xen-devel.lists.xenproject.org archive mirror
 help / color / mirror / Atom feed
* [PATCH 0/2] xtf: add launcher (+1 bugfix)
@ 2016-05-19 13:46 Roger Pau Monne
  2016-05-19 13:46 ` [PATCH 1/2] xtf: remove setting ROOT path in common.mk Roger Pau Monne
                   ` (2 more replies)
  0 siblings, 3 replies; 43+ messages in thread
From: Roger Pau Monne @ 2016-05-19 13:46 UTC (permalink / raw)
  To: xen-devel

Hello,

This series contains a bugfix for the build infrastructure and a basic 
launcher for XTF. Patches can also be found in the following git repo:

git://xenbits.xen.org/people/royger/xen-test-framework.git launcher_v1

Thanks, Roger.

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
http://lists.xen.org/xen-devel

^ permalink raw reply	[flat|nested] 43+ messages in thread

* [PATCH 1/2] xtf: remove setting ROOT path in common.mk
  2016-05-19 13:46 [PATCH 0/2] xtf: add launcher (+1 bugfix) Roger Pau Monne
@ 2016-05-19 13:46 ` Roger Pau Monne
  2016-05-19 13:52   ` Andrew Cooper
  2016-05-19 13:46 ` [PATCH 2/2] xtf: add a launcher script Roger Pau Monne
  2016-06-09 15:04 ` xenbits "official" repo for XTF (was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix) Ian Jackson
  2 siblings, 1 reply; 43+ messages in thread
From: Roger Pau Monne @ 2016-05-19 13:46 UTC (permalink / raw)
  To: xen-devel; +Cc: andrew.cooper3, Roger Pau Monne

Since it might be included from different paths that have different levels
of nestedness. Also all makefiles that include common.mk already define ROOT
on their own.

Signed-off-by: Roger Pau Monné <roger.pau@citrix.com>
---
Cc: andrew.cooper3@citrix.com
---
 build/common.mk | 1 -
 1 file changed, 1 deletion(-)

diff --git a/build/common.mk b/build/common.mk
index 13c468e..6d4c8e8 100644
--- a/build/common.mk
+++ b/build/common.mk
@@ -1,4 +1,3 @@
-ROOT := $(abspath $(CURDIR)/../..)
 DESTDIR ?= $(ROOT)/dist
 PREFIX ?= $(ROOT)
 CC ?= gcc
-- 
2.7.4 (Apple Git-66)


_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
http://lists.xen.org/xen-devel

^ permalink raw reply related	[flat|nested] 43+ messages in thread

* [PATCH 2/2] xtf: add a launcher script
  2016-05-19 13:46 [PATCH 0/2] xtf: add launcher (+1 bugfix) Roger Pau Monne
  2016-05-19 13:46 ` [PATCH 1/2] xtf: remove setting ROOT path in common.mk Roger Pau Monne
@ 2016-05-19 13:46 ` Roger Pau Monne
  2016-06-09 15:04 ` xenbits "official" repo for XTF (was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix) Ian Jackson
  2 siblings, 0 replies; 43+ messages in thread
From: Roger Pau Monne @ 2016-05-19 13:46 UTC (permalink / raw)
  To: xen-devel
  Cc: andrew.cooper3, anshul.makkar, Ian.Jackson, wei.liu2, Roger Pau Monne

Add a simple script that can list the tests relevant to the current
environment and run them. In it's current form it's functionality is quite
limited, and consists in one of this two options:

 - list: list tests relevant to the current environment. This information is
   fetched from each test test-info.json and the output of `xl info`.
 - test: launch a test (which can include creating several VMs) and report
   the results back. The results are printed on the terminal, and the error
   code is set appropriately (see backend/return_code.py for the list of
   possible error codes).

The path to the directory where the tests are stored is mandatory, and
should always be passed to the script. As an example, I've used the
following bash runes in order to launch all tests suitable for my
environment:

for test in `dist/bin/xtf-launcher.py -f dist/tests/ -l`
do
dist/bin/xtf-launcher.py -f dist/tests/ -t $test
echo "Test $test return code: $?"
done

Which yields the following (trimmed for convenience):

Parsing config from dist/tests/cpuid/test-pv64-cpuid.cfg
test-pv64-cpuid took 0s
Parsing config from dist/tests/cpuid/test-pv32pae-cpuid.cfg
test-pv32pae-cpuid took 0s
Parsing config from dist/tests/cpuid/test-hvm64-cpuid.cfg
libxl: warning: libxl_dm.c:1486:libxl__build_device_model_args_new: Could not find user xen-qemuuser-shared, starting QEMU as root
test-hvm64-cpuid took 0s
Parsing config from dist/tests/cpuid/test-hvm32pae-cpuid.cfg
libxl: warning: libxl_dm.c:1486:libxl__build_device_model_args_new: Could not find user xen-qemuuser-shared, starting QEMU as root
test-hvm32pae-cpuid took 0s
Parsing config from dist/tests/cpuid/test-hvm32pse-cpuid.cfg
libxl: warning: libxl_dm.c:1486:libxl__build_device_model_args_new: Could not find user xen-qemuuser-shared, starting QEMU as root
test-hvm32pse-cpuid took 0s
Parsing config from dist/tests/cpuid/test-hvm32-cpuid.cfg
libxl: warning: libxl_dm.c:1486:libxl__build_device_model_args_new: Could not find user xen-qemuuser-shared, starting QEMU as root
test-hvm32-cpuid took 0s
test-pv64-cpuid:	SUCCESS
test-hvm64-cpuid:	SUCCESS
test-hvm32-cpuid:	SUCCESS
test-hvm32pse-cpuid:	SUCCESS
test-hvm32pae-cpuid:	SUCCESS
test-pv32pae-cpuid:	SUCCESS
Test cpuid return code: 0
[...]

Note that the specific runes to interact with xl have been placed in a
separate file, and that other backends could be easily added provided that
config files suitable for them are also added to XTF tests. A new
--toolstack option should be added then in order to tell the launcher which
toolstack to use.

Signed-off-by: Roger Pau Monné <roger.pau@citrix.com>
---
Cc: andrew.cooper3@citrix.com
Cc: Ian.Jackson@eu.citrix.com
Cc: wei.liu2@citrix.com
Cc: anshul.makkar@citrix.com
---
 Makefile                               |   2 +-
 tools/launcher/Makefile                |  12 ++++
 tools/launcher/backends/__init__.py    |   0
 tools/launcher/backends/return_code.py |  28 +++++++++
 tools/launcher/backends/xl.py          |  59 +++++++++++++++++++
 tools/launcher/xtf-launcher.py         | 101 +++++++++++++++++++++++++++++++++
 6 files changed, 201 insertions(+), 1 deletion(-)
 create mode 100644 tools/launcher/Makefile
 create mode 100644 tools/launcher/backends/__init__.py
 create mode 100644 tools/launcher/backends/return_code.py
 create mode 100644 tools/launcher/backends/xl.py
 create mode 100644 tools/launcher/xtf-launcher.py

diff --git a/Makefile b/Makefile
index fd8c3e0..eda9a25 100644
--- a/Makefile
+++ b/Makefile
@@ -9,7 +9,7 @@ all:
 
