other platforms, particulary on Mysql.
>
> In fact, if you check to the libs and dlls for MySQL, Postgres, MS SQL,
> etc, thet are very old, from 2006. They do not use latest versions of
> servers and client libs. Here, more flecks pending ;).
>
>
> El lunes, 16 de julio de 2012 09:34:07 UTC+2, Andrzej Morgiewicz

> escribi�:

[Elvira]

Gerard,

The same experience as me ;).

I don´t know why they do not pay attention to he bugs. There are two possible reasons: a) they do not want to or b) the news system is not the best way to manage errors.

In last case, they should consider using a forum like Fivetech, Xailer or this group too. Also the company will save the dedicated server;). And here Patrick answers so quick.

Regards,

El martes, 17 de julio de 2012 11:34:07 UTC+2, Gerard de Vries escribió:
> We have been using SQLRDD with MSSQL Server for about 4 years now. In
> general, in my experience, SQLRDD is pretty stable, however, there are a few
> bugs. When I reported the bugs way back when, I was told everything worked
> fine, and that there were no bugs. So, I had to program around the bugs,
> which worked out okay. SQLRDD has definitely made the switch to working with
> SQL instead of DBF pretty easy though. All in all a very good product in my
> opinion, but I wish the bugs I reported had been handled differently.
>
> Gerard
>

> &quot;Patrick Mast&quot; schreef in bericht

> news:673248b3-6884-489d...@googlegroups.com...
>
> Hey Elivira.
>
> Thanks again for your thoughts ;-)
>
> You can also help SQLRDD by sending us reduced samples that shows the bugs
> so we can try to fix them. Please do not use any 3rd party GUI for the
> samples.
>
> Thank you for your support!
>
> Patrick
>
> On Monday, July 16, 2012 11:35:31 AM UTC+2, Elvira wrote:

> &gt; Hi Andrzej,
> &gt;
> &gt; Glad to see I am not alone with the problems
> &gt;
> &gt; This forum seems to be very efective, as Patrick answers very very quick.
> &gt; I can´t say the same to the multiple emails sent to him or the activity at
> &gt; the news server.
> &gt;
> &gt; It seems that Patrick only uses the product is Postgres but they do not
> &gt; test on other platforms, particulary on Mysql.
> &gt;
> &gt; In fact, if you check to the libs and dlls for MySQL, Postgres, MS SQL,
> &gt; etc, thet are very old, from 2006. They do not use latest versions of
> &gt; servers and client libs. Here, more flecks pending ;).
> &gt;
> &gt;
> &gt; El lunes, 16 de julio de 2012 09:34:07 UTC+2, Andrzej Morgiewicz
> &gt; escribió:
> &gt; &amp;gt; Thanks Elvira for your opinion.
> &gt; &amp;gt; I bought SQLRDD in 2009 and made a quick conversion of my simplest
> &gt; module and it seemed to work properly except of same issues ( for exemle
> &gt; memo fields were reported as character etc ) but at that time i had to
> &gt; leave further conversions and tests due to other works that had have been
> &gt; to do. You made me worried because i was convinced that SQLRDD is (almost)
> &gt; finished and now it will be quict an painless road to SQL :(. I hope that
> &gt; my question and yours answer will push same things ahead as could be seen
> &gt; in discussions on this forum :).
> &gt; &amp;gt; Andrzej Morgiewicz

[dlzc]

Dear Gerard and Elvria:

On Tuesday, July 17, 2012 6:58:46 AM UTC-7, Elvira wrote:
> Gerard,
>
> The same experience as me ;).
>
> I don´t know why they do not pay attention to

> the bugs.

They have asked, and always ask for simplified code, that shows the problem. They have been very responsive when someone can go the extra mile.

> There are two possible reasons: a) they do
> not want to or b) the news system is not the
> best way to manage errors.

c) they cannot decipher where in the "million lines" of code the problem lies, because the bug reporter does not do due diligence.

> And here Patrick answers so quick.

... but those that get a response from him ("need more input"), seem not to respond. I'd be very frustrated.

David A. Smith

[Patrick Mast]

Hey Elvira,

> The same experience as me ;).
> I don´t know why they do not pay attention to he bugs. There are two possible reasons: a) they do not want to or b) the news system is not the best way to manage errors.
> In last case, they should consider using a forum like Fivetech, Xailer or this group too. Also the company will save the dedicated server;). And here Patrick answers so quick.

Please Elvira, we need reduced samples that shows the bugs so we can fix them.
Thanks! ;-))

Patrick

[Patrick Mast]

Exactly ;-))

On Tuesday, July 17, 2012 4:16:57 PM UTC+2, dlzc wrote:
> Dear Gerard and Elvria:
>
> On Tuesday, July 17, 2012 6:58:46 AM UTC-7, Elvira wrote:

> &gt; Gerard,
> &gt;
> &gt; The same experience as me ;).
> &gt;
> &gt; I don´t know why they do not pay attention to
> &gt; the bugs.

>
> They have asked, and always ask for simplified code, that shows the problem. They have been very responsive when someone can go the extra mile.
>

> &gt; There are two possible reasons: a) they do
> &gt; not want to or b) the news system is not the
> &gt; best way to manage errors.
>
> c) they cannot decipher where in the &quot;million lines&quot; of code the problem lies, because the bug reporter does not do due diligence.
>
> &gt; And here Patrick answers so quick.
>
> ... but those that get a response from him (&quot;need more input&quot;), seem not to respond. I&#39;d be very frustrated.
>
> David A. Smith

On Tuesday, July 17, 2012 4:16:57 PM UTC+2, dlzc wrote:
> Dear Gerard and Elvria:
>
> On Tuesday, July 17, 2012 6:58:46 AM UTC-7, Elvira wrote:

> &gt; Gerard,
> &gt;
> &gt; The same experience as me ;).
> &gt;
> &gt; I don´t know why they do not pay attention to
> &gt; the bugs.

>
> They have asked, and always ask for simplified code, that shows the problem. They have been very responsive when someone can go the extra mile.
>

> &gt; There are two possible reasons: a) they do
> &gt; not want to or b) the news system is not the
> &gt; best way to manage errors.
>
> c) they cannot decipher where in the &quot;million lines&quot; of code the problem lies, because the bug reporter does not do due diligence.
>
> &gt; And here Patrick answers so quick.
>
> ... but those that get a response from him (&quot;need more input&quot;), seem not to respond. I&#39;d be very frustrated.
>
> David A. Smith

[Gerard de Vries]

There was one specific bug that I reported a few years ago, that had
something to do with using "deleted" records in SQL. I posted a very simple
little piece of code, with a description of the structure of the table. I
could not have made it easier to reproduce the error. And I was simply told
there were no bugs in SQLRDD. I don't know what else I could have done. So,
instead of hoping that the problem would be fixed I decided to code around
it.

I have to say that Patrick has always been very helpful and very fast to
answer emails and to try to fix bugs. But in this instance it wasn't
Patrick.

Kind regards,

Gerard

"Patrick Mast" schreef in bericht

news:a13b28f0-cc31-440d...@googlegroups.com...

[Gerard de Vries]

Dear David,

SQLRDD question are usually pretty specific, and cannot be answered by a lot
of people. From what I can remember, but it's been a couple of years, the
few times I've posted a potential problem, I have *not* been asked for more
specific code, but I've simply been told there are no bugs in SQLRDD. And
that was basically the end of the discussion.

Don't get me wrong, I love xHarbour, and I think SQLRDD is a great product,
but in the rare cases I have run into a problem with SQLRDD I've basically
hit a stone wall. I've had to figure out a solution i.e. workaround by
myself.

Kind regards,

Gerard

"dlzc" schreef in bericht
news:17463a1f-f05b-447f...@googlegroups.com...

[Patrick Mast]

Hey Gerard,

Thank you for your feedback. Please if you still have bugs, send me the reduced sample and I will do my very best to get them fixed.

