Conversation
ea5af32 to
52c2df9
Compare
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
htgoebel
left a comment
There was a problem hiding this comment.
Thanks for this pull-request.I now found time for reviewing it and added some inline-remarks.
Beside these:
- Please reword the commit message to follow our Commit Message Rules (esp. the prefix and to explain a bit more what and why this has been done. You did this already in the pull-request message ;-) For us the commit message is much more important.
- Please split the news-fragment and put it into the resp. commit.
(git rebase -i and git push -f are your friends)
I will then merge this as a branch with an appropriate commit message containing the issue number.
0060fea to
47b8bcb
Compare
|
@htgoebel Just checking, can you re-review this when you get a chance? Thanks so much! |
47b8bcb to
426de47
Compare
The common_prefix has already been normcase'd, so ensure that the path we are trying to remove the common_prefix from is also normcase'd so that replace can work. When defining name of the dependency get the basename from tpl[1] rather than using tpl[0] as tpl[0] may have not have the proper casing (it may have been normcase'd) which will cause issues on operating systems where filename are case-sensitive (eg Linux).
When a PKG is created the DEPENDENCY's need to be added to ensure the resulting archive has a reference to the relative file path of the DEPENDENCY and hence the bootloader will be able to load it.
Fix bug in pyi_copy_file where when files were copied, an extra bit of data was being added if the file didn't finish perfectly on the buffer size. fwrite writes the entire buffer, not just up until a null terminator character, so this meant the last fwrite would include data from the previous fwrite which corrupted files.
Use pyi_arch_status_new() (which uses calloc) to ensure archive->fp is NULL. Otherwise the value could be non-NULL in which case the archive will not be opened and bogus data will remain in archive->fp.
I missed this in 671770f.
426de47 to
3f4d5cf
Compare
96ca787 to
ae85dee
Compare
|
@coreydexter I intentionally removed these "noqa: F401" changes :-) |
|
@coreydexter I should have reasoned this: You added these comments to make the lint tests pass, wich basically is okay. But this is only test-code. And I count the information "renamed unchanged" more important than these errors. (Okay, what you most propably do not know: we only lint changed lines. So unless these lines will be touched again somewhen, lint will pass.) |
ae85dee to
3f4d5cf
Compare
|
@coreydexter So many thanks for solving this long-standing issue! |
|
@htgoebel the docs need fixing. They still say it's broken. |
Besides this, is there any way to automatically merge several spec files? |
|
Please open an issue if you have questions related to multipackage. |
Fix up the multipackage so that exe's can shared dependencies. Closes #1527
There are a few changes that needed to be done: