Thread: pgAdmin III

pgAdmin III

From
"Donald Fraser"
Date:
Hi Dave & pgAdmin III developers.
Just wanted to know if you wanted reports back about bugs in pgAdmin III.
I haven't seen any been reported or may be I just missed them?
 
Regards
Donald Fraser
 

Re: pgAdmin III

From
"Dave Page"
Date:
Hi Donald,
 
Yes please, we'd love to hear of any so we can get a release out soon!
 
IIRC you're quite good at finding them as well :-)
 
Regards, Dave.
-----Original Message-----
From: Donald Fraser [mailto:demolish@cwgsy.net]
Sent: 19 June 2003 15:56
To: [pgADMIN]
Subject: [pgadmin-support] pgAdmin III

Hi Dave & pgAdmin III developers.
Just wanted to know if you wanted reports back about bugs in pgAdmin III.
I haven't seen any been reported or may be I just missed them?
 
Regards
Donald Fraser
 

Re: pgAdmin III

From
Sven Köhler
Date:
> Just wanted to know if you wanted reports back about bugs in pgAdmin III.
> I haven't seen any been reported or may be I just missed them?

So i should add one:
If i try to view a table with pgAdminIII (by klicking on the sunglasses
or by doing a "select *" in the SQL window) pgAdmin crashes.
The crash is caused a memory access violation.

I'm using WinXP with postgresql under cygwin.
My DB is a Unicode DB, although i think, that it used to happen with an
ASCII DB to.

I tested
http://www.pgadmin.org/snapshots/pgadmin3-Alpha-Win32.zip
and
http://www.pgadmin.org/snapshots/win32/pgadmin3-20030619-Win32.zip

Both version crash :-(

BTW:
Does pgAdmin use the ODBC driver anymore? or is are client libs compiled
in statically?


Re: pgAdmin III

From
"Dave Page"
Date:

> -----Original Message-----
> From: Sven Köhler [mailto:skoehler@upb.de]
> Sent: 19 June 2003 17:17
> To: pgadmin-support@postgresql.org
> Subject: Re: [pgadmin-support] pgAdmin III
>
>
> > Just wanted to know if you wanted reports back about bugs
> in pgAdmin
> > III. I haven't seen any been reported or may be I just missed them?
>
> So i should add one:
> If i try to view a table with pgAdminIII (by klicking on the
> sunglasses
> or by doing a "select *" in the SQL window) pgAdmin crashes.
> The crash is caused a memory access violation.

Very odd. I'm using it in preference to pgAdmin II in production now and finding 200,000 row tables barely make it
blink.Is it a specific table, or schema? Also, are there any potential trouble-making columns in your tables such as
blobsor byte arrays? 

> I'm using WinXP with postgresql under cygwin.
> My DB is a Unicode DB, although i think, that it used to
> happen with an
> ASCII DB to.
>
> I tested http://www.pgadmin.org/snapshots/pgadmin3-Alpha-Win32.zip
> and http://www.pgadmin.org/snapshots/win32/pgadmin3-20030619-Win32.zip
>
> Both version crash :-(
>
> BTW:
> Does pgAdmin use the ODBC driver anymore? or is are client
> libs compiled
> in statically?

pgAdmin III has two classes pgConn and pgSet which which roughly emulate the MDAC functionality that pgAdmin II used.
Theseclasses make use of libpq (statically linked on Windows), not ODBC (this is mainly why the query engine is so much
betternow performance wise). 

Regards, Dave.

Re: pgAdmin III

From
Sven Köhler
Date:
>>So i should add one:
>>If i try to view a table with pgAdminIII (by klicking on the
>>sunglasses
>>or by doing a "select *" in the SQL window) pgAdmin crashes.
>>The crash is caused a memory access violation.
>
> Very odd. I'm using it in preference to pgAdmin II in production now and finding 200,000 row tables barely make it
blink.Is it a specific table, or schema? Also, are there any potential trouble-making columns in your tables such as
blobsor byte arrays? 

I have absolutely _no_ idea, why pgAdminIII crashes.
The tables i try to view consist of only one simple column - it was a
test-database i created, and even with those simple tables (just an
int-column) pgAdminIII crashes.
When using the SQL window, i can see the result in the background when
the windows' standard crash-dialog pops up.
It's perhaps something than happens afterwards.

> pgAdmin III has two classes pgConn and pgSet which which roughly emulate the MDAC functionality that pgAdmin II used.
Theseclasses make use of libpq (statically linked on Windows), not ODBC (this is mainly why the query engine is so much
betternow performance wise). 

Sweet - i didn't like the dependecy on the ODBC driver (i don't like
ODBC that much). I'm also glad that pgAdminIII comes with Unicode-support.


Re: pgAdmin III

From
"Dave Page"
Date:

> -----Original Message-----
> From: Sven Köhler [mailto:skoehler@upb.de]
> Sent: 19 June 2003 21:27
> To: Dave Page
> Cc: pgadmin-support@postgresql.org; Andreas Pflug
> Subject: Re: [pgadmin-support] pgAdmin III
>
>
> >>So i should add one:
> >>If i try to view a table with pgAdminIII (by klicking on the
> >>sunglasses
> >>or by doing a "select *" in the SQL window) pgAdmin crashes.
> >>The crash is caused a memory access violation.
> >
> > Very odd. I'm using it in preference to pgAdmin II in
> production now
> > and finding 200,000 row tables barely make it blink. Is it
> a specific
> > table, or schema? Also, are there any potential
> trouble-making columns
> > in your tables such as blobs or byte arrays?
>
> I have absolutely _no_ idea, why pgAdminIII crashes.
> The tables i try to view consist of only one simple column - it was a
> test-database i created, and even with those simple tables (just an
> int-column) pgAdminIII crashes.
> When using the SQL window, i can see the result in the
> background when
> the windows' standard crash-dialog pops up.
> It's perhaps something than happens afterwards.

Can you try switching logging to Debug (under File -> Options -> Logging) and recreating the error please? That might
helpus see where the error actually occurs. Oh, and please send the log of course :-) 

Thanks, Dave.

Re: pgAdmin III

From
Sven Köhler
Date:
> Can you try switching logging to Debug (under File -> Options -> Logging) and recreating the error please? That might
helpus see where the error actually occurs. Oh, and please send the log of course :-) 

As you wish.
I connected to the DB, selected a table, turned on logging and clicked
on the sunglasses.

here's the log:

2003-06-19 23:49:52 INFO   : Destroying an options dialogue
2003-06-19 23:49:55 INFO   : Creating pgConn object
2003-06-19 23:49:55 INFO   : Server name: localhost (resolved to: 127.0.0.1)
2003-06-19 23:49:55 INFO   : Opening connection with connection string:
hostaddr=127.0.0.1 dbname=test user=Sven port=5432
2003-06-19 23:49:55 INFO   : Setting client_encoding to 'UNICODE'
2003-06-19 23:49:55 INFO   : Creating EditGrid
2003-06-19 23:49:55 INFO   : Running query SELECT Oid, * FROM
public.test ORDER BY test
2003-06-19 23:49:55 QUERY  : Thread Query SELECT Oid, * FROM public.test
ORDER BY test
2003-06-19 23:49:55 INFO   : Creating pgSet object


Re: pgAdmin III

From
Sven Köhler
Date:
> Sven, what you didn't mention is that you're probably using the German
> language setting.
> There was a translation of "%d rows." into "%s Zeilen." which is
> obviously the highway to hell.

OK - so - why did pgAdmin crash? will the code be fixed? pgAdmin
shouldn't crash just because there's something wrong in the translation.
I might guess, that you're using printf which took the integer as a
string-pointer etc.
This is very unsafe and i wouldn't know any way to fix that :-( except
writing some kind of "safe" printf-function.


Re: pgAdmin III

From
Andreas Pflug
Date:
Sven Köhler wrote:

>>>Just wanted to know if you wanted reports back about bugs in pgAdmin III.
>>>
Of course we want!!!

>>So i should add one:
>>If i try to view a table with pgAdminIII (by klicking on the 
>>sunglasses 
>>or by doing a "select *" in the SQL window) pgAdmin crashes. 
>>The crash is caused a memory access violation.
>>
Sorry guys...
Sven, what you didn't mention is that you're probably using the German 
language setting.
There was a translation of "%d rows." into "%s Zeilen." which is 
obviously the highway to hell.
A new version is uploaded to snake.

>>I tested http://www.pgadmin.org/snapshots/pgadmin3-Alpha-Win32.zip
>>and http://www.pgadmin.org/snapshots/win32/pgadmin3-20030619-Win32.zip
>>
These are identical... Alpha always points towards the latest version.

Regards,
Andreas