Thanks!

Patrick

> There was one specific bug that I reported a few years ago, that had

> something to do with using &quot;deleted&quot; records in SQL. I posted a very simple

> little piece of code, with a description of the structure of the table. I
> could not have made it easier to reproduce the error. And I was simply told

> there were no bugs in SQLRDD. I don&#39;t know what else I could have done. So,

> instead of hoping that the problem would be fixed I decided to code around
> it.
>
> I have to say that Patrick has always been very helpful and very fast to

> answer emails and to try to fix bugs. But in this instance it wasn&#39;t
> Patrick.
>
> Kind regards,

[dlzc]

Dear Gerard de Vries:

On Friday, July 20, 2012 2:30:22 AM UTC-7, Gerard de Vries wrote:
> There was one specific bug that I reported a
> few years ago, that had something to do with
> using "deleted" records in SQL. I posted a
> very simple little piece of code, with a
> description of the structure of the table.

I reviewed the postings in this newsgroup, and I don't see any mention of this by you. I have not looked at the old newsgroup archive.

> I could not have made it easier to reproduce
> the error. And I was simply told there were
> no bugs in SQLRDD. I don't know what else I
> could have done.

When you have talked about using SQL here, you have discussed ODBC and such, intermediate steps that may block your attempts to succeed. Not sure if you reported those steps at the the same time, but might have an effect.

I was not intending to jump down your throat. Please do not feel that was my intent.

Many problems we all have are in unpublished assumptions. We guess or assume we have done things the right way, and forget we made those choices. And the root of problems can be in those choices. Which is where clear documentation comes in.

David A. Smith

[Gerard de Vries]

Dear Patrick,

Thank you very much for the offer. I'll do my best to dig up the code and
the tables I had problems with, but it may take me a little while, since it
was probably 2008 or 2009 that I had those problems.

Kind regards,

Gerard

"Patrick Mast" schreef in bericht

news:ff48211c-ddec-4d04...@googlegroups.com...

[Gerard de Vries]

Dear David,

I looked for those posts myself, and I did see something from back in 2009 I
think. But that was only something in this newsgroup. However, I probably
sent messages from Outlook Express or Outlook on my previous PC, directly to
one of the developers. I'm just not sure anymore, it's too long ago.

I would feel bad coming accross as unhappy about xHarbour, SQLRDD or the
support. xHarbour with SQLRDD has proven to be very stable, and I would
recommend it without hesitation.

Also, in general, I've learned a lot from reading the posts in this
newsgroup. It's a very valuable resource of information to me.

I'm going to try to reproduce the problems I had, and post about them here.

Kind regards,

Gerard



"dlzc" schreef in bericht

news:2cd1551a-f303-4b59...@googlegroups.com...

[Lucas]

Hi,

When I try to open a DBF that does not exist, it gives up a GPF.

Sample code:

cConnection:="HST="+cServer+";UID="+cUser+";PWD="+cPassword+";DTB="+cDatabase
nConnection:=SR_AddConnection(CONNECT_MYSQL,cConnection)

IF nConnection <= 0
MsgAlert("No se puede establecer la conexión. ","Verifique!")
QUIT
ENDIF



RddSetDefault("SQLRDD")
SR_UseDeleteds(.F.)


USE TABLENOEXIST NEW ALIAS "BIBLIO"


Thanks. I use xHarbour.org, BCC 5.82 and Fivewin 12.06

[Patrick Mast]

Hey Gerard,

> Thank you very much for the offer. I&#39;ll do my best to dig up the code and

> the tables I had problems with, but it may take me a little while, since it
> was probably 2008 or 2009 that I had those problems.

Thanks! ;-)

Patrick

[Patrick Mast]

Hey Lucas,

> When I try to open a DBF that does not exist, it gives up a GPF.

On first try I got an RTE:
http://dl.dropbox.com/u/390763/Screenshot/2012-07-24_1151.png

testing more here now..

> Sample code:
Thanks for the sample code, but this is not a reduced sample ;-)
A reduced sample means I can copy the code, put it in a PRG file and compile.

Yours uses variables that are not yet defined etc..

Now, for THIS sample, I understand what the problem is, but please, next time try to make the sample COPY/PASTE/RUN ready. Thanks for that!

Ok, testing further on trying to get the GPF.

You are using SQLRDD right, not SQLEX?

Patrick

[Lucas]

Hi Patrick,

As I told you at your news server, I use Fivewin 12.06, xHarbour.org, BCC 5.82 and MySql.

I am using only SQLRDD

I can´t put you my passwords and server IP. It is very easy to reproduce and get it.

Have you tested with Fivewin and MySQL?.

Thank you very much.Best regards

[Patrick Mast]

Hey Lucas,

> As I told you at your news server, I use Fivewin 12.06, xHarbour.org, BCC 5.82 and MySql.
> I am using only SQLRDD

Ok.

> I can´t put you my passwords and server IP. It is very easy to reproduce and get it.

Yes, but imagine if all our users send us half half tests. It will take us too much time to reconstruct your samples. The more easy you make it for us the faster we can test and deliver ;-)

> Have you tested with Fivewin and MySQL?.

Not yet. Though FiveWin should not be in the picture hee. We are testing SQLRDD. FiveWin should not interfere with our tests.

> Thank you very much.Best regards

Thank YOU for letting us know ;-)

Patrick

[Lucas]

Hey Patrick,

I did provide a full working sample at your forums:
http://uploaded.to/file/ixsob6vt

Maybe at your company they do not read them with care.

If you opened it, you will see the details you asked for in previous message ;).

Best regards.

[Elvira]

Lucas,

This behaviour seems very familiar to me...

I opened your file and there is a full sample that reproduces the gpf bug. Also, the file is from 16 th june. At xHarbour.com, they do not read the posts ;).

xHarbour.com always say SQLRDD work fine for them.

As I say before, the better thing to do with SQLRDD before it gets fully tested and debugged, is ask for the money back.

[Elvira]

Lucas,

I just read the whole thread you poted at news.xharbour.com abot the GPF.

It is very funny. They mixed questions and answer you about other topics.

Also, Patrick tells Luiz there is a bug. But here says other words.

You provide a full working bug and today Luiz says:

"Lucas

All samples should be self contained using only pure xharbour code

Regards
Luiz ".


You showed them the bugs and they don´t want to look into.

Those attitudes are not very serious. Not even proffesional. So that´s why the product is, somekind, a buggy fraud.

[cul...@gmail.com]

Elvira

I dont get gpfs on my machive, are you with the last clients dlls files, and wich database

Regards
Luiz

[Lucas]

Luiz,

Thanks for reply.

Have you downloaded http://uploaded.to/file/ixsob6vt ?

You will find a full working sample, also .exe file.

As I told you several times, I always use: xHarbour.org 1.2.1, Fivewin 12.06, Borland 5.82 and MySQL 5.09 via Internet.

Have you tested my sample with Fivewin?.

Thanks.

[dlzc]

Dear Elvira:

On Wednesday, July 25, 2012 2:52:53 AM UTC-7, Elvira wrote:
> Lucas,
>
> I just read the whole thread you poted at
> news.xharbour.com abot the GPF.
>
> It is very funny. They mixed questions and
> answer you about other topics.
>
> Also, Patrick tells Luiz there is a bug. But
> here says other words.
>
> You provide a full working bug and today Luiz
> says:
>
> "Lucas
>
> All samples should be self contained using
> only pure xharbour code
>
> Regards
> Luiz ".
>
>
> You showed them the bugs and they don´t
> want to look into.

Showing the bugs is not the need. We don't believe he came here to lie. But to end up at the problem, we need to know the steps. He is describing a behavior "no one else" sees. So we have to rub the puppy's nose in it.

> Those attitudes are not very serious. Not
> even proffesional. So that´s why the
> product is, somekind, a buggy fraud.

