When I create a new Drupal site I usually end up with at least one custom module and several community contributed modules. To get the site working as it should, many configuration values need to be set on the various modules. This makes deployment onto a fresh Drupal instance painstaking and error-prone.
I would like to give my custom module the ability to configure all the other modules. Either on install or on the click of a button on my custom module's administration page, all the necessary configuration values on the other modules would be programmatically set.
How would I best go about doing this?
AFAIK, there's no way to achieve what you mean easily. I tend to put as much as I can in hook_update_N() implementations and do frequent DB synchronisations as described in my answer to this question. However that does not work when you already have a live server with which you will have to merge data.
To that purpose, I use various tools according to the need. No one is perfect, but here's however a small collection of my favorite ones:
Features. This is a new concept and a new module. The idea is pretty awesome: it allows you to define a set of configuration/modules/settings and to export them as a feature. This feature will then be installed as if it were a module on the target site. This module does not export every possible setting, but it does however do a good job with the modules that need the hardest configuration, as CCK, Views, ImageCache and others... You can see a screencast demo (~10 mins) here.
Backup and migrate. This is a more radical approach: it simply dump and rebuild the entire database on a target system. It is good only if you need to overwrite the target system completely.
Node export. This allows to export (and import) nodes from a drupal installation to another one. It supports bulk operations but - unluckily - it does not support the migration of attached files and images.
Deploy. Because of the limitations of node export I once looked into using this module (still in development). I finally did not, and preferred to do a merge of the production and staging databases, but the concept seems very valid, as it allows to import/export complex data type via SOAP.
Taxonomy import/export. I suppose the name is self-explanatory. It uses files to achieve the tasks (XML or CSV).
Installation profiles (suggested by ctford) are useful when configuring new sites. They allow you to specify modules to enable, theme to default to etc on installation. They can be quite convenient because there is a command-line tool called Drush that automates the building of installation profiles. The downside is that the profiles are designed to be used on installation - not deployment of an individual module. It might be possible however to take the configuration code generated by Drush and call it when your module is enabled.
Finally, you can find a collection of tools for importing/exporting data here.
HTH!
have you looked at the "features" module? it is a new paradigm introduced as part of the open atrium distribution but also available as a stand-alone module. from their description:
"The features module enables the capture and management of features in Drupal. A feature is a collection of Drupal entities which taken together satisfy a certain use-case.
Features provides a UI and API for taking different site building components from modules with exportables and bundling them together in a single feature module. A feature module is like any other Drupal module except that it declares its components (e.g. views, contexts, CCK fields, etc.) in its .info file so that it can be checked, updated, or reverted programmatically."
http://drupal.org/project/features
Installation profiles are useful when configuring new sites. They allow you to specify modules to enable, theme to default to etc on installation. They can be quite convenient because there is a command-line tool called Drush that automates the building of installation profiles.
The downside is that the profiles are designed to be used on installation - not deployment of an individual module. It might be possible however to take the configuration code generated by Drush and call it when your module is enabled.
I know what you mean, it's a pain to set all modules up.
I'm sure you can investigate all 3rd party modules to see how configuration takes place and mimic that in your custom module, but I'd advise you against that...
The problem is that modules may change the way they store their settings from one revision to another, so whenever you update to a new version of any module you should do some reverse-engineering to see if your 'ultimate-one-click-configuration module' still works ok - which, if you ask me, is even more painful than manually configuring all modules for each project.
Just relax, take it easy, and enjoy Drupal :)
As the initialization is only required when Drupal is installed, I would think that a installation profile is the better solution; to keep a module that is not anymore used once that the installation is configured seems a little excessive, IMO.
Changing the installation profile used from a site, and make the new installation profile run its installation code isn't something that Drupal allows out-of-the-box. I would create a custom installation profile before creating the sites I need, and only for the features I know all the sites will share. For the other features, I would create separate custom modules I can later install, and eventually uninstall when the features they implement aren't anymore necessary.
Related
Usually I have to activate my available add-ons in order to make them working. But I find collective.geo.behaviour and my custom transmogrifier package seem working well without activating them. This make me wonder what is the trick behind the scenes. Will something go wrong if I keep using these addons without activating them?
Python packages that are installed for use in your Plone environment show up in your add-on list because they have Generic Setup profiles for addition to a Plone environment. Usually these profiles do things like set browser layers, add skin layers, add types or setup the catalog. They can also specify that the Generic Setup profile for some other add-on(s) should be run when this package is installed.
The two cases you mention here have different things going on:
crgis transmogrifier has a GS profile, but -- as far as I can tell based on examining its repository -- does not need one. It's GS profile does nothing. So, the install add-on choice will do nothing. Drop a note to the add-on author and tell them that.
Collective Geo Behavior's GS profile does nothing but specify that a couple of other add-on GS profiles be run. If you have already done the add-on installation for those, then this step does nothing. But, don't rely on that fact for future installations.
I think it's because these profiles don't do anything different than declare dependencies and the Code & ZCML are loaded at startup. So as long as you already have imported the dependencies listed in the profile you should be fine.