 .PHONY: install
 install:
-	@for D in $(wildcard tests/*); do \
+	@for D in $(wildcard tests/* tools/*); do \
 		[ ! -e $$D/Makefile ] && continue; \
 		$(MAKE) -C $$D install; \
 	done
diff --git a/tools/launcher/Makefile b/tools/launcher/Makefile
new file mode 100644
index 0000000..2c07604
--- /dev/null
+++ b/tools/launcher/Makefile
@@ -0,0 +1,12 @@
+ROOT := $(abspath $(CURDIR)/../..)
+
+include $(ROOT)/build/common.mk
+
+TOOLS := xtf-launcher.py
+BACKENDS := __init__.py xl.py return_code.py
+
+install: $(TOOLS) $(addprefix backends/,$(BACKENDS))
+	@mkdir -p $(DESTDIR)/bin
+	@mkdir -p $(DESTDIR)/bin/backends
+	install -m775 -p $(TOOLS) $(DESTDIR)/bin
+	install -m664 -p $(addprefix backends/,$(BACKENDS)) $(DESTDIR)/bin/backends/
diff --git a/tools/launcher/backends/__init__.py b/tools/launcher/backends/__init__.py
new file mode 100644
index 0000000..e69de29
diff --git a/tools/launcher/backends/return_code.py b/tools/launcher/backends/return_code.py
new file mode 100644
index 0000000..509460a
--- /dev/null
+++ b/tools/launcher/backends/return_code.py
@@ -0,0 +1,28 @@
+#
+# Error codes for XTF launcher
+#
+
+# All went fine.
+SUCCESS = 0
+
+# Bug in the launcher.
+ERROR_LAUNCHER = 1
+
+# Test cannot be completed.
+ERROR_SKIP = 2
+
+# Bug in the test code or environment.
+ERROR_ERROR = 3
+
+# Bug in the code under test.
+ERROR_FAILURE = 4
+
+def code_to_str(code):
+    err_trans = {
+        0   :   'SUCCESS',
+        1   :   'LAUNCHER ERROR',
+        2   :   'SKIP',
+        3   :   'ERROR',
+        4   :   'FAILURE',
+    }
+    return err_trans[code]
diff --git a/tools/launcher/backends/xl.py b/tools/launcher/backends/xl.py
new file mode 100644
index 0000000..c1ac111
--- /dev/null
+++ b/tools/launcher/backends/xl.py
@@ -0,0 +1,59 @@
+#
+# XL backend for XTF
+#
+
+import subprocess, sys
+import return_code
+
+# XXX: multiprocessing is _not_ available in python < 2.6
+from multiprocessing.pool import Pool
+
+from timeit import default_timer as timer
+
+def _xl_console_helper(domain_name):
+    output = subprocess.check_output(['xl', 'console', domain_name])
+    result_line = output.splitlines()[-1]
+    if "SUCCESS" in result_line:
+        return return_code.SUCCESS
+    elif "SKIP" in result_line:
+        return return_code.ERROR_SKIP
+    elif "ERROR" in result_line:
+        return return_code.ERROR_ERROR
+    elif "FAILURE" in result_line:
+        return return_code.ERROR_FALIURE
+
+    # Something went wrong, no test report line was found (or it's invalid).
+    return return_code.ERROR_LAUNCHER
+
+class xl:
+    """XL driver for XTF"""
+
+    def __init__(self):
+        self._thpool = Pool(processes=1)
+
+    def get_supported_environments(self):
+        output = subprocess.check_output(['xl', 'info'])
+        #print output
+        for line in output.splitlines():
+            #print line
+            if not line.startswith("xen_caps"):
+                continue
+            caps = line.split()[2:]
+            return caps
+        return None
+
+    def run_test(self, test_cfg, domain_name):
+        rc = subprocess.call(['xl', 'create', '-p', test_cfg])
+        if rc:
+            sys.stderr.write("Failed to create VM from config file %s\n" %
+                             test_cfg)
+            return return_code.ERROR_LAUNCHER
+        thread = self._thpool.apply_async(_xl_console_helper, (domain_name, ))
+        start = timer()
+        rc = subprocess.call(['xl', 'unpause', domain_name])
+        if rc:
+            sys.stderr.write("Failed to unpause VM %s\n" % domain_name)
+            return return_code.ERROR_LAUNCHER
+        rc = thread.get()
+        sys.stderr.write("%s took %ds\n" % (domain_name, timer() - start))
+        return rc
diff --git a/tools/launcher/xtf-launcher.py b/tools/launcher/xtf-launcher.py
new file mode 100644
index 0000000..0685c2f
--- /dev/null
+++ b/tools/launcher/xtf-launcher.py
@@ -0,0 +1,101 @@
+#!/usr/bin/env python
+#
+# Launcher script for XTF
+#
+
+# XXX: JSON module is _not_ available on Python < 2.6
+import os, sys, json
+import backends.xl
+import backends.return_code as return_code
+
+from optparse import OptionParser
+
+def exit_error(error_str):
+    sys.stderr.write(error_str)
+    sys.exit(return_code.ERROR_LAUNCHER)
+
+def check_envinronment(req_envs, avail_envs):
+    env_trans = {
+        'pv32pae'   :   'xen-3.0-x86_32p',
+        'pv64'      :   'xen-3.0-x86_64',
+        'hvm32'     :   'hvm-3.0-x86_32',
+        'hvm32pse'  :   'hvm-3.0-x86_32',
+        'hvm32pae'  :   'hvm-3.0-x86_32p',
+        'hvm64'     :   'hvm-3.0-x86_64',
+    }
+    for env in req_envs:
+        if env_trans[env] not in avail_envs:
+            return False
+    return True
+
+def list():
+    for folder in os.listdir(options.folder):
+        if not os.path.isdir(os.path.join(options.folder, folder)):
+            continue
+        test_info = os.path.join(options.folder, folder, 'test-info.json')
+        if not os.path.isfile(test_info):
+            sys.stderr.write("Unable to find %s, ignoring folder\n" % testinfo)
+            continue
+        with open(test_info) as data_file:
+            info = json.load(data_file)
+        envs = backend.get_supported_environments()
+        if not envs:
+            exit_error("Unable to get Xen supported environments\n")
+        avail = check_envinronment(info["environments"], envs)
+        if not avail:
+            sys.stderr.write("Skipping test %s: not supported\n" % folder)
+            continue
+        print folder
+    return return_code.SUCCESS
+
+def test():
+    test_folder = os.path.join(options.folder, options.test)
+    if not os.path.isdir(test_folder):
+        exit_error("%s is not a directory\n" % test_folder)
+    test_info = os.path.join(test_folder, 'test-info.json')
+    if not os.path.isfile(test_info):
+        exit_error("%s doesn't exist" % test_info)
+    with open(test_info) as data_file:
+        info = json.load(data_file)
+    result = {}
+    for env in info["environments"]:
+        test_cfg = os.path.join(test_folder,
+                                "test-%s-%s.cfg" % (env, options.test))
+        test_name = "test-%s-%s" % (env, options.test)
+        if not os.path.isfile(test_cfg):
+            exit_error("Unable to find test config file %s\n" % test_cfg)
+        result[test_name] = backend.run_test(test_cfg, test_name)
+    # Print all the results
+    result_code = 0
+    for key, value in result.iteritems():
+        print "%s:\t%s" % (key, return_code.code_to_str(value))
+        # Return the highest error code in case of error.
+        result_code = max(value, result_code)
+    return result_code
+
+parser = OptionParser()
+parser.add_option("-f", "--folder", dest="folder",
+                  help="FOLDER where XTF tests are located", metavar="FOLDER")
+parser.add_option("-t", "--test", dest="test",
+                  help="NAME of test to run", metavar="NAME")
+parser.add_option("-l", "--list", action="store_true", dest="list",
+                  help="List test that are applicable to current environment")
+(options, args) = parser.parse_args()
+
+if not options.folder:
+    exit_error("folder not given\n")
+
+if options.list and options.test:
+    exit_error("options -l and -t are mutually exclusive\n")
+
+if not os.path.isdir(options.folder):
+    exit_error("%s does not exist or is not a folder\n" % options.folder)
+
+backend = backends.xl.xl()
+
+if options.list:
+    rc = list()
+elif options.test:
+    rc = test()
+
+sys.exit(rc)
-- 
2.7.4 (Apple Git-66)


_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
http://lists.xen.org/xen-devel

^ permalink raw reply related	[flat|nested] 43+ messages in thread

* Re: [PATCH 1/2] xtf: remove setting ROOT path in common.mk
  2016-05-19 13:46 ` [PATCH 1/2] xtf: remove setting ROOT path in common.mk Roger Pau Monne
@ 2016-05-19 13:52   ` Andrew Cooper
  0 siblings, 0 replies; 43+ messages in thread
From: Andrew Cooper @ 2016-05-19 13:52 UTC (permalink / raw)
  To: Roger Pau Monne, xen-devel

On 19/05/16 14:46, Roger Pau Monne wrote:
> Since it might be included from different paths that have different levels
> of nestedness. Also all makefiles that include common.mk already define ROOT
> on their own.
>
> Signed-off-by: Roger Pau Monné <roger.pau@citrix.com>

I really need to work on improving the build system.  This one is
already creaking at the seams.

LGTM - Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com> and committed.

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
http://lists.xen.org/xen-devel

^ permalink raw reply	[flat|nested] 43+ messages in thread

* xenbits "official" repo for XTF (was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix)
  2016-05-19 13:46 [PATCH 0/2] xtf: add launcher (+1 bugfix) Roger Pau Monne
  2016-05-19 13:46 ` [PATCH 1/2] xtf: remove setting ROOT path in common.mk Roger Pau Monne
  2016-05-19 13:46 ` [PATCH 2/2] xtf: add a launcher script Roger Pau Monne
@ 2016-06-09 15:04 ` Ian Jackson
  2016-06-13  8:59   ` George Dunlap
                     ` (2 more replies)
  2 siblings, 3 replies; 43+ messages in thread
From: Ian Jackson @ 2016-06-09 15:04 UTC (permalink / raw)
  To: Andrew Cooper; +Cc: xen-devel, Anshul Makkar, Wei Liu, Roger Pau Monne

Roger Pau Monne writes ("[Xen-devel] [PATCH 0/2] xtf: add launcher (+1 bugfix)"):
> This series contains a bugfix for the build infrastructure and a basic 
> launcher for XTF. Patches can also be found in the following git repo:
> 
> git://xenbits.xen.org/people/royger/xen-test-framework.git launcher_v1

IMO we should give this thing a proper repo in the root of the xenbits
git namespace.

I'm slightly concerned that `xen-test-framework' is a bit vague.  It
sounds like a CI system (of which we already have at least three ...),
rather than a collection of tests based on tiny VMs, each designed to
make specific demands of the host environment.

Andrew, do you have other naming ideas ?  I don't object to "xtf" but
that's just an acronym, and the expansion is confusing.  I think
qualifying the somehow might be useful.

Others: do you have opinions ?  Is my concern ill-founded ?

Thanks,
Ian.

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
http://lists.xen.org/xen-devel

^ permalink raw reply	[flat|nested] 43+ messages in thread

* Re: xenbits "official" repo for XTF (was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix)
  2016-06-09 15:04 ` xenbits "official" repo for XTF (was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix) Ian Jackson
@ 2016-06-13  8:59   ` George Dunlap
  2016-06-13 10:10   ` Andrew Cooper
  2016-06-14 17:01   ` Konrad Rzeszutek Wilk
  2 siblings, 0 replies; 43+ messages in thread
From: George Dunlap @ 2016-06-13  8:59 UTC (permalink / raw)
  To: Ian Jackson
  Cc: Andrew Cooper, Anshul Makkar, Roger Pau Monne, Wei Liu, xen-devel

On Thu, Jun 9, 2016 at 4:04 PM, Ian Jackson <Ian.Jackson@eu.citrix.com> wrote:
> Roger Pau Monne writes ("[Xen-devel] [PATCH 0/2] xtf: add launcher (+1 bugfix)"):
>> This series contains a bugfix for the build infrastructure and a basic
>> launcher for XTF. Patches can also be found in the following git repo:
>>
>> git://xenbits.xen.org/people/royger/xen-test-framework.git launcher_v1
>
> IMO we should give this thing a proper repo in the root of the xenbits
> git namespace.
>
> I'm slightly concerned that `xen-test-framework' is a bit vague.  It
> sounds like a CI system (of which we already have at least three ...),
> rather than a collection of tests based on tiny VMs, each designed to
> make specific demands of the host environment.
>
> Andrew, do you have other naming ideas ?  I don't object to "xtf" but
> that's just an acronym, and the expansion is confusing.  I think
> qualifying the somehow might be useful.
>
> Others: do you have opinions ?  Is my concern ill-founded ?

I agree -- if someone had heard about osstest, but had forgotten the
name, they might well search for "xen test framework" and end up
rather frustrated.

Since the purpose of these tests is to verify very specific bits of
functionality (not unlike processor verification actually), maybe
something with some form of "verify" in the name?  Xen verification
framework / xvf?

 -George

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
http://lists.xen.org/xen-devel

^ permalink raw reply	[flat|nested] 43+ messages in thread

* Re: xenbits "official" repo for XTF (was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix)
  2016-06-09 15:04 ` xenbits "official" repo for XTF (was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix) Ian Jackson
  2016-06-13  8:59   ` George Dunlap
@ 2016-06-13 10:10   ` Andrew Cooper
  2016-06-13 14:03     ` George Dunlap
  2016-06-14 17:01   ` Konrad Rzeszutek Wilk
  2 siblings, 1 reply; 43+ messages in thread
From: Andrew Cooper @ 2016-06-13 10:10 UTC (permalink / raw)
  To: Ian Jackson; +Cc: xen-devel, Anshul Makkar, Wei Liu, Roger Pau Monne

On 09/06/16 16:04, Ian Jackson wrote:
> Roger Pau Monne writes ("[Xen-devel] [PATCH 0/2] xtf: add launcher (+1 bugfix)"):
>> This series contains a bugfix for the build infrastructure and a basic 
>> launcher for XTF. Patches can also be found in the following git repo:
>>
>> git://xenbits.xen.org/people/royger/xen-test-framework.git launcher_v1
> IMO we should give this thing a proper repo in the root of the xenbits
> git namespace.
>
> I'm slightly concerned that `xen-test-framework' is a bit vague.  It
> sounds like a CI system (of which we already have at least three ...),
> rather than a collection of tests based on tiny VMs, each designed to
> make specific demands of the host environment.
>
> Andrew, do you have other naming ideas ?  I don't object to "xtf" but
> that's just an acronym, and the expansion is confusing.  I think
> qualifying the somehow might be useful.
>
> Others: do you have opinions ?  Is my concern ill-founded ?

The XTF name has been around for 2 years now.

Out of curiosity, I searched for it on google, and found my written
documentation as the top hit.

https://www.google.co.uk/search?q=xen+test+framework&ie=utf-8&oe=utf-8&gws_rd=cr&ei=InxeV52HDYHc-QG0-qN4

The snippet summary includes the first line of the introduction, "This
is both a framework for creating microkernel-based tests, and a suite of
tests built using the framework", which should be sufficient
clarification for anyone with questions.

I am not completely averse to changing it, but I don't see an
alternative which is any better or clearer.

~Andrew

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
http://lists.xen.org/xen-devel

^ permalink raw reply	[flat|nested] 43+ messages in thread

* Re: xenbits "official" repo for XTF (was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix)
  2016-06-13 10:10   ` Andrew Cooper
@ 2016-06-13 14:03     ` George Dunlap
  2016-06-13 14:11       ` Ian Jackson
  0 siblings, 1 reply; 43+ messages in thread
From: George Dunlap @ 2016-06-13 14:03 UTC (permalink / raw)
  To: Andrew Cooper
  Cc: xen-devel, Anshul Makkar, Ian Jackson, Wei Liu, Roger Pau Monne

On Mon, Jun 13, 2016 at 11:10 AM, Andrew Cooper
<andrew.cooper3@citrix.com> wrote:
> On 09/06/16 16:04, Ian Jackson wrote:
>> Roger Pau Monne writes ("[Xen-devel] [PATCH 0/2] xtf: add launcher (+1 bugfix)"):
>>> This series contains a bugfix for the build infrastructure and a basic
>>> launcher for XTF. Patches can also be found in the following git repo:
>>>
>>> git://xenbits.xen.org/people/royger/xen-test-framework.git launcher_v1
>> IMO we should give this thing a proper repo in the root of the xenbits
>> git namespace.
>>
>> I'm slightly concerned that `xen-test-framework' is a bit vague.  It
>> sounds like a CI system (of which we already have at least three ...),
>> rather than a collection of tests based on tiny VMs, each designed to
>> make specific demands of the host environment.
>>
>> Andrew, do you have other naming ideas ?  I don't object to "xtf" but
>> that's just an acronym, and the expansion is confusing.  I think
>> qualifying the somehow might be useful.
>>
>> Others: do you have opinions ?  Is my concern ill-founded ?
>
> The XTF name has been around for 2 years now.
>
> Out of curiosity, I searched for it on google, and found my written
> documentation as the top hit.
>
> https://www.google.co.uk/search?q=xen+test+framework&ie=utf-8&oe=utf-8&gws_rd=cr&ei=InxeV52HDYHc-QG0-qN4
>
> The snippet summary includes the first line of the introduction, "This
> is both a framework for creating microkernel-based tests, and a suite of
> tests built using the framework", which should be sufficient
> clarification for anyone with questions.
>
> I am not completely averse to changing it, but I don't see an
> alternative which is any better or clearer.

Do you have any opinions on "Xen verification framework"?

If we weren't going to change it, then you should probably put a note
somewhere near the top of the introduction saying something like, "If
you're looking for the Xen automated push-gate software, you want
osstest[link]", so that people who want osstest but search for "xen
test framework" find what they're looking for.

 -George

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
http://lists.xen.org/xen-devel

^ permalink raw reply	[flat|nested] 43+ messages in thread

* Re: xenbits "official" repo for XTF (was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix)
  2016-06-13 14:03     ` George Dunlap
@ 2016-06-13 14:11       ` Ian Jackson
  2016-06-14 17:27         ` Andrew Cooper
  0 siblings, 1 reply; 43+ messages in thread
From: Ian Jackson @ 2016-06-13 14:11 UTC (permalink / raw)
  To: George Dunlap
  Cc: Andrew Cooper, Anshul Makkar, Roger Pau Monne, Wei Liu, xen-devel

George Dunlap writes ("Re: [Xen-devel] xenbits "official" repo for XTF (was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix)"):
> On Mon, Jun 13, 2016 at 11:10 AM, Andrew Cooper
> > I am not completely averse to changing it, but I don't see an
> > alternative which is any better or clearer.
> 
> Do you have any opinions on "Xen verification framework"?

The problem IMO is the word "framework" which implies some kind of
daemonic machinery.  The frameworky bit of XTF is quite small; mostly,
it is a suite of test cases.  "Xen Micro-VM Test Suite" maybe ?

(I think "verification" is inaccurate because to me it implies formal
methods, static analysis, or the like.)

> If we weren't going to change it, then you should probably put a note
> somewhere near the top of the introduction saying something like, "If
> you're looking for the Xen automated push-gate software, you want
> osstest[link]", so that people who want osstest but search for "xen
> test framework" find what they're looking for.

