Drupal 9 readiness meeting / 17 February 2020 [#3112536]

Meeting will happen in #d9readiness on drupal.slack.com.

Hello and welcome to this Drupal 9 readiness meeting!

This meeting:
➤ Is for core and contributed project developers as well as people who have integrations and services related to core. Site developers who want to stay in the know to keep up-to-date for the easiest Drupal 9 upgrade of their sites are also welcome.
➤ Usually happens every Monday at 19:00 UTC.
➤ Is done over chat.
➤ 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: `https://www.drupal.org/project/drupal/issues/3112536`
➤*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.

0️⃣ Who is here today? Comment in the thread below to introduce yourself.

mradcliffe	Matthew here.
Gábor Hojtsy (he/him)	Gábor, Drupal 9 coordinator
dww	Derek
dan2k3k4	Dan
greg.1.anderson	Greg, SF, sort of on vacation
chx (he/him)	Charlie
mikelutz	Mike, Migration subsystem maintainer.
shaal	Ofer Shaal, Umami / Drupal Rector
vijaycs85	Vijay
mixologic	howdy, here lates
damienmckenna	Damien, a little late.
thalles	thalles (edited)
longwave	Dave, catching up
larowlan	Lee
jaykandari	Jaideep here,Will be checking out the discussions of last meeting.Meeting timing seems a little off for +5.30 timezone peoples:pray::skin-tone-3:
wimleers (he/him)	:wave: Wim, catching up afterwards :slightly_smiling_face:

1️⃣ Do you have suggested topics you are looking to discuss? Post in this thread and we’ll open threads for them as appropriate.

dan2k3k4	"Showcase" - Has anyone got any working Drupal 9 sites?I'd like to try to upgrade https://drupal.ch - but it requires also patching at least 12 contrib modules (which will help them to get ready for D9 anyway :smile:) but I'm pretty busy for the next week or two at least (edited)
greg.1.anderson	Anyone have an update on how we're doing release-date wise? I am guessing we are too far off from the beta requirements for 28 Feb to be likely per https://www.drupal.org/core/release-cycle-overview
mradcliffe	The popper.js v1 to v2 issue ran into what might be a blocker with #3112800: Library dependencies are not taking library versions into account because there are way too many changes to do any sort of BC.
mradcliffe	#3108402: Update to Popper.js to 2.0.0
mradcliffe	#2550717: [JS] Replace jQuery.cookie with JS-cookie and provide a BC layer has some minor fixes need to get it back to Needs review and hopefully RTBC
chx (he/him)	Outreach. How are we going to reach Drupal 7 sites explaining upgrading to Drupal 9 is the thing to do, it's not a "tabula rasa" like Drupal 8 was. A good 70% of sites reporting back to drupal.org are still on D7 , more than four years after the release of D8.

2️⃣ “Showcase” - Has anyone got any working Drupal 9 sites?

