Cherry-picking commits from master to 2.066 branch (for RC2)#3851
Cherry-picking commits from master to 2.066 branch (for RC2)#3851AndrewEdwards merged 5 commits intodlang:2.066from
Conversation
[REG2.066a] Issue 13259 - [ICE] 'v.result' on line 191 in file 'todt.c'
[REG2.061] Issue 13252 - ParameterDefaultValueTuple affects other instantiations
Issue 13225 - [ICE] Access violation on invalid mixin template instantiation
fix Issue 13237 Wrong code with -inline -O
|
How many RCs are you planning? |
|
I am afraid it is not as much about planning - we can't make release with regressions :( And there is still quite a list here : http://wiki.dlang.org/Beta_Testing |
|
There are five outstanding regressions. Assuming no new ones are reported after the release of RC2 and all remaining regressions are addressed, I will be releasing one more. But as Dicebot said, this is purely based on the number of outstanding regressions. |
Issue 11453 - Compiling packages has a dependency on order of modules passed to the compiler
|
@AndrewEdwards This PR already contains #3840. And I cherry-picked #3834. |
|
Thanks @9rnsr. |
This is completely unrealistic. (I assume you mean known regressions). You can solve this by having shorter development cycles. How long has |
|
... the beta / rc been going on for now? At this rate you might as well branch master as 2.067-beta immediately after 2.066 goes out the door. (using phone with sensitive screen) |
Something like this is actually going to happen (1 month was initially planned but with current rate it can be reduced). There is really no good way out of it because releasing with known regressions is just too bad and only few developers pay attention to beta regression list (the fact that all existing ones are out of my competence does not help really) |
|
@ibuclaw, interestingly enough I did not see your comments here: 2.066 was to be released on 14 July according to my plans. However, the community spoke (at least the handful that normally does) and this is their desire. But, as you've suggested, I will more than likely be branching 2.067 after releasing 2.066 in order to get back on track. |
|
Would it be possible to get #3844 into 2.066? It might be the cause for some AA problems that are showing up in our testing server. |
Cherry-picking commits from master to 2.066 branch (for RC2)
No description provided.