Disable styles in WAWalkback?

Currently whenever a walkback occurs, the WAWalkback seems to inherit the
styles from the application that caused it. How would I go about un-styling
it as it looks very odd with our table styles applied to it :)

Thanks!

-Boris

--
+1.604.689.0322
DeepCove Labs Ltd.
4th floor 595 Howe Street
Vancouver, Canada V6C 2T5

[hidden email]

CONFIDENTIALITY NOTICE

This email is intended only for the persons named in the message
header. Unless otherwise indicated, it contains information that is
private and confidential. If you have received it in error, please
notify the sender and delete the entire message including any
attachments.

Thank you.

_______________________________________________
Seaside mailing list
[hidden email]
http://lists.squeakfoundation.org/cgi-bin/mailman/listinfo/seaside

Boris Popov wrote:
> Currently whenever a walkback occurs, the WAWalkback seems to inherit the
> styles from the application that caused it. How would I go about un-styling
> it as it looks very odd with our table styles applied to it :)
>  

This also occurs in other areas such as viewing source (via halos) if I
recall..  I can do that for one page
I've got that has centered text and all source will be centered as
well..  If there's a way to disable this, I'd
love to know how.  It's more an annoyance than anything else, but I
think it would be nice if there was
a way to override this behavior..

-- Rick
_______________________________________________
Seaside mailing list
[hidden email]
http://lists.squeakfoundation.org/cgi-bin/mailman/listinfo/seaside

On Jun 20, 2006, at 2:45 PM, Rick Flower wrote:

FWIW, we tend to have all CSS qualified by some class that's on a div  
that wraps around the entire page.  So:

<div class="global">
...
</div>

and then, instead of

table {color: red}

you have

.global table {color: red}

That way if you have something like the walkback page that isn't  
inside the normal global div, none of the styles apply.
It's also easy to do themes: just switch that div to global2 and you  
can have a whole different set of CSS come into effect.

Avi
_______________________________________________
Seaside mailing list
[hidden email]
http://lists.squeakfoundation.org/cgi-bin/mailman/listinfo/seaside

How does one obtain a working GLORP on Squeak these days?

(This is only minorly Seaside-relevant.)

Jeremy

_______________________________________________
Seaside mailing list
[hidden email]
http://lists.squeakfoundation.org/cgi-bin/mailman/listinfo/seaside

I think you load up Yanni's Postgres library, then GLORP from squeakmap.


On Jun 20, 2006, at 5:56 PM, Jeremy Shute wrote:

_______________________________________________
Seaside mailing list
[hidden email]
http://lists.squeakfoundation.org/cgi-bin/mailman/listinfo/seaside

> I think you load up Yanni's Postgres library, then GLORP from squeakmap.

When I do that, edit the Postgres configuration, and edit the PGConnection
class>>buildDefaultConnectionArgs to refer to my new configuration, the
debugger jumps to PGConnection(PGActiveObject)>>processEvent:, telling me
"Error: PGConnection has a defective state machine."

It is trying to find eventName #AuthenticationMD5Password in the
stateTransitionTable.  Except the instance variable "state" it is using is
already set to #CantHappen...  I'm having a very difficult time tracking
the logic, to be honest.

Anyone else have this experience?  PostgreSQL 8.1.4-1, PostgresV2 v.1.0,
and GLORP from SqueakMap?

Jeremy

GPG PUBLIC KEY: 0xA2B36CE5

_______________________________________________
Seaside mailing list
[hidden email]
http://lists.squeakfoundation.org/cgi-bin/mailman/listinfo/seaside

Jeremy Shute wrote:
> When I do that, edit the Postgres configuration, and edit the PGConnection
> class>>buildDefaultConnectionArgs to refer to my new configuration ...

That's the right thing to do. Just as a sanity check,
make sure psql can connect to your server. That's not likely
a problem for a local server, but a remote server may need
some configuration to allow remote connections.

> "Error: PGConnection has a defective state machine."

What this means is that the underlying state machine has not
correctly accounted for all possible scenarios. In this case,
it's probably an error condition that's left things in a
bad state. I see this error message sometimes when I save an
image with an open database connection, and then restart the
image. After I have proper code to reconnect to the database
on image restart, I rarely see the message anymore.

> It is trying to find eventName #AuthenticationMD5Password in the
> stateTransitionTable.  Except the instance variable "state" it is using is
> already set to #CantHappen...

You need to load the Cryptography package off SqueakMap,
if you want to use MD5 authentication. Otherwise, you can
only use PlainText authentication.

 > I'm having a very difficult time tracking
> the logic, to be honest.

It's a state machine design. Each message received from
the postgres server is treated as an event to the PGConnection
state machine. The event causes a state transition. Upon entering
the new state, the method named like "st<StateName>" is executed.

The state machine is pictured here:
     http://ca.geocities.com/zazu@.../pgsqueak/PGConnection.5.gif
That diagram is slightly out of date, if you look in the method:
     PGConnection>>buildStateTransitionTable
you'll see a representation of the real version used by the code.