It is what was left to us to maintain, by a *very* smart guy, whose current effort you appear to be trying to push. And when we ask *you* to provide simplified examples, you run away to snipe on another thread.

If you want the problem fixed, help. If you want to snipe, no one will take you seriously.

SQLRDD is only available as commercial xHarbour product. As such, you have guaranteed support through the xHarbour.com resources. This is not the channel for that support. Those channels are listed at the xHarbour.com website.

Not that you cannot get some help here, even from xHarbour developers (I am not). You can. But you seem more interested in denigrating, tearing down, rather than helping to fix. If you bought SQLRDD, they will have records you bought it, and can fix it. But they need your reduced code that shows how xHarbour (not xHarbour plus some third party package) breaks. They don't have teams of hundreds of employees whose job is to track down every fragmentary report.

David A. Smith

[Patrick Mast]

Hey Lucas,

Your gpf.exe also GPF's here. ;-)

If I reduce your sample to NOT using FiveWin and using SQLRDD for xHarbour Builder it does not GPF and gives me this:
http://dl.dropbox.com/u/390763/Screenshot/2012-07-25_2105.png

Now testing with BCC...

Patrick

[Patrick Mast]

Done testing with BCC58 and also no GPF, I still get the regular error message like the one above.

I will send you my test project tmrw in the morning ok?

Thanks.

Patrick

[Patrick Mast]

Hey Lucas,

This is the project I tested with:
https://dl.dropbox.com/u/4973801/SQLRDD-GPF.ZIP

Let me know if it also gives "regular" RTE on your end ok?

Patrick

[Lucas]

Hello Patrick,

Thanks for helping me.

Yes, in your exe I see the clipepr/xharbour error as the picture.

But if I compile with Fivewin, c:\fwh\samples\buildx.bat, I get the gpf.

But normal SQLRDD errors are shown ok.

[cul...@gmail.com]

How are you compiling, you should use xbuild to compile the app correctly

Regards
Luiz

[Lucas]

El jueves, 26 de julio de 2012 01:48:55 UTC+2, cul...@gmail.com escribió:
> How are you compiling, you should use xbuild to compile the app correctly
>

I do NOT use xharbour.com, I use xHarbour.org, Fivewin and BCC 5.82.

[Patrick Mast]

Hey Lucas,

> Thanks for helping me.
NP :)

> Yes, in your exe I see the clipepr/xharbour error as the picture.

Ok!

> But if I compile with Fivewin, c:\fwh\samples\buildx.bat, I get the gpf.
> But normal SQLRDD errors are shown ok.

FiveWin uses its own error system. Maybe you need to look there?
Try to trace where the GPF exactly occurs.

Patrick

[Patrick Mast]

Hey Lucas,

> I do NOT use xharbour.com, I use xHarbour.org, Fivewin and BCC 5.82.

Ok, than you don't NEED to use xBuild. Its just easier to use xBuild ;-)

> Have you tested my sample with Fivewin?

Sorry, we can not give support on FiveWin. As I said earlier, try to trace the errorsys of FiveWin to see where exactly the GPF occurs.

Patrick

[Lucas]

Hi Patrick,

It is not a Fivewin issue. I have nod modified FW error system. Also, for example the SQLRDD error "Attempt to write to an empty table without a previous Append Blank" it is reported fine in FWH.

In fact, you use Fivewin at Winfakt.

Errors are not always reported fine at SQLRDD.

This is an example of how to report an Error in Harbour:

#include "error.ch"

FUNCTION Main()

local oError

oError := ErrorNew()
oError:GenCode := EG_OPEN
oError:Severity := ES_ERROR
oError:subSystem := "Patrick Mast"
oError:SubCode := 2012
oError:Description := "Can not load xxx library"
oError:CanDefault := .F.
Eval(ErrorBlock(), oError)


RETURN (NIL)
//-------------------------------------------------------------------------------------------------------


It works perfect under Fivewin.

Hope it helps.

As I do not have the sources, I can go further. But there should be differences on how the error is reported when no table exist than "Attempt to write to an empty table without a previous Append Blank",;


Best regards

[Lucas]

Patrick,

I also use FastReportand it does report errors fine with Fivewin.

No issue at all.

[dlzc]

On Thursday, July 26, 2012 6:28:52 AM UTC-7, Lucas wrote:
>
> It is not a Fivewin issue.

It is, and you agreed that it only GPFs when FiveWin is included.

...

> It works perfect under Fivewin.

You get a GPF when you include FiveWin. You don't get a GPF when you use xHarbour alone. If FiveWin no longer works with xHarbour.org that has not been changed in more than 2 years, how is that something xHarbour can fix?

> Hope it helps.

How can it? Unless you are dealing with the FiveWin folks? Because they are not here.

David A. Smith

[cul...@gmail.com]

Lucas

Please post your compilation script

Regards
Luiz

[Patrick Mast]

Hey Lucas,

> It is not a Fivewin issue. I have nod modified FW error system.

As you could see in my reduced sample, its not a xHarbour or SQLRDD issue either.

> Also, for example
> the SQLRDD error &quot;Attempt to write to an empty table without a previous
> Append Blank&quot; it is reported fine in FWH.
Why do you bring this up? Sorry I don't understand.

> In fact, you use Fivewin at Winfakt.

Yes, I do.
Newer stuff is build with VXH though ;-)

> Errors are not always reported fine at SQLRDD.

If so, let me know via a reduced sample. I'm glad to let it fix if its broken.

> This is an example of how to report an Error in Harbour:

> #include &quot;error.ch&quot;

> FUNCTION Main()
> local oError
> oError := ErrorNew()
> oError:GenCode := EG_OPEN
> oError:Severity := ES_ERROR

> oError:subSystem := &quot;Patrick Mast&quot;
> oError:SubCode := 2012
> oError:Description := &quot;Can not load xxx library&quot;

> oError:CanDefault := .F.
> Eval(ErrorBlock(), oError)
> RETURN (NIL)
> //-------------------------------------------------------------------------------------------------------

I know ;-))
Why do you bring this up?
What are you trying to say?

> It works perfect under Fivewin.

I'm sure it is ;-)

> Hope it helps.
Help in what?

> As I do not have the sources, I can go further. But there should be differences

> on how the error is reported when no table exist than &quot;Attempt to
> write to an empty table without a previous Append Blank&quot;,;
Is this another issue you are addressing to us? You are mixing things here Lucas. Please stay focused. ;-)

Patrick

[Lucas]

Luiz,

I use fwh\samples\buildx to build samples.

I attach here at the end of this message.

> As you could see in my reduced sample, its not a xHarbour or SQLRDD issue either.

It is very strange. When I produce the SQLRDD error nº 1, which is Attempt to write to an empty table without a previous Append Blank, I do get normal RTE error with Fivewin´s errsys. No problem.

But other errors, such as opening a non existing table, do not produce an error, just a GPF.

So if it is a Fivein issue, I will get GPFS on each error. I don´t understand why in some cases I get an error and in others a GPF.


> &gt; In fact, you use Fivewin at Winfakt.
> Yes, I do.

At Winfakt and SQLRDD, when you try to open a non existing table, what happens?.


> &gt; This is an example of how to report an Error in Harbour:
> &gt; #include &amp;quot;error.ch&amp;quot;
> &gt; FUNCTION Main()
> &gt; local oError
> &gt; oError := ErrorNew()
> &gt; oError:GenCode := EG_OPEN
> &gt; oError:Severity := ES_ERROR
> &gt; oError:subSystem := &amp;quot;Patrick Mast&amp;quot;
> &gt; oError:SubCode := 2012
> &gt; oError:Description := &amp;quot;Can not load xxx library&amp;quot;
> &gt; oError:CanDefault := .F.
> &gt; Eval(ErrorBlock(), oError)
> &gt; RETURN (NIL)
> &gt; //-------------------------------------------------------------------------------------------------------

> I know ;-))
> Why do you bring this up?
> What are you trying to say?

