Hello all, it’s time for the weekly migration initiative meeting. The meeting will take place in slack in various threads
This meeting:
➤ Is for core migrate maintainers and developers and anybody else in the community with an interest in migrations
➤ Usually happens every Thursday and alternates between 1400 and 2100 UTC.
➤ Is done on the #migration channel in Drupal Slack (see www.drupal.org/slack for information).
➤ Happens in threads, which you can follow to be notified of new replies even if you don’t comment in the thread. You may also join the meeting later and participate asynchronously!
➤ Has a public agenda anyone can add to here: https://www.drupal.org/project/drupal/issues/3175894. See the parent issue for an idea of the typical agenda.
➤*Transcript will be exported and posted* to the agenda issue. For anonymous comments, start with a :bust_in_silhouette: emoji. To take a comment or thread off the record, start with a :no_entry_sign: emoji.
dinarcon |
Hello :wave: Olivero and Claro as new default themes. |
Gábor Hojtsy (he/him) |
Olivero for sure :slightly_smiling_face: also more multilingual migrations landing is great to see, thanks all for your work |
damienmckenna |
Looking forward to some of the improvements in the Migrate system. |
heddn |
lazy load of images! |
mikelutz (he/him) |
Here, but a little late today. |
irinaz |
Hi, I would like to invite those who are interested to a BoF at BADCamp tomorrow at 11 Pacific time for Migrate Feeds UI BoF https://hopin.to/events/badcamp-2020-sessions |
alison |
Alison here, catching up (specifically on documentation updates)while on a quick break during an awesome Migrate API training with @dinarcon at Virtual BADCamp Nope, but I'm excited for y'all's excitement :dog_excitement: |
quietone |
Hi. I just wan t to finish the multilingual migrate meta |
benjifisher |
IIUC, Olivero is going in as experimental, not the default theme. I am not sure about Claro, but it might be the default in Umami before it becomes the default in Standard. |
benjifisher |
I am now a maintainer of the guides for the Migrate API. Thanks, @masipila! |
benjifisher |
New page: https://www.drupal.org/docs/drupal-apis/migrate-api/process-pipelines (by me). There is just one example so far, but it is a start. I wonder if I should move it to the sub-guide on process plugins, or rearrange the menu of the main guide. |
benjifisher |
New page: https://www.drupal.org/docs/8/api/migrate-api/migrate-process-plugins/ch.... Thanks, @droces! (maybe not on Slack?) It is not yet in the menu, so basically impossible to find without a link. IMO it overuses *bold* and italic a little. |
benjifisher |
Reviews and updates are welcome. Documentation is a team activity! |
benjifisher |
@damienmckenna: Can you remind me of your plans for https://www.drupal.org/docs/drupal-apis/migrate-api/migrate-field-plugins? It is a sub-guide, and so far there is just one (overview) page. Maybe start by creating some stub pages? That will make it easier for others to contribute. Also, that is the only sub-guide that I do not have permissions to administer. Can you add me as a co-maintainer? |
damienmckenna |
I've added you. |
damienmckenna |
For the field plugins pages, I was hoping to have the first page be an overview, then provide an annotated example in a separate page; the initial page has kinda turned into a brain dump that needs reorganizing. |
alison |
Awesome new pages, Benji, thank you! |
benjifisher |
#3006750: Move memory management from MigrateExecutable to an event subscriber (edited) |
benjifisher |
Previously, I argued for some changes to make the new MemoryManager class more testable. I was outvoted. While reviewing the latest patch, I still think I was right, but now I have new reasons to recommend the restructuring. Let me see if I can change your mind, @heddn, @quietone, @mikelutz (he/him). |
benjifisher |
When we run low on memory, ...The MemoryManager class calls getMemoryUsageInBytes() (a wrapper for memory_get_usage()), saves the value in an event, and dispatches the event.The one event listener in the patch does not look at the stored usage value.If another module implements an event listener, it gets the stored value even though our event listener (or others) may have succeeded in reducing usage. |
benjifisher |
I think it makes more sense to let each event listener get the current memory usage, not whatever it was when the event was created and dispatched. |
benjifisher |
To do that, create a new class with the getMemoryUsageInBytes() and getMemoryLimit() methods (public). Save an object of that class in the event, instead of the current values. Then the event listener can check $event->getClassName()->getMemoryUsageInBytes() inistead of $event->getMemoryUsageInBytes() . |
benjifisher |
Instead of making a new class, we might add these methods to the existing Drupal\Component\Utility\Environment. |
heddn |
@benjifisher there's also a more global drupal issue for memory management. perhaps some of this could be resolved there? |
benjifisher |
This strategy gives event listeners current information, and it also makes it easier to test the MemoryManager class, as I said in my previous reviews. We just mock the Environment (or newly added) class and injuect that into the MemoryManager class when we test. |
benjifisher |
Link? |
benjifisher |
Anyway, it seems to me that we should not delegate this to a follow-up issue because once we add an event dispatcher, any changes would have to involve deprecating the old API. I want to get it right the first time. |
heddn |
well, maybe it isn't 100% what I was thinking, but #3000050: Replace \Drupal\Component\Utility\Environment::checkMemoryLimit with a PHP memory checking helper class and service is linked as a related issue |
benjifisher |
I missed that! Yes, I think that is the way to go. |
benjifisher |
The current patch already has a method that wraps memory_get_usage(), which is not in the original Environment class. IMO, we should postpone our issue on this one. Any other votes? |
heddn |
postpone! |
benjifisher |
I see that @mikelutz (he/him) already commented: #3000050-23. |
benjifisher |
#3095237: Migrate Drupal 7 date field "todate" value |
benjifisher |
I know next to nothing about field plugins, so help me out here! @heddn's latest comment:If we say the destination is datetime_range, but leave the plugin in datetime, that is going to confuse things. But we can't move it either, because of sites that didn't know they need datetime_range and don't have it installed. Not sure what to suggest/do here. |
heddn |
@benjifisher the destination module is used on field plugins to display the nice review form in migrate drupal ui. its what let you know if all your data is migrated |
benjifisher |
OK, I was about to ask what the effect of `destination_module` is. |
benjifisher |
It seems to me that there are bigger problems if datetime is enabled but not datetime_range. (edited) |
benjifisher |
Could we be setting fields to use a type from a module that is not enabled? That seems serious. |
heddn |
@benjifisher the 2 modules are distinct. Yes, we have problems. The field plugin really should be broken up into 2 plugins for this reason. But that's hard because there isn't a good way to know when to use one field plugin vs the other. |
benjifisher |
You mean, there is no good way to know until we start migrating entities that use the field, right? |
benjifisher |
It does seem complicated, but I think it is necessary. Should we mark the issue NW? (I think so.) |
benjifisher |
Can we say that when datetime_range is enabled, we use its field plugin? There must be some sort of priority or weight that determines which plugin is used. I hope. Then that plugin can extend the one from datetime or maybe pass, letting the next plugin take over. |
heddn |
@benjifisher not a bad idea. plugins have alter hooks. we could have datetime_range override the one from datetime. |
heddn |
and we should throw errors/exceptions if we try to stuff range data into datetime. |
benjifisher |
We have to do something if datetime_range is not enabled. At least log a migration message. Maybe throw an exception. |
benjifisher |
Can you add some of this to the issue and set it back to NW? |
benjifisher |
#2208401: [META] Remaining multilingual migration paths |
Gábor Hojtsy (he/him) |
It looks like that got unblocked (edited) |
benjifisher |
There are a few child issues that have been committed to 9.1 and are waiting for backports to 9.0. Once those are resolved (either committed to 9.0 or not) I think we can close the meta. Do we need to pay attention to the meta now? |
Gábor Hojtsy (he/him) |
I don’t know what exactly is left though |
Gábor Hojtsy (he/him) |
I think @quietone wrote that @wimleers (he/him) pointed out that there are still outstanding todos with all of those committed still? |
benjifisher |
#3008028 is postponed on #3112249, which is postponed on #3110669, which is RTBC. |
benjifisher |
... there are todos in core referring to this issue.I think that means that there are @todo comments in core. |
benjifisher |
$ ag -ri 2208401 corecore/modules/config_translation/migrations/state/config_translation.migrate_drupal.yml13: # See #2208401: [META] Remaining multilingual migration paths |
benjifisher |
We should not close the meta issue without removing these comments. I guess. |
Gábor Hojtsy (he/him) |
Yup |
quietone |
I want to check the state of the translation of user profiles before making a decision on closing the meta. Working on this is my top priority. |
Gábor Hojtsy (he/him) |
Thank you @quietone |
quietone |
I've done my research and closed the Meta! #2208401: [META] Remaining multilingual migration paths |
benjifisher |
@quietone: What about the # See #2208401: [META] Remaining multilingual migration paths comments in the YAML files? |
quietone |
That is in the followup, #3175729: Mark i18n migrations as finished |
dinarcon, Gábor Hojtsy, damienmckenna, heddn, mikelutz (he/him), irinaz, alison, quietone, benjifisher
Comments
Comment #2
benjifisherComment #3
benjifisherComment #11
quietone CreditAttribution: quietone as a volunteer commentedComment #12
quietone CreditAttribution: quietone as a volunteer commentedComment #13
benjifisherComment #14
benjifisherComment #15
benjifisher