Skip to contentSkip to conversation

SPDX: Add annotations to relationship

PATCH In review Saul Wold 6 participants 17 messages Oct 27, 2021 – Nov 16, 2021

lore
Share & tools
lore

Saul Wold

author
#0 lore raw

cover SPDX: Add annotations to relationship

Add annotations to relationships and refactor code to add create_annotation() function for code reuse.

Ensure that "cross" recipes are factored into isNative also.

v2: removed leftover and unused annotation per Joshua

Sau!

Saul Wold (3):
spdx.py: Add annotation to relationship
create-spdx: add create_annotation function
create-spdx: cross recipes are native also

 classes/create-spdx.bbclass | 22 ++++++++++++++--------
 lib/oe/spdx.py              |  6 +++++-
 2 files changed, 19 insertions(+), 9 deletions(-)
Signature · 1 line
2.31.1
author
#1 lore raw

PATCH v2 1/3 spdx.py: Add annotation to relationship

Having annotations on relationship can provide additional information about the relationship such as how it was derived.

---
 lib/oe/spdx.py | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)
lib/oe/spdx.py +5 -1
index 4416194..9e7ced5 100644
@@ -196,6 +196,7 @@ class SPDXRelationship(SPDXObject):
     relatedSpdxElement = _String()
     relationshipType = _String()
     comment = _String()
+    annotations = _ObjectList(SPDXAnnotation)
 
 
 class SPDXExternalReference(SPDXObject):
@@ -300,7 +301,7 @@ class SPDXDocument(SPDXObject):
     def from_json(cls, f):
         return cls(**json.load(f))
 
-    def add_relationship(self, _from, relationship, _to, *, comment=None):
+    def add_relationship(self, _from, relationship, _to, *, comment=None, annotation=None):
         if isinstance(_from, SPDXObject):
             from_spdxid = _from.SPDXID
         else:
@@ -320,6 +321,9 @@ class SPDXDocument(SPDXObject):
         if comment is not None:
             r.comment = comment
 
+        if annotation is not None:
+            r.annotations.append(annotation)
+
         self.relationships.append(r)
 
     def find_by_spdxid(self, spdxid):
Signature · 1 line
2.31.1
author
#2 lore raw

PATCH v2 2/3 create-spdx: add create_annotation function

In reply to Saul Wold

This allows code reuse and future usage with relationship annotations

---
 classes/create-spdx.bbclass | 18 ++++++++++++------
 1 file changed, 12 insertions(+), 6 deletions(-)
classes/create-spdx.bbclass +12 -6
index dd341db..72a9b55 100644
@@ -35,6 +35,17 @@ def get_doc_namespace(d, doc):
     namespace_uuid = uuid.uuid5(uuid.NAMESPACE_DNS, d.getVar("SPDX_UUID_NAMESPACE"))
     return "%s/%s-%s" % (d.getVar("SPDX_NAMESPACE_PREFIX"), doc.name, str(uuid.uuid5(namespace_uuid, doc.name)))
 
+def create_annotation(d, comment):
+    from datetime import datetime, timezone
+
+    creation_time = datetime.now(tz=timezone.utc).strftime("%Y-%m-%dT%H:%M:%SZ")
+    annotation = oe.spdx.SPDXAnnotation()
+    annotation.annotationDate = creation_time
+    annotation.annotationType = "OTHER"
+    annotation.annotator = "Tool: %s - %s" % (d.getVar("SPDX_TOOL_NAME"), d.getVar("SPDX_TOOL_VERSION"))
+    annotation.comment = comment
+    return annotation
+
 def recipe_spdx_is_native(d, recipe):
     return any(a.annotationType == "OTHER" and
       a.annotator == "Tool: %s - %s" % (d.getVar("SPDX_TOOL_NAME"), d.getVar("SPDX_TOOL_VERSION")) and
@@ -408,12 +419,7 @@ python do_create_spdx() {
     recipe.versionInfo = d.getVar("PV")
     recipe.SPDXID = oe.sbom.get_recipe_spdxid(d)
     if bb.data.inherits_class("native", d):
-        annotation = oe.spdx.SPDXAnnotation()
-        annotation.annotationDate = creation_time
-        annotation.annotationType = "OTHER"
-        annotation.annotator = "Tool: %s - %s" % (d.getVar("SPDX_TOOL_NAME"), d.getVar("SPDX_TOOL_VERSION"))
-        annotation.comment = "isNative"
-        recipe.annotations.append(annotation)
+        recipe.annotations.append(create_annotation(d, "isNative"))
 
     for s in d.getVar('SRC_URI').split():
         if not s.startswith("file://"):
Signature · 1 line
2.31.1
author
#3 lore raw

PATCH v2 3/3 create-spdx: cross recipes are native also

In reply to Saul Wold

Recipes that inherit cross should also be categorized as isNative

---
 classes/create-spdx.bbclass | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
classes/create-spdx.bbclass +2 -2
index 72a9b55..e8f476f 100644
@@ -418,7 +418,7 @@ python do_create_spdx() {
     recipe.name = d.getVar("PN")
     recipe.versionInfo = d.getVar("PV")
     recipe.SPDXID = oe.sbom.get_recipe_spdxid(d)
-    if bb.data.inherits_class("native", d):
+    if bb.data.inherits_class("native", d) or bb.data.inherits_class("cross", d):
         recipe.annotations.append(create_annotation(d, "isNative"))
 
     for s in d.getVar('SRC_URI').split():
@@ -610,7 +610,7 @@ python do_create_runtime_spdx() {
 
     deploy_dir_spdx = Path(d.getVar("DEPLOY_DIR_SPDX"))
     spdx_deploy = Path(d.getVar("SPDXRUNTIMEDEPLOY"))
-    is_native = bb.data.inherits_class("native", d)
+    is_native = bb.data.inherits_class("native", d) or bb.data.inherits_class("cross", d)
 
     creation_time = datetime.now(tz=timezone.utc).strftime("%Y-%m-%dT%H:%M:%SZ")
 
Signature · 1 line
2.31.1
#4 lore raw
In reply to Saul Wold

Hi all,

There are any plans or is it possible to backport the SBOM/SPDX to the dunfell branch?
Doing a quick look on it I see that it is not too intrusive and the most one is in
classes/package: Add extended packaged data 7ec54b174304e940ec66f21ac512f7b50fa637b3

Jose

Saul Wold <[email protected]> escreveu no dia quarta, 27/10/2021 à(s) 02:31:

Quoted 33 lines

Add annotations to relationships and refactor code to add create_annotation() function for code reuse.

Ensure that "cross" recipes are factored into isNative also.

v2: removed leftover and unused annotation per Joshua

Sau!

Saul Wold (3):
spdx.py: Add annotation to relationship
create-spdx: add create_annotation function
create-spdx: cross recipes are native also

 classes/create-spdx.bbclass | 22 ++++++++++++++--------
 lib/oe/spdx.py              |  6 +++++-
 2 files changed, 19 insertions(+), 9 deletions(-)

--
2.31.1

-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group. View/Reply Online (#157470):
https://lists.openembedded.org/g/openembedded-core/message/157470
Mute This Topic: https://lists.openembedded.org/mt/86616599/5052612
Group Owner: [email protected]
Unsubscribe: https://lists.openembedded.org/g/openembedded-core/unsub [ [email protected]]
-=-=-=-=-=-=-=-=-=-=-=-

Signature · 3 lines
Best regards,

José Quaresma
#5 lore raw
Quoted 4 lines · Jose Quaresma

Hi all,

There are any plans or is it possible to backport the SBOM/SPDX to the dunfell branch?

I'm going to yield to Saul as to whether he thinks this is desirable/possible or not.

Steve

Quoted 47 lines

Doing a quick look on it I see that it is not too intrusive and the most one is in
classes/package: Add extended packaged data 7ec54b174304e940ec66f21ac512f7b50fa637b3

Jose

Saul Wold <[email protected]> escreveu no dia quarta, 27/10/2021 à(s) 02:31:

Quoted 15 lines

Add annotations to relationships and refactor code to add create_annotation() function for code reuse.

Ensure that "cross" recipes are factored into isNative also.

v2: removed leftover and unused annotation per Joshua

Sau!

Saul Wold (3):
spdx.py: Add annotation to relationship
create-spdx: add create_annotation function
create-spdx: cross recipes are native also

Quoted 10 lines
 classes/create-spdx.bbclass | 22 ++++++++++++++--------
 lib/oe/spdx.py              |  6 +++++-
 2 files changed, 19 insertions(+), 9 deletions(-)

--
2.31.1

--
Best regards,

José Quaresma

-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.
View/Reply Online (#157476): https://lists.openembedded.org/g/openembedded-core/message/157476
Mute This Topic: https://lists.openembedded.org/mt/86616599/3617601
Group Owner: [email protected]
Unsubscribe: https://lists.openembedded.org/g/openembedded-core/unsub [[email protected]]
-=-=-=-=-=-=-=-=-=-=-=-

#6 lore raw
Quoted 8 lines · Steve Sakoman
Quoted 4 lines · Jose Quaresma

Hi all,

There are any plans or is it possible to backport the SBOM/SPDX to the dunfell branch?

I'm going to yield to Saul as to whether he thinks this is desirable/possible or not.

The packagedata changes are pretty invasive unfortunately and likely not something you're going to want in dunfell sadly.

Cheers,

Richard

#7 lore raw

Richard Purdie <[email protected]> escreveu no dia quinta, 28/10/2021 à(s) 21:58:

Quoted 6 lines · Steve Sakoman
Quoted 1 line

On Tue, Oct 26, 2021 at 10:41 PM Jose Quaresma <[email protected]>

Quoted 3 lines · wrote:
Quoted 3 lines

Hi all,

Quoted 9 lines
Quoted 1 line
Quoted 1 line

There are any plans or is it possible to backport the SBOM/SPDX to the

dunfell branch?

Quoted 3 lines

I'm going to yield to Saul as to whether he thinks this is desirable/possible or not.

The packagedata changes are pretty invasive unfortunately and likely not something you're going to want in dunfell sadly.

Thanks for the clarification.

Quoted 6 lines

Cheers,

Richard

Signature · 3 lines
Best regards,

José Quaresma
#8 lore raw
Quoted 16 lines · Jose Quaresma

Richard Purdie <[email protected]> escreveu no dia quinta, 28/10/2021 à(s) 21:58:

Quoted 12 lines · Steve Sakoman
Quoted 1 line

On Tue, Oct 26, 2021 at 10:41 PM Jose Quaresma <[email protected]>

Quoted 4 lines · wrote:
Quoted 4 lines

Hi all,

There are any plans or is it possible to backport the SBOM/SPDX to the

dunfell branch?

Quoted 3 lines

I'm going to yield to Saul as to whether he thinks this is desirable/possible or not.

Quoted 7 lines
Quoted 3 lines

The packagedata changes are pretty invasive unfortunately and likely not something you're going to want in dunfell sadly.

Thanks for the clarification.

I have been thinking a bit more about this. I did wonder if we should consider a mixin layer of some kind for it that could work with dunfell?

We could host it, it is just a question of writing the mixin layer and maintaining it.

Cheers,

Richard

#9 lore raw
Quoted 17 lines · Richard Purdie
Quoted 16 lines · Jose Quaresma

Richard Purdie <[email protected]> escreveu no dia quinta, 28/10/2021 à(s) 21:58:

Quoted 12 lines · Steve Sakoman
Quoted 1 line

On Tue, Oct 26, 2021 at 10:41 PM Jose Quaresma <[email protected]>

Quoted 3 lines · wrote:
Quoted 3 lines

Hi all,

There are any plans or is it possible to backport the SBOM/SPDX to the

dunfell branch?

Quoted 2 lines

I'm going to yield to Saul as to whether he thinks this is desirable/possible or not.

The packagedata changes are pretty invasive unfortunately and likely not something you're going to want in dunfell sadly.

Quoted 7 lines
Quoted 2 lines

Thanks for the clarification.

I have been thinking a bit more about this. I did wonder if we should consider a mixin layer of some kind for it that could work with dunfell?

We could host it, it is just a question of writing the mixin layer and maintaining it.

I don't think it's going to be possible with a pure mixin layer, since it relies on the extended package data?

Quoted 5 lines

Cheers,

Richard

#10 lore raw
Quoted 20 lines · Joshua Watt
Quoted 19 lines · Richard Purdie
Quoted 18 lines · Jose Quaresma

Richard Purdie <[email protected]> escreveu no dia quinta, 28/10/2021 à(s) 21:58:

Quoted 12 lines · Steve Sakoman
Quoted 1 line

On Tue, Oct 26, 2021 at 10:41 PM Jose Quaresma <[email protected]>

Quoted 3 lines · wrote:
Quoted 3 lines

Hi all,

There are any plans or is it possible to backport the SBOM/SPDX to the

dunfell branch?

Quoted 2 lines

I'm going to yield to Saul as to whether he thinks this is desirable/possible or not.

The packagedata changes are pretty invasive unfortunately and likely not something you're going to want in dunfell sadly.

Thanks for the clarification.

Quoted 8 lines
Quoted 5 lines

I have been thinking a bit more about this. I did wonder if we should consider a mixin layer of some kind for it that could work with dunfell?

We could host it, it is just a question of writing the mixin layer and maintaining it.

I don't think it's going to be possible with a pure mixin layer, since it relies on the extended package data?

I suspect that could perhaps be patched in through a layer though? You might choose to drop the compression piece or do it differently for the backport?

Cheers,

Richard

#11 lore raw
Quoted 22 lines · Richard Purdie
Quoted 21 lines · Joshua Watt
Quoted 20 lines · Richard Purdie
Quoted 7 lines · Jose Quaresma

Richard Purdie <[email protected]> escreveu no dia quinta, 28/10/2021 à(s) 21:58:

Quoted 5 lines · Steve Sakoman
Quoted 1 line

On Tue, Oct 26, 2021 at 10:41 PM Jose Quaresma <[email protected]>

Quoted 2 lines · wrote:
Quoted 2 lines

Hi all,

Quoted 9 lines
Quoted 7 lines
Quoted 1 line
Quoted 1 line

There are any plans or is it possible to backport the SBOM/SPDX to the

dunfell branch?

Quoted 2 lines

I'm going to yield to Saul as to whether he thinks this is desirable/possible or not.

The packagedata changes are pretty invasive unfortunately and likely not something you're going to want in dunfell sadly.

Thanks for the clarification.

I have been thinking a bit more about this. I did wonder if we should consider a mixin layer of some kind for it that could work with dunfell?

Quoted 6 lines
Quoted 4 lines
Quoted 2 lines

We could host it, it is just a question of writing the mixin layer and maintaining it.

I don't think it's going to be possible with a pure mixin layer, since it relies on the extended package data?

I suspect that could perhaps be patched in through a layer though? You might choose to drop the compression piece or do it differently for the backport?

I'm not sure if a layer could hook in well enough to get the data needed...  maybe worth an experiment though

With a backport, I would probably either use GZip compression or no compression. The zstd compression was designed as a drop in replacement for Gzip if we wanted to go that route.

Quoted 5 lines

Cheers,

Richard

author
#12 lore raw
Quoted 26 lines · Joshua Watt
Quoted 24 lines · Richard Purdie
Quoted 23 lines · Joshua Watt
Quoted 22 lines · Richard Purdie
Quoted 7 lines · Jose Quaresma

Richard Purdie <[email protected]> escreveu no dia quinta, 28/10/2021 à(s) 21:58:

Quoted 5 lines · Steve Sakoman
Quoted 1 line

On Tue, Oct 26, 2021 at 10:41 PM Jose Quaresma <[email protected]>

Quoted 2 lines · wrote:
Quoted 2 lines

Hi all,

Quoted 9 lines
Quoted 7 lines
Quoted 1 line
Quoted 1 line

There are any plans or is it possible to backport the SBOM/SPDX to the

dunfell branch?

Quoted 2 lines

I'm going to yield to Saul as to whether he thinks this is desirable/possible or not.

The packagedata changes are pretty invasive unfortunately and likely not something you're going to want in dunfell sadly.

Thanks for the clarification.

I have been thinking a bit more about this. I did wonder if we should consider a mixin layer of some kind for it that could work with dunfell?

We could host it, it is just a question of writing the mixin layer and maintaining it.

Quoted 8 lines
Quoted 4 lines
Quoted 2 lines

I don't think it's going to be possible with a pure mixin layer, since it relies on the extended package data?

I suspect that could perhaps be patched in through a layer though? You might choose to drop the compression piece or do it differently for the backport?

I'm not sure if a layer could hook in well enough to get the data needed...  maybe worth an experiment though

Yeah, I am not sure an mixin could track the changes for package.bbclass

Quoted 3 lines

With a backport, I would probably either use GZip compression or no compression. The zstd compression was designed as a drop in replacement for Gzip if we wanted to go that route.

I will say that we did something similar with Hardknott for WRLinux, but did not propose it upstream as Hardknott was knot going to be supported longer term.

Having the spdx class standalone with the correctly backported changes seems to be working

Sau!

Quoted 5 lines
Quoted 5 lines

Cheers,

Richard

Signature · 1 line
Sau!
#13 lore raw
Quoted 54 lines · On Tuesday, 9 November 2021 08:01:38 NZDT Saul Wold wrote:
Quoted 39 lines · Joshua Watt
Quoted 38 lines · Richard Purdie
Quoted 36 lines · Joshua Watt
Quoted 24 lines · Richard Purdie
Quoted 23 lines · Jose Quaresma

Richard Purdie <[email protected]> escreveu no dia quinta,

28/10/2021 à(s) 21:58:

Quoted 18 lines · Steve Sakoman
Quoted 2 lines

On Tue, Oct 26, 2021 at 10:41 PM Jose Quaresma
<[email protected]>

Quoted 4 lines · wrote:
Quoted 4 lines

Hi all,

There are any plans or is it possible to backport the SBOM/SPDX to the

dunfell branch?

Quoted 2 lines

I'm going to yield to Saul as to whether he thinks this is desirable/possible or not.

The packagedata changes are pretty invasive unfortunately and likely not
something you're going to want in dunfell sadly.

Quoted 8 lines
Quoted 1 line

Thanks for the clarification.

I have been thinking a bit more about this. I did wonder if we should consider a
mixin layer of some kind for it that could work with dunfell?

We could host it, it is just a question of writing the mixin layer and maintaining it.

I don't think it's going to be possible with a pure mixin layer, since it relies on the extended package data?

Quoted 7 lines
Quoted 4 lines

I suspect that could perhaps be patched in through a layer though? You might
choose to drop the compression piece or do it differently for the backport?

I'm not sure if a layer could hook in well enough to get the data needed... maybe worth an experiment though

Yeah, I am not sure an mixin could track the changes for package.bbclass

Quoted 3 lines

With a backport, I would probably either use GZip compression or no compression. The zstd compression was designed as a drop in replacement for Gzip if we wanted to go that route.

Quoted 6 lines

I will say that we did something similar with Hardknott for WRLinux, but did not propose it upstream as Hardknott was knot going to be supported longer term.

Having the spdx class standalone with the correctly backported changes seems to be working

FYI Andres and I have done this backport to dunfell - should I post it? That said, I did just take the hit on some of the invasive parts (e.g. LICENSE value changes). I think given regulatory requirements this is important for lots of folks, so we probably need to do something here. Happy to be part of it.

Cheers
Paul

#14 lore raw

Hi Paul,

Great to hear it.

I think the SPDX/SBOM will be useful for a bunch of users/companies that currently use the yocto LTS dunfell.
It will be an awesome improvement if we have the SPDX/SBOM supported and even more given that the dunfel
branch support has been extended for more than 2 years, until Apr. 2024.

I would be very happy if I could help in any way.

Jose

Paul Eggleton <[email protected]> escreveu no dia segunda, 15/11/2021 à(s) 22:44:

Quoted 81 lines · On Tuesday, 9 November 2021 08:01:38 NZDT Saul Wold wrote:
Quoted 6 lines · Joshua Watt
Quoted 5 lines · Richard Purdie
Quoted 4 lines · Joshua Watt
Quoted 3 lines · Richard Purdie
Quoted 2 lines · Jose Quaresma
Quoted 1 line

Richard Purdie <[email protected]> escreveu no

dia

Quoted 22 lines
Quoted 22 lines
Quoted 22 lines
Quoted 22 lines
Quoted 22 lines
Quoted 22 lines

quinta,

28/10/2021 à(s) 21:58:

Quoted 18 lines · Steve Sakoman
Quoted 2 lines

On Tue, Oct 26, 2021 at 10:41 PM Jose Quaresma
<[email protected]>

Quoted 4 lines · wrote:
Quoted 4 lines

Hi all,

There are any plans or is it possible to backport the SBOM/SPDX to the

dunfell branch?

Quoted 2 lines

I'm going to yield to Saul as to whether he thinks this is desirable/possible or not.

The packagedata changes are pretty invasive unfortunately and likely not
something you're going to want in dunfell sadly.

Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 1 line

Thanks for the clarification.

I have been thinking a bit more about this. I did wonder if we should consider a
mixin layer of some kind for it that could work with dunfell?

We could host it, it is just a question of writing the mixin layer

and

Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 1 line

maintaining it.

I don't think it's going to be possible with a pure mixin layer,

since

Quoted 16 lines
Quoted 9 lines
Quoted 6 lines
Quoted 1 line

it relies on the extended package data?

I suspect that could perhaps be patched in through a layer though? You might
choose to drop the compression piece or do it differently for the backport?

I'm not sure if a layer could hook in well enough to get the data needed... maybe worth an experiment though

Yeah, I am not sure an mixin could track the changes for package.bbclass

Quoted 3 lines

With a backport, I would probably either use GZip compression or no compression. The zstd compression was designed as a drop in replacement for Gzip if we wanted to go that route.

Quoted 6 lines

I will say that we did something similar with Hardknott for WRLinux, but did not propose it upstream as Hardknott was knot going to be supported longer term.

Having the spdx class standalone with the correctly backported changes seems to be working

FYI Andres and I have done this backport to dunfell - should I post it? That
said, I did just take the hit on some of the invasive parts (e.g. LICENSE value changes). I think given regulatory requirements this is important for
lots of folks, so we probably need to do something here. Happy to be part of
it.

Cheers
Paul

Signature · 3 lines
Best regards,

José Quaresma
author
#15 lore raw
In reply to Paul Eggleton
Quoted 59 lines · Paul Eggleton
Quoted 58 lines · On Tuesday, 9 November 2021 08:01:38 NZDT Saul Wold wrote:
Quoted 39 lines · Joshua Watt
Quoted 38 lines · Richard Purdie
Quoted 36 lines · Joshua Watt
Quoted 24 lines · Richard Purdie
Quoted 23 lines · Jose Quaresma

Richard Purdie <[email protected]> escreveu no dia quinta,

28/10/2021 à(s) 21:58:

Quoted 18 lines · Steve Sakoman
Quoted 2 lines

On Tue, Oct 26, 2021 at 10:41 PM Jose Quaresma
<[email protected]>

Quoted 4 lines · wrote:
Quoted 4 lines

Hi all,

There are any plans or is it possible to backport the SBOM/SPDX to the

dunfell branch?

Quoted 2 lines

I'm going to yield to Saul as to whether he thinks this is desirable/possible or not.

The packagedata changes are pretty invasive unfortunately and likely not
something you're going to want in dunfell sadly.

Quoted 8 lines
Quoted 1 line

Thanks for the clarification.

I have been thinking a bit more about this. I did wonder if we should consider a
mixin layer of some kind for it that could work with dunfell?

We could host it, it is just a question of writing the mixin layer and maintaining it.

I don't think it's going to be possible with a pure mixin layer, since it relies on the extended package data?

Quoted 7 lines
Quoted 4 lines

I suspect that could perhaps be patched in through a layer though? You might
choose to drop the compression piece or do it differently for the backport?

I'm not sure if a layer could hook in well enough to get the data needed... maybe worth an experiment though

Yeah, I am not sure an mixin could track the changes for package.bbclass

Quoted 3 lines

With a backport, I would probably either use GZip compression or no compression. The zstd compression was designed as a drop in replacement for Gzip if we wanted to go that route.

I will say that we did something similar with Hardknott for WRLinux, but did not propose it upstream as Hardknott was knot going to be supported longer term.

Quoted 9 lines
Quoted 2 lines

Having the spdx class standalone with the correctly backported changes seems to be working

FYI Andres and I have done this backport to dunfell - should I post it? That said, I did just take the hit on some of the invasive parts (e.g. LICENSE value changes). I think given regulatory requirements this is important for lots of folks, so we probably need to do something here. Happy to be part of it.

Hi Paul, Andres:

We talked about this during the Tech Call this morning and the consensus was that this work should be done in a mix-in style layer so that it could be used by multiple releases.

The LICENSE value changes could be handled by a single file with LICENSE_<package> style overrides in the mix-in layer, or by a set of bbappends in the mix-in layer.

Did you include the compression changes or convert that back to basic XZ compression?

We realize that this make for more work, but it's the problem of backporting a feature to the release vs having the feature in a separate mix-in.

Hope this is clear.

Sau!

Quoted 6 lines

Cheers
Paul

Signature · 1 line
Sau!
author
#16 lore raw
Quoted 71 lines · Saul Wold
Quoted 59 lines · Paul Eggleton
Quoted 58 lines · On Tuesday, 9 November 2021 08:01:38 NZDT Saul Wold wrote:
Quoted 39 lines · Joshua Watt
Quoted 38 lines · Richard Purdie
Quoted 36 lines · Joshua Watt
Quoted 24 lines · Richard Purdie
Quoted 23 lines · Jose Quaresma

Richard Purdie <[email protected]> escreveu no dia quinta,

28/10/2021 à(s) 21:58:

Quoted 18 lines · Steve Sakoman
Quoted 2 lines

On Tue, Oct 26, 2021 at 10:41 PM Jose Quaresma
<[email protected]>

Quoted 4 lines · wrote:
Quoted 4 lines

Hi all,

There are any plans or is it possible to backport the SBOM/SPDX to the

dunfell branch?

Quoted 2 lines

I'm going to yield to Saul as to whether he thinks this is desirable/possible or not.

The packagedata changes are pretty invasive unfortunately and likely not
something you're going to want in dunfell sadly.

Quoted 8 lines
Quoted 1 line

Thanks for the clarification.

I have been thinking a bit more about this. I did wonder if we should consider a
mixin layer of some kind for it that could work with dunfell?

We could host it, it is just a question of writing the mixin layer and maintaining it.

I don't think it's going to be possible with a pure mixin layer, since it relies on the extended package data?

Quoted 7 lines
Quoted 4 lines

I suspect that could perhaps be patched in through a layer though? You might
choose to drop the compression piece or do it differently for the backport?

I'm not sure if a layer could hook in well enough to get the data needed...  maybe worth an experiment though

Yeah, I am not sure an mixin could track the changes for package.bbclass

Quoted 3 lines

With a backport, I would probably either use GZip compression or no compression. The zstd compression was designed as a drop in replacement for Gzip if we wanted to go that route.

I will say that we did something similar with Hardknott for WRLinux, but did not propose it upstream as Hardknott was knot going to be supported longer term.

Quoted 9 lines
Quoted 2 lines

Having the spdx class standalone with the correctly backported changes seems to be working

FYI Andres and I have done this backport to dunfell - should I post it? That said, I did just take the hit on some of the invasive parts (e.g. LICENSE value changes). I think given regulatory requirements this is important for lots of folks, so we probably need to do something here. Happy to be part of it.

Quoted 6 lines

Hi Paul, Andres:

We talked about this during the Tech Call this morning and the consensus was that this work should be done in a mix-in style layer so that it could be used by multiple releases.

The LICENSE value changes could be handled by a single file with LICENSE_<package> style overrides in the mix-in layer, or by a set of bbappends in the mix-in layer.

minor correct: LIENCE_pn-<package_name>

Quoted 15 lines

Did you include the compression changes or convert that back to basic XZ compression?

We realize that this make for more work, but it's the problem of backporting a feature to the release vs having the feature in a separate mix-in.

Hope this is clear.

Sau!

Quoted 6 lines

Cheers
Paul

Signature · 1 line
Sau!