Re: Renaming `log4j-core`
> On Jun 25, 2023, at 11:32 PM, Piotr P. Karwasz > wrote: > > Hi Tim, > > On Sat, 24 Jun 2023 at 22:00, Tim Perry wrote: >> >> Can we publish log4j-core so it spits out a deprecation warning and just >> pulls in log4j-impl or log4j-runtime or whatever? > > This is probably the best solution. How can it be implemented? > We could also consider adding other dependencies to `log4j-core` that > were split into separate modules, like `log4j-smtp`. I’m not sure I understand what you are getting at here. Ralph
Re: Renaming `log4j-core`
Hi Ralph, On Mon, 26 Jun 2023 at 20:33, Ralph Goers wrote: > > On Jun 25, 2023, at 11:32 PM, Piotr P. Karwasz > > wrote: > > > > Hi Tim, > > > > On Sat, 24 Jun 2023 at 22:00, Tim Perry wrote: > >> > >> Can we publish log4j-core so it spits out a deprecation warning and just > >> pulls in log4j-impl or log4j-runtime or whatever? > > > > This is probably the best solution. > > How can it be implemented? We replace `log4j-core` with an empty JAR file with `log4j-impl` as dependency. > > We could also consider adding other dependencies to `log4j-core` that > > were split into separate modules, like `log4j-smtp`. > > I’m not sure I understand what you are getting at here. My idea was to add as dependencies of `log4j-core` all the 3.x modules that were split off it. This way users would have the same `log4j-core` as in 2.x. I am however not convinced it could work, since 2.x optional dependencies would become non-optional deps. Piotr
Re: Renaming `log4j-core`
> On Jun 26, 2023, at 12:17 PM, Piotr P. Karwasz > wrote: > > Hi Ralph, > > On Mon, 26 Jun 2023 at 20:33, Ralph Goers wrote: >>> On Jun 25, 2023, at 11:32 PM, Piotr P. Karwasz >>> wrote: >>> >>> Hi Tim, >>> >>> On Sat, 24 Jun 2023 at 22:00, Tim Perry wrote: Can we publish log4j-core so it spits out a deprecation warning and just pulls in log4j-impl or log4j-runtime or whatever? >>> >>> This is probably the best solution. >> >> How can it be implemented? > > We replace `log4j-core` with an empty JAR file with `log4j-impl` as > dependency. I can’t say I am thrilled with this idea but I also can’t point to any concrete problems it would cause without testing. > >>> We could also consider adding other dependencies to `log4j-core` that >>> were split into separate modules, like `log4j-smtp`. >> >> I’m not sure I understand what you are getting at here. > > My idea was to add as dependencies of `log4j-core` all the 3.x modules > that were split off it. This way users would have the same > `log4j-core` as in 2.x. I am however not convinced it could work, > since 2.x optional dependencies would become non-optional deps. I think that would be a horrible idea. It defeats the whole point of splitting them out. Ralph
Re: Renaming `log4j-core`
> On Jun 26, 2023, at 1:38 PM, Ralph Goers wrote: > > > >> On Jun 26, 2023, at 12:17 PM, Piotr P. Karwasz >> wrote: >> >> Hi Ralph, >> >> On Mon, 26 Jun 2023 at 20:33, Ralph Goers wrote: On Jun 25, 2023, at 11:32 PM, Piotr P. Karwasz wrote: Hi Tim, > On Sat, 24 Jun 2023 at 22:00, Tim Perry wrote: >> >> Can we publish log4j-core so it spits out a deprecation warning and just >> pulls in log4j-impl or log4j-runtime or whatever? > > This is probably the best solution. >>> >>> How can it be implemented? >> >> We replace `log4j-core` with an empty JAR file with `log4j-impl` as >> dependency. > > I can’t say I am thrilled with this idea but I also can’t point to any > concrete problems it would cause without testing It would need a class which prints a depreciation warning at startup as well. Preferably with directions to fix. I’m not thrilled by this, but at least it helps them along. Possibly better to stick with log4j-core
Re: Renaming `log4j-core`
> On Jun 26, 2023, at 2:57 PM, Tim Perry wrote: > > > >> On Jun 26, 2023, at 1:38 PM, Ralph Goers wrote: >> >> >> >>> On Jun 26, 2023, at 12:17 PM, Piotr P. Karwasz >>> wrote: >>> >>> Hi Ralph, >>> >>> On Mon, 26 Jun 2023 at 20:33, Ralph Goers >>> wrote: > On Jun 25, 2023, at 11:32 PM, Piotr P. Karwasz > wrote: > > Hi Tim, > >> On Sat, 24 Jun 2023 at 22:00, Tim Perry wrote: >>> >>> Can we publish log4j-core so it spits out a deprecation warning and just >>> pulls in log4j-impl or log4j-runtime or whatever? >> >> This is probably the best solution. How can it be implemented? >>> >>> We replace `log4j-core` with an empty JAR file with `log4j-impl` as >>> dependency. >> >> I can’t say I am thrilled with this idea but I also can’t point to any >> concrete problems it would cause without testing > > It would need a class which prints a depreciation warning at startup as well. > Preferably with directions to fix. > > I’m not thrilled by this, but at least it helps them along. > > Possibly better to stick with log4j-core Yeah - I think it is better to fix the web site. Separate Log4j-API to its own site. Then see where we are. Ralph