I don´t know your sorce code. Maybe there are internal differences on how errors are managed or produced.




Buildx.bat:

@ECHO OFF
CLS
ECHO ÚÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄ¿
ECHO ³ FiveWin for xHarbour 12.06 xHarbour development power ³Ü
ECHO ³ (c) FiveTech, 1993-2011 for Microsoft Windows 9X/NT/200X/ME/XP/Vista/7 ³Û
ECHO ÀÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÙÛ
ECHO ÿ ßßßßßßßßßßßßßßßßßßßßßßßßßßßßßßßßßßßßßßßßßßßßßßßßßßßßßßßßßßßßßßßßßßßßßßßßßßßßß

if A%1 == A GOTO :SINTAX
if NOT EXIST %1.prg GOTO :NOEXIST

ECHO Compiling...

if "%FWDIR%" == "" set FWDIR=c:\develop\fwh
if "%XHDIR%" == "" set XHDIR=c:\develop\xharbour
rem if "%2" == "/b" set GT=gtwin
rem if not "%2" == "/b" set GT=gtgui
set GT=gtgui

set hdir=%XHDIR%
set hdirl=%hdir%\lib
set bcdir=c:\develop\bcc582
set fwh=%FWDIR%

%hdir%\bin\harbour %1 /n /i%fwh%\include;%hdir%\include /w /p %2 %3 > comp.log
IF ERRORLEVEL 1 GOTO COMPILEERRORS
@type comp.log

echo -O2 -e%1.exe -I%hdir%\include -I%bcdir%\include %1.c > b32.bc
%bcdir%\bin\bcc32 -M -c -v @b32.bc
:ENDCOMPILE

IF EXIST %1.rc %bcdir%\bin\brc32 -r -I%bcdir%\include %1

echo %bcdir%\lib\c0w32.obj + > b32.bc
echo %1.obj, + >> b32.bc
echo %1.exe, + >> b32.bc
echo %1.map, + >> b32.bc
echo %fwh%\lib\Fivehx.lib %fwh%\lib\FiveHC.lib + >> b32.bc
echo %hdirl%\rtl.lib + >> b32.bc
echo %hdirl%\vm.lib + >> b32.bc
echo %hdirl%\%GT%.lib + >> b32.bc
echo %hdirl%\lang.lib + >> b32.bc
echo %hdirl%\macro.lib + >> b32.bc
echo %hdirl%\rdd.lib + >> b32.bc
echo %hdirl%\dbfntx.lib + >> b32.bc
echo %hdirl%\dbfcdx.lib + >> b32.bc
echo %hdirl%\dbffpt.lib + >> b32.bc
echo %hdirl%\hbsix.lib + >> b32.bc
echo %hdirl%\debug.lib + >> b32.bc
echo %hdirl%\common.lib + >> b32.bc
echo %hdirl%\pp.lib + >> b32.bc
echo %hdirl%\pcrepos.lib + >> b32.bc
echo %hdirl%\ct.lib + >> b32.bc
echo %hdirl%\zlib.lib + >> b32.bc
echo %hdirl%\hbzip.lib + >> b32.bc
echo %hdirl%\png.lib + >> b32.bc

rem Uncomment these two lines to use Advantage RDD
rem echo %hdir%\lib\rddads.lib + >> b32.bc
rem echo %hdir%\lib\Ace32.lib + >> b32.bc

echo libmysql.lib + >> b32.bc
echo libpq.lib + >> b32.bc
echo odbccp32.lib + >> b32.bc
echo sqlbcc58.lib + >> b32.bc


echo %bcdir%\lib\cw32.lib + >> b32.bc
echo %bcdir%\lib\import32.lib + >> b32.bc
echo %bcdir%\lib\uuid.lib + >> b32.bc
echo %bcdir%\lib\psdk\odbc32.lib + >> b32.bc
echo %bcdir%\lib\psdk\rasapi32.lib + >> b32.bc
echo %bcdir%\lib\psdk\nddeapi.lib + >> b32.bc
echo %bcdir%\lib\psdk\msimg32.lib + >> b32.bc
echo %bcdir%\lib\psdk\psapi.lib + >> b32.bc
echo %bcdir%\lib\psdk\gdiplus.lib + >> b32.bc
echo %bcdir%\lib\psdk\iphlpapi.lib, >> b32.bc





IF EXIST %1.res echo %1.res >> b32.bc

rem uncomment this line to use the debugger and comment the following one
if %GT% == gtwin %bcdir%\bin\ilink32 -Gn -Tpe -s -v @b32.bc
IF ERRORLEVEL 1 GOTO LINKERROR
if %GT% == gtgui %bcdir%\bin\ilink32 -Gn -aa -Tpe -s -v @b32.bc
IF ERRORLEVEL 1 GOTO LINKERROR
ECHO * Application successfully built *
%1
GOTO EXIT
ECHO

rem delete temporary files
@del %1.c

:COMPILEERRORS
@type comp.log
ECHO * Compile errors *
GOTO EXIT

:LINKERROR
ECHO * Linking errors *
GOTO EXIT

:SINTAX
ECHO SYNTAX: Build [Program] {-- No especifiques la extensi¢n PRG
ECHO {-- Don't specify .PRG extension
GOTO EXIT

:NOEXIST
ECHO The specified PRG %1 does not exist

:EXIT

[cul...@gmail.com]

Lucas

This is the problem, you are linking libs on wrong order

this is the correct order

FiveHC.lib
FiveHX.lib
gtgui.lib
sqlbcc58.lib
libmysql.lib
vm.lib
rtl.lib
macro.lib
pp.lib
common.lib
lang.lib
rdd.lib
dbfntx.lib
dbfnsx.lib
dbfcdx.lib
dbffpt.lib
debug.lib
pcrepos.lib
hsx.lib
hbsix.lib
ct.lib
zlib.lib
codepage.lib
cw32.lib
import32.lib
odbc32.lib

Regards
Luiz

[Lucas]

Luiz,

You missed png.lib, hbzip.lib and odbccp32.lib.

Anyway, I change as follows and I still get the GPF.

@ECHO OFF
CLS
ECHO ÚÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄ¿

ECHO ³ FiveWin for xHarbour 11.11 - Nov. 2011 xHarbour development power ³Ü

echo %fwh%\lib\Fivehc.lib %fwh%\lib\FiveHx.lib + >> b32.bc

echo %hdirl%\gtgui.lib + >> b32.bc
echo sqlbcc58.lib + >> b32.bc

echo libmysql.lib + >> b32.bc
echo libpq.lib + >> b32.bc

echo %hdirl%\vm.lib + >> b32.bc
echo %hdirl%\rtl.lib + >> b32.bc

echo %hdirl%\macro.lib + >> b32.bc

echo %hdirl%\pp.lib + >> b32.bc
echo %hdirl%\common.lib + >> b32.bc

echo %hdirl%\lang.lib + >> b32.bc

echo %hdirl%\rdd.lib + >> b32.bc
echo %hdirl%\dbfntx.lib + >> b32.bc

echo %hdirl%\dbfnsx.lib + >> b32.bc

echo %hdirl%\dbfcdx.lib + >> b32.bc
echo %hdirl%\dbffpt.lib + >> b32.bc

echo %hdirl%\debug.lib + >> b32.bc

echo %hdirl%\pcrepos.lib + >> b32.bc

echo %hdirl%\hsx.lib + >> b32.bc

echo %hdirl%\hbsix.lib + >> b32.bc

echo %hdirl%\ct.lib + >> b32.bc
echo %hdirl%\zlib.lib + >> b32.bc

echo %hdirl%\codepage.lib + >> b32.bc

echo %hdirl%\hbzip.lib + >> b32.bc
echo %hdirl%\png.lib + >> b32.bc

rem Uncomment these two lines to use Advantage RDD
rem echo %hdir%\lib\rddads.lib + >> b32.bc
rem echo %hdir%\lib\Ace32.lib + >> b32.bc

echo odbccp32.lib + >> b32.bc

[Enrico Maria Giordano]

> This is the problem, you are linking libs on wrong order
>
> this is the correct order
>
> FiveHC.lib
> FiveHX.lib

Why? The following order is the official and recommended by Fivetechsoft (it
is in their compile batch):

FiveHX.lib
FiveHC.lib

EMG

--
EMAG Software Homepage: http://www.emagsoftware.it
The EMG's ZX-Spectrum Page: http://www.emagsoftware.it/spectrum
The Best of Spectrum Games: http://www.emagsoftware.it/tbosg
The EMG Music page: http://www.emagsoftware.it/emgmusic

[cul...@gmail.com]

Lucas

Please download xharbour from http://free.xharbour.com and use the
xbuild utility to compile sqlrdd samples

Regards
Luiz

On 27 jul, 06:01, Lucas <lucasdebelt...@gmail.com> wrote:
> Luiz,
>
> You missed png.lib, hbzip.lib and odbccp32.lib.
>
> Anyway, I change as follows and I still get the GPF.
>
> @ECHO OFF
> CLS

> ECHO здддддддддддддддддддддддддддддддддддддддддддддддддддддддддддддддддддддддддддд©
> ECHO Ё FiveWin for xHarbour 11.11 - Nov. 2011          xHarbour development power Ёэ
> ECHO Ё (c) FiveTech, 1993-2011     for Microsoft Windows 9X/NT/200X/ME/XP/Vista/7 Ёш
> ECHO юддддддддддддддддддддддддддддддддддддддддддддддддддддддддддддддддддддддддддддыш
> ECHO Ъ ъъъъъъъъъъъъъъъъъъъъъъъъъъъъъъъъъъъъъъъъъъъъъъъъъъъъъъъъъъъъъъъъъъъъъъъъъъъъъ

[Lucas]

Luiz,

I can´t sorry. That build does not work with Fivewin.

Also, I need to use a 3rd party lib which is only for BCC.


In SQLRDD source code, can you show fragments of code of how the error on opening a non existant table is created?.

Thanks

[cul...@gmail.com]

Hi

The build from free.xharbour.com work with fivewin

Regards
Luiz

[Elvira]

Lucas,

I advise you stop loosing your time with such a pseudo company.

They won´t check your sample and they do not want to fix the product.

They always said: It is working for us, it is Fivewin´s fault, as you have seen.

Please, use a real alternative such as Eagle1 or TDolphin.

xHarbour.com is a scam that sells unfinished products with no support.

Run away as long as you can ;).

[Patrick Mast]

Not even worth replying to ;-))

[Elvira]

Because you don´t have either proofs or arguments.

Why don´t you just fix your products?. Or put on the web before purchasing "Beta- Under development".

[karlos...@gmail.com]

El sábado, 28 de julio de 2012 11:30:41 UTC-6, Elvira escribió:
> Because you don´t have either proofs or arguments.
>
>
>
> Why don´t you just fix your products?. Or put on the web before purchasing "Beta- Under development".

Siento no poder opinar en ingles.

disculpen el irrespeto de dirigirme en un idioma diferente a los participante.

espero que mi opinion interesanten y sea traducida, si no, no hay drama.

Yo como cliente de xharbour entiendo y confirmo que el producto xharbour.com es completamente funcional en la parte compilador, y el sqlrdd. (no asi en VXH) muchos de los problemas que se presenta en sqlrdd con los browses es relativo al manejo del skipper, scrollbar y la necesidad de saber en todo momento la cantidad de registros ( normalemente en modo local se usa la función ordkeycount()) lo cual implica que se necesita hacer un recorrido por el cursor obtenido de una consulta para saber el total de registro, lo cual provoca perdida de perfomance, es por ello que creo que xhb ha optado por retornar un valor de 0 para esta funcion, ahora todos los browse gui fallan por que ellos esperan un valor mayor a cero para calcular la posicion relativa de un registro, lo cual es irrelevante en colsultas donde las mismas consltas representan un grupo de registros. Es importante recalcar que un simple browse() en modo consola funciona correctamente en xhb con sqlrdd, esto nos da luz sobre donde puede estar el problema.

el mismo delphi no usa calculo alguno para la posicion relativa de un registro en un cursos y el thumbar de las scrollbar no funciona completamente, solamente funciona up, down, y el thum siempre esta al medio, esto cuaado el ambiente es cliente servidor.

ahora, he realizado la prueba de apertura de una tabla que no existe en un servidor y cuando en el ejemplo se incluye fwh como gui ocurre un gpf lo cual indica que algo no adecuado esta sucediendo en errorsys, adema ocurre otra situacion:

con la version mas actual de xhb y fwh actual:

en un ejemplo cualquiera, se proboca un error, no aparece la ventana de error de fwh, sino que ocurre un gpf, y es que la llamada a la funcion NEXMEM la cual es usda en errorsys esta generando un en si misma un error, haciendo recursivo la llamada a errorsys. En la version borlan no existe ese problema. esto se corrije agregando el codigo de la funcion que mas abajo adjunto en el proyecto y ahora si salta correctamente el errorsys de fwh. el porque de este error no lo se.

se que hay algunas cosas por pulir en xharbour.com por ejemplo;

La funcion API FindResource en fwh cuando se combina con xhb.com y el proyecto es grande, falla en encontrar un recursos (BITMAP) que esta incrustado en el EXE, lo interesante esque si los recursos estan en una dll, no se da el problema.

ahora con xharbou.org y bcc este error no existe, este es un error que a mi en particular me impide usar xhb.com

se que la falla no es fwh, ya que tanto antonio como daniel y yo hemos dado lo mejor para encontrar la falla, y todo apunta a un fallo en xcc. el cual pido encarecidamente actualizen, ya pellesc ya va por la version 7, y xc esta basado en 2.70

En cuanto al soporte se que Ella y Patrick, hacen lo posible pero y son muy capaces, y cuentan con mucha tecnica, en especial Ella, de la cual cada respuesta es una luz en la oscuridad, muchas veces dando mas info de la necesaria.

a ellos, mis respetos.

#pragma BEGINDUMP

#include <windows.h>
#include <hbapi.h>

HB_FUNC( NEXTMEM ) // --> nHardwareMemory
{
MEMORYSTATUSEX mst;

memset( &mst, 0, sizeof( MEMORYSTATUSEX ) );
mst.dwLength = sizeof( MEMORYSTATUSEX );

GlobalMemoryStatusEx( &mst );

hb_retnll( mst.ullTotalPhys );
}

HB_FUNC( NAVAILMEM ) // --> nHardwareMemory
{
MEMORYSTATUSEX mst;

memset( &mst, 0, sizeof( MEMORYSTATUSEX ) );
mst.dwLength = sizeof( MEMORYSTATUSEX );

GlobalMemoryStatusEx( &mst );

hb_retnll( mst.ullAvailPhys );
}
#pragma ENDDUMP

[Patrick Mast]

Thank you for your thoughts!

[Ella Stern]

Carlos:


Gracias por la respuesta documentada.

Para disfrutar de las ventajas ofrecidas por OOP, se nececitan separar totalmente los datos de la interfaz gráfica.

El hecho es que ni en ADO.NET no hay un metodo parecido con RecCount(), y el programador debe refinar la vinculación entre el origen de datos y la rejilla.

El conjunto de clases DataTable, DataSource y DataGrid diseñado por Augusto opera lo más apropiado con LINQ.

Muchas gracias por la orientación del Augusto, y por el soporte de Patrick!


Ella

[Ron Pinkas]

>> FiveHC.lib
>> FiveHX.lib
>
> Why? The following order is the official and recommended by Fivetechsoft
> (it is in their compile batch):
>
> FiveHX.lib
> FiveHC.lib

The order of these 2 libs should not make any difference as they should not
have any duplicates between them. I believe Luiz was addressing other
entries in the list, but I admit I did not check what Luiz disorder was
trying to address, nor if he is correct or not.

Ron

[Enrico Maria Giordano]

> > FiveHX.lib
> > FiveHC.lib
>
> The order of these 2 libs should not make any difference as they should
> not
> have any duplicates between them.

Right, they shouldn't.

[Elvira]

Hello Ron,

Glad to see here. It is always a pleasure to talk with one of the xHarbour fondunders. Hope also you keep participating and helping Andy to support the compiler, once Viktor and others left xHarbour forever.

First, my framework is xHarbour, latest Fivewin, BCC 5.82 and MySQL 5.x. The program has over 20 dfbs with an average of 150 records. Here I start with trouble, as Luiz and Patrick do not like Fivewin and the pre-definied reply is: “it is a Fivewin issue, claim Fivetech”.

But it is not. For testing I use the same data content and same program of course. With DBFCDX or ADSRDD I do not get any issue. Everything is working fine. No issues at all!!.

The first issue is that I can´t use xBrowse anymore. With previous SQLRDD versions it was possible. I think it has to do with Ord…. functions. That worked before and now does not work anymore. RDDADS works perfect with Ord…. And I can´t remove the browsing system of our line of software for obvious.

I do experience random GPFS with no error.log or sqlerror.txt. For example, I just go to a REPORT, starts building it with the Fivewin´s TREPORT system (I get the message of building page 1), and it crashes. Very hard to trace it. Next week it may be working. Same test with same data content in DBFCDX or RDDADS works perfect. No issues at all.
In fact, Lucas showed a sample and he was just ignored. xHarbour.com refused to test with Fivewin.
As you do not provide any source code, I can´t just trace it. I understand the internals should kept private. But the rest of the functions and modules should be available.

Next, from time to time I experience sudden slowness, not related to huge tables in particular (remember, average is 150 records). Closing the software and starting again solves it.

Data is not properly updated. For example, when I have various PCs adding a customer at almost the same time, they always get the same number:
DbGoBottom()
nNUMERO = FIELD->NUMERO+1



Also, I detected that SUM command is very slow.

I have a table with 100 fields, two records and a total size of 16 kb. Browsing, reporting it or doing a do while takes too much time. Only two records and almost 65 seconds.


Those kind of errors show that the product is not tested enough. The design concept of the product hooked me and we decided to choose it. But the truth is that the product lacks proper testing and debug.

[Ron Pinkas]

Elvira,

> Glad to see here. It is always a pleasure to talk with one of the xHarbour
> fondunders.
> Hope also you keep participating and helping Andy to support the compiler,
> once Viktor
> and others left xHarbour forever.

Thx, but Victor had never left xHarbour, as he never was involved with
xHarbour. Not sure whom are you referring to by "others", but I will be
happy to respond to specifics.

> First, my framework is xHarbour, latest Fivewin, BCC 5.82 and MySQL 5.x.
> The program has over 20 dfbs with an average of 150 records. Here I start
> with trouble,
> as Luiz and Patrick do not like Fivewin and the pre-definied reply is: “it
> is a Fivewin
>issue, claim Fivetech”.

Please try to avoid speculations, it is not that they don't like FWH, that
is not what they told you. What they did try to explain is that in our
testing we MUST EXCLUDE all 3rd party products. If you CAN NOT REPLICATE a
problem withOUT using some 3rd part product then it is fair to claim that we
can not be THE RESPONSIBLE PARTY.

Let me ask you, what would you do if one of your clients told you that your
program keep failing on machines that have let's say uTorrent, but that it
works perfectly when uTorrent is not on the machine. What would you tell
them when they tell you that NO OTHER PROGRAM has problem with uTorrent only
your program so it must be that your product is defective. When you try to
explain that as they know your product works correctly without uTorrent so
that maybe the problem is with uTorrent, they tell you that you have a
predefined answer that you don't like uTorrent. Even if you did have the
resources and the interest to test with uTorrent what could you really do if
you do not have their source code?

> But it is not. For testing I use the same data content and same program
> of course. With
> DBFCDX or ADSRDD I do not get any issue. Everything is working fine. No
> issues at
> all!!.

Similarly I can claim that SQLRDD works perfectly with xHarbour as you saw
yourself, everything is working fine. No issue at all!!!

> The first issue is that I can´t use xBrowse anymore. With previous SQLRDD
> versions it
> was possible. I think it has to do with Ord…. functions. That worked
> before and now
> does not work anymore. RDDADS works perfect with Ord…. And I can´t remove
> the
> browsing system of our line of software for obvious.

What is the exact Error Message? Have you looked at xBrowse code and tried
to ISOLATE the problem?

If any Ord*() is not working properly it should be very easy to demonstrate
in a reduced sample. Obviously, you don't have to involve FWH in order to
test any of the functionality of a DB product.

> I do experience random GPFS with no error.log or sqlerror.txt. For
> example, I just go to a
> REPORT, starts building it with the Fivewin´s TREPORT system (I get the
> message of
> building page 1), and it crashes. Very hard to trace it. Next week it may
> be working.
> Same test with same data content in DBFCDX or RDDADS works perfect. No
> issues at

> all.In fact, Lucas showed a sample and he was just ignored. xHarbour.com

> refused to
> test with Fivewin.

Elvira, please try to stay focused. SQLRDD is an RDD, it doesn't require any
GUI to TEST its functionality. If something is not working properly, or is
broken then it should be visible withOUT FWH as well!!! All you or Lucas
have to do is the BASIC DUE DILIGENCE of creating a SMALL, REDUCED, SELF
CONTAINED, sample. This is the most efficient way to find bugs anyway, and I
practice it myself everyday that I program.

> As you do not provide any source code, I can´t just trace it. I understand
> the internals
> should kept private. But the rest of the functions and modules should be
> available.

Well this is a completely unrelated issue. Please let's focus on the
TECHNICAL issues.

> Next, from time to time I experience sudden slowness, not related to huge
> tables in
> particular (remember, average is 150 records). Closing the software and
> starting again
> solves it.

If this was an SQLRDD problem then clearly this should be easy to reproduce
in a reduced sample. Please think as a programmer. the RDD doesn't know nor
care what you use for GUI, and if it is SLOW then it must be slow also
withOUT FWH.

> Data is not properly updated. For example, when I have various PCs adding
> a
> customer at almost the same time, they always get the same number:
> DbGoBottom()
> nNUMERO = FIELD->NUMERO+1

Here is a GREAT EXAMPLE of BAD CODE, which demonstrates that you are not yet
aware of some satellites of problems and solutions for Concurrent Usage of
DB, by multiple users. The above code is simply not M/U safe. Instead some
sort of granulated, concurrency protected logic, should be used. You should
not code based on ASSUMPTIONs. Instead you should create code that
EXPLICITLY address the issue at end. For example:

OBTAIN EXCLUSIVE ACCESS to some resource.
OBTAIN VALUE and INCREMENT Numerator.
RELEASE EXCLUSIVE ACCESS

The above pseudo code is the only way to guarantee functionality of
INCREMENTAL NUMERATORs under concurrency demands.

> Also, I detected that SUM command is very slow.

Surely it should be easily demonstrated with a REDUCED, SELF CONTAINED,
sample. Please provide a sample to support your observation.

> I have a table with 100 fields, two records and a total size of 16 kb.
> Browsing, reporting
> it or doing a do while takes too much time. Only two records and almost 65
> seconds.

Surely it should be easily demonstrated with a REDUCED, SELF CONTAINED,
sample. Please provide a sample to support your observation.

