Re: nd naming

classic Classic list List threaded Threaded
11 messages Options
Reply | Threaded
Open this post in threaded view
|

Re: nd naming

Robert Buels
OK, after a long hiatus let's try to get some conclusions about
separating the namespaces of table names in Chado.

The options that have been raised so far are:

1.) Add prefixes to table names, e.g. 'nd_' for tables in the natural
diversity module.  This is ugly and makes the natdiv modules stand out
like a sore thumb, but it's straightforward and cross-platform

2.) Put the natdiv module (and presumably other modules eventually)
into its own separate Postgres schema is cleaner, only slightly less
straightforward, but it's not cross-platform.  For example,
Bio::Chado::Schema runs its unit tests on a SQLite backend, and this
would break that.  I really like being able to deploy (most of) Chado to
SQLite.

3.) Do not attempt to separate the table namespaces at this point, and
just rename the natdiv assay table to avoid colliding with the MAGE
module's assay table.  This is just postponing an eventual reckoning.

After this long with no movement, I'm now voting for option 3.  I
suggest renaming the natdiv's assay table back to 'experiment', since
it's almost as good a name as 'assay'.  The rate at which modules are
added to Chado is quite slow, so perhaps on second thought this problem
of table name collisions is not such a high priority.

Does anyone else have an opinion?  I'm planning to remove the nd_ table
prefixes and rename natdiv assay to experiment.

Rob


Hilmar Lapp wrote:

>
> On May 10, 2010, at 7:44 PM, Robert Buels wrote:
>
>> If we were to decide that it's Postgres-only, I would change my tune
>> to "Pg schemas for everything!".
>
>
> I'm all for cross-platform compatibility, but in fact Chado very early
> on made decision to be PostgreSQL-only.
>
> Widening that up to other RDBMS platforms is really a rather recent
> thing. It's a positive development, but let's not forget that a lot of
> functionality is programmed in PL/pgSQL, which isn't cross-platform by
> definition.
>
> Also, are you sure that different ORMs can't translate namespace into
> schemas for one RDBMS and prefixes for another? (Bioperl-db does
> something like this actually for BioSQL, though admittedly maintaining
> this is not pretty. But then, bioperl-db uses a home-grown ORM.)
>
> Finally, there is a quite a bit of software that relies on the rule that
> the primary key or foreign key to a table is the table name plus "_id".
> If you give that up, there'll better be a lot of testing. For what gain?
>
>     -hilmar


------------------------------------------------------------------------------
This SF.net email is sponsored by Sprint
What will you do first with EVO, the first 4G phone?
Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
_______________________________________________
Gmod-schema mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gmod-schema
Reply | Threaded
Open this post in threaded view
|

Re: nd naming

Hilmar Lapp-3
I like option #3 too. Tables named identically in different modules should be reconciled or renamed so they are distinct.

-hilmar

Sent from away

On Jul 1, 2010, at 10:41 PM, Robert Buels <[hidden email]> wrote:

> OK, after a long hiatus let's try to get some conclusions about
> separating the namespaces of table names in Chado.
>
> The options that have been raised so far are:
>
> 1.) Add prefixes to table names, e.g. 'nd_' for tables in the natural
> diversity module.  This is ugly and makes the natdiv modules stand out
> like a sore thumb, but it's straightforward and cross-platform
>
> 2.) Put the natdiv module (and presumably other modules eventually)
> into its own separate Postgres schema is cleaner, only slightly less
> straightforward, but it's not cross-platform.  For example,
> Bio::Chado::Schema runs its unit tests on a SQLite backend, and this
> would break that.  I really like being able to deploy (most of) Chado to
> SQLite.
>
> 3.) Do not attempt to separate the table namespaces at this point, and
> just rename the natdiv assay table to avoid colliding with the MAGE
> module's assay table.  This is just postponing an eventual reckoning.
>
> After this long with no movement, I'm now voting for option 3.  I
> suggest renaming the natdiv's assay table back to 'experiment', since
> it's almost as good a name as 'assay'.  The rate at which modules are
> added to Chado is quite slow, so perhaps on second thought this problem
> of table name collisions is not such a high priority.
>
> Does anyone else have an opinion?  I'm planning to remove the nd_ table
> prefixes and rename natdiv assay to experiment.
>
> Rob
>
>
> Hilmar Lapp wrote:
>>
>> On May 10, 2010, at 7:44 PM, Robert Buels wrote:
>>
>>> If we were to decide that it's Postgres-only, I would change my tune
>>> to "Pg schemas for everything!".
>>
>>
>> I'm all for cross-platform compatibility, but in fact Chado very early
>> on made decision to be PostgreSQL-only.
>>
>> Widening that up to other RDBMS platforms is really a rather recent
>> thing. It's a positive development, but let's not forget that a lot of
>> functionality is programmed in PL/pgSQL, which isn't cross-platform by
>> definition.
>>
>> Also, are you sure that different ORMs can't translate namespace into
>> schemas for one RDBMS and prefixes for another? (Bioperl-db does
>> something like this actually for BioSQL, though admittedly maintaining
>> this is not pretty. But then, bioperl-db uses a home-grown ORM.)
>>
>> Finally, there is a quite a bit of software that relies on the rule that
>> the primary key or foreign key to a table is the table name plus "_id".
>> If you give that up, there'll better be a lot of testing. For what gain?
>>
>>    -hilmar
>
>
> ------------------------------------------------------------------------------
> This SF.net email is sponsored by Sprint
> What will you do first with EVO, the first 4G phone?
> Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
> _______________________________________________
> Gmod-schema mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/gmod-schema

