| # |
Jul 11th 2017, 14:42 |
 admad |
if one is already using 3.x i wouldn't recommend wasting time updating the deprecated methods until 4.0. |
| # |
Jul 11th 2017, 14:41 |
jeremyharris |
I see your point though. the getter/setters are a lot of changes; but having to do ALL of those immediately is much more daunting than doing the other things first, imo |
| # |
Jul 11th 2017, 14:41 |
neothermic |
true |
| # |
Jul 11th 2017, 14:41 |
admad |
@neothermic sure, but no point changing to already deprecated methods and waste time fixing them again in future |
| # |
Jul 11th 2017, 14:40 |
jeremyharris |
I just found it’s generally easier to take small steps - also mentally getting rewarded is a benefit |
| # |
Jul 11th 2017, 14:40 |
neothermic |
@admad Hundreds of thousands of lines of code can't be changed in a night ;) |
| # |
Jul 11th 2017, 14:40 |
admad |
@jeremyharris why stay behind the curve when it's possible to get on top right now? |
| # |
Jul 11th 2017, 14:39 |
neothermic |
I'm possibly going to be using the shim plugin to go into 3.x, as there's otherwise too many changes to make, and then slowly change things to not need the shim |
| # |
Jul 11th 2017, 14:39 |
jeremyharris |
you don’t need to worry about those until 4.x though, right? |
| # |
Jul 11th 2017, 14:38 |
admad |
@neothermic i say directly form 2.x to 3.4, that way you wont have to waste time again changing things like request->param() to request->getParam() as former was deprecated |
| # |
Jul 11th 2017, 14:36 |
jeremyharris |
well consider 3.x trial by fire for writing your tests |
| # |
Jul 11th 2017, 14:36 |
neothermic |
There's many reasons why our 1.3 -> 2.x upgrade took ~3 months |
| # |
Jul 11th 2017, 14:36 |
jeremyharris |
:) |
| # |
Jul 11th 2017, 14:36 |
jeremyharris |
uh oh |
| # |
Jul 11th 2017, 14:36 |
neothermic |
`run your tests` yeaaaaahhhh... |
| # |
Jul 11th 2017, 14:36 |
jeremyharris |
you only have to read a small doc at a time, run your tests, move on |
| # |
Jul 11th 2017, 14:35 |
jeremyharris |
that’s what I did. most of the minor versions go really quick, but it ensures you don’t miss anything |
| # |
Jul 11th 2017, 14:35 |
neothermic |
so from 2.x to 3.0, and then up each 3.x to latest? |
| # |
Jul 11th 2017, 14:35 |
jeremyharris |
I recently upgraded a small to medium app from 2.6 to 3.4, minor at a time |
| # |
Jul 11th 2017, 14:34 |
jeremyharris |
probably best to go minor versions at a time. it’s a lot less to think about and the migration docs are pretty solid on them |
| # |
Jul 11th 2017, 14:34 |
neothermic |
Yeah, I'm just debating how far into 3.x I go from 2.x; I possibly wont' go right to 3.6 if there's depreciations (for 4.0) in it |
| # |
Jul 11th 2017, 14:33 |
jeremyharris |
yeah, much easier to fix the deprecations along the way. the 3.x change will be massive though, so plan time for it. well worth it after the fact |
| # |
Jul 11th 2017, 14:32 |
neothermic |
So I've got a cleaner upgrade path |
| # |
Jul 11th 2017, 14:32 |
neothermic |
but this is why I also want to get past 2.8 :slightly_smiling_face: |
| # |
Jul 11th 2017, 14:28 |
neothermic |
Yeah, the 2.x -> 3.x change is something we're going to have to tackle in our codebase at some point before 2.x is moonlighted |
| # |
Jul 11th 2017, 14:27 |
inoas |
there is no reason to change everything... the 2.x -> 3.x break and the middleware change is already a huge change alltogether |
| # |
Jul 11th 2017, 14:26 |
inoas |
$options schema validator |
| # |
Jul 11th 2017, 14:26 |
inoas |
NullInflector, AutoTables warning/exception, strict types, etc |
| # |
Jul 11th 2017, 14:26 |
inoas |
and I would just focus on making it less "trappy" |
| # |
Jul 11th 2017, 14:26 |
inoas |
I like how cakephp3 is |
| # |
Jul 11th 2017, 14:26 |
jeremyharris |
yeah that’s been the best thing the framework has done in its maturing - move away from too much magic |
| # |
Jul 11th 2017, 14:26 |
inoas |
hmic well at leas that's a "4.0 must have" feature... unless all that gibberish |
| # |
Jul 11th 2017, 14:25 |
hmic |
like the autotables warning lately |
| # |
Jul 11th 2017, 14:25 |
hmic |
there should be an option in debug: "show me all the magic i'm consuming" |
| # |
Jul 11th 2017, 14:24 |
neothermic |
Yeah, sometimes magic is handy. sometimes you work against the magic :) |
| # |
Jul 11th 2017, 14:18 |
jeremyharris |
was a Good Idea At The Time |
| # |
Jul 11th 2017, 14:18 |
jeremyharris |
if I recall correctly I’m responsible for some of that mess :P |
| # |
Jul 11th 2017, 14:17 |
neothermic |
yikes :slightly_smiling_face: |
| # |
Jul 11th 2017, 14:15 |
jeremyharris |
here’s the hunk if you want to figure it out: https://api.cakephp.org/1.3/source-class-FormHelper.html#255-282 |
| # |
Jul 11th 2017, 14:14 |
admad |
yes, 1.x did weird internal magic to decide which action/url to use based on request data |
| # |
Jul 11th 2017, 14:14 |
neothermic |
that might explain some of the redundant ones in this code |