mglaman,
@mglaman@phpc.social avatar

MATT, WHAT HAVE YOU DONE?

At @midcamp, I gave a lightning talk about the greatest and worst idea I have had for four years.

An interoperability layer for to keep running legacy code.

mglaman/drupal-bridge is coming soon!

what if you could migrate from D7 -> D10 without rewriting all of your code? what if we only needed to migrate data and refactor legacy code over time? how many Drupal 7 sites could this unblock?

Picture of Drupal 7 taxnomy_menu module installed on Drupal 10
Screenshot of Drupal 7 module's database table being created (okay, this kind of just worked without trickery)

ambientimpact,
bsnodgrass,
@bsnodgrass@drupal.community avatar

@mglaman @midcamp great idea!!

phenaproxima,
@phenaproxima@drupal.community avatar

@mglaman @midcamp Nice. I'd had this same idea kicking around, but I never bothered to try it.

jon,

@mglaman @midcamp

Why stop there? Some poor souls still have Drupal 6... Drupal 5... 4... 3... 2... 💥

mglaman,
@mglaman@phpc.social avatar

@jon 💀

rimu,
@rimu@mastodon.nzoss.nz avatar

@mglaman Isnt migrating the data the main problem for most sites?

mglaman,
@mglaman@phpc.social avatar

@rimu that's a good question. I honestly don't think so. I think it's having to rewrite all of the custom code. Could be a mix a both.

rimu,
@rimu@mastodon.nzoss.nz avatar

@mglaman Yeah, it varies. I have one site that has 10 years of custom code. A dozen content types, lots of fields and 30k nodes. It's a monster. Your idea could help a lot but probably not enough to avoid upgrading to Backdrop.

Every other Drupal site I've ever made had only a few lines of custom code and a whole lot of theming work (templates and CSS). Migrating the content and recreating the theme is the issue there.

mglaman,
@mglaman@phpc.social avatar

@rimu moving to Backdrop requires a migration still, yeah? Or is it literally drop-in?

rimu,
@rimu@mastodon.nzoss.nz avatar

@mglaman Mostly a drop in, like upgrading from D6 to D7 was. Also porting a Drupal module to Backdrop is mostly automated by the Coder module.

A pretty attractive option.

marvil07,

@mglaman @rimu Assuming this is a good idea 😉 , it makes sense to focus on the custom code, given Migrate API can help a lot on the other side.

cellear,

@mglaman @rimu

A survey of holdout D7 site owners is desperately needed to answer questions like these. Who are they, and why haven't they upgraded?

"Having to rewrite custom code" is also my prediction for the main impediment, though "baffled by Composer" might be in there somewhere as well.

stpaultim,
@stpaultim@fosstodon.org avatar

@mglaman @midcamp

This is getting interesting.

geerlingguy,
@geerlingguy@mastodon.social avatar

@stpaultim would be huge, especially if you could get like 80% of the top 100 D7 modules working without completely breaking.

mglaman,
@mglaman@phpc.social avatar

@geerlingguy @stpaultim LET THE GAME BEGIN!

Yeah. I figured if we could say "just migrate to the new data tables but your existing code will just work" it'd be huge.

db_select(term_data) catches that table name and translates to new entity base table for example

ultimike,
@ultimike@drupal.community avatar

@mglaman is a mad scientist. In a good way. I think.

cspitzlay,

@mglaman @midcamp

Will it support the content destruction kit (https://www.drupal.org/project/cdk) which requires the D7 version of Bad Judgement (https://www.drupal.org/project/bad_judgement)?

  • All
  • Subscribed
  • Moderated
  • Favorites
  • php
  • ngwrru68w68
  • GTA5RPClips
  • osvaldo12
  • everett
  • khanakhh
  • rosin
  • InstantRegret
  • DreamBathrooms
  • thenastyranch
  • magazineikmin
  • Youngstown
  • mdbf
  • slotface
  • kavyap
  • JUstTest
  • modclub
  • cisconetworking
  • cubers
  • anitta
  • Leos
  • Durango
  • ethstaker
  • normalnudes
  • tester
  • megavids
  • tacticalgear
  • provamag3
  • lostlight
  • All magazines