------------------------------------------------------------------------------
This SF.net email is sponsored by Sprint
What will you do first with EVO, the first 4G phone?
Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
_______________________________________________
Gmod-schema mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gmod-schema
Reply | Threaded
Open this post in threaded view
|

Re: nd naming

Sook Jung
In reply to this post by Robert Buels
Hi Rob,

The option 3 sounds good to me too. What could be the new name for
assay table if we do go for the option 3?
Since it involves stock (phenotype/genotype assay, cross, field
collection, etc), how about 'stockassay'?

I've been looking at the schema too and I think we really need a place
to store aliases of stock. Can we add stock_synonym just like
feature_synonym? We could use stock_dbxref but synonyms not always
have appropriate db..

Thanks

Sook

On Thu, Jul 1, 2010 at 10:41 PM, Robert Buels <[hidden email]> wrote:

> OK, after a long hiatus let's try to get some conclusions about
> separating the namespaces of table names in Chado.
>
> The options that have been raised so far are:
>
> 1.) Add prefixes to table names, e.g. 'nd_' for tables in the natural
> diversity module.  This is ugly and makes the natdiv modules stand out
> like a sore thumb, but it's straightforward and cross-platform
>
> 2.) Put the natdiv module (and presumably other modules eventually)
> into its own separate Postgres schema is cleaner, only slightly less
> straightforward, but it's not cross-platform.  For example,
> Bio::Chado::Schema runs its unit tests on a SQLite backend, and this
> would break that.  I really like being able to deploy (most of) Chado to
> SQLite.
>
> 3.) Do not attempt to separate the table namespaces at this point, and
> just rename the natdiv assay table to avoid colliding with the MAGE
> module's assay table.  This is just postponing an eventual reckoning.
>
> After this long with no movement, I'm now voting for option 3.  I
> suggest renaming the natdiv's assay table back to 'experiment', since
> it's almost as good a name as 'assay'.  The rate at which modules are
> added to Chado is quite slow, so perhaps on second thought this problem
> of table name collisions is not such a high priority.
>
> Does anyone else have an opinion?  I'm planning to remove the nd_ table
> prefixes and rename natdiv assay to experiment.
>
> Rob
>
>
> Hilmar Lapp wrote:
>>
>> On May 10, 2010, at 7:44 PM, Robert Buels wrote:
>>
>>> If we were to decide that it's Postgres-only, I would change my tune
>>> to "Pg schemas for everything!".
>>
>>
>> I'm all for cross-platform compatibility, but in fact Chado very early
>> on made decision to be PostgreSQL-only.
>>
>> Widening that up to other RDBMS platforms is really a rather recent
>> thing. It's a positive development, but let's not forget that a lot of
>> functionality is programmed in PL/pgSQL, which isn't cross-platform by
>> definition.
>>
>> Also, are you sure that different ORMs can't translate namespace into
>> schemas for one RDBMS and prefixes for another? (Bioperl-db does
>> something like this actually for BioSQL, though admittedly maintaining
>> this is not pretty. But then, bioperl-db uses a home-grown ORM.)
>>
>> Finally, there is a quite a bit of software that relies on the rule that
>> the primary key or foreign key to a table is the table name plus "_id".
>> If you give that up, there'll better be a lot of testing. For what gain?
>>
>>     -hilmar
>
>
> ------------------------------------------------------------------------------
> This SF.net email is sponsored by Sprint
> What will you do first with EVO, the first 4G phone?
> Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
> _______________________________________________
> Gmod-schema mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/gmod-schema
>



--
Sook Jung, PhD
Assistant Research Professor of Bioinformatics
Dept of Horticulture and Landscape Architecture
Washington State University
45 Johnson Hall, Pullman, WA 99164-6414
Email:[hidden email]

------------------------------------------------------------------------------
This SF.net email is sponsored by Sprint
What will you do first with EVO, the first 4G phone?
Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
_______________________________________________
Gmod-schema mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gmod-schema
Reply | Threaded
Open this post in threaded view
|

Re: nd naming

Robert Buels
Sook Jung wrote:
> The option 3 sounds good to me too. What could be the new name for
> assay table if we do go for the option 3?
> Since it involves stock (phenotype/genotype assay, cross, field
> collection, etc), how about 'stockassay'?

I was thinking of changing 'assay' to 'experiment' (which I think is a
name it had at a previous point in development, but my memory is hazy
after months), since it has to do with more things than just stocks.

Other opinions?

> I've been looking at the schema too and I think we really need a place
> to store aliases of stock. Can we add stock_synonym just like
> feature_synonym? We could use stock_dbxref but synonyms not always
> have appropriate db..

This sounds OK to me, it would go in the stock module.  Not really
related to the natdiv effort, probably should just go into trunk.

Rob


------------------------------------------------------------------------------
This SF.net email is sponsored by Sprint
What will you do first with EVO, the first 4G phone?
Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
_______________________________________________
Gmod-schema mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gmod-schema
Reply | Threaded
Open this post in threaded view
|

Re: nd naming

Bob MacCallum
I'm keen to see stability in table names (for forthcoming
Bio::Chado::Schema development) as I am sure Rob is.

The name change is unfortunate, but doesn't really bother us. Maybe
the mage table could change instead (arrayassay?) :-)

cheers,
Bob
(+Seth)


On Fri, Jul 2, 2010 at 5:14 AM, Robert Buels <[hidden email]> wrote:

> Sook Jung wrote:
>> The option 3 sounds good to me too. What could be the new name for
>> assay table if we do go for the option 3?
>> Since it involves stock (phenotype/genotype assay, cross, field
>> collection, etc), how about 'stockassay'?
>
> I was thinking of changing 'assay' to 'experiment' (which I think is a
> name it had at a previous point in development, but my memory is hazy
> after months), since it has to do with more things than just stocks.
>
> Other opinions?
>
>> I've been looking at the schema too and I think we really need a place
>> to store aliases of stock. Can we add stock_synonym just like
>> feature_synonym? We could use stock_dbxref but synonyms not always
>> have appropriate db..
>
> This sounds OK to me, it would go in the stock module.  Not really
> related to the natdiv effort, probably should just go into trunk.
>
> Rob
>
>
> ------------------------------------------------------------------------------
> This SF.net email is sponsored by Sprint
> What will you do first with EVO, the first 4G phone?
> Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
> _______________________________________________
> Gmod-schema mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/gmod-schema
>

------------------------------------------------------------------------------
This SF.net email is sponsored by Sprint
What will you do first with EVO, the first 4G phone?
Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
_______________________________________________
Gmod-schema mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gmod-schema
Reply | Threaded
Open this post in threaded view
|

Re: nd naming

Robert Buels
Bob MacCallum wrote:
> The name change is unfortunate, but doesn't really bother us. Maybe
> the mage table could change instead (arrayassay?) :-)

I don't think we can change the name of an existing table right now,
unless we provided migration scripts for existing data.  Chado needs an
infrastructure for supporting schema upgrades, I think
http://search.cpan.org/perldoc?DBIx::Migration::Directories could be a
good choice for that.

So, in summary, we can't change an existing table name until we have
migration support, so we should probably change the new natdiv table's
name instead.

Rob


------------------------------------------------------------------------------
This SF.net email is sponsored by Sprint
What will you do first with EVO, the first 4G phone?
Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
_______________________________________________
Gmod-schema mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gmod-schema
Reply | Threaded
Open this post in threaded view
|

Re: nd naming

Scott Cain
Right; I think it certainly makes more sense to change the name of a
table that is still in development versus a table that has been in the
core for several years :-)

Scott



On Fri, Jul 2, 2010 at 12:23 PM, Robert Buels <[hidden email]> wrote:

> Bob MacCallum wrote:
>> The name change is unfortunate, but doesn't really bother us. Maybe
>> the mage table could change instead (arrayassay?) :-)
>
> I don't think we can change the name of an existing table right now,
> unless we provided migration scripts for existing data.  Chado needs an
> infrastructure for supporting schema upgrades, I think
> http://search.cpan.org/perldoc?DBIx::Migration::Directories could be a
> good choice for that.
>
> So, in summary, we can't change an existing table name until we have
> migration support, so we should probably change the new natdiv table's
> name instead.
>
> Rob
>
>
> ------------------------------------------------------------------------------
> This SF.net email is sponsored by Sprint
> What will you do first with EVO, the first 4G phone?
> Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
> _______________________________________________
> Gmod-schema mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/gmod-schema
>



--
------------------------------------------------------------------------
Scott Cain, Ph. D.                                   scott at scottcain dot net
GMOD Coordinator (http://gmod.org/)                     216-392-3087
Ontario Institute for Cancer Research

------------------------------------------------------------------------------
This SF.net email is sponsored by Sprint
What will you do first with EVO, the first 4G phone?
Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
_______________________________________________
Gmod-schema mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gmod-schema
Reply | Threaded
Open this post in threaded view
|

Re: nd naming

Robert Buels
In reply to this post by Sook Jung
Sook, please try to keep the discussion on the public list.  I think the
easiest thing would be for you to write the table definition SQL
(including COMMENT ON statements) and send it to Scott for integration
into trunk.

You can mostly cut and paste one of the other X_synonym table defs, I
think, like this one:
http://gmod.svn.sourceforge.net/viewvc/gmod/schema/trunk/chado/modules/sequence/sequence.sql?revision=21981&view=markup
starting at line 633.

Rob


Sook Jung wrote:
> Hi Rob,
> 'Experiment' sounds good since it goes with other tables like protocol
> and reagent. So are you going to change the names? Can you add
> stock_synonym to the trunk also or can you tell me where to submit it?
> Thanks
> Sook
>


------------------------------------------------------------------------------
This SF.net email is sponsored by Sprint
What will you do first with EVO, the first 4G phone?
Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
_______________________________________________
Gmod-schema mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gmod-schema
Reply | Threaded
Open this post in threaded view
|

Re: nd naming

Sook Jung
Hi Scott,
I'm attaching the SQL for the stock_synonym. Could you integrate this
into trunk? Thanks
Sook

On Fri, Jul 2, 2010 at 12:37 PM, Robert Buels <[hidden email]> wrote:

> Sook, please try to keep the discussion on the public list.  I think the
> easiest thing would be for you to write the table definition SQL (including
> COMMENT ON statements) and send it to Scott for integration into trunk.
>
> You can mostly cut and paste one of the other X_synonym table defs, I think,
> like this one:
> http://gmod.svn.sourceforge.net/viewvc/gmod/schema/trunk/chado/modules/sequence/sequence.sql?revision=21981&view=markup
> starting at line 633.
>
> Rob
>
>
> Sook Jung wrote:
>>
>> Hi Rob,
>> 'Experiment' sounds good since it goes with other tables like protocol
>> and reagent. So are you going to change the names? Can you add
>> stock_synonym to the trunk also or can you tell me where to submit it?
>> Thanks
>> Sook
>>
>
>


--
Sook Jung, PhD
Assistant Research Professor of Bioinformatics
Dept of Horticulture and Landscape Architecture
Washington State University
45 Johnson Hall, Pullman, WA 99164-6414
Email:[hidden email]

------------------------------------------------------------------------------
This SF.net email is sponsored by Sprint
What will you do first with EVO, the first 4G phone?
Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
_______________________________________________
Gmod-schema mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gmod-schema

stock_synonym.txt (2K) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: nd naming

Sook Jung
Hi Naama,
Sorry for the late reply. It sounds good - now I checked my document
again and I actually used stockprop to store various aliases.. It
works better since there are really a lot of different types of
synonyms for stock.
Thanks
Sook


On Tue, Aug 10, 2010 at 9:25 AM, Naama Menda <[hidden email]> wrote:

> hey Sook,
>
> Why not store synonyms in stockprop, and link the publication using
> stockprop_pub ?
>
> The type of synonym (internal, current, etc. ) should be coded in the cvterm
> table (see column 'type_id' in stockprop)
>
> This way the synonyms are kept more generic, and we don't have to add new
> columns for different types of synonyms that may arise.
>
>  Let me know what you think
>
> -Naama
>
>
>
> On Fri, Jul 2, 2010 at 2:04 PM, Sook Jung <[hidden email]> wrote:
>>
>> Hi Scott,
>> I'm attaching the SQL for the stock_synonym. Could you integrate this
>> into trunk? Thanks
>> Sook
>>
>> On Fri, Jul 2, 2010 at 12:37 PM, Robert Buels <[hidden email]> wrote:
>> > Sook, please try to keep the discussion on the public list.  I think the
>> > easiest thing would be for you to write the table definition SQL
>> > (including
>> > COMMENT ON statements) and send it to Scott for integration into trunk.
>> >
>> > You can mostly cut and paste one of the other X_synonym table defs, I
>> > think,
>> > like this one:
>> >
>> > http://gmod.svn.sourceforge.net/viewvc/gmod/schema/trunk/chado/modules/sequence/sequence.sql?revision=21981&view=markup
>> > starting at line 633.
>> >
>> > Rob
>> >
>> >
>> > Sook Jung wrote:
>> >>
>> >> Hi Rob,
>> >> 'Experiment' sounds good since it goes with other tables like protocol
>> >> and reagent. So are you going to change the names? Can you add
>> >> stock_synonym to the trunk also or can you tell me where to submit it?
>> >> Thanks
>> >> Sook
>> >>
>> >
>> >
>>
>>
>>
>> --
>> Sook Jung, PhD
>> Assistant Research Professor of Bioinformatics
>> Dept of Horticulture and Landscape Architecture
>> Washington State University
>> 45 Johnson Hall, Pullman, WA 99164-6414
>> Email:[hidden email]
>>
>>
>> ------------------------------------------------------------------------------
>> This SF.net email is sponsored by Sprint
>> What will you do first with EVO, the first 4G phone?
>> Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
>> _______________________________________________
>> Gmod-schema mailing list
>> [hidden email]
>> https://lists.sourceforge.net/lists/listinfo/gmod-schema
>>
>
>



--
Sook Jung, PhD
Assistant Research Professor of Bioinformatics
Dept of Horticulture and Landscape Architecture
Washington State University
45 Johnson Hall, Pullman, WA 99164-6414
Email:[hidden email]

------------------------------------------------------------------------------
This SF.net email is sponsored by

Make an app they can't live without
Enter the BlackBerry Developer Challenge
http://p.sf.net/sfu/RIM-dev2dev 
_______________________________________________
Gmod-schema mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gmod-schema
Reply | Threaded
Open this post in threaded view
|

Re: nd naming

Naama Menda
In reply to this post by Bob MacCallum
Seth or Bob (don't remember who generated the schema graph) ,
the Natural Diversity SQL is updated and merged with the trunk.
Could you regenerate the graph in the wiki?
http://gmod.org/wiki/Talk:Chado_Natural_Diversity_Module_Working_Group#April_2010

Thanks!
-Naama


Naama Menda
Boyce Thompson Institute for Plant Research
Tower Rd
Ithaca NY 14853
USA

(607) 254 3569
Sol Genomics Network
http://solgenomics.net/
[hidden email]


On Fri, Jul 2, 2010 at 10:17 AM, Bob MacCallum <[hidden email]> wrote:
I'm keen to see stability in table names (for forthcoming
Bio::Chado::Schema development) as I am sure Rob is.

The name change is unfortunate, but doesn't really bother us. Maybe
the mage table could change instead (arrayassay?) :-)

cheers,
Bob
(+Seth)


On Fri, Jul 2, 2010 at 5:14 AM, Robert Buels <[hidden email]> wrote:
> Sook Jung wrote:
>> The option 3 sounds good to me too. What could be the new name for
>> assay table if we do go for the option 3?
>> Since it involves stock (phenotype/genotype assay, cross, field
>> collection, etc), how about 'stockassay'?
>
> I was thinking of changing 'assay' to 'experiment' (which I think is a
> name it had at a previous point in development, but my memory is hazy
> after months), since it has to do with more things than just stocks.
>
> Other opinions?
>
>> I've been looking at the schema too and I think we really need a place
>> to store aliases of stock. Can we add stock_synonym just like
>> feature_synonym? We could use stock_dbxref but synonyms not always
>> have appropriate db..
>
> This sounds OK to me, it would go in the stock module.  Not really
> related to the natdiv effort, probably should just go into trunk.
>
> Rob
>
>
> ------------------------------------------------------------------------------
> This SF.net email is sponsored by Sprint
> What will you do first with EVO, the first 4G phone?
> Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
> _______________________________________________
> Gmod-schema mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/gmod-schema
>

------------------------------------------------------------------------------
This SF.net email is sponsored by Sprint
What will you do first with EVO, the first 4G phone?
Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
_______________________________________________
Gmod-schema mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gmod-schema


------------------------------------------------------------------------------
Sell apps to millions through the Intel(R) Atom(Tm) Developer Program
Be part of this innovative community and reach millions of netbook users
worldwide. Take advantage of special opportunities to increase revenue and
speed time-to-market. Join now, and jumpstart your future.
http://p.sf.net/sfu/intel-atom-d2d
_______________________________________________
Gmod-schema mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gmod-schema