[gentoo-dev] Making more repoman checks fatal

[gentoo-dev] Making more repoman checks fatal

[Patrick Lauer]

Right now repoman is relatively permissive - it whines about many things, but 
treats many issues as warning. 
The result is that many ebuilds get committed with 'minor' cosmetic issues 
which then someone more OCD than the original committer cleans up, making 
pretty much everyone involved more unhappy.

There's no reason to not error out on, for example, an invalid RESTRICT. Just 
printing a message is relatively useless.


Thus I suggest making the following warnings proper errors:

(Taken from current repoman 'qawarnings' set)

"changelog.missing",
"changelog.notadded",
"digest.assumed",
"digest.unused",
"ebuild.notadded",
"ebuild.nesteddie",
"DESCRIPTION.toolong",
"RESTRICT.invalid",
"ebuild.minorsyn",
"ebuild.badheader",
"metadata.warning",
"LIVEVCS.stable",
"LIVEVCS.unmasked",

Most of these have few or no occurrences in the current tree, so changing the 
default from warn to error won't get in the way of the normal workflow.

(A few of them, like DESCRIPTION.toolong, still have about a dozen leftovers, 
but that should be easy to fix)

Have fun,

Patrick

Re: [gentoo-dev] Making more repoman checks fatal

[Alexander Berntsen]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA256

FWIW: I'm in the "warnings are pointless, either we care about
something (so make it an error), or we don't (so get rid of it)".

- -- 
Alexander
bernalex@gentoo.org
https://secure.plaimi.net/~alexander
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2

iF4EAREIAAYFAlTh6lUACgkQRtClrXBQc7UXHgD9EG6jZBwJ/wvBY0E+XJvuCB8Q
D4Al3H2n2XyYXllCmYUA/1UHAmxBpb/rbNGPE3w/I3JmxE/yNoa57bxEIxDwAREB
=DD9H
-----END PGP SIGNATURE-----

Re: [gentoo-dev] Making more repoman checks fatal

[Alexander Berntsen]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA256

On 16/02/15 14:02, Alexander Berntsen wrote:
> FWIW: I'm in the "warnings are pointless, either we care about 
> something (so make it an error), or we don't (so get rid of it)".
s/\./ camp./

(I accidentally a word...)
- -- 
Alexander
bernalex@gentoo.org
https://secure.plaimi.net/~alexander
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2

iF4EAREIAAYFAlTh6uIACgkQRtClrXBQc7WOPQEAs4AmG2HHxbjWe54j8zHfN9S8
0IpQPJ68mYli2adDNH4BAIox20ink9c2Ntd/Nsk80jyqWpKsNFd0UcJzLALnDu/I
=ldpG
-----END PGP SIGNATURE-----

Re: [gentoo-dev] Making more repoman checks fatal

[NP Hardass]

[-- Attachment #1: Type: text/plain, Size: 1419 bytes --]

On Feb 16, 2015 8:01 AM, "Patrick Lauer" <patrick@gentoo.org> wrote:
>
> Right now repoman is relatively permissive - it whines about many things,
but
> treats many issues as warning.
> The result is that many ebuilds get committed with 'minor' cosmetic issues
> which then someone more OCD than the original committer cleans up, making
> pretty much everyone involved more unhappy.
>
> There's no reason to not error out on, for example, an invalid RESTRICT.
Just
> printing a message is relatively useless.
>
>
> Thus I suggest making the following warnings proper errors:
>
> (Taken from current repoman 'qawarnings' set)
>
> "changelog.missing",
> "changelog.notadded",
> "digest.assumed",
> "digest.unused",
> "ebuild.notadded",
> "ebuild.nesteddie",
> "DESCRIPTION.toolong",
> "RESTRICT.invalid",
> "ebuild.minorsyn",
> "ebuild.badheader",
> "metadata.warning",
> "LIVEVCS.stable",
> "LIVEVCS.unmasked",
>
> Most of these have few or no occurrences in the current tree, so changing
the
> default from warn to error won't get in the way of the normal workflow.
>
> (A few of them, like DESCRIPTION.toolong, still have about a dozen
leftovers,
> but that should be easy to fix)
>
> Have fun,
>
> Patrick
>

I would also like to put forward the idea of a pedantic flag, like with
GCC, To make all warnings fatal. Or at minimum, make a lesser flag where
certain flags that are not normally fatal, would become fatal.

[-- Attachment #2: Type: text/html, Size: 1945 bytes --]

Re: [gentoo-dev] Making more repoman checks fatal

[Andreas K. Huettel]

> Thus I suggest making the following warnings proper errors:
> 
> (Taken from current repoman 'qawarnings' set)
> 
> "changelog.missing",
> "changelog.notadded",
> "digest.assumed",
> "digest.unused",
> "ebuild.notadded",
> "ebuild.nesteddie",
> "DESCRIPTION.toolong",
> "RESTRICT.invalid",
> "ebuild.minorsyn",
> "ebuild.badheader",
> "metadata.warning",
> "LIVEVCS.stable",
> "LIVEVCS.unmasked",
> 

Yes please. And in addition "dependency.perlcore"


-- 
Andreas K. Huettel
Gentoo Linux developer
perl, office, comrel, council

Re: [gentoo-dev] Making more repoman checks fatal

[Pacho Ramos]

El lun, 16-02-2015 a las 21:00 +0800, Patrick Lauer escribió:
[...]
> 

I agree

Re: [gentoo-dev] Making more repoman checks fatal

[Mike Frysinger]

[-- Attachment #1: Type: text/plain, Size: 328 bytes --]

On 16 Feb 2015 21:00, Patrick Lauer wrote:
> Thus I suggest making the following warnings proper errors:

some of these are because they produce false positives.  at least these bugs 
probably need to be fixed first:
	https://bugs.gentoo.org/405017
	https://bugs.gentoo.org/488836
	https://bugs.gentoo.org/533460
-mike

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 819 bytes --]

Re: [gentoo-dev] Making more repoman checks fatal

[Rafael Goncalves Martins]

On Mon, Feb 16, 2015 at 11:19 AM, Mike Frysinger <vapier@gentoo.org> wrote:
> On 16 Feb 2015 21:00, Patrick Lauer wrote:
>> Thus I suggest making the following warnings proper errors:
>
> some of these are because they produce false positives.  at least these bugs
> probably need to be fixed first:
>         https://bugs.gentoo.org/405017
>         https://bugs.gentoo.org/488836
>         https://bugs.gentoo.org/533460
> -mike

I think that just the last bug is still valid.

[]s

-- 
Rafael Goncalves Martins
Gentoo Linux developer
http://rafaelmartins.eng.br/

Re: [gentoo-dev] Making more repoman checks fatal

[Andrew Savchenko]

[-- Attachment #1: Type: text/plain, Size: 1384 bytes --]

On Mon, 16 Feb 2015 21:00:16 +0800 Patrick Lauer wrote:
> Right now repoman is relatively permissive - it whines about many things, but 
> treats many issues as warning. 
> The result is that many ebuilds get committed with 'minor' cosmetic issues 
> which then someone more OCD than the original committer cleans up, making 
> pretty much everyone involved more unhappy.
> 
> There's no reason to not error out on, for example, an invalid RESTRICT. Just 
> printing a message is relatively useless.
> 
> 
> Thus I suggest making the following warnings proper errors:
> 
> (Taken from current repoman 'qawarnings' set)
[...]
> "ebuild.minorsyn",

Not this one, please. It gives tons of false warnings, e.g.:
  ebuild.minorsyn               1
   media-video/kino/kino-1.3.4.ebuild: Unquoted Variable on line: 95

It also have problems with nested "\"\"" constructs. I can't
remember example right now.

> "ebuild.badheader",
> "metadata.warning",
> "LIVEVCS.stable",
> "LIVEVCS.unmasked",
> 
> Most of these have few or no occurrences in the current tree, so changing the 
> default from warn to error won't get in the way of the normal workflow.
> 
> (A few of them, like DESCRIPTION.toolong, still have about a dozen leftovers, 
> but that should be easy to fix)

I don't see any reasons to make this one fatal.

Best regards,
Andrew Savchenko

[-- Attachment #2: Type: application/pgp-signature, Size: 819 bytes --]

Re: [gentoo-dev] Making more repoman checks fatal

[Brian Dolbec]

On Mon, 16 Feb 2015 21:00:16 +0800
Patrick Lauer <patrick@gentoo.org> wrote:

> Right now repoman is relatively permissive - it whines about many
> things, but treats many issues as warning. 
> The result is that many ebuilds get committed with 'minor' cosmetic
> issues which then someone more OCD than the original committer cleans
> up, making pretty much everyone involved more unhappy.
> 
> There's no reason to not error out on, for example, an invalid
> RESTRICT. Just printing a message is relatively useless.
> 
> 
> Thus I suggest making the following warnings proper errors:
> 
> (Taken from current repoman 'qawarnings' set)
> 
> "changelog.missing",
> "changelog.notadded",
> "digest.assumed",
> "digest.unused",
> "ebuild.notadded",
> "ebuild.nesteddie",
> "DESCRIPTION.toolong",
> "RESTRICT.invalid",
> "ebuild.minorsyn",
> "ebuild.badheader",
> "metadata.warning",
> "LIVEVCS.stable",
> "LIVEVCS.unmasked",
> 
> Most of these have few or no occurrences in the current tree, so
> changing the default from warn to error won't get in the way of the
> normal workflow.
> 
> (A few of them, like DESCRIPTION.toolong, still have about a dozen
> leftovers, but that should be easy to fix)
> 
> Have fun,
> 
> Patrick
> 

While I can agree to the principal of this, Unfortunately I maintain
several pkgs that I would love to clean out old versions of
that trigger several of these, but can't.  (But I am getting closer to
being able to clean them)

If I recall correctly they are:

  "ebuild.minorsyn",
  "ebuild.badheader",

plus a couple other odd ones

Portage ebuild triggers minorsyn due to a false positive quoting issue.


-- 
Brian Dolbec <dolsen>

Re: [gentoo-dev] Making more repoman checks fatal

[Mike Gilbert]

On Mon, Feb 16, 2015 at 8:00 AM, Patrick Lauer <patrick@gentoo.org> wrote:
> Thus I suggest making the following warnings proper errors:
>
> (Taken from current repoman 'qawarnings' set)
>
> "changelog.missing",
> "changelog.notadded",

These two are pretty much irrelevant now that repoman auto-generates
ChangeLog, so making them fatal doesn't really matter.

> "digest.unused",

This pretty much never pops up since repoman regens the Manifest
before committing anyway.

> "ebuild.notadded",

This automatically becomes fatal when you run repoman in commit mode.
It is a bit pointless to even run this check before that point.

> "DESCRIPTION.toolong",

I am of the opinion that strictly enforcing a length on DESCRIPTION is
stupid. Keep this as a warning please.

> "ebuild.minorsyn",

That should not be fatal. There is little value, and there are false
positives in any case.

Re: [gentoo-dev] Making more repoman checks fatal

[hasufell]

Patrick Lauer:
> "ebuild.badheader",

That would break repoman for the majority of overlays. You don't really
expect overlay maintainers to follow gentoo copyright, do you?

Really... before repoman is fixed, none of this will happen (or people
will just run a hacked repoman version).

Re: [gentoo-dev] Making more repoman checks fatal

[Mike Frysinger]

[-- Attachment #1: Type: text/plain, Size: 1094 bytes --]

On 16 Feb 2015 11:45, Rafael Goncalves Martins wrote:
> On Mon, Feb 16, 2015 at 11:19 AM, Mike Frysinger wrote:
> > On 16 Feb 2015 21:00, Patrick Lauer wrote:
> >> Thus I suggest making the following warnings proper errors:
> >
> > some of these are because they produce false positives.  at least these bugs
> > probably need to be fixed first:
> >         https://bugs.gentoo.org/405017
> >         https://bugs.gentoo.org/488836
> >         https://bugs.gentoo.org/533460
> 
> I think that just the last bug is still valid.

the first is still broken:
$ cd dev-python/boto/
$ cvs up
$ repoman full

RepoMan scours the neighborhood...

Note: use --include-dev (-d) to check dependencies for 'dev' profiles

RepoMan sez: "If everyone were like you, I'd be out of business!"

$ rm *.ebuild
$ cvs up >&/dev/null
$ repoman full

RepoMan scours the neighborhood...
  ebuild.badheader              17
   dev-python/boto/boto-2.11.0.ebuild: Invalid Gentoo Copyright on line: 1
   dev-python/boto/boto-2.19.0.ebuild: Invalid Gentoo Copyright on line: 1
   ...
-mike

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 819 bytes --]