I'll adjust this to use the new macro as soon as it is upstream.
--
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
These changes do both to be on the save side.
On the longer term the question is if the Guile object really should not be
stripped. `eu-strip` should be able to strip those, too. Not sure if we want to
switch over to that in general.
--
You are receiving this because you are subscribed to
Merged #1783 into master.
--
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/rpm-software-management/rpm/pull/1783#event-5355924327___
Rpm-maint mailing list
@ffesti pushed 1 commit.
c1d631c94bf078936fcfe07b136344f924a112cd Require package names to be valid
provides
--
You are receiving this because you are subscribed to this thread.
View it on GitHub:
@ffesti commented on this pull request.
> @@ -19,6 +19,7 @@
#undef HTDATATYPE
#define ALLOWED_CHARS_NAME ".-_+%{}"
+#define ALLOWED_FIRSTCHARS_NAME "_%{}"
Ok, after reading `rpmCharCheck()` 3 more times: Looks like `%{}`are actually
legal in dependency names and only create a warning. So
@ffesti pushed 1 commit.
2019abae2439c5d3d4e250098b093e648a1f72d7 Require package names to be valid
provides
--
You are receiving this because you are subscribed to this thread.
View it on GitHub:
@ffesti commented on this pull request.
> @@ -19,6 +19,7 @@
#undef HTDATATYPE
#define ALLOWED_CHARS_NAME ".-_+%{}"
+#define ALLOWED_FIRSTCHARS_NAME "_%{}"
OK, so this is not by accident but to not blow up for unexpanded macros that
may just not be available at that point in time.
--
@ffesti pushed 1 commit.
a5e67695d970c964089470001f7fad1d99354fac Require package names to be valid
provides
--
You are receiving this because you are subscribed to this thread.
View it on GitHub:
It's quite possible that this is a bug. Can you still please verify that these
files are not owned by another package by running `rpm -qf /usr/bin/strip`?
Thanks!
--
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
Merged #1793 into master.
--
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/rpm-software-management/rpm/pull/1793#event-5434455706___
Rpm-maint mailing list
Closed #1765 via ee5d150aea19ebd10569cd805917acc583719e49.
--
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
Merged #1780 into master.
--
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/rpm-software-management/rpm/pull/1780#event-5372346573___
Rpm-maint mailing list
I talked to mjw about this. eu-strip would probably be able to strip even these
Guile files. I don't really know whether they should be stripped or not,
though. Excluding the Golang (source) files should really not matter but this
is to be on the save side.
And yes this whole brp business
Merged #1784 into master.
--
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/rpm-software-management/rpm/pull/1784#event-5372460114___
Rpm-maint mailing list
Closed #1782 via #1784.
--
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/rpm-software-management/rpm/issues/1782#event-5372460130___
Rpm-maint mailing list
Yeah, not beautiful but just good enough for now.
--
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/rpm-software-management/rpm/pull/1784#issuecomment-929018182___
Ok, more thorough investigation turns up a few more problems than expected:
- Macro parameters are split by ` splitQuoted()`
https://github.com/rpm-software-management/rpm/blob/master/rpmio/macro.c#L930
- It uses Unit Separator (ASCII 0x1f) as separator and ignores quotes
- Non UTF-8 characters
Skipping over `"\ "` and `"\^t"` in `splitQuoted()` makes `%autosetup -n auto\
foo` work. Still need to figure out what this all breaks. It is also not
compatible with `shescape` which uses single quotes.
--
You are receiving this because you are subscribed to this thread.
Reply to this email
OK, to be more clear with this: We currently don't have any (working) means to
have spaces in macro parameters (with the exception of built-in macros which
are different in all kind of wonderful ways).
I wonder if we actually need two ways for doing that. One that is stripped
during parsing
This is caused by RPM no longer using the user and group information from out
side of the change root. Earlier versions were not properly using the inside
files in all cases.
The uid and gid of these files are set before they are moved in place. So the
shadow group does not exist yet. While
Oh, right. The patch does only move the `rpmugInit` call to later. I first
thought it now does that whenever entering the chroot and was wondering why the
commit message is more vocal about that.
I still have this vivid memory of rpm no longer using the the files from
outside the chroot. But I
Only allow alphanumeric or _ as first character.
Also check the name of Obsoletes.
Resolves: #1694
You can view, comment on, or merge this pull request online at:
https://github.com/rpm-software-management/rpm/pull/1778
-- Commit Summary --
* Require package names to be valid provides
--
Reopened #1644.
--
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/rpm-software-management/rpm/issues/1644#event-5341789680___
Rpm-maint mailing list
Yes, we need quoting at more places. Reopened #1644.
--
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
Not sure if we can do a quoting macro. This patch has the benefit of running in
lua and though work on the original content of the file name. A macro will
always see the macro expanded version of the parameter. Not sure if there is a
way around that with the current macro system - especially as
This needs the same kind of quoting as #1445 but at other places. %patches is
not actually used by %autosetup.
--
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
No point in allowing NULL only for one of the arguments.
Thanks to ex0z3 (https://github.com/ex0z3) for reporting!
Resolves: #1782
You can view, comment on, or merge this pull request online at:
https://github.com/rpm-software-management/rpm/pull/1784
-- Commit Summary --
*
Looks good to me.
--
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/rpm-software-management/rpm/pull/1783#issuecomment-924876305___
Rpm-maint mailing list
by enclosing them in single quotes ().
Replace sigle quotes within with \ as needed for sh
Resolves: #1445
You can view, comment on, or merge this pull request online at:
https://github.com/rpm-software-management/rpm/pull/1781
-- Commit Summary --
*
See #1781
--
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/rpm-software-management/rpm/issues/1644#issuecomment-923858687___
Rpm-maint mailing list
Yes, definitely a duplicate. Closing here - but working on the original issue.
--
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
Closed #1644.
--
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/rpm-software-management/rpm/issues/1644#event-5334505537___
Rpm-maint mailing list
It won't solve the issue though, as the group still doesn't exist when the
files are created. Guess something like `echo "shadow:x:404:" >> /etc/group` in
%pre might do the trick.
--
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on
Not 100% sure if this no machine business is 100% correct. But it
looks like a better defense against `strip` failing than relying on the file
name alone.
You can view, comment on, or merge this pull request online at:
https://github.com/rpm-software-management/rpm/pull/1780
-- Commit
as far as possible. This allows setting these tags during the build using the
dynamic spec feature.
Move
copyTagsDuringParse
requiredTags
optionalTags
isMemberInEntry()
checkForValidArchitectures()
checkForRequired()
checkForDuplicates()
fillOutMainPackage()
copyInheritedTags()
from
This is the major part of what is needed for #1240
--
Reply to this email directly or view it on GitHub:
https://github.com/rpm-software-management/rpm/pull/2646#issuecomment-1718917225
You are receiving this because you are subscribed to this thread.
Message ID:
The BuildRequires case is just one example of Preamble content. The dynamic
spec feature will soon drop you into the preamble of the main package for each
parsed file (PR coming soon). The other option (which is possible now) is to
expand a macro directly where the Autobuild directive is
Yeah, this needs to be non dynamic. But it can use dynamic spec in the
implementation of the different build types. So this gets confusing quickly.
--
Reply to this email directly or view it on GitHub:
https://github.com/rpm-software-management/rpm/pull/2620#issuecomment-1691676788
You are
This needs some design document or at least section in the commit message. This
looks awfully like it could also be done with a simple macro or lua script. So
we need some explanation why this isn't done that way.
--
Reply to this email directly or view it on GitHub:
Issue is less the availability of tools but which one to use - and which
dependencies we want to drag in.
There is https://github.com/spdx with several relevant repositories. There also
are the tools around http://reuse.software https://git.fsfe.org/reuse
--
Reply to this email directly or
Avoid the now deprecated RSA and DSA data types and use the generic EVP_PKEY
Resolves: #2294
You can view, comment on, or merge this pull request online at:
https://github.com/rpm-software-management/rpm/pull/2723
-- Commit Summary --
* Move OpenSSL code to newer API
-- File Changes --
@ffesti pushed 5 commits.
7258c44f688c7712af9cdcdb12227d3820e77879 Move checks and package
initialization after build
80feaf69bf841293daa0707bccc546707ca7968e Remove checks during parsing of
packages
2581fecd67178574f5e813e8c97fb5c21045d93f Always start parsing in the preable
of the main
OK, everything that didn't get moved to its own ticket should be addressed.
Renamed the constant to NOFINALIZE, fixed warnings and added to more test cases
with parsing errors.
I am a bit confused that giving Summary two times is only a warning and
doesn't break the build. BUt that's not
@ffesti pushed 6 commits.
a830cc6c8c009080a6d78b621c5206f2c4059bb2 Drop NVR parameter to make them
easier to reuse
6bb9f49f4380c8ec96ea5ffe104116f97e693c6b Make functions available to be moved
later on
7594a223af9cdd4a33f6857ecdc22df5ba84ed6b Move checks and package
initialization after
Also `gpg-pubkey` is not the prefix of the name but the actual name. Otherwise
`rpm -q gpg-pubkey` wouldn't work. The other gibberish is the hash of the key
stuffed in version and release to make it unique.
--
Reply to this email directly or view it on GitHub:
Right place for this would be the `rpmkeys` tool. There actually already is a -
commented out - stub for `--list-keys` and `--delete-key` waiting for someone
to come along and implement them.
--
Reply to this email directly or view it on GitHub:
@ffesti pushed 1 commit.
009daa3ab584b0e271f08d717c19daaa18de3eed Move OpenSSL code to newer API
--
View it on GitHub:
https://github.com/rpm-software-management/rpm/pull/2723/files/f23f894620350dbe35384eb22c88fbca51d0502e..009daa3ab584b0e271f08d717c19daaa18de3eed
You are receiving this
@ffesti commented on this pull request.
> if (!constructDSASigningKey(key))
goto done;
if (!constructDSASignature(sig))
goto done;
-if (DSA_do_verify(hash, hashlen, sig->dsa_sig, key->dsa_key) == 1)
+pkey_ctx = EVP_PKEY_CTX_new(key->evp_pkey, NULL);
+
OK, given that the `BUILDHOST` can already be set via `%_buildhost` what is
still missing here? Or can we just close this ticket?
--
Reply to this email directly or view it on GitHub:
https://github.com/rpm-software-management/rpm/issues/2603#issuecomment-1766142951
You are receiving this
Closed #2085 as completed.
--
Reply to this email directly or view it on GitHub:
https://github.com/rpm-software-management/rpm/issues/2085#event-10767718055
You are receiving this because you are subscribed to this thread.
Message ID:
___
Rpm-maint
Sorry for the late answer! The spec file runs the rpm test suite which needs
fakechroot. It's possible that it is either not available or not working
properly on RISC-V Linux. Note that the current release do use a different
technology for isolating the test suite.
I am closing this as it is
We are not planning to add such an option. While we have stricter requirement
for the backward compatibility of SRPMs than for normal packages they do not
include backward compatibility for all time. Closing.
--
Reply to this email directly or view it on GitHub:
Closed #2727 as completed.
--
Reply to this email directly or view it on GitHub:
https://github.com/rpm-software-management/rpm/issues/2727#event-10766717419
You are receiving this because you are subscribed to this thread.
Message ID:
___
Rpm-maint
So I think going back to the old behaviour is just not worth the trouble - even
if it may be easy to do code wise. So all that's left is cleaning up the code a
little bit. Not keeping a ticket open for stuff like that - otherwise we had
tickets for each line of RPM code...