Gábor Hojtsy (he/him)	@dan2k3k4 suggested this topic
Gábor Hojtsy (he/him)	I would like to underline that the alpha is not supposed to be used on public sites, also no upgrade path is provided and ALL RISKS ARE YOURS.
Gábor Hojtsy (he/him)	@dan2k3k4 wrote:I’d like to try to upgrade https://drupal.ch - but it requires also patching at least 12 contrib modules (which will help them to get ready for D9 anyway :smile:) but I’m pretty busy for the next week or two at least(edited)
Gábor Hojtsy (he/him)	I know of https://matthewgrasmick.com/posts/screencast-updating-drupal-9-15-minutes
Gábor Hojtsy (he/him)	is not a live site :slightly_smiling_face:
Gábor Hojtsy (he/him)	Matthew updated his simple blog to Drupal 9 alpha1 locally
dan2k3k4	oh yeah totally agree with the alpha, so maybe it might not be a good idea to start talking about it publically - in any case for drupal.ch it has no members and is just 2-3 node pages so it's "low risk"... I wouldn't recommend upgrading anything that's actually "client related"
Gábor Hojtsy (he/him)	@dan2k3k4 yeah we added the “try upgrade sites” part to the release notes, so people start surfacing issues now rather than later
Gábor Hojtsy (he/him)	I think @berdir had a great point that we should not encourage it too much because people may get carried away in issue queues
Gábor Hojtsy (he/him)	I personally think attempting local updates is a good use of time :slightly_smiling_face:
dan2k3k4	hmm yeah
dan2k3k4	especially if it ends up being: "oh look 12 contrib modules I can patch" - and I'd be more motivated to patch those modules then some random module I find on say the deprecation table that I've never used before
berdir	You can not patch modules for that, not for the initial d9 compatibility definition
Gábor Hojtsy (he/him)	@berdir yeah @grasmash features a composer version alias trick in his video for that + patches
berdir	I see, that works, but that's pretty tedious :slightly_smiling_face: Anyway, I fully agree, you want to work on modules that your site is using. But just installing upgrade_status on a local/test site is imho going to give you much more useful results to work with than trying to actually update to D9.
berdir	also, I'd argue that at this point, what needs the most help isn't contrib, it's core, contrib still has a few months to catch up until the release is really here, and there is absolutely no benefit to updating before a stable release, there are zero new features. but for core, there's a considerable time pressure to be able to release beta asap
berdir	additionally to what upgrade_status reports, we are also tracking all contrib modules that we use in a simple spreadsheet. I'm considering to share that, currently has some internal info in it too, but progress is really good. from those ~110 projects, almost 50 have a d9 compatible release or development snapshot, another 30 have a complete and RTBC patch many of them need 8.8 which is why they're "blocked"
Gábor Hojtsy (he/him)	@berdir re installing upgrade status rather than trying Drupal 9 I think @grasmash went to try it out to validate the marketing :smile:
Gábor Hojtsy (he/him)	also same reason we did the Johanna case study for Dries for Amsterdam last year :slightly_smiling_face:
Gábor Hojtsy (he/him)	to explain the hands on steps so its evidently much easier than Drupal 7 to 8 :slightly_smiling_face:
berdir	yeah, sure, I get that you want to try that to validate the marketing claims. but I maintain that really is too early for trying to actually update anything but the most trivial site. And I'm the crazy person who launched large news portals on Drupal 8 alpha :slightly_smiling_face:
berdir	because unlike Drupal 8 alpha, there are no benefits from being in Drupal 9 early, and even though we support updating from 8.8, we should make it really clear that for 99% of the sites out there, the by far best option is updating to 8.9 and latest versions of all your contribs and then start trying to update to 9 (edited)
Gábor Hojtsy (he/him)	@berdir I hear you, will consider this when forming messaging :slightly_smiling_face:

3️⃣ Update on how we’re doing release-date wise

Gábor Hojtsy (he/him)	@greg.1.anderson raised this great question :slightly_smiling_face:
Gábor Hojtsy (he/him)	he also wroteI am guessing we are too far off from the beta requirements for 28 Feb to be likely per https://www.drupal.org/core/release-cycle-overview
greg.1.anderson	I'll admit I have not been tracking this super closely myself; it seems that the deprecation removals are steadily ramping down, but there are still a bunch left, and other things as well
greg.1.anderson	Reference: #3079680: [META] Requirements for tagging Drupal 9.0.0-beta1
greg.1.anderson	But, yay, Drupal 9.0.0-alpha1 :tada: (edited)
Gábor Hojtsy (he/him)	@greg.1.anderson yeah https://www.drupal.org/project/issues/search?projects=&project_issue_fol... is 3 more than one page, so 53 issues that would need to get done by end of next week, so it is starting to look like we’ll not make end of next week and thus we will not make the June 3 release date either
Gábor Hojtsy (he/him)	@greg.1.anderson it is absolutely true that there was renewed activity with the alpha1
mradcliffe	That issue should probably link to the tag.
mradcliffe	Or I am not seeing it correctly.
Gábor Hojtsy (he/him)	so its not impossible, but it would take a lot of people all at once to be focused on this, which does not seem realistic ATM :slightly_smiling_face: we can be more certain by the end of this week, when we wanted to reasses the situation
Gábor Hojtsy (he/him)	@mradcliffe good point, will add link
chx (he/him)	53 issues... realistically, the august 5 target date will require a push
chx (he/him)	DDD?
chx (he/him)	DDD is just on time for a final push for the Aug 5 target date.
Gábor Hojtsy (he/him)	@chx (he/him) the next beta target date would be end of April so DDD would be a big event to get together and make big moves :slightly_smiling_face:
Gábor Hojtsy (he/him)	That said, I’ll also be in London in a month alongside @alexpott and @lauriii and possibly @catch :slightly_smiling_face:
Gábor Hojtsy (he/him)	for DrupalCamp London :slightly_smiling_face:
chx (he/him)	that's quite fortunate to start the wave and then we can pick up the leftovers in Ghent
chx (he/him)	that looks way more realistic than somehow push fifty three core issues through by end of next week
chx (he/him)	(I just checked a page really is 50 issues.) (edited)
mradcliffe	Yes.

