|
|
 | | From: | Paulo Pereira [APKomp Brasil] | | Subject: | How can people survive without begintrans() working properly? | | Date: | Mon, 6 Dec 2004 15:15:18 -0200 |
|
|
 | Dear Sirs
When I see many people dreaming about a lot of features like .net, linux,
etc. I only think:
--If I only could use begintrans()?
--I remember that after a lot of effort we manged to convince dBASE there
was a BIG BUG in begintran(): Under the same transaction one cannot alter
twice the value of the same field and record, using browse or grid.
This was quite straighforward in versions before 5.7. This is a so basic
feature, that I wonder what kind of priorities dBI is having.
My wish is: rational priority.
Without BEGINTRANS working well, all our programming in PLUS is stuck.
What a of a limitation you dare to have: NO USE OF BROWSE OR GRID for
serious programming. Without begintrans, there is no serious programming. Is
this asking too much? We have been waiting for years waiting...
dBI, are you going to hear us this time? It is hard to believe that this is
a limitation of BDE, Cannot people in DELPHI do this. Who knows!!!!
Believe, the question is not PLUS is good or bad, the questionn is that PLUS
is not WORKABLE yet. When the transactions do not work well, the database is
not workable yet.
I am only asking for PLUS to be workable! This is my wish: dBASE PLUS please
BE. Just what I want: dBASE PLUS to there be.... PLUS has not been born yet:
This is the reality!
In bynaries I will post again, the example (teste2.wfm) that proves that
begintran does not work.
Please, do not ask us anything about why that is not working, blah, blah...
This is a so tremendous BUG!!!
Otherwise you are going to disapoint us...
Best Regards
Paulo Pereira
P.S.: I repeat: I only want PLUS there to be.
|
|
| | From: | Marty Kay | | Subject: | Re: How can people survive without begintrans() working properly? | | Date: | Mon, 6 Dec 2004 16:10:58 -0500 |
|
|
| I found a QAID: 3677 which sounds like this bug.
Is this the one you are referring to?
- Marty Kay (dBI) -
> --If I only could use begintrans()?
>
> --I remember that after a lot of effort we manged to convince dBASE there
> was a BIG BUG in begintran(): Under the same transaction one cannot alter
> twice the value of the same field and record, using browse or grid.
>
> This was quite straighforward in versions before 5.7. This is a so basic
> feature, that I wonder what kind of priorities dBI is having.
>
> My wish is: rational priority.
>
> Without BEGINTRANS working well, all our programming in PLUS is stuck.
|
|
| | From: | - | | Subject: | Re: How can people survive without begintrans() working properly? | | Date: | Mon, 6 Dec 2004 17:00:30 -0800 |
|
|
| This bug goes back to VdB 7.0 - BeginTrans() is still limited to memory and
could only be of use with small transactions for any rollback in the data
linked set... Remember that due to RI bugs we never could really fully test
this to the extent of non linked related table dependencies....
--
-charlesm
"Marty Kay" wrote in message
news:Gq4vKg92EHA.736@news-server...
> I found a QAID: 3677 which sounds like this bug.
> Is this the one you are referring to?
>
> - Marty Kay (dBI) -
>
>
>
> > --If I only could use begintrans()?
> >
> > --I remember that after a lot of effort we manged to convince dBASE
there
> > was a BIG BUG in begintran(): Under the same transaction one cannot
alter
> > twice the value of the same field and record, using browse or grid.
> >
> > This was quite straighforward in versions before 5.7. This is a so basic
> > feature, that I wonder what kind of priorities dBI is having.
> >
> > My wish is: rational priority.
> >
> > Without BEGINTRANS working well, all our programming in PLUS is stuck.
>
>
|
|
| | From: | Paulo Pereira [APKomp Brasil] | | Subject: | Re: How can people survive without begintrans() working properly? | | Date: | Wed, 8 Dec 2004 16:12:38 -0200 |
|
|
| Yes. This wish is already in BUG report. I am sorry but I could not find
(with this new layout) where the list of bugs to be fixed is and therefore I
do not know if it is that number.
However the kernel of
my wish is to be able to work with PLUS. I see so many peolple dreaming
about such things, and I only want to be capable to do the same I could in
5.7. Believe you or not, AS A WHOLE, 5.7 is much more powerful. PLUS allow
many more advanced things and healed a lot of little bugs that appear when
working with XP, etc., but it just does not work in the basics.
I waited patiently, but I had to say this. Who knows this new Direction has
no idea about was is going on. May be they think that they have a product
that works but they do not. A product cannot fail in the very basics.
Transactions are the very basic. GRID and BROWSE do not work with
transations this is too basic... Please, if you want to confirm see
teste2.wfm in binaries...
Regards
Paulo Pereira
|
|
| | From: | Rafael Carrazco | | Subject: | Re: How can people survive without begintrans() working properly? | | Date: | Mon, 6 Dec 2004 13:34:02 -0700 |
|
|
| Hi Paulo,
more than a wish this is a bug that need to be fixed.
Rafa
"Paulo Pereira [APKomp Brasil]" wrote in message
news:y0mT3h72EHA.1700@news-server...
> Dear Sirs
>
> When I see many people dreaming about a lot of features like .net, linux,
> etc. I only think:
>
> --If I only could use begintrans()?
>
> --I remember that after a lot of effort we manged to convince dBASE there
> was a BIG BUG in begintran(): Under the same transaction one cannot alter
> twice the value of the same field and record, using browse or grid.
>
> This was quite straighforward in versions before 5.7. This is a so basic
> feature, that I wonder what kind of priorities dBI is having.
>
> My wish is: rational priority.
>
> Without BEGINTRANS working well, all our programming in PLUS is stuck.
>
> What a of a limitation you dare to have: NO USE OF BROWSE OR GRID for
> serious programming. Without begintrans, there is no serious programming.
Is
> this asking too much? We have been waiting for years waiting...
>
> dBI, are you going to hear us this time? It is hard to believe that this
is
> a limitation of BDE, Cannot people in DELPHI do this. Who knows!!!!
>
> Believe, the question is not PLUS is good or bad, the questionn is that
PLUS
> is not WORKABLE yet. When the transactions do not work well, the database
is
> not workable yet.
>
> I am only asking for PLUS to be workable! This is my wish: dBASE PLUS
please
> BE. Just what I want: dBASE PLUS to there be.... PLUS has not been born
yet:
> This is the reality!
>
> In bynaries I will post again, the example (teste2.wfm) that proves that
> begintran does not work.
> Please, do not ask us anything about why that is not working, blah,
blah...
> This is a so tremendous BUG!!!
> Otherwise you are going to disapoint us...
>
> Best Regards
> Paulo Pereira
>
> P.S.: I repeat: I only want PLUS there to be.
>
>
|
|
| | From: | Rafael Carrazco | | Subject: | Re: How can people survive without begintrans() working properly?: | | Date: | Mon, 6 Dec 2004 23:19:11 -0700 |
|
|
| people can survive eating, sleeping, excersicing etc etc
sorry , could't resist! :-D
i hope your wish/bug be fixed , Im not using begintrans now, I used, but
errors were happening so......
Rafa
"Paulo Pereira [APKomp Brasil]" wrote in message
news:y0mT3h72EHA.1700@news-server...
> Dear Sirs
>
> When I see many people dreaming about a lot of features like .net, linux,
> etc. I only think:
>
> --If I only could use begintrans()?
>
> --I remember that after a lot of effort we manged to convince dBASE there
> was a BIG BUG in begintran(): Under the same transaction one cannot alter
> twice the value of the same field and record, using browse or grid.
>
> This was quite straighforward in versions before 5.7. This is a so basic
> feature, that I wonder what kind of priorities dBI is having.
>
> My wish is: rational priority.
>
> Without BEGINTRANS working well, all our programming in PLUS is stuck.
>
> What a of a limitation you dare to have: NO USE OF BROWSE OR GRID for
> serious programming. Without begintrans, there is no serious programming.
Is
> this asking too much? We have been waiting for years waiting...
>
> dBI, are you going to hear us this time? It is hard to believe that this
is
> a limitation of BDE, Cannot people in DELPHI do this. Who knows!!!!
>
> Believe, the question is not PLUS is good or bad, the questionn is that
PLUS
> is not WORKABLE yet. When the transactions do not work well, the database
is
> not workable yet.
>
> I am only asking for PLUS to be workable! This is my wish: dBASE PLUS
please
> BE. Just what I want: dBASE PLUS to there be.... PLUS has not been born
yet:
> This is the reality!
>
> In bynaries I will post again, the example (teste2.wfm) that proves that
> begintran does not work.
> Please, do not ask us anything about why that is not working, blah,
blah...
> This is a so tremendous BUG!!!
> Otherwise you are going to disapoint us...
>
> Best Regards
> Paulo Pereira
>
> P.S.: I repeat: I only want PLUS there to be.
>
>
|
|
| | From: | Ken Mayer [dBVIPS] | | Subject: | Re: How can people survive without begintrans() working properly? | | Date: | Mon, 06 Dec 2004 09:28:15 -0800 |
|
|
| Paulo Pereira [APKomp Brasil] wrote:
> Dear Sirs
>
> When I see many people dreaming about a lot of features like .net, linux,
> etc. I only think:
>
> --If I only could use begintrans()?