--
Reply to this
Closed #2319 as completed.
--
Reply to this email directly or view it on GitHub:
https://github.com/rpm-software-management/rpm/issues/2319#event-10766742033
You are receiving this because you are subscribed to this thread.
Message ID:
___
Rpm-maint
I am still a bit puzzled on what this can do that you can't do with a simple
macro definition. I mean I have a rough idea. But may be the docs should make
a bit more of an effort to explain what to use in which case.
--
Reply to this email directly or view it on GitHub:
Closed #1768 as completed.
--
Reply to this email directly or view it on GitHub:
https://github.com/rpm-software-management/rpm/issues/1768#event-10843533706
You are receiving this because you are subscribed to this thread.
Message ID:
___
Rpm-maint
Looks like the conclusion in 2021 was that `Obsoleted-by:` is not the solution.
Closing.
--
Reply to this email directly or view it on GitHub:
https://github.com/rpm-software-management/rpm/issues/1768#issuecomment-1790663062
You are receiving this because you are subscribed to this thread.
Configure traditionally sets it to %{_prefix}/com which RPM has followed so
far. But this directory is not used anywhere and everybody changes the location
to /var/lib. As we are only changing the macro and not the configure default
this should be relatively save to do.
Resolves: #2092
You can
I agree that there should be more control over what files go where and the
current means given to the packagers are not that great. Having more special
code for creating sub packages is not something we want, though. We'd rather
give the packager the means to do that in the spec - or the build
Closed #1448 as completed.
--
Reply to this email directly or view it on GitHub:
https://github.com/rpm-software-management/rpm/issues/1448#event-10843864415
You are receiving this because you are subscribed to this thread.
Message ID:
___
Rpm-maint
Merged #2737 into master.
--
Reply to this email directly or view it on GitHub:
https://github.com/rpm-software-management/rpm/pull/2737#event-10816892415
You are receiving this because you are subscribed to this thread.
Message ID:
___
Rpm-maint
Thanks for the fixes!
--
Reply to this email directly or view it on GitHub:
https://github.com/rpm-software-management/rpm/pull/2737#issuecomment-1786683577
You are receiving this because you are subscribed to this thread.
Message ID: ___
Rpm-maint
Looks actually quite nice and surprisingly straight forward.
I wonder if we should do an announcement if we really want to drop %patchN so
distributions have a bit more time to prepare.
--
Reply to this email directly or view it on GitHub:
May be people like this version better.
--
Reply to this email directly or view it on GitHub:
https://github.com/rpm-software-management/rpm/pull/2728#issuecomment-1781371354
You are receiving this because you are subscribed to this thread.
Message ID:
@ffesti pushed 1 commit.
a8b4e5be21cae2f980e75bc528bff0405ec64689 Reword spec.md change
--
View it on GitHub:
https://github.com/rpm-software-management/rpm/pull/2728/files/39973be09ef2f0217177fd2afc65cabd1cd3945d..a8b4e5be21cae2f980e75bc528bff0405ec64689
You are receiving this because you are
@ffesti pushed 1 commit.
557112f44791d499b3a5208329a01846af63f480 Use uniform formatting for SEE ALSO
sections
--
View it on GitHub:
https://github.com/rpm-software-management/rpm/pull/2732/files/2449e94632bab36a52bd5cafbff2ac03e86790ed..557112f44791d499b3a5208329a01846af63f480
You are
I don't have a strong opinion. Guess it depends on what pages you sample. I
just switched to bold.
--
Reply to this email directly or view it on GitHub:
https://github.com/rpm-software-management/rpm/pull/2732#issuecomment-1778994409
You are receiving this because you are subscribed to this
@ffesti pushed 6 commits.
28dbca6a15efdef33863c1412b331279e9ae2853 Drop NVR parameter to make them
easier to reuse
fb0d81ef9cb886ec31e34c64b13ade85e023b062 Make functions available to be moved
later on
7d3d23cd7c13de952be577f47dfad5bfd0bdda74 Move checks and package
initialization after
@ffesti pushed 6 commits.
7ec8dd2235d6d684f590d3bb92ec6adfa75d583d Drop NVR parameter to make them
easier to reuse
65be4c1e70b7b7745b582dc59be9de1832f550d7 Make functions available to be moved
later on
1894a121973c1f93decc64e6a81a9a790d9288a3 Move checks and package
initialization after
OK, I addressed the comments above:
Re-Added a check for the NVR tags, renamed RPMSPEC_DONTFINALIZE, added a test
case and fixed issues that the test case turned up.
--
Reply to this email directly or view it on GitHub:
Yes, this thought has occurred to me, too. I have not addressed this here as it
is mainly an issue of the original dynamic spec change. But it is something we
need to address.
Funny enough we could actually allow %prep to create later build scripts. Ofc
this doesn't work right now. Also there
As noted in #2646 dynamic spec parts can create/declare tags and sections that
cannot be used that late and we should not allow that and error out. First
rough list:
- all build sections
- including %generatebuildrequires
- BuildRequires
- BuildArch except noarch
--
Reply to this email
Merged #2670 into master.
--
Reply to this email directly or view it on GitHub:
https://github.com/rpm-software-management/rpm/pull/2670#event-10410724809
You are receiving this because you are subscribed to this thread.
Message ID:
___
Rpm-maint
Looks good and pretty straight forward. May be should get a test case.
It might also be a good idea to point out that these are called even when
running without --root. Technically we omit the chroot call in this case but
the hoooks are ofc executed no matter what (as they should be)
--
Reply
Merged #2675 into master.
--
Reply to this email directly or view it on GitHub:
https://github.com/rpm-software-management/rpm/pull/2675#event-10460704680
You are receiving this because you are subscribed to this thread.
Message ID:
___
Rpm-maint
Over are the good old days when one could rely on the garbage collection to
just close the files for you. Had to do something very similar recently in my
own Python project.
Thanks for the patch!
--
Reply to this email directly or view it on GitHub:
@ffesti pushed 1 commit.
32b12aec2d81690f271cd1cde8b8bf72c358229a Move OpenSSL code to newer API from
version 3.0
--
View it on GitHub:
https://github.com/rpm-software-management/rpm/pull/2723/files/009daa3ab584b0e271f08d717c19daaa18de3eed..32b12aec2d81690f271cd1cde8b8bf72c358229a
You are
OK, turns out this is code based on OpenSSL 3.0 which is from 2021. So it is a
bit new. Otoh it no longer is the default variant to be built and the next
release shouldn't be backported to some ancient enterprise distribution.
--
Reply to this email directly or view it on GitHub:
See #2455
--
Reply to this email directly or view it on GitHub:
https://github.com/rpm-software-management/rpm/issues/2454#issuecomment-1757463275
You are receiving this because you are subscribed to this thread.
Message ID: ___
Rpm-maint mailing list
@ffesti converted this issue into discussion #2715.
--
Reply to this email directly or view it on GitHub:
https://github.com/rpm-software-management/rpm/issues/2443#event-10617291792
You are receiving this because you are subscribed to this thread.
Message ID:
Thank would be great, it's not that I have a clue about OpenSSL either. I just
banged it with a hammer until it seemed to work.
--
Reply to this email directly or view it on GitHub:
https://github.com/rpm-software-management/rpm/pull/2723#issuecomment-1772906920
You are receiving this because
Offer a generic way to place icons in the proper system dir
Resolves: #2196
You can view, comment on, or merge this pull request online at:
https://github.com/rpm-software-management/rpm/pull/2729
-- Commit Summary --
* Add %_iconsdir macro
-- File Changes --
M macros.in (1)
--
Closed #2411 as completed.
--
Reply to this email directly or view it on GitHub:
https://github.com/rpm-software-management/rpm/issues/2411#event-10661746538
You are receiving this because you are subscribed to this thread.
Message ID:
___
Rpm-maint
The issue here is that these tags are arrays and need to be surrounded by
square brackets. The CLI params already do that for you. They also include the
flags and version while the using the tags directly only gives you the name
part of the tripplets. You can just use `%{conflictnevrs}` and
You can view, comment on, or merge this pull request online at:
https://github.com/rpm-software-management/rpm/pull/2705
-- Commit Summary --
* Reword to make sentence easier to understand
* Add docs/README.md
-- File Changes --
M docs/CMakeLists.txt (1)
A docs/README.md (30)
Documented how to build the reference manual locally but we really should
actually do that in our cmake files.
--
Reply to this email directly or view it on GitHub:
https://github.com/rpm-software-management/rpm/issues/2259#issuecomment-1750545445
You are receiving this because you are
I also don't have too strong of an opinion on that. Only the comment that most
build systems may be pretty surprised if suddenly packages from more than one
build arch show up. May be we should keep things as they are and just clean up
the code.
--
Reply to this email directly or view it on
Yeah, giving the packager more control here is something we generally want. The
dependency generators are a bit clunky as soon as they don't exactly what one
wants/needs.
--
Reply to this email directly or view it on GitHub:
Improved wording of the commit message
--
Reply to this email directly or view it on GitHub:
https://github.com/rpm-software-management/rpm/pull/2732#issuecomment-1777100699
You are receiving this because you are subscribed to this thread.
Message ID:
Closed #2603 as completed.
--
Reply to this email directly or view it on GitHub:
https://github.com/rpm-software-management/rpm/issues/2603#event-10752520725
You are receiving this because you are subscribed to this thread.
Message ID:
___
Rpm-maint
Guess not. Feel free to re-open if we missed something and there is really
something that needs doing.
--
Reply to this email directly or view it on GitHub:
https://github.com/rpm-software-management/rpm/issues/2603#issuecomment-1777011289
You are receiving this because you are subscribed to
While there is no generally agreed standard for this we should at least use
only one. Comma separated lists are commonly used. Lets just pick *italic*
for man pages.
Thanks to Frank Dana https://github.com/ferdnyc; for pointing this out.
Resolves: #2731
You can view, comment on, or merge this
We should probably put those build macros into separate macro files - one per
build type. That makes it easier for people to add their own and to hand them
over to other communities, if we want to get rid of them upstream.
Also the macros file is too big already.
--
Reply to this email
and make the project REUSE compatible.
See https://reuse.software/
This will more explicit what license each file is under and will allow to
quickly review the overall license of the whole project.
--
Reply to this email directly or view it on GitHub:
It closes a section. So it would end the `%prep` section. Idea is that the
debuginfo magic wold then be evaluated after the `prep` section. Basically the
same as if there was another section inbetween.
--
Reply to this email directly or view it on GitHub:
OK, I guess I start with writing down how debuginfo works right now:
There are quite a few macros that
[control](https://github.com/rpm-software-management/rpm/blob/master/macros.in#L452)
and
[implement](https://github.com/rpm-software-management/rpm/blob/master/macros.in#L135)
the debuginfo
Well, `%install` following the preamble is technically a different issue.
`%prep` is setting `%buildsubdir` via `%setup` and everything that makes use of
that needs to be after `%prep`. This is not limited to the debuginfo magic -
although it is especially non-obvious there.
--
Reply to this
The issue here is that we actually want some macros in the SPEC file. I wonder
if we could just add `%__spec_pre` and `%__spec_post` macros that are expanded
at the beginning and the end of the spec file. We have something similar for
the build scripts already. This would give rpm itself but
For now build root policy scripts are just shipped in the
[scripts/](https://github.com/rpm-software-management/rpm/tree/master/scripts)
and it is left to the distributions to run them in ` %__os_install_post`
([Fedora as an
1001 - 1100 of 1575 matches
Mail list logo