> Those kind of errors show that the product is not tested enough. The
> design concept of
> the product hooked me and we decided to choose it. But the truth is that
> the product
> lacks proper testing and debug.

Let me disagree. I believe that your observation are based on intuition and
emotions rather then verifiable technical review. Since the product is in
daily use by many large operations it is fair to assume that your
observations are simply incorrect.

Ron

[Elvira]

Ron,

>Please try to avoid speculations, it is not that they don't like FWH, that
>is not what they told you. What they did try to explain is that in our
>testing we MUST EXCLUDE all 3rd party products. If you CAN NOT REPLICATE a
>problem withOUT using some 3rd part product then it is fair to claim that we
>can not be THE RESPONSIBLE PARTY.

Curious, with Fivewin and RDDADS, the same code, no problem.
More clear is impossible, except one no whant to face it,

Let me ask you, what would you do if one of your clients told you that your
program keep failing on machines that have let's say uTorrent, but that it
works perfectly when uTorrent is not on the machine.

Sorry, FWH is not uTorrent.

But if I work with Firefox and a customer says "when I open Crome, I get an error", surely I will check with Crome.
All are excuses. Patrick uses FWH for Winfakt. It is a question of egos, sorry, Fivetech against xVH, Linares against you.

>Similarly I can claim that SQLRDD works perfectly with xHarbour as you saw
>yourself, everything is working fine.

Yes, there are many issues. Also, they are confirmed by other people, even in this post.

>What is the exact Error Message? Have you looked at xBrowse code and tried
>to ISOLATE the problem?

No errors. The behaviour is incorrect as OrdKeyNo() always returns 0 in SQLRDD.

> I do experience random GPFS with no error.log or sqlerror.txt. For
> example, I just go to a
> REPORT, starts building it with the Fivewin´s TREPORT system (I get the
> message of
> building page 1), and it crashes. Very hard to trace it. Next week it may
> be working.
> Same test with same data content in DBFCDX or RDDADS works perfect. No
> issues at
> all.In fact, Lucas showed a sample and he was just ignored. xHarbour.com
> refused to
> test with Fivewin.

>Elvira, please try to stay focused. SQLRDD is an RDD, it doesn't require any
>GUI to TEST its functionality. If something is not working properly, or is
>broken then it should be visible withOUT FWH as well!!!

Same code, same GUI, same content, no problem at all with DBFCDX or RDDADS. Again it is talking with a wall, sorry,

>All you or Lucas
>have to do is the BASIC DUE DILIGENCE of creating a SMALL, REDUCED, SELF
>CONTAINED, sample. This is the most efficient way to find bugs anyway, and I
>practice it myself everyday that I program.

If those bugs were possible to trace, I would not be here complaining.

>> As you do not provide any source code, I can´t just trace it. I understand
>> the internals
>> should kept private. But the rest of the functions and modules should be
>> available.

>Well this is a completely unrelated issue. Please let's focus on the
>TECHNICAL issues.

It i very very connected.

How can I trace this?
#Include "SQLRDD.Ch"
#Include "MySQL.Ch"
#include "dbinfo.ch"

REQUEST SQLRDD, SR_MYSQL
REQUEST DbfCDX, DbfFpt

FUNCTION Main()

Local cServer:= "dolphintest.sitasoft.net"
Local cDatabase:="dolphin_man"
Local cUser:="test_dolphin"
Local cPassword:= "123456"
Local cConnection, nConnection := 0
local cInfo



RddSetDefault("SQLRDD")

cConnection:="HST="+cServer+";UID="+cUser+";PWD="+cPassword+";DTB="+cDatabase
nConnection:=SR_AddConnection(CONNECT_MYSQL,cConnection)

IF nConnection <= 0
? "No se puede establecer la conexión."
QUIT
ENDIF

SR_UseDeleteds(.F.)

USE COLEGIOS NEW //It does exsist but not created by SQLRDD dbcreate

USE COLEGIOS_noexist NEW //It does not exist

? "test here"

DBGOTOP()

DbCloseAll()

SR_End()

RETURN (NIL)
//-------------------------------------------------------------------------------------------------------

>> Next, from time to time I experience sudden slowness, not related to huge
>> tables in
>> particular (remember, average is 150 records). Closing the software and
>> starting again
>> solves it.

>If this was an SQLRDD problem then clearly this should be easy to reproduce
>in a reduced sample.

No, it is not. It is random. Not connected with a specific table or data. It has to do with internal poor manage by SQLRDD

>> Data is not properly updated. For example, when I have various PCs adding
>> a
>> customer at almost the same time, they always get the same number:
>> DbGoBottom()
>> nNUMERO = FIELD->NUMERO+1

>Here is a GREAT EXAMPLE of BAD CODE, which demonstrates that you are not yet
>aware of some satellites of problems and solutions for Concurrent Usage of
>DB, by multiple users. The above code is simply not M/U safe. Instead some
>sort of granulated, concurrency protected logic, should be used. You should
>not code based on ASSUMPTIONs. Instead you should create code that
>EXPLICITLY address the issue at end. For example:

>OBTAIN EXCLUSIVE ACCESS to some resource.
>OBTAIN VALUE and INCREMENT Numerator.
>RELEASE EXCLUSIVE ACCESS

Well, the table is in shared mode and can´t be opened in EXCLUSIVE.
Yes, the solution is to create an independant table to store counters. But SQLRDD does not refresh and update content propelly. ADS does it.

>> Those kind of errors show that the product is not tested enough. The
>> design concept of
>> the product hooked me and we decided to choose it. But the truth is that
>> the product
>> lacks proper testing and debug.

>Let me disagree. I believe that your observation are based on intuition and
>emotions rather then verifiable technical review. Since the product is in
>daily use by many large operations it is fair to assume that your
>observations are simply incorrect.

My results are my own experience using the product and producing a very hard debug as we do with our software to achieve the highest quality and almost no bugs.

Also, talking with other users of SQLRDD confirmed my results: the product is not finished, many bugs are pending, the support is not as good as the one provided by Fiveteh and they do not test their product, if so they would have discovered those issues before releasing the product.

If you want more testimonials, just go to fivetechsoft.com/forums

[Elvira]

[Ron Pinkas]

>>Please try to avoid speculations, it is not that they don't like FWH, that
>>is not what they told you. What they did try to explain is that in our
>>testing we MUST EXCLUDE all 3rd party products. If you CAN NOT REPLICATE a
>>problem withOUT using some 3rd part product then it is fair to claim that
>>we
>>can not be THE RESPONSIBLE PARTY.
>
> Curious, with Fivewin and RDDADS, the same code, no problem.
> More clear is impossible, except one no whant to face it,

If you accept that logic as solid then why not the opposite? What if I said:
Curious, withOUT Fivewin using SQLRDD, the same code, no problem.
More clear is impossible, except one does not want to face it,

> > Let me ask you, what would you do if one of your clients told you that
> > your
> > program keep failing on machines that have let's say uTorrent, but that
> > it
> > works perfectly when uTorrent is not on the machine.
>
> Sorry, FWH is not uTorrent.

The issue is NOT the quality of FWH or any other 3rd party product. Feel
free to replace uTorrent with any other name, and then PLEASE ANSWER my
question.

> But if I work with Firefox and a customer says "when I open Crome, I get
> an error", surely
> I will check with Crome.
> All are excuses. Patrick uses FWH for Winfakt. It is a question of egos,
> sorry, Fivetech
> against xVH, Linares against you.

What a sad display of prejudice. FYI I believe FWH is a great product, and
Antonio is a great programmer. The only ego problem I see so far is yours.

>>Similarly I can claim that SQLRDD works perfectly with xHarbour as you saw
>>yourself, everything is working fine.
>
> Yes, there are many issues. Also, they are confirmed by other people, even
> in this post.

Then PLEASE provide a reduced, self contained. sample - it will speak far
louder and clearer then anything else you or I may say.

>>What is the exact Error Message? Have you looked at xBrowse code and tried
>>to ISOLATE the problem?
>
> No errors. The behaviour is incorrect as OrdKeyNo() always returns 0 in
> SQLRDD.

I believe this to be a documented limitation of the SQL backend engine.

>> I do experience random GPFS with no error.log or sqlerror.txt. For
>> example, I just go to a

>> REPORT, starts building it with the Fivewin愀 TREPORT system (I get the

>> message of
>> building page 1), and it crashes. Very hard to trace it. Next week it may
>> be working.
>> Same test with same data content in DBFCDX or RDDADS works perfect. No
>> issues at
>> all.In fact, Lucas showed a sample and he was just ignored. xHarbour.com
>> refused to
>> test with Fivewin.
>
>>Elvira, please try to stay focused. SQLRDD is an RDD, it doesn't require
>>any
>>GUI to TEST its functionality. If something is not working properly, or is
>>broken then it should be visible withOUT FWH as well!!!
>
> Same code, same GUI, same content, no problem at all with DBFCDX or
> RDDADS.
> Again it is talking with a wall, sorry,

Same code, NO GUI, same content, no problem with SQLRDD. So how come you
don't see that talking to you is indeed a practice in patience. You keep
repeating the same emotional, non technical, baseless assertions, and yet
you keep believing that I am at fault.

>>All you or Lucas
>>have to do is the BASIC DUE DILIGENCE of creating a SMALL, REDUCED, SELF
>>CONTAINED, sample. This is the most efficient way to find bugs anyway, and
>>I
>>practice it myself everyday that I program.
>
> If those bugs were possible to trace, I would not be here complaining.

I did not expect you to trace the bugs, only to DEMONSTRATE them.

Ron

[Claudio Herszage]

Elvira

It should take you no more than 2 minutes to test your routine without FWH, did you do it?
I guess not...

> DbGoBottom()
> nNUMERO = FIELD->NUMERO+1

Learning the basics of programming will take you a little more than 2 minutes, but you can anyway give it a try.

Claudio H

[karlos...@gmail.com]

This code run on in local and network, but is not recommend in client/server
envieroment dbgobottom() skip the cursors or query record by record from top to bottom, is slow exterm.

DbGoBottom()
nNUMERO = FIELD->NUMERO+1

sorry for my very bad english.

salu2
carlos vargas

[Elvira]

Claudio,

Here you have a reduced sample that focus the problem of not refreshing content. Very simple to show it.

Of course, the solution is to create a conters dbf. But ADS or DBFCDX or Leto does not require it.



Ron,

Abot the error at Report, it was due that using a picture minor than length of the field produces a GPF instead of ******** as normal DBCFX or ADS.

Maybe this is the correct behaviour of a buggy product ;).

[Ron Pinkas]

> Here you have a reduced sample that focus the problem of not refreshing
> content. Very simple to show it.
>
> Of course, the solution is to create a conters dbf. But ADS or DBFCDX or
> Leto does not require it.

Sorry you are mistaken, the code you provided is NOT guaranteed to work on
any RDD, because it has NO CONCURRENCY PROTECTION!

> Abot the error at Report, it was due that using a picture minor than
> length of the field produces a GPF instead of ******** as normal DBCFX or
> ADS.
>
> Maybe this is the correct behaviour of a buggy product ;).

Here you go again trying to be offensive instead of doing the more
reasonable, most simple, and constructive thing to do, which is to post a
REDUCED, SELF CONTAINED, SAMPLE.

Ron

[Elvira]

> Here you go again trying to be offensive instead of doing the more
>
> reasonable, most simple, and constructive thing to do, which is to post a
>
> REDUCED, SELF CONTAINED, SAMPLE.

Sorry, you ignored previous samples.

I did the report with Fivewin and I was able also to track with Fivewin.

I can´t lost my time building samples that will be ignored.

It was an SQLRDD error, as it does not handle pictures lower than field size.

[Elvira]

> Here you go again trying to be offensive instead of doing the more
>
> reasonable, most simple, and constructive thing to do, which is to post a
>
> REDUCED, SELF CONTAINED, SAMPLE.



Sorry, you ignored previous samples.

I did the report with Fivewin and I was able also to track and solve with Fivewin.

I can´t lost my time building samples that will be ignored. As you said before, Fivewin is like emule. You also could say it is a virus ;)

[Patrick Mast]

Hey Elvira,

> > Here you go again trying to be offensive instead of doing the more
> > reasonable, most simple, and constructive thing to do, which is to post a
> > REDUCED, SELF CONTAINED, SAMPLE.

> Sorry, you ignored previous samples.
> I did the report with Fivewin and I was able also to track and solve with Fivewin.

And I was able to show you that without FiveWin linked in, the error showed up normally.

> I can´t lost my time building samples that will be ignored. As you said before, Fivewin is like emule.
> You also could say it is a virus ;)

We did not ignore it, I took MY time to proof to you that it works without linkin in Fivewin. YOU seem to ignore the FACTS.

> It was an SQLRDD error, as it does not handle pictures lower than field size.

Hmm, is this something new you are talking about? Any samples to show it?

Thanks.

Patrick

[Ron Pinkas]

Elvira,

>> Here you go again trying to be offensive instead of doing the more
>>
>> reasonable, most simple, and constructive thing to do, which is to post a
>>
>> REDUCED, SELF CONTAINED, SAMPLE.
>
>
>
> Sorry, you ignored previous samples.

Please don't misrepresent the truth, I NEVER ignored any sample you posted.
Rather it is you that REPEATEDLY ignored our advice and requests to provide
a REDUCED, SELF CONTAINED, SAMPLE that does NOT involve any 3rd party code,
nor any C code.

> I did the report with Fivewin and I was able also to track with Fivewin.
>
> I can�t lost my time building samples that will be ignored.

1. As per above no valid sample has ever been ignored.
2. What gives you the right expect US to waste our time, doing the minimal
good faith work that should be done by you?

> It was an SQLRDD error, as it does not handle pictures lower than field
> size.

This is another example of your lacking background when it comes to systems
and programming concepts. Let me help you understand.

1. An RDD is NOT INVOLVED in ANYWAY in the implementation of the PICTURE
clause.
2. An RDD ONLY returns a certain VALUE, and/or perform a certain DB
OPERATION in response to a specific RDD FUNCTION.
3. The PICTURE functionality does NOT KNOW, neither cares HOW or WHERE it
got the VALUE it is supposed to process.
4. If you experience any problem or error in the processing of a PICTURE it
can NOT be directly related to ANY RDD OPERATION.

Now please try to PUT ASIDE your obvious animosity, and READ CAREFULLY my
reply, and try to process the info I provided using your LOGIC, and then
please lets work together to resolve the REAL programming problems you
experienced.

Ron

[toy...@gmail.com]

Anybody,

Just sharing my thoughts on these conundrums.

We bought SQLRDD way back in 2006. We had some issues that even without the explicit help from the creators, we managed to solve them. Some are GPFs triggered by some data that weren't there in the beginning. But we still managed to solve them or have a workaround solution.

We have different implementation on our use of SQLRDD. So, basically, our sharing with the issues might not be significant.

Every great product comes great investments from the creators. Since our purpose for the SQLRDD is porting the existing system, we did not plan to use them for a very long time, so, we shoot buying the Enterprise Edition. That way, we help in a little way to at least share with the development cost.

But, until now, we are still using SQLRDD/xHarbour/Fivewin combo because we think that it had proven somehow to be a good product.

For future plans? We still will use SQLRDD/xHarbour/Fivewin TOGETHER with other good products for the web. I'm afraid that if SQLRDD/xHarbour will not be there in the near future, say 20 years from now, the new programmers will not be in a difficult position to continue maintaining the system we left them.

As of writing this, I still have a good and positive commendation to SQLRDD.

Thank you.

Toyet