This method keeps appearing in my image, causing a dirty ST80 package.
Is anyone else seeing this? No senders, but I keep needing to revert the package to know what I've got dirty in the image. Also seeing some spurious class-category "changes" causing dirty packages, even though I'm not doing any recategorizing.. |
On 2 December 2013 20:18, Chris Muller <[hidden email]> wrote:
> This method keeps appearing in my image, causing a dirty ST80 package. > Is anyone else seeing this? No senders, but I keep needing to revert > the package to know what I've got dirty in the image. I can't say I've seen this. > Also seeing some spurious class-category "changes" causing dirty > packages, even though I'm not doing any recategorizing.. I make sure nowadays that I don't issue commits with shuffled class categories, but I certainly have done so in the past. Monticello considers an order change in the categories to be significant, but I vaguely remember Bert committing something to mitigate the issue. frank |
It appears it was due to a corrupt ancestry in 45Deprecated. 45Deprecated-fbs.3 was the version included in Squeak4.5-13352 at the FTP site, but since no future package had that version as an ancestor, everything was merged during the update, resulting in spurious dirties. I reposted Squeak4.5-13352 with 45Deprecated-fbs.5 loaded and that seems to fix it. On Mon, Dec 2, 2013 at 4:04 PM, Frank Shearar <[hidden email]> wrote: On 2 December 2013 20:18, Chris Muller <[hidden email]> wrote: |
On 02 Dec 2013, at 22:42, Chris Muller <[hidden email]> wrote:
Thanks for taking the time to find & fix the problem (especially since it's my fault). However, we should never replace/alter versioned artifacts. I would rather you had deleted the file and put a new file with a different name on the site. Even calling it 4.4-rc0 would work for me, but definitely not 4.5-13352. frank
|
Free forum by Nabble | Edit this page |