Message128645
| Author | eric.araujo |
|---|---|
| Recipients | Kain94, Martijn.van.Oosterhout, amaury.forgeotdarc, belopolsky, collinwinter, eric.araujo, eric.smith, giampaolo.rodola, gvanrossum, pitrou, r.david.murray |
| Date | 2011-02-16.10:45:49 |
| SpamBayes Score | 3.136136e-08 |
| Marked as misclassified | No |
| Message-id | <1297853150.62.0.924713228377.issue10968@psf.upfronthosting.co.za> |
| In-reply-to |
| Content | |
|---|---|
One concern expressed on a duplicate report by Martijn van Oosterhout: > Note this is a behaviour change. Under the old scheme (Foo is a class) > > Foo.timerclass = Timer > > created a method, whereas now it will just assign the class as an > attribute. To work around this you had to use _Timer. Will that dummy > class remain as an alias to avoid breaking code (in 2.7 at least)? Stable versions (2.7, 3.1, soon 3.2) won’t get this change. They may get a doc patch to warn people about the fake class/factory thing. In the py3k documentation for threading, some of the fake classes are documented as factories (Event) and other ones as classes (Timer); do you people think there would be harm in cleaning up all those outdated shenanigans for 3.3, with due versionchanged notes in the doc? |
|
| History | |||
|---|---|---|---|
| Date | User | Action | Args |
| 2011-02-16 10:45:50 | eric.araujo | set | recipients: + eric.araujo, gvanrossum, collinwinter, amaury.forgeotdarc, belopolsky, pitrou, eric.smith, giampaolo.rodola, r.david.murray, Kain94, Martijn.van.Oosterhout |
| 2011-02-16 10:45:50 | eric.araujo | set | messageid: <1297853150.62.0.924713228377.issue10968@psf.upfronthosting.co.za> |
| 2011-02-16 10:45:50 | eric.araujo | link | issue10968 messages |
| 2011-02-16 10:45:49 | eric.araujo | create | |