@pmatilai does using a build directory that is outside of the source directory
(something like `~/repos-build/rpm`) help? IIRC this is best practice for at
least some build systems.
--
Reply to this email directly or view it on GitHub:
@dmnks commented on this pull request.
> + *
+ * | Type | Description | Mode chars|
+ * ||---|---|
+ * | `ufdio` | no compression| `r,w,a,b,+,x` |
+ * | `gzdio` | gzip | `r,w,a` |
+ * | `bzdio` | bzip2
@dmnks commented on this pull request.
> + *
+ * | Type | Description | Mode chars|
+ * ||---|---|
+ * | `ufdio` | no compression| `r,w,a,b,+,x` |
+ * | `gzdio` | gzip | `r,w,a` |
+ * | `bzdio` | bzip2
@dmnks commented on this pull request.
> + *
+ * | Type | Description | Mode chars|
+ * ||---|---|
+ * | `ufdio` | no compression| `r,w,a,b,+,x` |
+ * | `gzdio` | gzip | `r,w,a` |
+ * | `bzdio` | bzip2
@dmnks commented on this pull request.
> + *
+ * | Type | Description | Mode chars|
+ * ||---|---|
+ * | `ufdio` | no compression| `r,w,a,b,+,x` |
+ * | `gzdio` | gzip | `r,w,a` |
+ * | `bzdio` | bzip2
@dmnks commented on this pull request.
> @@ -138,8 +138,14 @@ in the RPM sources.
## Payload
-The Payload is currently a gzipped cpio archive. The cpio
-archive type used is SVR4 with a CRC checksum.
+The Payload is currently a cpio archive, gzipped by default. The cpio archive
+type
@dmnks commented on this pull request.
> */
FD_t Fdopen(FD_t ofd, const char * fmode);
/** \ingroup rpmio
* fopen(3) clone.
+ *
+ * Supports compression.
+ *
+ * The `fmode` parameter is based on that of `fopen(3)`, but may also include a
+ * compression method (`type` and `flags`) to
OK, I made the following changes:
* Put the mode string description into the `Fopen()` API docs as that's where
it really belongs
* Add a short note about the payload compression macros to the Payload section
in the RPM format page.
--
Reply to this email directly or view it on GitHub:
@dmnks pushed 2 commits.
3c61d5ff61a05cb03af7ea2ca3e39be1468fb68b Add mode string documentation to
F*open()
02799839e685ee4445d46104e5551ffee9de8e5d Add payload compression macro
documentation
--
View it on GitHub:
OK, added test case and some docs.
In addition I stripped the find-lang patch as this needs to g into it's own PR
and was only there as a demonstration piece.
The patch now uses a `SPECS` sub dir in the `buildsubdir` to allow detecting
whether the feature is supported.
I also removed the code to
@ffesti pushed 5 commits.
4320abf21b4d228b446fb679f0fa2e9f992e5985 Split actual parsing of the spec into
its own function
d8f830518e70e1bae31fab10c75720185446c5ba Allow starting new spec parts with
PART_EMPTY
29708e32c8e2fb9f2b758d97fda24f23e1506ca6 Read in spec pieces generated during
These look like they were different blocks needed at different places at some
point. But I fail to see why this split is still necessary.
You can view, comment on, or merge this pull request online at:
https://github.com/rpm-software-management/rpm/pull/2243
-- Commit Summary --
* Remove
Of course, what we really should aim to is eliminating the human factor from
release creation as much as possible. Once you have a script which grabs a
clean git checkout, runs doc pre-generation automatically and then produces the
dist tarball from that known state that's not accessible to us
Oh, yep:
> [pmatilai︎localhost _build]$ make package_source
Run CPack packaging tool for source...
CPack: Create package using TBZ2
CPack: Install projects
CPack: - Install directory: /home/pmatilai/repos/rpm
CPack Error: Problem copying file:
/home/pmatilai/repos/rpm/_build/tests/testing/usr/
My initial impression from cpack is that it merrily copies whatever random crap
may be in your tree, unless you have it excluded. It's entirely possible I'm
mistaken and/or that it's possible to it more selective, but until that time I
trust git for the task about a 100 times more :sweat_smile:
15 matches
Mail list logo