_______________________________________________
Seaside mailing list
[hidden email]
http://lists.squeakfoundation.org/cgi-bin/mailman/listinfo/seaside

I'm familiar with state machines and other automata.  The event is listed
in the debugger as PGAuthentication(type=5,salt=a ByteArray(249 138 60
148))

I tried loading Crypgography from SqueakMap, same error.  How do I get it
to authenticate in plaintext?

Jeremy

_______________________________________________
Seaside mailing list
[hidden email]
http://lists.squeakfoundation.org/cgi-bin/mailman/listinfo/seaside

Just to make sure, does your pg_hba.conf accept plaintext passwords, instead of md5 ?

Wouter



_______________________________________________
Seaside mailing list
[hidden email]
http://lists.squeakfoundation.org/cgi-bin/mailman/listinfo/seaside

Changing "md5" to "password", I get 49 tables.  It's alive (and running
tests)!  *Cackle*

What cryptographic options actually work?  Should I try "crypt" instead of
"md5"?  Was "md5" EXPECTED to work?  Is there anything I can do to help
figure out what's wrong?

Jeremy




GPG PUBLIC KEY: 0xA2B36CE5

_______________________________________________
Seaside mailing list
[hidden email]
http://lists.squeakfoundation.org/cgi-bin/mailman/listinfo/seaside

Jeremy Shute wrote:
> What cryptographic options actually work?  Should I try "crypt" instead of
> "md5"?  Was "md5" EXPECTED to work?  Is there anything I can do to help
> figure out what's wrong?

Only plaintext and md5 are implemented. Well, it's documented
now, in the mailing list :)

"md5" should work, if you have the Cryptography package loaded
prior to loading the postgres package. If you do it the other
way around, you may need to force a recompile of the postgres
package for the MD5 stuff to be recogized. IIRC, you can use
the Monticello browser to reload the package, which will cause
a recompile.

_______________________________________________
Seaside mailing list
[hidden email]
http://lists.squeakfoundation.org/cgi-bin/mailman/listinfo/seaside

> "md5" should work, if you have the Cryptography package loaded
> prior to loading the postgres package. If you do it the other
> way around, you may need to force a recompile of the postgres
> package for the MD5 stuff to be recogized. IIRC, you can use
> the Monticello browser to reload the package, which will cause
> a recompile.

Yes, I explicitly forced a recompile, because I loaded Cryptography after
PostgresV2.  Unfortunately, when switching back to "md5" encryption, it
still fails with the same error.  Hmmm...

Also, with an open transcript, I notice that GLORP has a bug, where in:

DatabaseTable>>primaryKeyUniqueConstraintName

  ^self name, '_UNIQ'.

...which should be '_PK' for PostgreSQL (which implicitly creates these,
according to the NOTIFICATION I am seeing in the Transcript).  Even then,
the error still occurs.  It seems I cannot drop primary key constraints?
Should I just ignore the errors from GLORP?

Jeremy

_______________________________________________
Seaside mailing list
[hidden email]
http://lists.squeakfoundation.org/cgi-bin/mailman/listinfo/seaside

Jeremy Shute wrote:
> Yes, I explicitly forced a recompile, because I loaded Cryptography after
> PostgresV2.  Unfortunately, when switching back to "md5" encryption, it
> still fails with the same error.  Hmmm...
>
> Also, with an open transcript, I notice that GLORP has a bug, where in:

If the unit tests for postgres work when you enable "md5",
then the postgres connection is working.

Can you give a recap of what problems you're having?
I thought you'd said you could connect with "plaintext",
but not "md5" (even after loading Cryptography). But your
last message seems to be saying you're having problems
with Glorp when using "md5" that don't appear when you
use "plaintext". Code snippets and (trimmed) stack dumps
will help to understand what you're trying to do, and
what errors you're seeing.

_______________________________________________
Seaside mailing list
[hidden email]
http://lists.squeakfoundation.org/cgi-bin/mailman/listinfo/seaside

PROBLEM 1
=========

I mentioned a problem with the state machine previously, when the database
is set up in "md5" mode.  This error occurred both before and after
downloading Cryptography and recompiling PostgresV2 and trying to connect
in "md5" mode.  Therefore, everything I say from this point forward is
qualified by the fact that connections are made in "plaintext" mode...



PROBLEM 2
=========

reset
        | session admin |
        session := self newSession.

        "First, we would like to drop all the database tables.  This destroys the
data!!!"
        session accessor dropTables: ((TableSorter for: (session system)
allTables) sort reversed).

        "Now, we would like to recreate all the tables."
        session inTransactionDo: [
                session system allTables do: [:each |
                        session accessor createTable: each ifError: [:error |
                                Transcript show: error messageText
                        ]
                ]
        ].

        "Now, we would like to ensure that at least one login exists."
        admin := TestLogin new.
        admin
                login: 'my_admin_name';
                password: 'my_admin_password'.
        session beginUnitOfWork.
        session register: admin.
        session commitUnitOfWork



Sending this message appears to succeed from the Workspace, except that it
prints the following in the transcript:

ERROR:  constraint "login_uniq" does not exist
[PGConnection(1277)] NOTIFICATION: PGNoticeResponse(value='NOTICE:  CREATE
TABLE / PRIMARY KEY will create implicit index "login_pk" for table
"login"
')

...which is highlighting the fact that Postgres is implicitly making
"login_pk" and GLORP is attempting to explicitly delete it later as
"login_uniq".  Changing the '_UNIQ' to '_PK' in the GLORP code still does
not allow the drop.  Is this expected, and something the season
professional ignores???