4️⃣ Outreach. How are we going to reach Drupal 7 sites explaining upgrading to Drupal 9 is the thing to do, it’s not a “tabula rasa” like Drupal 8 was. A good 70% of sites reporting back to drupal.org are still on D7 , more than four years after the release of D8.

Gábor Hojtsy (he/him)	@chx (he/him) raised this
Gábor Hojtsy (he/him)	so Drupal 9 is a tabula rasa from Drupal 7, much like Drupal 8 was, so that directly cannot be avoided for users we can convince to move, but we are focusing our communication on telling the story that Drupal 8 onwards is NOT a reinvention of Drupal
Gábor Hojtsy (he/him)	my open source State of Drupal 9 slide deck starts off by explaining how Drupal innovates (not with major releases), and how major releases are a cleanup now not clean slates http://hojtsy.hu/blog/2019-jun-07/state-drupal-9-may-2019-session-recording
mixologic	Im curious what we know about our existing d7 users.
chx (he/him)	What I meant is that when D8 was released there were no contrib modules and knowledge either
chx (he/him)	Drupal 9 is different
vijaycs85	I know, it could be asking a lot, but is it worth considering some form of estimation tool that would provide approx effort from 7 to 9?
vijaycs85	and get some anonymous contrib/core version details to find the weak spots?
chx (he/him)	@mixologic excellent question! Do we know anything :slightly_smiling_face: ? Outreach has always been one of my biggest headaches.
mradcliffe	I haven't migrated my own site from Drupal 7 because I don't have a content migration path to Drupal 8 or 9 yet.
mixologic	As far as the update stats go, I can tell you that they are at best a blurry trend, but that we should not put a huge amount of weight on the actual numbers themselves.
mikelutz	I’m not sure the messaging need to change dramatically right now. If I’m completely honest, I would recommend people migrate from D7 to 8.9 until 9.1 comes out.
mixologic	We all likely have a lot of theories about who/what those d7 users represent, and what the effort and obstacles are to moving forward. Im curious if anybody else might have some real data (i.e. Acquia's customers/Pantheon etc)
chx (he/him)	webchick used to run informal polls on twitter about obstacles in upgrading to drupal 8
mradcliffe	Maybe coming up with a standard survey that companies can ask / check-in with their customers directly?
mixologic	For example, from my perspective, many d7 sites are enormously complicated with a lot of custom work in both the theme side and the custom code side, with a lot of complex business logic embedded in places that maybe they shouldnt have been embedded.
mixologic	That certainly is the case for drupal.org
mradcliffe	And getting the same site on Drupal 8 requires time to re-implement views and layouts.
chx (he/him)	https://twitter.com/webchick/status/1113139464849592320 here's one
chx (he/him)	Yes -- a lot of Drupal 7 sites just can't upgrade to Drupal 8 they essentially need to rebuild the entire site from ground up.
mixologic	Sure. but to be honest, Im willing to bet that of the d7 sites that are remaining in our stats, that 98% of those d7 site owners wouldnt know who webchick is or follow her on Twitter etc.
mradcliffe	I agree. Sadly not everyone follows webchick.
mixologic	So we end up with results that are heavily biased towards users who are already insiders, basically.
mradcliffe	If we could get results that are statistically significant, then we can extrapolate and make a strategy.
mradcliffe	Is there a next step here with regard to outreach?
mixologic	To me its, "find more people to reach out to, somehow"
chx (he/him)	I guess @Gábor Hojtsy (he/him) could shop around at Acquia maybe?
chx (he/him)	I heard Acquia has a few customers running Drupal :wink:
mixologic	Right, have any companies done internal polls of their customers?
chx (he/him)	Excellent question.
Gábor Hojtsy (he/him)	we (Acquia) did polling on our partners but we got (to me) shockingly low participation — on questions specific to migrating away from Drupal 7, so the pressure is probably not felt that much yet(?)
mixologic	But this loops back to the original question: Outreach. If we did have a list of site owners we could communicate to, what is it we're hoping to communicate? What is the outcome we're hoping for?
mradcliffe	I would like to find out 1. What are blockers to getting to Drupal 8 (or 9)? 2. What site owners' awareness of resources for getting to Drupal 8 (or 9) with the goal of introducing methods to site-owners for 2 by polling them and trying to get focus for things to work on for 1.
mradcliffe	if 50% of responders were blocked on 1-3 areas, then maybe that's a big call to make critical / important.
greg.1.anderson	One thing to keep in mind is that all of these Drupal sites that are still on 7 have had the opportunity to upgrade to 8, and haven't done so yet. The one main thing that is different about the approaching release of 9 is that also brings closer the approaching EOL of 7. So, I think the main messaging would be:There's more contrib nowDrupal 9 is more compatible with Drupal 8 -- Drupal is moving away from the upgrade patterns of the past, which required a lot of rework / migrations.
mikelutz	Even if the migration system were perfect (and I know it isn’t) you still need to rebuild your theme from scratch, which means you likely need a developer, and you are probably going to pay more than you did to have your d7 site built in the first place.
greg.1.anderson	Yeah, the 7-to-9 migration isn't going to be much different than the 8-to-9 migration, but it might be some consolation that future major upgrades will be easier
greg.1.anderson	Contrib will continue to be an issue, though. If modules are using deprecated methods, and are not well maintained, their availability could still be delayed.
mixologic	Right. I think that, given a great poll with great participation, we might find out that the responders are blocked on things that we dont really have the ability to affect, like 1. complexity of existing site 2. cost to upgrade that site 3. value propisition for what they're going to get for that upgrade cost.
greg.1.anderson	Even if a module uses no deprecated methods, someone still needs to notice and make a new release that declares compatibility with 9.
Gábor Hojtsy (he/him)	@greg.1.anderson it is definitely a core of the messaging that with 8 and 9 you get on a train that has stops and not need to be disassembled and moved to another track in pieces by hand >D
Gábor Hojtsy (he/him)	(Dries literally had slides about that in Seattle :D)
Gábor Hojtsy (he/him)	@greg.1.anderson re contrib, I think it could be a good opportunity to move ownership of some, we’ll need to facilitate that
mikelutz	This is why I’m thinking there is no need to push people to D9 on the day it is released. While the upgrade path for contrib is much much easier than it ever has been, I still think it will be 9.1 before contrib is really really caught up and ready.
Gábor Hojtsy (he/him)	@mikelutz yeah it would still be a really hard sell to tell people to update to Drupal 8.9 first, while 9.1 will still be well within the support time of Drupal 7
Gábor Hojtsy (he/him)	however easy it is to move from Drupal 8.9 to 9.1 it is still more work than not needing to move at all because you just jump to 9.1 straight
Gábor Hojtsy (he/him)	https://www.ou.edu/deptcomm/dodjcc/groups/99A2/curve.JPG(14 kB)https://www.ou.edu/deptcomm/dodjcc/groups/99A2/curve.JPG
Gábor Hojtsy (he/him)	so the early majority would not arrive IMHO before 9.1 either way
mikelutz	I know, and I realize the messaging is that it’s all easy and doesn’t matter, just get abord the modern drupal train. I do think the reality is that most sites are probably going to be best holding off for 9.1, or at least late in the 9.0 cycle to give contrib more time to be caught up, and a couple bugfix releases of D9 out.
mikelutz	Whether they are coming from D7 or D8. Of course, D7 to 9.1 is far harder because a d7->d8/9 migration might have a 6+ month development time
mikelutz	At the end of the day, it’s going to come down to how much of contrib has a D9 ready release, and when will there be a critical mass of contrib that’s ready.
andypost	What I most often hear from D7 users (mostly small sites) iscomplexity - D7 does not require a lot technical knowledge and composer is main blocker on their way (they used to download tarball and run)missing modules - there's no straight upgrade path for most of small modules which are used by small sitesD8 is for big sites - this mantra all over interwebs
andypost	Also Drupal 7 was packaged by a lot of linux distros and I see no sane way to do that for 8
longwave	Absolutely agree with @mixologic and @chx (he/him) around rebuilding and custom code; also, it is a hard sell to customers when they see large costs and little benefit - as far as they are concerned their Drupal 7 sites still work, so why upgrade?
Gábor Hojtsy (he/him)	yeah the long terms support program will probably keep a lot of sites
andypost	I bet stats for core is clearly shows that remaining D7 sites (mostly) are this ones which have no technical owners/team behind it and just "hope" that Drupal will start think about this small sites
Gábor Hojtsy (he/him)	yeah there is nothing collected that would show anything like that clearly :smile:
andypost	I mean amount of D7 sites with high complexity which is not migrated to 8 is not comparable to small ones
andypost	Another outreach is required to convince "hosting panels" to ship D8 but that's the same as packaging D8 to rpm/deb - require lots of effort
Gábor Hojtsy (he/him)	and keeping it up to date especially :slightly_smiling_face:
longwave	I wonder if we can get something into either next 7.x release or Drupal 7 update status that points out to users that end of life is coming? Many end users probably have no direct contact with drupal.org and little idea that support will be ending (LTS excepted)
mixologic	Maybe not in d7, but thats exactly why we've been discussing doing something like that in d8: #2940739: Project messaging channel in core initiative
Gábor Hojtsy (he/him)	@longwave yeah that is a good idea
andypost	@mixologic I think we need this for D7 to get telemetry per module/theme, I recall there was issue about it...
mixologic	@andypost? per module/theme? do you mean submodules vs project install info?
andypost	yes, exactly
andypost	OTOH maybe there's way to get stats from Google checking for known D7 files, like tools used to find "gedon-affected" ones
mixologic	I mean, we have that data, we just havent put it out publicly in a digestable format, but thats not really relevant to this discussion as far as I can tell
mixologic	What would that tell us?
andypost	Why not relevant? it will tell at least how many d7 sites using standard theme (garland or bartik) (edited)
mixologic	Oh, okay, that is interesting. If they are using garland or bartik we could make assumptions about the complexity of the site.
mixologic	I hadnt considered that.
mixologic	I'll look that up right now
mixologic	(it might still not be very useful data because IIRC the updates module will say if a module is enabled but for d7 themes, you could have several enabled, but only be using one)
mixologic	So, this isnt all that informative, because sites will report on any theme that is enabled.
mixologic	about 215k sites have either garland or bartik, or both enabled in d7.
mikelutz	Actually that’s less than I would have expected.
mixologic	out of about 710k sites (edited)

5️⃣ The popper.js v1 to v2 issue ran into a possible blocker of library versioning because the two versions are so different.

Gábor Hojtsy (he/him)	@mradcliffe raised this topic
Gábor Hojtsy (he/him)	Library versioning is at #3112800: Library dependencies are not taking library versions into account
Gábor Hojtsy (he/him)	And the Popper update is #3108402: Update to Popper.js to 2.0.0
mradcliffe	js libraries take the same namespace, and there doesn't seem like a way to make things backwards-compatible. Is it okay to put this burden on any one implementing popper by trying to ease that burden with documented error messages in the change record?
Gábor Hojtsy (he/him)	how wide do we expect the implementors be?
mradcliffe	If any, it would probably be a custom implementation and not a contrib. implementation.
mradcliffe	So I don't know if there's a way to track that.
Gábor Hojtsy (he/him)	@mradcliffe I think its totally ok to put it on the integrator to do this, I don’t know how much uptake did pooper had, and we are apparently spending a lot of cycles figuring it out, so I hope it is noticable :smile:
berdir	so if I understand that correctly then popperjs was only added to core in 8.8. there is a contrib module that provided the same, but it had like 6 installations. so really doesn't look like a common use case. I think we were at some point investigating to use it in paragraphs, but I wasn't directly involved in that
berdir	but it being in being in core could certainly make that more interesting now, js library dependencies in contrib module are a pain :wink:

6️⃣ “Replace jQuery.cookie with JS-cookie and provide a BC layer” needs some minor fixes, review and RTBC

Gábor Hojtsy (he/him)	#2550717: [JS] Replace jQuery.cookie with JS-cookie and provide a BC layer
Gábor Hojtsy (he/him)	was raised by @mradcliffe
mradcliffe	I'll probably get to this some time this week if someone else doesn't do so.
Gábor Hojtsy (he/him)	it will still need reviewers :slightly_smiling_face: anyone up for that?

7️⃣ Events!

7️⃣ .1️⃣ https://www.fldrupal.camp/ is next Fri-Sun and will have people working on Rector (automated fixes for Drupal 8->9) and Olivero theme (edited)

Gábor Hojtsy (he/him)	See @shaal, @mherchel and @Putra Bonaccorsi there
Gábor Hojtsy (he/him)	Shaal for drupal-rector, the other two folks for Olivero (new Drupal 9 frontend theme)
mherchel	:partyparrot:
shaal	:tada:
mherchel	Super excited for this
shaal	We'll have plenty of work for both back-end and front-end developers :slightly_smiling_face:

7️⃣ .3️⃣ https://www.midcamp.org/ is March 18-21 with Drupal 9 content and contribution (edited)

Gábor Hojtsy (he/him)	@mikelutz will present “Preparing custom and contributed code for Drupal 9”
Gábor Hojtsy (he/him)	@Dan will present “Automate Drupal 9 upgrades: Addressing deprecated code as a community”
mikelutz	Yeah, if anyone has anything specific they want to get the word out on in mine, let me know. I’m primarily going to talk about finding and fixing deprecated custom code, but I’m hoping to have time to briefly touch on how to check on the status of your contributed modules and help move them along if they aren’t ready yet.
drumm	I’ll be there, maybe sprinting on what’s left of semver or maybe issue forks.