Have you looked at the database object?
database.beginTrans()
database.commit()
database.rollback()
all should work.
Ken
--
/(Opinions expressed are purely my own, not those of dataBased
Intelligence, Inc.)/
*Ken Mayer* [dBVIPS]
/Golden Stag Productions/
dBASE at goldenstag dot net
http://www.goldenstag.net/GSP
http://www.goldenstag.net/dbase
|
|
| | From: | Paulo Pereira [APKomp Brasil] | | Subject: | Re: How can people survive without begintrans() working properly? | | Date: | Mon, 6 Dec 2004 16:49:05 -0200 |
|
|
| This is a multi-part message in MIME format.
------=_NextPart_000_004B_01C4DBB3.7F6FF890
Content-Type: text/plain;
charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
No they don't (when with browse or grid).
You can see in the example in binaries.
Both the old functions BEGINTRANS() and the new database.BeginTrans() do =
not
work:
One cannot modify the same value twice (getting in and out) inside the =
same transaction...
They have exactly the same behaviour with BROWSE AND GRID
respectively.
I know it is hard to believe, but that is just like that.
One thing may be unbelievable. That does not mean that it is not =
true....
Regards!
Paulo Pereira
------=_NextPart_000_004B_01C4DBB3.7F6FF890
Content-Type: text/html;
charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
charset=3Diso-8859-1">
No they don't (when with browse or =
grid).
You=20
can see in the example in binaries.
Both the old functions =
BEGINTRANS() and=20
the new database.BeginTrans() do not
work:
size=3D2>
One cannot modify the =
same value=20
twice (getting in and out) inside the same =
transaction...
They have exactly the same behaviour =
with BROWSE=20
AND GRID
respectively.
I know it is hard to believe, but that =
is just=20
like that.
One thing may be unbelievable. That does not mean that it =
is not=20
true....
Regards!
Paulo Pereira
------=_NextPart_000_004B_01C4DBB3.7F6FF890--
|
|
| | From: | Ken Mayer [dBVIPS] | | Subject: | Re: How can people survive without begintrans() working properly? | | Date: | Mon, 06 Dec 2004 13:37:47 -0800 |
|
|
| Paulo Pereira [APKomp Brasil] wrote:
> No they don't (when with browse or grid).
> You can see in the example in binaries.
> Both the old functions BEGINTRANS() and the new database.BeginTrans() do not
> work:
>
> One cannot modify the same value twice (getting in and out) inside the
> same transaction...
>
> They have exactly the same behaviour with BROWSE AND GRID
> respectively.
>
> I know it is hard to believe, but that is just like that.
> One thing may be unbelievable. That does not mean that it is not true....
I never said anything was "unbelievable" (I am not sure why you have to
put words in people's messages ... you read into things what people are
not saying) ... I just don't understand the problem. If you edit a row,
it's locked, and if you read the OLH, it says that it remains locked
until the transaction is commited or rolledback. That's working as
designed as I understand it.
Ken
--
/(Opinions expressed are purely my own, not those of dataBased
Intelligence, Inc.)/
*Ken Mayer* [dBVIPS]
/Golden Stag Productions/
dBASE at goldenstag dot net
http://www.goldenstag.net/GSP
http://www.goldenstag.net/dbase
|
|
| | From: | Paulo Pereira [APKomp Brasil] | | Subject: | Re: How can people survive without begintrans() working properly? | | Date: | Mon, 6 Dec 2004 23:08:17 -0200 |
|
|
| I did not say that you said that it was unbelivable. I just consider it very
easy to be understood like that...
1) Of course that when you are under one transaction you must be able to do
in a grid
1.1) To change a value of a cell
1.2) To change a value of another cell
1.3) To change the value of the first cell again
1.4) To change the value of the second cell again
1.5) To change the value of the first cell again
1.6) To change the value of the second cell again
.......
all this under the same transaction as it is the same user doing it.
.....
It has beem always like this in versions 5.* and that is the logic for
transactions. The same user can change the value of the same cell as many
times as he wants under the same transaction. If after all this he does
rollback, the very first value of each one is assumed.
.....
The problem of the grid and browse is that they consider one as a different
user when one returns to a former cell: THAT IS A HUGE BUG. That is against
all fundamentals. That kills any database...
I hope this time to have been clear. I apologise if my language is too
storng, but I have been waiting for so many years for a such simple thing!
Regards
Paulo Pereira
|
|
|