PROBLEM 3
=========

The above routine, despite appearing to succeed, is not inserting the row
with 'my_admin_name' and 'my_admin_password'.  This, I am still
diagnosing, because the GLORP SUnit tests share the connection string and
write to the database just fine.



Jeremy

_______________________________________________
Seaside mailing list
[hidden email]
http://lists.squeakfoundation.org/cgi-bin/mailman/listinfo/seaside

Cross PROBLEM 3 off.  Using #constructAllClasses instead of #allClassNames
was the solution to that one.  The other ones are still (minor) problems
that I'd like to help fix if I could...

Jeremy

_______________________________________________
Seaside mailing list
[hidden email]
http://lists.squeakfoundation.org/cgi-bin/mailman/listinfo/seaside

Hi all, having an issue if anyone can help.  I'm rendering a table, some of
the data necessary for the table takes a while to get because of a web
service call, so I'm using scriptalicious to do an Ajax.Updater to live
update just that portion, while rendering what I have immediately.  Problem
is, seaside seems to be making all the callback synchronously, when I need
them to be async.  Anyone know if it's possible to have multiple callbacks
executing asynchronously?

_______________________________________________
Seaside mailing list
[hidden email]
http://lists.squeakfoundation.org/cgi-bin/mailman/listinfo/seaside

Jeremy Shute wrote:
> PROBLEM 1
> =========
>
> I mentioned a problem with the state machine previously, when the database
> is set up in "md5" mode.  This error occurred both before and after
> downloading Cryptography and recompiling PostgresV2 and trying to connect
> in "md5" mode.  Therefore, everything I say from this point forward is
> qualified by the fact that connections are made in "plaintext" mode...

Dunno what could be wrong then. Maybe you can try loading into
a fresh image, just Cryptography and PostgresV2 (no Glorp yet),
to isolate the "md5" connection issue.

> PROBLEM 2
> =========
> ...which is highlighting the fact that Postgres is implicitly making
> "login_pk" and GLORP is attempting to explicitly delete it later as
> "login_uniq".  Changing the '_UNIQ' to '_PK' in the GLORP code still does
> not allow the drop.  Is this expected, and something the season
> professional ignores???

I'm not a Glorp user, so I can be of much help here. But,
my guess is that the error (constraint "login_uniq" does not exist)
is causing a transaction rollback.

_______________________________________________
Seaside mailing list
[hidden email]
http://lists.squeakfoundation.org/cgi-bin/mailman/listinfo/seaside

Jeremy Shute wrote:
> Cross PROBLEM 3 off.  Using #constructAllClasses instead of #allClassNames
> was the solution to that one.  The other ones are still (minor) problems
> that I'd like to help fix if I could...
>
>  
Jeremy -- did you ever find a solution to #2?  I'm seeing the same thing
in my updated test code
now that I've consolidated all of my descriptors -- I see the following
logged to my Postgresql
status log :

NOTICE:  CREATE TABLE will create implicit sequence
"auditlog_org_id_seq" for serial column "auditlog.org_id"
NOTICE:  CREATE TABLE / PRIMARY KEY will create implicit index
"auditlog_pk" for table "auditlog"
NOTICE:  CREATE TABLE / PRIMARY KEY will create implicit index
"categories_pk" for table "categories"

I believe these are being done during the create loop I've got for all
of my test tables (which looks pretty much
like the code you posted earlier for dropping and then re-creating
tables) and it keeps me from having ANY
tables in Postgres which then causes my unit tests to fail -- below is
the list of available tables in PSQL :

myscrip=> \d
No relations found.

Is it safe to assume that the above notices are confusing Glorp as Yanni
suggested?
_______________________________________________
Seaside mailing list
[hidden email]
http://lists.squeakfoundation.org/cgi-bin/mailman/listinfo/seaside

Actually, I see these warnings but my tables appear.  If your tables are
not appearing AND you're getting those error messages, hmmm...  :-(

I will have time to load a fresh image and try cryptography again over the
long weekend, I think.

Jeremy




GPG PUBLIC KEY: 0xA2B36CE5

_______________________________________________
Seaside mailing list
[hidden email]
http://lists.squeakfoundation.org/cgi-bin/mailman/listinfo/seaside

Jeremy Shute wrote:
> Actually, I see these warnings but my tables appear.  If your tables are
> not appearing AND you're getting those error messages, hmmm...  :-(
>
> I will have time to load a fresh image and try cryptography again over the
> long weekend, I think.
>
>  
I think I'll try removing any tables with primary keys et-al and see if
they magically start re-appearing
again..

_______________________________________________
Seaside mailing list
[hidden email]
http://lists.squeakfoundation.org/cgi-bin/mailman/listinfo/seaside