Also I think you could have packages without a profile, if you don't have a dependency or need to register something.
I'm trying to write a module that requires either module a or module b (not both, though having both is fine). It just dependent on a OR b. Is there a way to put this in the .info file or enforce this OR dependency programatically?
You can't natively in any current version of Drupal, there is a discussion here which talks about adding a enhances[] and enhancedby[] property for .info files (amongst a few other new properties), but it has not been put into core as far as I can tell.
If you're using Drupal 6 you might have a look at the Module Supports module which was created in response to that discussion.
If you're using Drupal 7 the discussion might get you half way there, and if not there are a few patches on that page that will at least show you roughly what you need to do to patch Drupal core yourself to do get it working.
The only way I've seen other modules implement this is by reporting problems after a module is installed and not including the modules as dependent. I would assume this would be similar to modules like ckeditor or jquery_ui which require non-drupal code to be put in place to operate properly.
I downloaded someone's Drupal module (non core) I changed 1 function in there. I want my changes to not be over written by future module updates however still able to enjoy new updates.
How do you do this?
Thanks.
Generally if you have changed something, this means you are no longer having the version that can be maintained by the original developer (unless he applies the same change in his/her module).
But your options are:
making original module more flexible / expanding it: If you are familiar with the concept of Drupal's hooks, you can use them to build your own module and that way modify the way this module behaves. If it has no hooks in the feature you need, you can add them and ask author to merge the changes so when someone installs your module, it seamlessly integrates with his/her module.
using version control system: Alternatively you can check the updates in a different way, not from Drupal itself. You can use Git to download that module (drupal.org modules directory provides you with Git access), and use it to merge new versions to your code - then what you need is to be more familiar with Git versioning system.
Better to use hooks, but if there are no other solutions ...
Simplest solution : create a diff, apply a patch manually after module's automatic update.
http://jungels.net/articles/diff-patch-ten-minutes.html
Or You could edit try to edit Your module's info file, overriding the configuration of the version & the datestamp of Your module, to avoid automatic updates.
If your changed function fix a bug or something else, is better to create a path and propose your modification to the maintainer.
Otherwise, if you just modify the function to help your project, you should also create a patch and apply it after every update of your module.
Another solution, but less elegant, its to clone the module and create your own with the modified function (like this, your module can be out of updates).
As a typical 'integrator' programmer customising Plone, what should I know about the ZMI to help me code more effectively? What are the settings, tools, pitfalls, shortcuts and dark corners that will save me time and help me write better code?
Edit: take it as read that I am coding on the filesystem, using GenericSetup profiles to make settings changes. I know making changes in the ZMI is a bad idea and generally steer clear. But occasionally the ZMI sure is useful: for inspecting a workflow, or examining a content item's permissions, or installing only one part of a profile via portal_setup. Is there really nothing worth knowing about the ZMI? Or are there other useful little tidbits in there?
There are a few places in the ZMI that I find myself returning to for diagnostic information:
/Control_Panel/Database: Select a ZODB mountpoint. Cache Parameters tab shows how much of your designated ZODB cache size has been used. Activity tab shows how many objects are being loaded to cache and written over time.
/Control_Panel/DebugInfo/manage: Lots of info, including showing what request each thread is serving at the current moment. The 'Cache detail' and 'Cache extreme detail' links give info on what classes of objects are currently in the ZODB cache.
Components tab of the Plone site root: Quick way to see what local adapters and utilities are registered. DON'T HIT THE APPLY BUTTON!
Undo tab of most objects: See who has committed transactions affecting the object lately.
Security tab: See what permissions are actually in effect for an object. You really don't want to change permissions here 90% of the time; it's too hard to keep track of where permissions are set and they are liable to be reset by workflow. Use the Sharing tab in the Plone UI to assign local roles instead. (The one exception is that I often find it handy to enable the add permission for a particular type in specific contexts.) In Zope 2.12, there is a new feature on this tab to enter a username and see what permissions and roles would be in effect for that user, which is handy.
Catalog tab of portal_catalog: See what index data and metadata is stored for a particular path. (Can also remove bogus entries from the index.)
Index tab of portal_catalog: Select an index, then click its Browse tab to get an overview of what keys are indexed and which items are associated with each key.
The key thing to know is that while many ZMI tools provide quick, through-the-web customization, the customizations that you make this way are hard to export out of the database. So, they don't move easily from development to production environments or from one deployment to another.
Ideally, a new developer should use the ZMI to explore and find points of intervention. Then, learn how to implement the same changes in policy add ons (products) that move from one deployment to another much more reproducibly.
If you want to write code for Plone, it's best to avoid the ZMI. The concept of doing things through the ZMI is very limited and discouraged - more and more things are not available in there and it will go away at some point.
The actual Plone control panels offer you most of the configuration options you can use. For anything else the file system is the best place to look.
I agree with the other posters that you shouldn't configure too much via the ZMI, as it's not in version control and you can easily lose track of the changes.
But the ZMI is still very useful for debugging and to see specific site configurations.
Here are some tools in the ZMI that I regularly consult:
portal_javascripts: To turn debugging on off. Checking which
scripts are there, what are their
conditions for rendering and are they
found?
portal_css: Basically the same as portal_javascripts but for
stylesheets.
portal_types: To see what a type's properties are. Can it be
created globally? What types can you
create inside it? What is its default
view? Etc.
portal_catalog: What indexes are there? What metadata is there in the
catalog? You can clear and rebuild
the catalog and even browse the
catalog.
portal_workflow: What states/transitions/permissions are
there in a certain workflow? What workflow is active on a
certain type?
portal_properties/site_properties: View and set site-wide properties. A
lot of these settings are in the
plone_control_panel (i.e outside of
the ZMI), but here they are on one
page and the ZMI is quicker to
navigate.
portal_skins: See which skins folders are installed. See the
ordering of the skin layers (via the properties tab). You can also edit
the templates, stylesheets and javascripts in the skins directories.
Not recommended! But useful for debugging.
portal_setup: Some very big and complex Plone websites can break if
you just willy-nilly add/remove/reinstall add-ons. Often it's safer to just run a specific GenericSetup update. For example, if you have added a new portlet, rather
import the specific (portlets.xml) step via portal_setup (the import tab), then reinstalling the whole product.
portal_actions: Configure which actions are visible/present.
portal_quickinstaller: Quickly reinstall, uninstall add-ons. Often
quicker and more lightweight than
loading Plone Control Panel's
equivalent.
acl_users: Sometimes when using an add-on like LDAPUserFolder, you'll
have to dig around in acl_users to configure and test it. You can also
create users here, although it's better to do this via the Plone
Control Panel (i.e not in ZMI)
There are many more tools and things to tweak (and break your site with) in the ZMI, but the above ones are what I use 90% of the time.
The portal_historiesstorage tool can eat a lot of disk space. Any content type set to save revisions saves them here, and by default Plone keeps all revisions (see the portal_purgepolicy tool).
I want all revisions on the production Data.fs, but after taking a copy for development the first thing I do is purge portal_historiesstorage. The procedure is:
Go to your Plone site in the ZMI
Delete the portal_historiesstorage tool
Go to portal_setup, Import tab
Under 'Select Profile or Snapshot' choose 'CMFEditions'
Select the step with handler Products.GenericSetup.tool.importToolset
Uncheck 'Include dependencies?'
Hit 'Import selected steps' to re-add portal_historiesstorage
Pack the Data.fs and delete the resulting Data.fs.old from the filesystem
On my 3G Data.fs, this little sequence removes 2.5G!
I have only ever done this on a development Data.fs. Without advice from someone who really knows, I don't recommend doing this on your production site.
There is usually no reason for an integrator or a developer to touch the ZMI other for possible maintenance tasks. Almost any customization can be done using Python or a GenericSetup profile. The advantage of profiles are: repeatability - being able to maintain on the filesystem - being able to put files under revision control.
Being able to work and configure stuff through the ZMI is partly working against Plone - especially when Plone is doing extra stuff under the hood. So the only recommendation can be: stay of the ZMI if you can. The ZMI is not a suitable replacement for using the Plone UI and should only be touch if you really know what you are doing.
Yep, the ZMI is for the occasional maintenance task or, when pressed, a quick-and-dirty CSS or template tweak. It's not meant for any real "coding" work, and in the context of Plone is best thought of as an odd and minimally useful leftover from Zope history.
portal_actions is also useful for more flexible top level navigation. but again best configured via gnericsetup.
I've been reading about drupal install profiles, and I'm wondering if there's much of a difference between using a packaged install profile vs. installing core + manually installing the modules listed in the install profiles?
I'd like to do the latter (manually installing each) to control the versions of each module installed, which I can't control with a packaged install profile that may not have been maintained.
But should I or will I be opening the door to something I'm not aware of? Shouldn't the 2 be identical, just one automated and other is manual?
What kiamlaluno said, plus the fact that installation profiles may perform custom configuration of settings on install, might construct custom views/content-types/etc (especially by means of features.module, which you can see heavy use of in OpenAtrium), and might provide other custom code in distro-specific module.
The short answer is, no you can't just replicate an install profile by downloading a clean drupal with all those modules -- best bet is to use the install profile. If you're worried about module versions, just make sure you're using a profile that's actively maintained.
The difference is that an installation profile includes the right version of all the modules it needs.
This means that differently from manually installing each module, you don't need to verify the correct version of the module X that effectively works together the module Y; there are few cases where one module doesn't work well when version A of another module is installed, and you need to install version B of the same module, if you don't want problems.
An installation profile can have a custom installation page that allows you to change some parameters of your site; it also allows the installation profile author to define a patch that needs to be applied a module, in order to fix a bug of the module, or to make it work better with another module.
If you need to set a site to work for a particular purpose, installation profiles are useful for you as they allow you to set the site correctly without to know all the details about how a Drupal site needs to be set.
I believe you can specify the versions of the modules you want to install see