Re: [RDD] Strange Grid Behaviour

2013-12-05 Thread Cowboy
On Thursday 05 December 2013 11:41:36 am Fred Gleason wrote:
> An arguably better design

 There's a design ?

 Shirley, you jest !?

 :)

-- 
Cowboy

http://cowboy.cwf1.com

Boren's Laws:
(1) When in charge, ponder.
(2) When in trouble, delegate.
(3) When in doubt, mumble.

___
Rivendell-dev mailing list
Rivendell-dev@lists.rivendellaudio.org
http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev


Re: [RDD] Strange Grid Behaviour

2013-12-05 Thread Fred Gleason
On Dec 5, 2013, at 11:25 58, Wayne Merricks  
wrote:

> Potentially to "fix" this you have a few approaches:
> 
> * Set the initial SQL create statements to include a MyISAM table create 
> statement for the SERVICES table
> * Refactor the SERVICES table so that there isn't 248 columns per record.

Which would also break SQL portability (which I guess is not really an issue, 
as I am aware of no one running RD with an alternative RDBMS).


> Refactoring is probably the way to go but its not a small amount of work.  
> Off the top of my head you could separate off all 168 clock fields into their 
> own table, something like:
> 
> CLOCK_NAME: Name of the clock, same as CLOCK1 - 168 at the moment
> SLOT: value 1 - 168 to replace the CLOCK1 - 168 columns
> SERVICE: Name of the service this clock belongs to

An arguably better design in terms of data normalization.  As you observe 
though, it wouldn’t be a trivial amount of work.  :)

Cheers!


|-|
| Frederick F. Gleason, Jr. |   Chief Developer   |
|   |   Paravel Systems   |
|-|
|  All science is either physics or stamp collecting. |
|  -- Ernest Rutherford   |
|-|

___
Rivendell-dev mailing list
Rivendell-dev@lists.rivendellaudio.org
http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev


Re: [RDD] Strange Grid Behaviour

2013-12-05 Thread Wayne Merricks

Hi,

My understanding is something to do with a row length limit in InnoDB.  
If I remember correctly its because MySQL 5.5 ships with InnoDB by 
default as opposed to MyISAM.


InnoDB has advantages in that it is supposed to be more robust in terms 
of not losing data in weird circumstances (power outages etc), hence the 
reason for the switch.


Potentially to "fix" this you have a few approaches:

* Set the initial SQL create statements to include a MyISAM table 
create statement for the SERVICES table
* Refactor the SERVICES table so that there isn't 248 columns per 
record.


Refactoring is probably the way to go but its not a small amount of 
work.  Off the top of my head you could separate off all 168 clock 
fields into their own table, something like:


CLOCK_NAME: Name of the clock, same as CLOCK1 - 168 at the moment
SLOT: value 1 - 168 to replace the CLOCK1 - 168 columns
SERVICE: Name of the service this clock belongs to

Sounds simple but the amount of stuff you'd need to change (as a guess, 
I've not looked into the code):


* Log Generation
* Log Manager for the Grids
* Install SQL table creation for SERVICES and a new table

Theres probably a lot more that I've not thought about.

By the way, even with this change SERVICES will still have 80 columns.  
I don't like that for readability but theres no real reason to worry 
about it.  I quite like the way the Wordpress options table is 
structured with the whole: id, setting name, setting value.  Plus from a 
programming view you can just loop through the results and shove them in 
a hash map for easier retrieval (instead of having a big array and 
having to remember element 150 = clock 69.


Again this would make no difference other than to satisfy my own need 
for weird ways of structuring/coding things.


Regards,

Wayne



On 2013-12-05 04:30, Karl Koscher wrote:
Yeah, Ive ran in to this before as well. Which raises a point: why 
are

service grids stored with clock names instead of IDs?

On Wed, Dec 4, 2013 at 4:13 PM, Pedro Picoto  wrote:


I suffered the same a few weeks ago. Wayne Merricks suggested me
this:

Check that your SERVICES table is set to MyISAM and not INNODB
(which is the default I think in Mysql 5.5+)

$ mysql -h rivserver_ip_or_name -u root -p
$> use Rivendell;
$> show create table SERVICES;

Youll get a big mess of SQL splurged out, scroll down to here:

`CLOCK166` char(64) DEFAULT NULL,
`CLOCK167` char(64) DEFAULT NULL,
PRIMARY KEY (`NAME`)
) ENGINE=MyISAM DEFAULT CHARSET=latin1 |

If its not MyISAM try this:

$> alter table SERVICES ENGINE=MyISAM;

Then redo your grids and see if it sticks.

IT WORKED!

Backup the DB prior doing anything!!!

___
Rivendell-dev mailing list
Rivendell-dev@lists.rivendellaudio.org [1]
http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev
[2]




Links:
--
[1] mailto:Rivendell-dev@lists.rivendellaudio.org
[2] http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev
[3] mailto:pedro.pic...@gmail.com


___
Rivendell-dev mailing list
Rivendell-dev@lists.rivendellaudio.org
http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev


Re: [RDD] Strange Grid Behaviour

2013-12-05 Thread Fred Gleason
On Dec 4, 2013, at 23:30 02, Karl Koscher  wrote:

> Which raises a point: why are service grids stored with clock names instead 
> of IDs?

And this is a problem why?

Cheers!


|-|
| Frederick F. Gleason, Jr. |   Chief Developer   |
|   |   Paravel Systems   |
|-|
|  All science is either physics or stamp collecting. |
|  -- Ernest Rutherford   |
|-|

___
Rivendell-dev mailing list
Rivendell-dev@lists.rivendellaudio.org
http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev


Re: [RDD] Strange Grid Behaviour

2013-12-04 Thread Karl Koscher
Yeah, I've ran in to this before as well. Which raises a point: why are
service grids stored with clock names instead of IDs?


On Wed, Dec 4, 2013 at 4:13 PM, Pedro Picoto  wrote:

>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
> *I suffered the same a few weeks ago. Wayne Merricks suggested me this:
> Check that your SERVICES table is set to MyISAM and not INNODB (which is
> the default I think in Mysql 5.5+) $ mysql -h rivserver_ip_or_name -u root
> -p $> use Rivendell; $> show create table SERVICES; You'll get a big mess
> of SQL splurged out, scroll down to here: `CLOCK166` char(64) DEFAULT NULL,
> `CLOCK167` char(64) DEFAULT NULL, PRIMARY KEY (`NAME`) ) ENGINE=MyISAM
> DEFAULT CHARSET=latin1 | If its not MyISAM try this: $> alter table
> SERVICES ENGINE=MyISAM; Then redo your grids and see if it sticks. *
> *It worked!*
>
> Backup the DB prior doing anything!!!
>
>
> ___
> Rivendell-dev mailing list
> Rivendell-dev@lists.rivendellaudio.org
> http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev
>
>
___
Rivendell-dev mailing list
Rivendell-dev@lists.rivendellaudio.org
http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev


[RDD] Strange Grid Behaviour

2013-12-04 Thread Pedro Picoto
*I suffered the same a few weeks ago. Wayne Merricks suggested me this:
Check that your SERVICES table is set to MyISAM and not INNODB (which is
the default I think in Mysql 5.5+) $ mysql -h rivserver_ip_or_name -u root
-p $> use Rivendell; $> show create table SERVICES; You'll get a big mess
of SQL splurged out, scroll down to here: `CLOCK166` char(64) DEFAULT NULL,
`CLOCK167` char(64) DEFAULT NULL, PRIMARY KEY (`NAME`) ) ENGINE=MyISAM
DEFAULT CHARSET=latin1 | If its not MyISAM try this: $> alter table
SERVICES ENGINE=MyISAM; Then redo your grids and see if it sticks. *
*It worked!*

Backup the DB prior doing anything!!!
___
Rivendell-dev mailing list
Rivendell-dev@lists.rivendellaudio.org
http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev


[RDD] Strange Grid Behaviour

2013-12-04 Thread Civl Tech

Hi All,

I've ran into a strange issue using grids in RDlogmanager. I set up a 
grid for generating a log for filling every day, and hour, with music. 
After about a week several clocks have become 'emptied'. Adding a clock 
into the newly vacated block will add a clock. But reopening the Edit 
Grid dialogue shows the hour still vacant.
I've read through the operations guide, and I haven't found any leads as 
to what might be causing this? Any ideas?


- Kurtis
___
Rivendell-dev mailing list
Rivendell-dev@lists.rivendellaudio.org
http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev