There's a bug in setting the default title for an entity where the data is stored in the chado.stock table. Currently it's assigned [stock__name] which is not a valid token resulting in an empty title for published entities. It should be assigning [schema__name] which is the equivalent.

Copied from Drupal issue queue: https://www.drupal.org/node/2784501

In the event that a site developer "syncs" content types that they no longer need/want the only way to remove the nodes without removing the data in Chado is to manually remove the records from the corresponding chado_xxxxx table and then clean up orphaned nodes. This is a bit unfriendly. We should have a way to "unsync"

Copied from Drupal isssue queue: https://www.drupal.org/node/2713125

Various bug fixes have been made to Tripal 2 since Tripal 3 development began. These bug fixes should be merged into the Tripal 3 API where applicable. Due to the massive changes this will have to be done on a per commit bases...

According to this comparison this will be a doozy of a job :-(
7.x-3.x...7.x-2.x

Let's try to keep track of it as we go in this issue.

@laceysanderson reported some progress on this. See Drupal issue queue for further details.

prints out too much to the screen when there are too many items.

Copied issue from the Drupal issue queue:

The webservices can be found directly on the Apollo instances, but they should be globally available:

http://icebox.lbl.gov/Apollo2/WebServices/

You should want IOServices -> write

An example script written in groovy is here:

https://github.com/GMOD/Apollo/blob/master/docs/web_services/examples/gr...

Some more doc here:

http://webapollo.readthedocs.org/en/latest/Web_services.html

Copied issue from Drupal issue queue: https://www.drupal.org/node/2754831

During an upgrade of a Chado v1.2 database to v1.3 an error occured that prevent the upgrade using the Tripal upgrader. The view so.unorient_intra_transposition uses the feature table and when that table is altered to change ints to bigints for Chado v1.3 the upgrade crashes. This View is created in Chado v1.11 but is not present in Chado v1.2. Therefore the upgrade script to Chado v1.3 does not consider this view and the upgrade fails with an error.

On tripal v3 after editing a content type that has a sequence field the checksum disappears... It should be deprecated

Hello,

When running the tripal job to prepare chado, I got this error. The error occurred only on the first run. Running the job a second time does not cause the same error and the job completes successfully.

$drush trp-run-jobs --username=username_here --root=/data/www/html/drupal-7.54
Tripal Job Launcher
Running as user 'username_here'
-------------------
Calling: tripal_chado_prepare_chado(a:0:{}, 1080)

NOTE: Populating of chado_cvterm_mapping table is performed using a database transaction.
If the load fails or is terminated prematurely then the entire set of
insertions/updates is rolled back and will not be found in the database

Examining analysis...
Examining contact...
Examining control...
Examining feature...
Examining featuremap...
Examining genotype...
Examining library...
Examining organism...
Examining phenotype...
Examining phylotree...
Examining project...
Examining protocol...
Examining protocolparam...
Examining pub...
Examining stock...
Examining stockcollection...
Examining studyfactor...
Examining synonym...

Done.
Exception: Exception: SQLSTATE[25P02]: In failed sql transaction: 7  [error]
ERROR:  current transaction is aborted, commands ignored until end of
transaction block: SELECT expire, value FROM {semaphore} WHERE name =
:name; Array
(
[:name] => field_info:bundle:TripalTerm:TripalTerm
)
in
/data/www/html/drupal-7.54/sites/all/modules/tripal/tripal_chado/includes/tripal_chado.setup.inc:243
Stack trace:
#0
/data/www/html/drupal-7.54/sites/all/modules/tripal/tripal/api/tripal.jobs.api.inc(383):
tripal_chado_prepare_chado('a:0:{}', '1080')
#1
/data/www/html/drupal-7.54/sites/all/modules/tripal/tripal/tripal.drush.inc(173):
tripal_launch_job(0, NULL)
#2 phar:///bin/drush/includes/command.inc(422):
drush_tripal_trp_run_jobs()
#3 phar:///bin/drush/includes/command.inc(231):
_drush_invoke_hooks(Array, Array)
#4 phar:///bin/drush/includes/command.inc(199): drush_command()
#5 phar:///bin/drush/lib/Drush/Boot/BaseBoot.php(67):
drush_dispatch(Array)
#6 phar:///bin/drush/includes/preflight.inc(66):
Drush\Boot\BaseBoot->bootstrap_and_dispatch()
#7 phar:///bin/drush/includes/startup.inc(458): drush_main()
#8 phar:///bin/drush/includes/startup.inc(365): drush_run_main(false,
'/', 'Phar detected. ...')
#9 phar:///bin/drush/drush(114): drush_startup(Array)
#10 /bin/drush(10): require('phar:///bin/dru...')
#11 {main} in
/data/www/html/drupal-7.54/sites/all/modules/tripal/tripal_chado/includes/tripal_chado.setup.inc:332
Stack trace:
#0
/data/www/html/drupal-7.54/sites/all/modules/tripal/tripal/api/tripal.jobs.api.inc(383):
tripal_chado_prepare_chado('a:0:{}', '1080')
#1
/data/www/html/drupal-7.54/sites/all/modules/tripal/tripal/tripal.drush.inc(173):
tripal_launch_job(0, NULL)
#2 phar:///bin/drush/includes/command.inc(422):
drush_tripal_trp_run_jobs()
#3 phar:///bin/drush/includes/command.inc(231):
_drush_invoke_hooks(Array, Array)
#4 phar:///bin/drush/includes/command.inc(199): drush_command()
#5 phar:///bin/drush/lib/Drush/Boot/BaseBoot.php(67):
drush_dispatch(Array)
#6 phar:///bin/drush/includes/preflight.inc(66):
Drush\Boot\BaseBoot->bootstrap_and_dispatch()
#7 phar:///bin/drush/includes/startup.inc(458): drush_main()
#8 phar:///bin/drush/includes/startup.inc(365): drush_run_main(false,
'/', 'Phar detected. ...')
#9 phar:///bin/drush/drush(114): drush_startup(Array)
#10 /bin/drush(10): require('phar:///bin/dru...')
#11 {main}

When importing an organism, a link to the taxonomy record should be added to the organism_dbxref table.

Copied from Drupal issue queue: https://www.drupal.org/node/2499301

Andrew expressed a desire for public cvterm pages that could be linked to in views and listing. These would provide addition information to users and could include term hierarchies, etc.

NOTE: currently we have administrative pages only for cv/cvterms and even then we don't have a view (only edit) page.

A publication has been added to both an organism and analysis. They are present when editing the field, but they don't show up on the page when viewing.

You see the following errors at the top of the page when editing the Relationship field (Home » Administration » Structure » Tripal Content Types » Organism » Manage fields):

Notice: Undefined index: #entity in sbo__relationship_widget->form() (line 18 of /var/www/XXX/sites/all/modules/tripal/tripal_chado/includes/TripalFields/sbo__relationship/sbo__relationship_widget.inc).
Notice: Undefined index: relationships in sbo__relationship->settingsForm() (line 549 of /var/www/XXX/sites/all/modules/tripal/tripal_chado/includes/TripalFields/sbo__relationship/sbo__relationship.inc).
Notice: Undefined index: relationships in sbo__relationship->settingsForm() (line 567 of /var/www/XXX/sites/all/modules/tripal/tripal_chado/includes/TripalFields/sbo__relationship/sbo__relationship.inc).
Notice: Undefined index: relationships in sbo__relationship->settingsForm() (line 578 of /var/www/XXX/sites/all/modules/tripal/tripal_chado/includes/TripalFields/sbo__relationship/sbo__relationship.inc).
Notice: Undefined index: relationships in sbo__relationship->settingsForm() (line 592 of /var/www/XXX/sites/all/modules/tripal/tripal_chado/includes/TripalFields/sbo__relationship/sbo__relationship.inc).

Following http://tripal.info/tutorials/v3.x/example-genomic-site/organism, there appears to be a bug adding the image.

If I fill in everything as instructed including clicking the "upload" button after browsing for the image then on submit the entity is not created, the form is reset and I see the following error:

Could not save the entity: SQLSTATE[42P01]: Undefined table: 7 ERROR: relation "field_data_data__image" does not exist LINE 1: INSERT INTO field_data_data__image (entity_type, entity_id, ... ^
Cannot save entity

If I fill out the form, browse for the image but don't click upload before submitting the form then I see the same symptoms as stated above.

If I don't add an image but fill out the other fields as instructed then the entity is created but the page I'm redirected to (the entity view page) is a white error page with the following PDO Exception:

PDOException: SQLSTATE[42P01]: Undefined table: 7 ERROR: relation "field_data_data__image" does not exist LINE 3: field_data_data__image t ^: SELECT t.* FROM {field_data_data__image} t WHERE (entity_type = :db_condition_placeholder_0) AND (entity_id IN (:db_condition_placeholder_1)) AND (language IN (:db_condition_placeholder_2)) AND (deleted = :db_condition_placeholder_3) ORDER BY delta ASC; Array ( [:db_condition_placeholder_0] => TripalEntity [:db_condition_placeholder_1] => 2 [:db_condition_placeholder_2] => und [:db_condition_placeholder_3] => 0 ) in field_sql_storage_field_storage_load() (line 412 of /var/www/dev/tripal3/modules/field/modules/field_sql_storage/field_sql_storage.module).

I'm able to create other Tripal Content without images (e.g. project) just fine.

I recieve the following chado views error repeated 7 times whenever I clear the cache.

Notice: Undefined index: type in tripal_chado_views_views_data() (line 221 of /var/www/XXX/sites/all/modules/tripal/tripal_chado_views/tripal_chado_views.views.inc).

Copied from Drupal issue: https://www.drupal.org/node/2825357

From @guignonv:

When using several Chado instances using the Multi-Chado module, on installation of a new instance the common schemata 'genetic_code', 'so' and 'frange' are dropped and re-created.
The PostgreSQL user account used to create the new Chado instance will be the owner of those schemata. The default Drupal account or event other PostgreSQL account used for other instances may not be allowed to access those schemata anymore.
This appends in tripal_core/includes/tripal_core.chado_install.inc:tripal_core_reset_chado_schema().

I don't have the best solution yet but I see several ways to fix it...

#1: we keep those schemata if they exist but then, updating Chado may be an issue if those schema are not updated.

#2: after installation of a new Chado instance, we reassign the ownership to Drupal default user. Then comes the "how do we know a new installation is done?". I'm thinking of 2 possible hooks to solve that: one would be to simply add a "hook_chado_installed" when the installation process is done. Another more flexible possibility would be to add a hook to Tripal jobs API to make it fire hooks when a job is completed so other modules can do some post-processing. Here the Multi-Chado module would reassign the appropriate ownership if the job run was a successful creation of a new Chado instance.

#3: Give a way to alter those common schemata names so the original ones won't be removed by the installation process and the installation SQL scripts could also be altered to they reflect those changes. For instance, if I install Chado in a "private_chado" schema, I would make its PostgreSQL functions use the schemata private_genetic_code, and such...

Other point of views are welcome!

Tripal Version: 7.x-3.0-beta1

I followed the tutorial (http://tripal.info/tutorials/v3.x/installation/tripal) and the step "prepare both Drupal and Chado" produced many errors when the Tripal job was launched. Furthermore, despite all the errors, my Tripal3 site now says "Your site is prepared." in green.

I've attached the entire log from running the Tripal job.
tripal3.preparechado.txt

It seems to be having trouble loading ontologies and later selecting terms.

• TPUB:0000002 isn't in the dbxref table.
• Tripal publication cv has no cvterms inserted.
• chado.tripal_obo_temp table does exist.
• cvterms do exist for the following cvs (number of terms in brackets).
  An additional type for the item, typically used for adding more specific types from external vocabularies in microdata syntax. This is a relationship between something and a class that the thing is in. (1), analysis_property (1), chado_properties (1), EDAM (12), ero (1), featuremap_property (8), featuremap_units (5), featurepos_property (2), IAO (2), iao (1), library_property (1), library_type (6), local (19), ncbitaxon (1), nd_experiment_types (2), nd_geolocation_property (1), null (1), obi (1), ogi (1), organism_property (11), project_property (1), rdfs (2), sbo (3), schema (6), sequence (2), SIO (3), swo (1), synonym_type (4), taxonomic_rank (3), tripal_analysis (2), tripal_phylogeny (4), uo (1).

Copied issue from Drupal issue queue: https://www.drupal.org/node/2477443

When creating a new Bulk Loader template record, I initially found the order of inputs a little confusing. It might have been more clear if the "Type of Field" entry was moved below the record naming fields with a line or some other separator drawn between the record-related and field-related fields. This would make it clear that you are first entering the record information, then the field information.

All that being said, it might be nice to have record and field creation on different pages. This would also allow the other record options (e.g. "Action to take when Loading Record", "Additional Select Options", etc... sections) to be presented during record creation. Even if not separate pages, having access to those additional record options during record creation would be nice, though it would make the single record-and-first-field creation page a bit long.

Overall the bulk loading system is really impressive though!

This issue is being copied over from the Drupal issue queue: https://www.drupal.org/node/2796857

The GFF3 spec has a section describing discontinuous features which allows a feature to have the same ID and span multiple district regions. This is not supported by the Chado Perl based loader and it's not directly supported by Chado either. However if we want to fully handle the GFF3 spec we should probably consider supporting them. One idea may be to add an option that the user loading the file would have to check that gives their consent to have the Id's renamed automatically for discontinuous features. But if the parent is not specified then a new parent feature may need to be created. It seems Maker may generate CDS feature in this way but a parent is provided. For these we could just rename the IDs.

Copied from the Drupal issue queue: https://www.drupal.org/node/2696487

Using:

$featuremap = chado_expand_var($featuremap, 'table', 'stock_relationship', array('return_array' => TRUE));

Then the relationship from the object_id is an array or an object depending on how many relationship records there are.

$featuremap->featuremap_stock->stock_id->stock_relationship->object_id

Shouldn't the 'return_array' argument be applied recursively?

You see the following errors at the top of the page when editing the Publication field (Home » Administration » Structure » Tripal Content Types » Organism » Manage fields):

Notice: Undefined index: #entity in schema__publication_widget->form() (line 16 of /var/www/XXX/sites/all/modules/tripal/tripal_chado/includes/TripalFields/schema__publication/schema__publication_widget.inc).
Notice: Trying to get property of non-object in schema__publication_widget->form() (line 29 of /var/www/XXX/sites/all/modules/tripal/tripal_chado/includes/TripalFields/schema__publication/schema__publication_widget.inc).

Issue copied from Drupal Issue queue: https://www.drupal.org/node/2764463

As per the comment made on this thread in the mailing list: https://sourceforge.net/p/gmod/mailman/message/35214518/

We should explore parallel loading of GFF/FASTA files to improve loading time.

Some other potential ideas to speed performance: http://stackoverflow.com/questions/12206600/how-to-speed-up-insertion-performance-in-postgresql

Error: Call to undefined function dpm() in tripal_field_instance_settings_form_alter_process() (line 433 of /var/www/html/hops/sites/all/modules/tripal/tripal/includes/tripal.fields.inc).

Copied from the Drupal issue queue: https://www.drupal.org/node/2694039

We are missing documentation in the User's guide for the Materialized Views interface and the Custom Tables interface: http://tripal.info/node/121

When viewing the list of Tripal content types the published/not published filter doesn't work. Reported by TreeGenes

Issue copied from Drupal issue queue: https://www.drupal.org/node/2499303

It was brought up at the June Tripal Developers meeting that Sook/Jing need a means to bulk load publications that are not in PubMed. This can be done by creating a spreadsheet format and matching Tripal Bulk Loader template but this won't be very flexible on the fields required/expected for a journal article. Stephen suggested a BibTex loader might be a better alternative since this can be exported from EndNote and is more flexible on the fields required.

@laceysanderson said:
I've started work on a BibTeX loader. Progress can be found here: https://github.com/UofS-Pulse-Binfo/tripal_bibtex
Currently it's a little rough though it does work for my site. It's called via drush and doesn't yet have a UI. That said -it's better than nothing and if any one else needs this functionality it's available. If you run into problems file issues on github and I will fix them as soon as possible.

I'm going to mark this issue as postponed until I can finish the module sufficiently to re-discuss bringing it into core of the publication module.

Copied from the Drupal issue queue: https://www.drupal.org/node/2691787

It was reported on the mailing list (https://sourceforge.net/p/gmod/mailman/message/34924927/) that if the regular expression used for importing a FASTA file is incorrect that an error message occurs. This should be handled gracefully with a more intuitive notice given to the user. Here is the message recieved:

WD trp-fasta: [error]
array_keys() expects parameter 1 to be array, string given tripal_core.tripal.api.inc:117 [warning]
[site http://default] [TRIPAL ] [TRP-FASTA] Array
strtr() expects parameter 1 to be string, array given syslog.module:115 [warning]
syslog() expects parameter 1 to be long, string given syslog.module:118 [warning]
html_entity_decode() expects parameter 1 to be string, array given unicode.inc:463 [warning]
WD trp-fasta:

Copying issue from Drupal Issue queue: https://www.drupal.org/node/2757465

During a major upgrade (i.e. from D6 -> D7) all of the Tripal modules are typically disabled. Some of the API functions are using drupal_write_record but unfortunately this function doesn't work when modules are disabled because it requires the module's schema be available (which it's not when disabled). We should therefore always use db_insert or db_update in API functions as those may sometimes get used in upgrade functions.

We need to clean all of the API functions in Tripal 3.x to not use drupal_write_record. It may not be necessary to spend the time to do this in 2.x because the upgrade is working.

Copied from Drupal issue queue: https://www.drupal.org/node/2314235

There has been some requests for more drush commands to support administrative functions that are currently only available using the web interface. Some folks want the ability to do as much as possible on the command-line. This has the advantage of making Tripal scriptable as well.

Development of all of these drush commands isn't a 7.x-2.x beta blocker but perhaps development of a drush command naming convention, and style is a beta-blocker. This design can be used to convert existing drush commands to a consistent format and allow for consistent development of future commands.

When I disable a Tripal pane in manage display, it still shows up in the table of contents on the page. However when I click the link, the disabled pane doesn't show up in the right side. For example, the following screenshot shows the manage display page for a SNP with all but the summary and relationship panes disabled:

However, as you can see in the following screenshot, I still have all the links for the disabled panes in my table of contents:

This is not fixed by clearing the cache.

Issue copied from the Drupal issue queue: https://www.drupal.org/node/2765823

The AGL importer for the Pub module is no longer working. A very cursory look at the problems seems to indicate that the response from the Agricola Z39.50 compliant server is not what is expected (http://agricola.nal.usda.gov/help/z3950.html).

My Perspective:
My users are strongly against the long page of content... Thus the default Tripal2 display was really well suited to them :-) I like the new Tripal3 default but would like to only show one pane by default not all of them...

More general suggestion:
It would be good to have the ability to configure in "Manage Display" whether a Tripal pane is shown on page load or not. This way admin could choose which panes to display to their user and thus better control their first impression. The default could remain as it is (all panes on) or be changed to only summary shown by default (more Tripal2-like) based on community input. I'm good either way :-)

Issue copied from Drupal issue queue: https://www.drupal.org/node/2053069

It has been requested that we add the ability to load a GFF3 file and sync the features all at one time. We can do this if we add a textbox on the GFF3 loader form that lets the user specify which feature types should be synced. Then after the GFF3 file is completely loaded, features that match those sequence types could be synced.

Copied from Drupal Issue queue: https://www.drupal.org/node/2702831

Would be great if I could add a REST track to JBRowse by pointing to a TRIPAL instance:

The API is: http://gmod.org/wiki/JBrowse_Configuration_Guide#JBrowse_REST_Feature_St...

example:

{
"label": "Orange Genes",
"key": "Orange Genes",
"type": "JBrowse/View/Track/HTMLFeatures",
"storeClass": "JBrowse/Store/SeqFeature/REST",
"baseUrl": "https://www.citrusgenomedb.org/rest/api/base",
"query": {
"organism": "ruby red grapefruit"
}
}

Trying to install Tripal3 beta3 on CentOS 7 failed with following error.

drush pm-enable tripal

The following extensions will be enabled: tripal, php
Do you really want to continue? (y/n): y
PHP Parse error: syntax error, unexpected '.', expecting ',' or ';' in /var/www/html/sites/all/modules/tripal/tripal/includes/TripalFields/TripalField.inc on line 18
Drush command terminated abnormally due to an unrecoverable error. [error]
Error: syntax error, unexpected '.', expecting ',' or ';' in
/var/www/html/sites/all/modules/tripal/tripal/includes/TripalFields/TripalField.inc, line 18

The following small typos were reported to me so I'm leaving a note here for reference :-)

On Import Vacabulary Page: Home » Administration » Tripal » Vocabularies ([tripal3site]/admin/tripal/terms/import):
Once you select a file to import, the form element description for "Remote URL" is missing the closing bracket on the e.g.

Please enter a URL for the online OBO file. The file will be downloaded and parsed. (e.g. http://www.obofoundry.org/ro/ro.obo

On Add Tripal Content Type Page: Home » Administration » Structure » Tripal Content Types » Add Tripal Content Type ([tripal3site]/admin/structure/bio_data):
Select a term and click "Lookup Term" then select a chado table to reveal the "Are all records in the XXX table of type YYY?" section. In the form element description it says

Select "Yes" if the XXX table houses more that just data of type YYY.

Gotcha 1)

When creating an analysis (http://tripal.info/tutorials/v3.x/example-genomic-site/analysis), the tutorial specifies the Time Executed as Feb 1, 2011. However, the time executed field in Tripal 3 only provides 2014-2020 as options in the year drop-down. Either we should change the tutorial or configure this field to have a textfield for the year.

When the tripal_ds module is enabled the backwards compatibility support for using old templates doesn't work.

Copied from Drupal issue queue: https://www.drupal.org/node/2265611

When adding cvterms using KEGG/GO/InterProScan those terms should be annotated with the evidence level of those assignments so that viewers do not assume they are fully curated terms. Using the GO evidence codes (at http://www.geneontology.org/GO.evidence.shtml) may be a good set to follow.

Working through the tutorial (http://tripal.info/tutorials/v3.x/example-genomic-site/genomic-data), I load the mRNA FASTA and add in the "Advanced Options" for automatically creating the database references. During loading the Tripal job doesn't report any errors yet when I go to the mRNA page after publishing content, the database reference link is <a href="https://phytozome.jgi.doe.gov/phytomine/portal.do?externalid=PAC:Phytozome:">Phytozome:</a>

When I poke in the database, this is what I see:

SELECT 
     f.feature_id, 
     f.name, 
     f.uniquename,
     fd.dbxref_id, 
     db.name as db_name, 
     dbxref.accession as dbxref_accession 
  FROM chado.feature f 
    LEFT JOIN chado.feature_dbxref fd ON fd.feature_id=f.feature_id 
    LEFT JOIN chado.dbxref ON dbxref.dbxref_id=fd.dbxref_id 
    LEFT JOIN chado.db ON db.db_id=dbxref.db_id 
  WHERE f.type_id=422 
  ORDER BY feature_id ASC, db.name ASC;

 feature_id |       name        |  uniquename  | dbxref_id |  db_name   | dbxref_accession 
------------+-------------------+--------------+-----------+------------+------------------
          3 | orange1.1g015632m | PAC:18136217 |     77549 | GFF_source | phytozome6
          3 | orange1.1g015632m | PAC:18136217 |     77551 | Phytozome  | 
         20 | orange1.1g015645m | PAC:18136218 |     77549 | GFF_source | phytozome6
         20 | orange1.1g015645m | PAC:18136218 |     77551 | Phytozome  | 
         37 | orange1.1g015615m | PAC:18136219 |     77549 | GFF_source | phytozome6
         37 | orange1.1g015615m | PAC:18136219 |     77551 | Phytozome  | 
         53 | orange1.1g015662m | PAC:18136220 |     77549 | GFF_source | phytozome6
         53 | orange1.1g015662m | PAC:18136220 |     77551 | Phytozome  | 
         68 | orange1.1g017341m | PAC:18136221 |     77549 | GFF_source | phytozome6
         68 | orange1.1g017341m | PAC:18136221 |     77551 | Phytozome  | 
         82 | orange1.1g018514m | PAC:18136222 |     77549 | GFF_source | phytozome6
         82 | orange1.1g018514m | PAC:18136222 |     77551 | Phytozome  | 
         98 | orange1.1g022520m | PAC:18136223 |     77549 | GFF_source | phytozome6
         98 | orange1.1g022520m | PAC:18136223 |     77551 | Phytozome  | 
        110 | orange1.1g022799m | PAC:18136224 |     77549 | GFF_source | phytozome6
        110 | orange1.1g022799m | PAC:18136224 |     77551 | Phytozome  | 
        123 | orange1.1g022797m | PAC:18136225 |     77549 | GFF_source | phytozome6
        123 | orange1.1g022797m | PAC:18136225 |     77551 | Phytozome  | 
(18 rows)

Also, just in case it's relevant, the feature.dbxref_id is empty.

I've tried this process three times on two fresh installs and this result is repeatable.

Hello,

When upgrading from Tripal v2 to v3-beta3 I got this error during installation of tripal_chado.
However, running the command a second time doesn't throw any errors and executes successfully.

$drush en tripal_chado

DatabaseSchemaObjectExistsException: Table cache_image already    [error]
in /data/www/html/drupal-7.54/includes/database/schema.inc:660
Stack trace:
#0
/data/www/html/drupal-7.54/includes/database/database.inc(2776):
DatabaseSchema->createTable('cache_image', Array)
#1
/data/www/html/drupal-7.54/includes/common.inc(7114):
db_create_table('cache_image', Array)
#2
/data/www/html/drupal-7.54/includes/module.inc(479):
drupal_install_schema('image')
#3 phar:///bin/drush/commands/core/drupal/environment_7.inc(143):
module_enable(Array)
#4 phar:///bin/drush/commands/pm/pm.drush.inc(1167):
drush_module_enable(Array)
#5 phar:///bin/drush/includes/command.inc(422):
drush_pm_enable('tripal_chado')
#6 phar:///bin/drush/includes/command.inc(231):
_drush_invoke_hooks(Array, Array)
#7 phar:///bin/drush/includes/command.inc(199):
drush_command('tripal_chado')
#8 phar:///bin/drush/lib/Drush/Boot/BaseBoot.php(67):
drush_dispatch(Array)
#9 phar:///bin/drush/includes/preflight.inc(66):
Drush\Boot\BaseBoot->bootstrap_and_dispatch()
#10 phar:///bin/drush/includes/startup.inc(458): drush_main()
#11 phar:///bin/drush/includes/startup.inc(365):
drush_run_main(false, '/', 'Phar detected. ...')
#12 phar:///bin/drush/drush(114): drush_startup(Array)
#13 /bin/drush(10): require('phar:///bin/dru...')
#14 {main}

Hello,

After upgrading to Tripal 3, the organism, feature and analysis pages are not displaying the left sidebar they used to. Is that the intended behavior?

Tripal 2	Tripal 3

Thanks!

When following the Tripal 3 tutorial (http://tripal.info/tutorials/v3.x/installation/tripal), you

1. Install Tripal Core
2. Install Tripal Chado
3. Install Chado schema & prepare chado/drupal (created basic content types including organism)
4. Install Tripal DS

Following this workflow I don't get the default arrangement of Tripal Panes for all the original basic content types created via preparing your Chado/Drupal database. Tripal panes are available as a group formatter and the layout is available but I would have to build the hierarchy myself.

However, any new content types I create (e.g. genetic_marker) will have the default layout. In other words, Any content types existing before Tripal DS is enabled are not getting a default arrangement; where as, content types created after Tripal DS is enabled are.

This is incongruent with the tutorial which shows the organism and analysis as having Tripal Panes when following the current order of the tutorial doesn't actually produce that effect.

Do we want to apply default grouping on pre-existing Tripal Content Types? We would run the risk of overriding custom settings if we did... That said, if we don't then what about those admin that don't enable Tripal DS to begin with thinking they want to customize everything themselves and then change their mind after realizing how much work it is ;-). Thoughts?

I need the ability to hide empty panes from the display completely. For example, if there are no publications for a given page then that pane shouldn't show up in either the table of contents or alongside the other panes --this is the current functionality in Tripal2. This should likely be a configuration option on the "Manage Display" since some groups have opted for the "display everything even if it's empty" approach with their Tripal2 sites.

NOTE: I can't do this by simply hiding the field in "Manage Display" for a given entity type because I want it to show up on pages that do have data in that field.

Copied from Drupal issue queue: https://www.drupal.org/node/2052833

A request has been made via the mailing list for a GFF3 exporter:
http://sourceforge.net/mailarchive/message.php?msg_id=31206560

Copying from Drupal issue queue: https://www.drupal.org/node/2069953

A recent email on the Tripal mailing list.

Hi,

I am working with some Tripal jobs which sometimes fail, e.g. due to bad input files etc. I found myself inserting
debug output via print sometimes. When a job fails I get "Error Message Job has terminated unexpectedly."
in the job details, and running jobs on the console (via tripal_launch_jobs.php) doesn't yield output either. I assume
that the job runs into one or the other 'watchdog()' exceptions, but the output of this doesn't seem to get forwarded.
I think the best solution would be to store the error message and display it via the job interface, because often it might
give valuable hints on how to fix the files. Would that be possible?

Michael

@laceysanderson said:

Quick Update on this:
The newest dev version of Tripal 7.x-2.x logs all tripal_report_error()'s to Apache's Error log.

My current plan for Job logging is:

1. Implement basic start/end with job_id logging to /var/log/messages using syslog()
2. Create a job_logging API function that each job can implement to give status messages excluding job progress bars. These messages would be both printed to the screen and printed to a file in the /tmp directory where one file would exist per job. These files could then be displayed on the job details page to give people a UI for their jobs output.
3. Start implementing this API for the main jobs (ie: bulk loader, OBO, FASTA & GFF3 loaders).

We could also implement a drupal-style progress bar on the job details page using the percent complete but this would be an advanced feature for later I think.

As far as timeline, I'm thinking that 1 & 2 could maybe be implemented for the release candidate but #3 for sure would have to wait and be implemented over time.

After discussion, there are concerns with using /tmp to store job-specific logs. We feel this concern can be alleviated by allowing admin to configure where thse log files end up but defaulting to /tmp.

Additionally, we should take care of log clean-up through Tripal utilizing the Drupal cron.

Stephen also had an interesting suggestion that the Dojo Toolkit (Javascript Library) might provide good support for loading of long log files by allowing files to be loaded in sections as the user scrolls.

Additionally, it has been decided that this functionality will make it into the next release of Tripal and thus is no longer a release blocker.

Currently if you change the page title format in the entity edit form, you have to edit each and every single entity/page of that type to see the new title take effect. Since Tripal Content often has a large number of pages, this isn't feasible. We should add a Tripal Job which can be scheduled from the entity edit page that will programmatically update each entity/page to the new title.

This is a feature that has been requested by multiple groups for Tripal 2 so we should try to get it into Tripal 3.

Copied from Drupal issue queue: https://www.drupal.org/node/1803920

Currently we can create custom tables using the Tripal web interface, but there is no way to associate the custom table as a referring table to another table. We should adjust the code to make that connection for custom tables by examining the FK relationships, or add a field to let the user specify. I think the first option is better.

Copying from Drupal issue queue: https://www.drupal.org/node/1816142

When loading records to the featureloc table the loading can be extremely slow. This is the case when the srcfeature_id is for a large chromosome already exists, is being selected for by the loader and has the 'residues' field fully populated with DNA sequence. It takes a while to retrieve and transfer that large of a record so it slows the loader to a crawl. In, the case of a select, perhaps we should only return the field requested and the PK and FK ids for later "referral" joins. The bulk loader already supports joining of tables on non-FK fields so, in this event the non-FK field is automatically added when needed so limiting the fields returned to just the requested field, the PK and FK should be fine (I believe).

You see the following errors at the top of the page when editing the Contact field (Home » Administration » Structure » Tripal Content Types » Project » Manage fields):

Notice: Undefined index: #entity in chado_linker__contact_widget->form() (line 17 of /var/www/dev/XXX/sites/all/modules/tripal/tripal_chado/includes/TripalFields/chado_linker__contact/chado_linker__contact_widget.inc).
Notice: Trying to get property of non-object in chado_linker__contact_widget->form() (line 30 of /var/www/XXX/sites/all/modules/tripal/tripal_chado/includes/TripalFields/chado_linker__contact/chado_linker__contact_widget.inc).
notice: Undefined index: #entity in chado_linker__contact_widget->form() (line 17 of /var/www/XXX/sites/all/modules/tripal/tripal_chado/includes/TripalFields/chado_linker__contact/chado_linker__contact_widget.inc).Notice: Trying to get property of non-object in chado_linker__contact_widget->form() (line 30 of /var/www/XXX/sites/all/modules/tripal/tripal_chado/includes/TripalFields/chado_linker__contact/chado_linker__contact_widget.inc).

Copied from Drupal issue queue: https://www.drupal.org/node/2832554

When a user submits a job and goes to the jobs page to look at it they see 'execute' and 'cancel' links but those may not be appropriate given the job state. Those links should disappear or be present as appropriate.