"CI (continuous integration)" is the keyword that many people will
have for osstest.

I would suggest

  (This is not the Xen Project's CI / Continuous Integration /
   automated push gate system.  For that, see
   <a href="wiki">osstest</a>.)

or something.

Ian.

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
http://lists.xen.org/xen-devel

^ permalink raw reply	[flat|nested] 43+ messages in thread

* Re: xenbits "official" repo for XTF (was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix)
  2016-06-09 15:04 ` xenbits "official" repo for XTF (was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix) Ian Jackson
  2016-06-13  8:59   ` George Dunlap
  2016-06-13 10:10   ` Andrew Cooper
@ 2016-06-14 17:01   ` Konrad Rzeszutek Wilk
  2 siblings, 0 replies; 43+ messages in thread
From: Konrad Rzeszutek Wilk @ 2016-06-14 17:01 UTC (permalink / raw)
  To: Ian Jackson
  Cc: Andrew Cooper, Anshul Makkar, Roger Pau Monne, Wei Liu, xen-devel

On Thu, Jun 09, 2016 at 04:04:59PM +0100, Ian Jackson wrote:
> Roger Pau Monne writes ("[Xen-devel] [PATCH 0/2] xtf: add launcher (+1 bugfix)"):
> > This series contains a bugfix for the build infrastructure and a basic 
> > launcher for XTF. Patches can also be found in the following git repo:
> > 
> > git://xenbits.xen.org/people/royger/xen-test-framework.git launcher_v1
> 
> IMO we should give this thing a proper repo in the root of the xenbits
> git namespace.
> 
> I'm slightly concerned that `xen-test-framework' is a bit vague.  It
> sounds like a CI system (of which we already have at least three ...),
> rather than a collection of tests based on tiny VMs, each designed to
> make specific demands of the host environment.

xtf-os.git ?

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
http://lists.xen.org/xen-devel

^ permalink raw reply	[flat|nested] 43+ messages in thread

* Re: xenbits "official" repo for XTF (was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix)
  2016-06-13 14:11       ` Ian Jackson
@ 2016-06-14 17:27         ` Andrew Cooper
  2016-06-15  8:31           ` George Dunlap
  0 siblings, 1 reply; 43+ messages in thread
From: Andrew Cooper @ 2016-06-14 17:27 UTC (permalink / raw)
  To: Ian Jackson, George Dunlap
  Cc: xen-devel, Anshul Makkar, Wei Liu, Roger Pau Monne

On 13/06/16 15:11, Ian Jackson wrote:
> George Dunlap writes ("Re: [Xen-devel] xenbits "official" repo for XTF (was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix)"):
>> On Mon, Jun 13, 2016 at 11:10 AM, Andrew Cooper
>>> I am not completely averse to changing it, but I don't see an
>>> alternative which is any better or clearer.
>> Do you have any opinions on "Xen verification framework"?
> The problem IMO is the word "framework" which implies some kind of
> daemonic machinery.

I disagree.  Framework implies building blocks.

I can't see any implication or suggestion of daemonic machinery here.

> The frameworky bit of XTF is quite small; mostly,
> it is a suite of test cases.  "Xen Micro-VM Test Suite" maybe ?

At the moment, the framework is larger than the suite, although I don't
necessarily expect this to remain the case.  Having said that, I would
argue that the "framework" is the more important part; It is the bit
with the complicated logic to build the same code into multiple
different VM types, and APIs to make test reporting easy and consistent.

>
> (I think "verification" is inaccurate because to me it implies formal
> methods, static analysis, or the like.)

I concur.  "verification" is definitely a less accurate description.

>
>> If we weren't going to change it, then you should probably put a note
>> somewhere near the top of the introduction saying something like, "If
>> you're looking for the Xen automated push-gate software, you want
>> osstest[link]", so that people who want osstest but search for "xen
>> test framework" find what they're looking for.
> "CI (continuous integration)" is the keyword that many people will
> have for osstest.
>
> I would suggest
>
>   (This is not the Xen Project's CI / Continuous Integration /
>    automated push gate system.  For that, see
>    <a href="wiki">osstest</a>.)
>
> or something.

I am happy to add text to that effect.

~Andrew



_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
http://lists.xen.org/xen-devel

^ permalink raw reply	[flat|nested] 43+ messages in thread

* Re: xenbits "official" repo for XTF (was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix)
  2016-06-14 17:27         ` Andrew Cooper
@ 2016-06-15  8:31           ` George Dunlap
  2016-06-20 17:03             ` Ian Jackson
  0 siblings, 1 reply; 43+ messages in thread
From: George Dunlap @ 2016-06-15  8:31 UTC (permalink / raw)
  To: Andrew Cooper
  Cc: xen-devel, Anshul Makkar, Ian Jackson, Wei Liu, Roger Pau Monne

On Tue, Jun 14, 2016 at 6:27 PM, Andrew Cooper
<andrew.cooper3@citrix.com> wrote:
> On 13/06/16 15:11, Ian Jackson wrote:
>> George Dunlap writes ("Re: [Xen-devel] xenbits "official" repo for XTF (was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix)"):
>>> On Mon, Jun 13, 2016 at 11:10 AM, Andrew Cooper
>>>> I am not completely averse to changing it, but I don't see an
>>>> alternative which is any better or clearer.
>>> Do you have any opinions on "Xen verification framework"?
>> The problem IMO is the word "framework" which implies some kind of
>> daemonic machinery.
>
> I disagree.  Framework implies building blocks.
>
> I can't see any implication or suggestion of daemonic machinery here.

I agree with Andrew here -- "framework" is used in many circumstances
to indicate a pre-existing structure or set of components which can be
used to build something;  "web development framework" clearly has
nothing to do with daemons.  So "framework" is (from what I gather,
not having actually looked at the code) the exact word to describe the
code.

 -George

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
http://lists.xen.org/xen-devel

^ permalink raw reply	[flat|nested] 43+ messages in thread

* Re: xenbits "official" repo for XTF (was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix)
  2016-06-15  8:31           ` George Dunlap
@ 2016-06-20 17:03             ` Ian Jackson
  2016-06-20 17:10               ` Konrad Rzeszutek Wilk
                                 ` (2 more replies)
  0 siblings, 3 replies; 43+ messages in thread
From: Ian Jackson @ 2016-06-20 17:03 UTC (permalink / raw)
  To: George Dunlap
  Cc: Andrew Cooper, Anshul Makkar, Roger Pau Monne, Wei Liu, xen-devel

We are in danger of getting stuck on this naming question.  I would
like everyone to put forward some suggestions for the name of thisr
toplevel epo on xenbits.

Hopefully we can find one that Andrew likes and that's acceptable to
the committers.

I suggest
  xen-microvm-test-framework
  xen-microvm-test-suite
  xtf-microvm-suite

Ian.

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
http://lists.xen.org/xen-devel

^ permalink raw reply	[flat|nested] 43+ messages in thread

* Re: xenbits "official" repo for XTF (was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix)
  2016-06-20 17:03             ` Ian Jackson
@ 2016-06-20 17:10               ` Konrad Rzeszutek Wilk
  2016-06-21 10:17               ` Wei Liu
  2016-06-21 10:36               ` David Vrabel
  2 siblings, 0 replies; 43+ messages in thread
From: Konrad Rzeszutek Wilk @ 2016-06-20 17:10 UTC (permalink / raw)
  To: Ian Jackson
  Cc: Wei Liu, Andrew Cooper, George Dunlap, xen-devel, Anshul Makkar,
	Roger Pau Monne

>   xtf-microvm-suite

+1

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
http://lists.xen.org/xen-devel

^ permalink raw reply	[flat|nested] 43+ messages in thread

* Re: xenbits "official" repo for XTF (was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix)
  2016-06-20 17:03             ` Ian Jackson
  2016-06-20 17:10               ` Konrad Rzeszutek Wilk
@ 2016-06-21 10:17               ` Wei Liu
  2016-06-21 10:36               ` David Vrabel
  2 siblings, 0 replies; 43+ messages in thread
From: Wei Liu @ 2016-06-21 10:17 UTC (permalink / raw)
  To: Ian Jackson
  Cc: Wei Liu, Andrew Cooper, George Dunlap, xen-devel, Anshul Makkar,
	Roger Pau Monne

On Mon, Jun 20, 2016 at 06:03:42PM +0100, Ian Jackson wrote:
> We are in danger of getting stuck on this naming question.  I would
> like everyone to put forward some suggestions for the name of thisr
> toplevel epo on xenbits.
> 
> Hopefully we can find one that Andrew likes and that's acceptable to
> the committers.
> 
> I suggest
>   xen-microvm-test-framework

Either this or the original name looks ok to me.

>   xen-microvm-test-suite
>   xtf-microvm-suite
> 
> Ian.

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
http://lists.xen.org/xen-devel

^ permalink raw reply	[flat|nested] 43+ messages in thread

* Re: xenbits "official" repo for XTF (was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix)
  2016-06-20 17:03             ` Ian Jackson
  2016-06-20 17:10               ` Konrad Rzeszutek Wilk
  2016-06-21 10:17               ` Wei Liu
@ 2016-06-21 10:36               ` David Vrabel
  2016-06-21 10:42                 ` Ian Jackson
  2016-07-01 19:04                 ` Ian Jackson
  2 siblings, 2 replies; 43+ messages in thread
From: David Vrabel @ 2016-06-21 10:36 UTC (permalink / raw)
  To: Ian Jackson, George Dunlap
  Cc: Andrew Cooper, Anshul Makkar, Wei Liu, xen-devel, Roger Pau Monne

On 20/06/16 18:03, Ian Jackson wrote:
> We are in danger of getting stuck on this naming question.  I would
> like everyone to put forward some suggestions for the name of thisr
> toplevel epo on xenbits.
> 
> Hopefully we can find one that Andrew likes and that's acceptable to
> the committers.
> 
> I suggest
>   xen-microvm-test-framework
>   xen-microvm-test-suite
>   xtf-microvm-suite

"xtf"

It seems unfair to give Andrew's project a clunky (repo) name because
osstest is not sufficiently discoverable.

Perhaps you should consider renaming osstest to xen-ci-system instead?

David

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
http://lists.xen.org/xen-devel

^ permalink raw reply	[flat|nested] 43+ messages in thread

* Re: xenbits "official" repo for XTF (was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix)
  2016-06-21 10:36               ` David Vrabel
@ 2016-06-21 10:42                 ` Ian Jackson
  2016-07-01 19:04                 ` Ian Jackson
  1 sibling, 0 replies; 43+ messages in thread
From: Ian Jackson @ 2016-06-21 10:42 UTC (permalink / raw)
  To: David Vrabel
  Cc: Wei Liu, Andrew Cooper, George Dunlap, xen-devel, Anshul Makkar,
	Roger Pau Monne

David Vrabel writes ("Re: [Xen-devel] xenbits "official" repo for XTF (was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix)"):
> It seems unfair to give Andrew's project a clunky (repo) name because
> osstest is not sufficiently discoverable.
> 
> Perhaps you should consider renaming osstest to xen-ci-system instead?

Sure, how about "xen-test-framework" ?  :-P

Ian.

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
http://lists.xen.org/xen-devel

^ permalink raw reply	[flat|nested] 43+ messages in thread

* Re: xenbits "official" repo for XTF (was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix)
  2016-06-21 10:36               ` David Vrabel
  2016-06-21 10:42                 ` Ian Jackson
@ 2016-07-01 19:04                 ` Ian Jackson
  2016-07-06 11:19                   ` Lars Kurth
  1 sibling, 1 reply; 43+ messages in thread
From: Ian Jackson @ 2016-07-01 19:04 UTC (permalink / raw)
  To: Lars Kurth
  Cc: Wei Liu, Andrew Cooper, George Dunlap, David Vrabel, xen-devel,
	Anshul Makkar, Roger Pau Monne

David Vrabel writes ("Re: [Xen-devel] xenbits "official" repo for XTF (was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix)"):
> On 20/06/16 18:03, Ian Jackson wrote:
> > Hopefully we can find one that Andrew likes and that's acceptable to
> > the committers.
> > 
> > I suggest
> >   xen-microvm-test-framework
> >   xen-microvm-test-suite
> >   xtf-microvm-suite
> 
> "xtf"

iwj@xenbits:~$ ls ~xen/git
kemari                         qemu-xen-4.1-testing.git
libvirt.git                    qemu-xen-4.2-testing.git
linux-2.6-xen.git              qemu-xen-4.3-testing.git
linux-pvops.git                qemu-xen-4.4-testing.git
livepatch-build-tools.git      qemu-xen-4.5-testing.git
mini-os.git                    qemu-xen-4.6-testing.git
osstest                        qemu-xen.git
osstest.git                    qemu-xen-traditional.git
ovmf.git                       qemu-xen-unstable.git
people                         raisin.git
pvdrivers                      rumpuser-xen.git
qemu-upstream-4.2-testing.git  seabios.git
qemu-upstream-4.3-testing.git  staging
qemu-upstream-4.4-testing.git  xcp
qemu-upstream-4.5-testing.git  xenalyze.git
qemu-upstream-4.6-testing.git  xenclient
qemu-upstream-unstable.git     xen.git
qemu-xen-3.3-testing.git       xen.git-aside
qemu-xen-3.4-testing.git       xenrt-citrix
qemu-xen-4.0-testing.git       xentesttools
iwj@xenbits:~$

> It seems unfair to give Andrew's project a clunky (repo) name because
> osstest is not sufficiently discoverable.

Is that a complaint that it's too long ?

I could live with "xtf", although I think it's rather too short.

Lars, can you please advise what process we need to use to come to
closure on this decision ?

Ian.

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
http://lists.xen.org/xen-devel

^ permalink raw reply	[flat|nested] 43+ messages in thread

* Re: xenbits "official" repo for XTF (was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix)
  2016-07-01 19:04                 ` Ian Jackson
@ 2016-07-06 11:19                   ` Lars Kurth
  2016-07-06 14:22                     ` Konrad Rzeszutek Wilk
  0 siblings, 1 reply; 43+ messages in thread
From: Lars Kurth @ 2016-07-06 11:19 UTC (permalink / raw)
  To: Ian Jackson
  Cc: Wei Liu, Andrew Cooper, George Dunlap, David Vrabel, xen-devel,
	Anshul Makkar, Roger Pau Monne



On 01/07/2016 20:04, "Ian Jackson" <ian.jackson@eu.citrix.com> wrote:

>David Vrabel writes ("Re: [Xen-devel] xenbits "official" repo for XTF
>(was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix)"):
>> On 20/06/16 18:03, Ian Jackson wrote:
>> > Hopefully we can find one that Andrew likes and that's acceptable to
>> > the committers.
>> > 
>> > I suggest
>> >   xen-microvm-test-framework
>> >   xen-microvm-test-suite
>> >   xtf-microvm-suite
>> 
>> "xtf"
>
>iwj@xenbits:~$ ls ~xen/git
>kemari                         qemu-xen-4.1-testing.git
>libvirt.git                    qemu-xen-4.2-testing.git
>linux-2.6-xen.git              qemu-xen-4.3-testing.git
>linux-pvops.git                qemu-xen-4.4-testing.git
>livepatch-build-tools.git      qemu-xen-4.5-testing.git
>mini-os.git                    qemu-xen-4.6-testing.git
>osstest                        qemu-xen.git
>osstest.git                    qemu-xen-traditional.git
>ovmf.git                       qemu-xen-unstable.git
>people                         raisin.git
>pvdrivers                      rumpuser-xen.git
>qemu-upstream-4.2-testing.git  seabios.git
>qemu-upstream-4.3-testing.git  staging
>qemu-upstream-4.4-testing.git  xcp
>qemu-upstream-4.5-testing.git  xenalyze.git
>qemu-upstream-4.6-testing.git  xenclient
>qemu-upstream-unstable.git     xen.git
>qemu-xen-3.3-testing.git       xen.git-aside
>qemu-xen-3.4-testing.git       xenrt-citrix
>qemu-xen-4.0-testing.git       xentesttools
>iwj@xenbits:~$
>
>> It seems unfair to give Andrew's project a clunky (repo) name because
>> osstest is not sufficiently discoverable.
>
>Is that a complaint that it's too long ?
>
>I could live with "xtf", although I think it's rather too short.

Let me just summarise my understanding of the discussion:
A) The label xtf has been around for a while, thus changing it may be
confusing
B) Generally there is concern about not being able to discover test
related technology easily in the git tree
C) More verbose repository names seem to be a little long
D) None of the other proposals on the table seem to be much clearer / more
accurate

I couldn't tell from the thread who really is in favour. As far as I can
tell

1 committer is in favour of using xtf.git
2 committer have voiced concrete concerns

A number of other committers and Linux maintainers have provided some
input, but have not been very clear about their position.

Looking at the above, it occurs to me that, this whole area seems to be a
little inconsistent anyway and could do with a little house-keeping. We
have
- osstest.git
- there also is osstest/*.git which seems to be odd and seems to have been
inactive for a while (not very clear to me what these do)
- and we have old and inactive xentesttools/*.git
- and we are adding a new repo for XTF


Maybe, moving everything test related under testing/* or test/* would be
sensible, but that would cause some disruption (not sure how bad that
would be). It would address A, B and D. It wouldn't make C much worse than
now.

We already follow a similar pattern for out-of-tree via pvdrivers/*

It does in fact occur to me that some of the older inactive repos should
be archived somehow: candidates seem to be kemari/*, xentesttools/*,
xenclient/*, xcp/*, ... pollute the namespace. If we are concerned about
the namespace, we should address this at some point. There are also some
inactive top level git repos such as linux-2.5-xen.git and quite a few
inactive people repos.

>Lars, can you please advise what process we need to use to come to
>closure on this decision ?

Really we need to come to some sort of proposal. My gut feel is that maybe
this would most quickly resolved in a short IRC meeting, where we discuss
this issue and reduce it down to a concrete proposal (IRC log could then
be posted) and then put it to a vote, if need be. Otherwise this
discussion will drag on for a while as e-mail.

Lars

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel

^ permalink raw reply	[flat|nested] 43+ messages in thread

* Re: xenbits "official" repo for XTF (was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix)
  2016-07-06 11:19                   ` Lars Kurth
@ 2016-07-06 14:22                     ` Konrad Rzeszutek Wilk
  2016-07-07 11:10                       ` Lars Kurth
  0 siblings, 1 reply; 43+ messages in thread
From: Konrad Rzeszutek Wilk @ 2016-07-06 14:22 UTC (permalink / raw)
  To: Lars Kurth, boris.ostrovsky
  Cc: Wei Liu, Andrew Cooper, George Dunlap, David Vrabel, xen-devel,
	Ian Jackson, Anshul Makkar, Roger Pau Monne

> Looking at the above, it occurs to me that, this whole area seems to be a
> little inconsistent anyway and could do with a little house-keeping. We
> have
> - osstest.git
> - there also is osstest/*.git which seems to be odd and seems to have been
> inactive for a while (not very clear to me what these do)
> - and we have old and inactive xentesttools/*.git

They are not inactive. I have some patches for it and we are using
it (Boris and me).

> - and we are adding a new repo for XTF
> 
> 
> Maybe, moving everything test related under testing/* or test/* would be
> sensible, but that would cause some disruption (not sure how bad that
> would be). It would address A, B and D. It wouldn't make C much worse than
> now.

I don't mind the xentesttools going under a 'test' directory. 

> 
> We already follow a similar pattern for out-of-tree via pvdrivers/*
> 
> It does in fact occur to me that some of the older inactive repos should
> be archived somehow: candidates seem to be kemari/*, xentesttools/*,
> xenclient/*, xcp/*, ... pollute the namespace. If we are concerned about
> the namespace, we should address this at some point. There are also some
> inactive top level git repos such as linux-2.5-xen.git and quite a few
> inactive people repos.

The Linux ones, the linux-pvops.git could go.

> 
> >Lars, can you please advise what process we need to use to come to
> >closure on this decision ?
> 

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel

^ permalink raw reply	[flat|nested] 43+ messages in thread

* Re: xenbits "official" repo for XTF (was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix)
  2016-07-06 14:22                     ` Konrad Rzeszutek Wilk
@ 2016-07-07 11:10                       ` Lars Kurth
  2016-07-07 11:26                         ` Andrew Cooper
  0 siblings, 1 reply; 43+ messages in thread
From: Lars Kurth @ 2016-07-07 11:10 UTC (permalink / raw)
  To: Konrad Rzeszutek Wilk, Andrew Cooper, David Vrabel, Ian Jackson,
	George Dunlap
  Cc: Lars Kurth, Wei Liu, Anshul Makkar, xen-devel, Boris Ostrovsky,
	Roger Pau Monne


> On 6 Jul 2016, at 15:22, Konrad Rzeszutek Wilk <konrad.wilk@oracle.com> wrote:
> 
>> Looking at the above, it occurs to me that, this whole area seems to be a
>> little inconsistent anyway and could do with a little house-keeping. We
>> have
>> - osstest.git
>> - there also is osstest/*.git which seems to be odd and seems to have been
>> inactive for a while (not very clear to me what these do)
>> - and we have old and inactive xentesttools/*.git
> 
> They are not inactive. I have some patches for it and we are using
> it (Boris and me).

That is good to know.

>> - and we are adding a new repo for XTF
>> 
>> 
>> Maybe, moving everything test related under testing/* or test/* would be
>> sensible, but that would cause some disruption (not sure how bad that
>> would be). It would address A, B and D. It wouldn't make C much worse than
>> now.
> 
> I don't mind the xentesttools going under a 'test' directory. 

@Ian, @George: Any other views regarding say osstest and the other osstest/* directories? 

@Andrew: would something like test/xtf.git work

Regards
Lars
P.S.: Parking the question about old inactive repos for now
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel

^ permalink raw reply	[flat|nested] 43+ messages in thread

* Re: xenbits "official" repo for XTF (was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix)
  2016-07-07 11:10                       ` Lars Kurth
@ 2016-07-07 11:26                         ` Andrew Cooper
  2016-07-07 13:59                           ` Ian Jackson
  2016-07-07 15:22                           ` Konrad Rzeszutek Wilk
  0 siblings, 2 replies; 43+ messages in thread
From: Andrew Cooper @ 2016-07-07 11:26 UTC (permalink / raw)
  To: Lars Kurth, Konrad Rzeszutek Wilk, David Vrabel, Ian Jackson,
	George Dunlap
  Cc: Lars Kurth, Wei Liu, Anshul Makkar, xen-devel, Boris Ostrovsky,
	Roger Pau Monne

On 07/07/16 12:10, Lars Kurth wrote:
>> On 6 Jul 2016, at 15:22, Konrad Rzeszutek Wilk <konrad.wilk@oracle.com> wrote:
>>
>>> Looking at the above, it occurs to me that, this whole area seems to be a
>>> little inconsistent anyway and could do with a little house-keeping. We
>>> have
>>> - osstest.git
>>> - there also is osstest/*.git which seems to be odd and seems to have been
>>> inactive for a while (not very clear to me what these do)
>>> - and we have old and inactive xentesttools/*.git
>> They are not inactive. I have some patches for it and we are using
>> it (Boris and me).
> That is good to know.
>
>>> - and we are adding a new repo for XTF
>>>
>>>
>>> Maybe, moving everything test related under testing/* or test/* would be
>>> sensible, but that would cause some disruption (not sure how bad that
>>> would be). It would address A, B and D. It wouldn't make C much worse than
>>> now.
>> I don't mind the xentesttools going under a 'test' directory. 
> @Ian, @George: Any other views regarding say osstest and the other osstest/* directories? 
>
> @Andrew: would something like test/xtf.git work

It would, although given a straight choice I would prefer
xen-test-framework.git over its abbreviation.

~Andrew

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel

^ permalink raw reply	[flat|nested] 43+ messages in thread

* Re: xenbits "official" repo for XTF (was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix)
  2016-07-07 11:26                         ` Andrew Cooper
@ 2016-07-07 13:59                           ` Ian Jackson
  2016-07-07 14:02                             ` David Vrabel
  2016-07-07 14:11                             ` Andrew Cooper
  2016-07-07 15:22                           ` Konrad Rzeszutek Wilk
  1 sibling, 2 replies; 43+ messages in thread
From: Ian Jackson @ 2016-07-07 13:59 UTC (permalink / raw)
  To: Andrew Cooper
  Cc: Lars Kurth, Wei Liu, Lars Kurth, George Dunlap, David Vrabel,
	xen-devel, Boris Ostrovsky, Anshul Makkar, Roger Pau Monne

Andrew Cooper writes ("Re: [Xen-devel] xenbits "official" repo for XTF (was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix)"):
> On 07/07/16 12:10, Lars Kurth wrote:
> > @Andrew: would something like test/xtf.git work

I would live with that.

> It would, although given a straight choice I would prefer
> xen-test-framework.git over its abbreviation.

This conversation is in danger of going round in circles.

If we can't get consensus on something like this, the sensible thing
to do would be to vote.  Our governance docs don't really cope with
this kind of multi-answer question; they only do yes/no.

If we need to vote I suggest we use Condorcet.  That is:
 * Each committers gets to nominate up to 3 possible names for this
   proposed new repo.
 * Each committer ranks all the nominated possibilities in order.
 * We use Condorcet to establish whether there is one dominating
   nominee.  (If there is no single Condorcet winner, we remain
   deadlocked but at least we will have a small set of plausible
   possibilities to choose from.)

Condorcet is good because it necessarily demonstrates the underlying
legitimacy, according to the electorate, of the chosen option.  It
also copes well with large numbers of options.  (Tactical nomination
and tactical voting are ineffective.)

Ian.

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel

^ permalink raw reply	[flat|nested] 43+ messages in thread

* Re: xenbits "official" repo for XTF (was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix)
  2016-07-07 13:59                           ` Ian Jackson
@ 2016-07-07 14:02                             ` David Vrabel
  2016-07-07 14:58                               ` Ian Jackson
  2016-07-07 14:11                             ` Andrew Cooper
  1 sibling, 1 reply; 43+ messages in thread
From: David Vrabel @ 2016-07-07 14:02 UTC (permalink / raw)
  To: Ian Jackson, Andrew Cooper
  Cc: Lars Kurth, Wei Liu, Lars Kurth, George Dunlap, xen-devel,
	Boris Ostrovsky, Anshul Makkar, Roger Pau Monne

On 07/07/16 14:59, Ian Jackson wrote:
> Andrew Cooper writes ("Re: [Xen-devel] xenbits "official" repo for XTF (was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix)"):
>> On 07/07/16 12:10, Lars Kurth wrote:
>>> @Andrew: would something like test/xtf.git work
> 
> I would live with that.
> 
>> It would, although given a straight choice I would prefer
>> xen-test-framework.git over its abbreviation.
> 
> This conversation is in danger of going round in circles.
> 
> If we can't get consensus on something like this, the sensible thing
> to do would be to vote.  Our governance docs don't really cope with
> this kind of multi-answer question; they only do yes/no.
> 
> If we need to vote I suggest we use Condorcet.  That is:
>  * Each committers gets to nominate up to 3 possible names for this
>    proposed new repo.
>  * Each committer ranks all the nominated possibilities in order.
>  * We use Condorcet to establish whether there is one dominating
>    nominee.  (If there is no single Condorcet winner, we remain
>    deadlocked but at least we will have a small set of plausible
>    possibilities to choose from.)
> 
> Condorcet is good because it necessarily demonstrates the underlying
> legitimacy, according to the electorate, of the chosen option.  It
> also copes well with large numbers of options.  (Tactical nomination
> and tactical voting are ineffective.)

Really?  Is it that difficult to accept that the original project author
gets to choose the name?

My recommendation to Andrew would be to host the project somewhere else
-- somewhere without this absurd amount of bureaucracy.  github is
popular these days.

David

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel

^ permalink raw reply	[flat|nested] 43+ messages in thread

* Re: xenbits "official" repo for XTF (was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix)
  2016-07-07 13:59                           ` Ian Jackson
  2016-07-07 14:02                             ` David Vrabel
@ 2016-07-07 14:11                             ` Andrew Cooper
  2016-07-07 17:17                               ` Lars Kurth
  1 sibling, 1 reply; 43+ messages in thread
From: Andrew Cooper @ 2016-07-07 14:11 UTC (permalink / raw)
  To: Ian Jackson
  Cc: Lars Kurth, Wei Liu, Lars Kurth, George Dunlap, David Vrabel,
	xen-devel, Boris Ostrovsky, Anshul Makkar, Roger Pau Monne

On 07/07/16 14:59, Ian Jackson wrote:
> Andrew Cooper writes ("Re: [Xen-devel] xenbits "official" repo for XTF (was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix)"):
>> On 07/07/16 12:10, Lars Kurth wrote:
>>> @Andrew: would something like test/xtf.git work
> I would live with that.
>
>> It would, although given a straight choice I would prefer
>> xen-test-framework.git over its abbreviation.
> This conversation is in danger of going round in circles.

Right - let me draw a line in the sand.

The current name is xen-test-framework, and that has been around and
used for several years now.  XTF is also used frequently as an abbreviation.

A concern has been raised about the correctness of this name, but no
alternatives have been suggested which come close to being a plausible
replacement.  (All suggestions are less accurate or descriptive than the
current name).

As the author, I chose the current name, and chose it as the most
appropriate name I could think of.  If others feel strongly that it
isn't, then the onus is on them to choose a specific alternative and
argue as to why it is more appropriate.  Nothing along this line has
occurred.

~Andrew

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel

^ permalink raw reply	[flat|nested] 43+ messages in thread

* Re: xenbits "official" repo for XTF (was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix)
  2016-07-07 14:02                             ` David Vrabel
@ 2016-07-07 14:58                               ` Ian Jackson
  2016-07-07 15:00                                 ` Ian Jackson
  0 siblings, 1 reply; 43+ messages in thread
From: Ian Jackson @ 2016-07-07 14:58 UTC (permalink / raw)
  To: David Vrabel
  Cc: Lars Kurth, Wei Liu, Lars Kurth, Andrew Cooper, George Dunlap,
	Ian Jackson, xen-devel, Boris Ostrovsky, Anshul Makkar,
	Roger Pau Monne

David Vrabel writes ("Re: [Xen-devel] xenbits "official" repo for XTF (was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix)"):
> Really?  Is it that difficult to accept that the original project author
> gets to choose the name?

Yes.

https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=614907#113

As I say, I think osstest has a far better claim to the name `Xen Test
Framework'.  But I haven't called it that because I have an
understanding that namespace landgrabs are antisocial.

Ian.

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel

^ permalink raw reply	[flat|nested] 43+ messages in thread

* Re: xenbits "official" repo for XTF (was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix)
  2016-07-07 14:58                               ` Ian Jackson
@ 2016-07-07 15:00                                 ` Ian Jackson
  0 siblings, 0 replies; 43+ messages in thread
From: Ian Jackson @ 2016-07-07 15:00 UTC (permalink / raw)
  To: David Vrabel, Andrew Cooper, Lars Kurth, Konrad Rzeszutek Wilk,
	George Dunlap, Lars Kurth, Boris Ostrovsky, Wei Liu, xen-devel,
	Anshul Makkar, Roger Pau Monne

Ian Jackson writes ("Re: [Xen-devel] xenbits "official" repo for XTF (was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix)"):
> David Vrabel writes ("Re: [Xen-devel] xenbits "official" repo for XTF (was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix)"):
> > Really?  Is it that difficult to accept that the original project author
> > gets to choose the name?
> 
> Yes.
> 
> https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=614907#113

Sorry, I should have given the authoritative link:

https://lists.debian.org/debian-devel-announce/2012/07/msg00002.html

> As I say, I think osstest has a far better claim to the name `Xen Test
> Framework'.  But I haven't called it that because I have an
> understanding that namespace landgrabs are antisocial.

Ian.

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel

^ permalink raw reply	[flat|nested] 43+ messages in thread

* Re: xenbits "official" repo for XTF (was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix)
  2016-07-07 11:26                         ` Andrew Cooper
  2016-07-07 13:59                           ` Ian Jackson
@ 2016-07-07 15:22                           ` Konrad Rzeszutek Wilk
  1 sibling, 0 replies; 43+ messages in thread
From: Konrad Rzeszutek Wilk @ 2016-07-07 15:22 UTC (permalink / raw)
  To: Andrew Cooper
  Cc: Lars Kurth, Wei Liu, Lars Kurth, George Dunlap, David Vrabel,
	xen-devel, Ian Jackson, Boris Ostrovsky, Anshul Makkar,
	Roger Pau Monne

On Thu, Jul 07, 2016 at 12:26:09PM +0100, Andrew Cooper wrote:
> On 07/07/16 12:10, Lars Kurth wrote:
> >> On 6 Jul 2016, at 15:22, Konrad Rzeszutek Wilk <konrad.wilk@oracle.com> wrote:
> >>
> >>> Looking at the above, it occurs to me that, this whole area seems to be a
> >>> little inconsistent anyway and could do with a little house-keeping. We
> >>> have
> >>> - osstest.git
> >>> - there also is osstest/*.git which seems to be odd and seems to have been
> >>> inactive for a while (not very clear to me what these do)
> >>> - and we have old and inactive xentesttools/*.git
> >> They are not inactive. I have some patches for it and we are using
> >> it (Boris and me).
> > That is good to know.
> >
> >>> - and we are adding a new repo for XTF
> >>>
> >>>
> >>> Maybe, moving everything test related under testing/* or test/* would be
> >>> sensible, but that would cause some disruption (not sure how bad that
> >>> would be). It would address A, B and D. It wouldn't make C much worse than
> >>> now.
> >> I don't mind the xentesttools going under a 'test' directory. 
> > @Ian, @George: Any other views regarding say osstest and the other osstest/* directories? 
> >
> > @Andrew: would something like test/xtf.git work
> 
> It would, although given a straight choice I would prefer
> xen-test-framework.git over its abbreviation.

+1

> 
> ~Andrew

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel

^ permalink raw reply	[flat|nested] 43+ messages in thread

* Re: xenbits "official" repo for XTF (was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix)
  2016-07-07 14:11                             ` Andrew Cooper
@ 2016-07-07 17:17                               ` Lars Kurth
  2016-07-08 13:06                                 ` Andrew Cooper
  2016-07-15 10:21                                 ` Ian Jackson
  0 siblings, 2 replies; 43+ messages in thread
From: Lars Kurth @ 2016-07-07 17:17 UTC (permalink / raw)
  To: Andrew Cooper, Ian Jackson
  Cc: Wei Liu, Lars Kurth, George Dunlap, David Vrabel, xen-devel,
	Boris Ostrovsky, Anshul Makkar, Roger Pau Monne

Alright,

it appears we are at an impasse here. Not hosting the code on xenbits as
suggested by David, seems to be the worst solution and will benefit
no-one. 

> If we can't get consensus on something like this, the sensible thing
> to do would be to vote. Our governance docs don't really cope with
> this kind of multi-answer question; they only do yes/no.

I am not convinced that we need a formal process in this case. Our
governance has the mechanism to referee, when there is disagreement. For
code changes the referee would be the maintainer/committer which owns a
piece of code and the mechanism would work by withholding an ACK. For
unowned changes the referee would be the project lead: but we have none
and in fact we want none.

The next level up is the Advisory Board: but I really don't want to go to
the AB with a bike-shed issue like this.


In particular as WE DO ACTUALLY HAVE CONSENSUS for a compromise by the two
main people disagreeing.

> On 20/06/16 18:03, Ian Jackson wrote:

> I could live with "xtf", although I think it's rather too short.



> On 07/07/2016 12:26, "Andrew Cooper" <andrew.cooper3@citrix.com> wrote:
>> On 07/07/16 12:10, Lars Kurth wrote:
>> @Andrew: would something like test/xtf.git work

> It would, although given a straight choice I would prefer
> xen-test-framework.git over its abbreviation.


So let's just go with "./xtf.git" and make use of the "Description" field
in http://xenbits.xen.org/gitweb/ to add a bit more verbosity. Adding
something such as "Xen Test Framework and Suite for creating
microkernel-based tests". This is accurate and searchable.

It is no worse than "raisin.git", "osstest.git", and other top-level repos.

Maybe we can make improve the description for "./osstest.git": something
along the lines of "Xen Test Framework and Suite, used for Open Source Xen
Continuous Integration that also acts as push gate" or something like it.
That would be more accurate than what we have now.

Compromise
A.1) Create "xtf.git" and use "Xen Test Framework and Suite for creating
microkernel-based tests" in Description field

A.2) Update description for osstest.git to "Xen Test Framework and Suite,
used for Open Source Xen Continuous Integration that also acts as push
gate"

> Out of curiosity, I searched for it on google, and found my written
> documentation as the top hit.
>
> 
>https://www.google.co.uk/search?q=xen+test+framework&ie=utf-8&oe=utf-8&gws
>_rd=cr&ei=InxeV52HDYHc-QG0-qN4

That has nothing to do with the repo name. The reason why google finds
this page is because
http://xenbits.xen.org/people/andrewcoop/xen-test-framework/ exists, but
no equivalent page exists for OSSTEST.


Improvements to web searchability for "xen test framework" to ensure that
searches for both frameworks lead somewhere sensible
B.1) http://xenbits.xen.org/people/andrewcoop/xen-test-framework/ should
be move under docs and re-named to "XTF: Xen Test Framework and Suite for
creating microkernel-based tests"

B.2) Add a similar page under docs for OSSTEST with a similarly verbose
title, e.g. "OSSTEST: Xen Test Framework and Suite for Open Source Xen
Continuous Integration"

That should address everyones concern, as far as I can tell from the the
e-mail thread. If anyone disagrees, please shout within the next few days.

Best Regards
Lars
P.S.: I moved fixing some of our governance issues towards the top of my
TODO list


On 07/07/2016 15:11, "Andrew Cooper" <andrew.cooper3@citrix.com> wrote:

>On 07/07/16 14:59, Ian Jackson wrote:
>> Andrew Cooper writes ("Re: [Xen-devel] xenbits "official" repo for XTF
>>(was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix)"):
>>> On 07/07/16 12:10, Lars Kurth wrote:
>>>> @Andrew: would something like test/xtf.git work
>> I would live with that.
>>
>>> It would, although given a straight choice I would prefer
>>> xen-test-framework.git over its abbreviation.
>> This conversation is in danger of going round in circles.
>
>Right - let me draw a line in the sand.
>
>The current name is xen-test-framework, and that has been around and
>used for several years now.  XTF is also used frequently as an
>abbreviation.
>
>A concern has been raised about the correctness of this name, but no
>alternatives have been suggested which come close to being a plausible
>replacement.  (All suggestions are less accurate or descriptive than the
>current name).
>
>As the author, I chose the current name, and chose it as the most
>appropriate name I could think of.  If others feel strongly that it
>isn't, then the onus is on them to choose a specific alternative and
>argue as to why it is more appropriate.  Nothing along this line has
>occurred.
>
>~Andrew

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel

^ permalink raw reply	[flat|nested] 43+ messages in thread

* Re: xenbits "official" repo for XTF (was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix)
  2016-07-07 17:17                               ` Lars Kurth
@ 2016-07-08 13:06                                 ` Andrew Cooper
  2016-07-08 13:42                                   ` Lars Kurth
  2016-07-12 15:13                                   ` George Dunlap
  2016-07-15 10:21                                 ` Ian Jackson
  1 sibling, 2 replies; 43+ messages in thread
From: Andrew Cooper @ 2016-07-08 13:06 UTC (permalink / raw)
  To: Lars Kurth, Ian Jackson
  Cc: Wei Liu, Lars Kurth, George Dunlap, David Vrabel, xen-devel,
	Boris Ostrovsky, Anshul Makkar, Roger Pau Monne

On 07/07/16 18:17, Lars Kurth wrote:
> Alright,
>
> it appears we are at an impasse here. Not hosting the code on xenbits as
> suggested by David, seems to be the worst solution and will benefit
> no-one. 
>
>> If we can't get consensus on something like this, the sensible thing
>> to do would be to vote. Our governance docs don't really cope with
>> this kind of multi-answer question; they only do yes/no.
> I am not convinced that we need a formal process in this case. Our
> governance has the mechanism to referee, when there is disagreement. For
> code changes the referee would be the maintainer/committer which owns a
> piece of code and the mechanism would work by withholding an ACK. For
> unowned changes the referee would be the project lead: but we have none
> and in fact we want none.
>
> The next level up is the Advisory Board: but I really don't want to go to
> the AB with a bike-shed issue like this.
>
>
> In particular as WE DO ACTUALLY HAVE CONSENSUS for a compromise by the two
> main people disagreeing.
>
>> On 20/06/16 18:03, Ian Jackson wrote:
>> I could live with "xtf", although I think it's rather too short.
>
>
>> On 07/07/2016 12:26, "Andrew Cooper" <andrew.cooper3@citrix.com> wrote:
>>> On 07/07/16 12:10, Lars Kurth wrote:
>>> @Andrew: would something like test/xtf.git work
>> It would, although given a straight choice I would prefer
>> xen-test-framework.git over its abbreviation.
>
> So let's just go with "./xtf.git" and make use of the "Description" field
> in http://xenbits.xen.org/gitweb/ to add a bit more verbosity. Adding
> something such as "Xen Test Framework and Suite for creating
> microkernel-based tests". This is accurate and searchable.
>
> It is no worse than "raisin.git", "osstest.git", and other top-level repos.
>
> Maybe we can make improve the description for "./osstest.git": something
> along the lines of "Xen Test Framework and Suite, used for Open Source Xen
> Continuous Integration that also acts as push gate" or something like it.
> That would be more accurate than what we have now.
>
> Compromise
> A.1) Create "xtf.git" and use "Xen Test Framework and Suite for creating
> microkernel-based tests" in Description field
>
> A.2) Update description for osstest.git to "Xen Test Framework and Suite,
> used for Open Source Xen Continuous Integration that also acts as push
> gate"
>
>> Out of curiosity, I searched for it on google, and found my written
>> documentation as the top hit.
>>
>>
>> https://www.google.co.uk/search?q=xen+test+framework&ie=utf-8&oe=utf-8&gws
>> _rd=cr&ei=InxeV52HDYHc-QG0-qN4
> That has nothing to do with the repo name. The reason why google finds
> this page is because
> http://xenbits.xen.org/people/andrewcoop/xen-test-framework/ exists, but
> no equivalent page exists for OSSTEST.
>
>
> Improvements to web searchability for "xen test framework" to ensure that
> searches for both frameworks lead somewhere sensible
> B.1) http://xenbits.xen.org/people/andrewcoop/xen-test-framework/ should
> be move under docs and re-named to "XTF: Xen Test Framework and Suite for
> creating microkernel-based tests"
>
> B.2) Add a similar page under docs for OSSTEST with a similarly verbose
> title, e.g. "OSSTEST: Xen Test Framework and Suite for Open Source Xen
> Continuous Integration"
>
> That should address everyones concern, as far as I can tell from the the
> e-mail thread. If anyone disagrees, please shout within the next few days.
>
> Best Regards
> Lars
> P.S.: I moved fixing some of our governance issues towards the top of my
> TODO list

I have no problem with Ian's earlier suggestion:

> "CI (continuous integration)" is the keyword that many people will
> have for osstest.
>
> I would suggest
>
>   (This is not the Xen Project's CI / Continuous Integration /
>    automated push gate system.  For that, see
>    <a href="wiki">osstest</a>.)
>
> or something.

Adding something like that to the XTF documentation is perfectly fine. 
I also have no problem with the other xtf changes in descriptions/etc
suggested.


However, OSSTest has always been known as OSSTest (including all
references in the automated emails), and not as a xen test framework. 
Taking any steps to make OSSTest retroactively searchable as a xen test
framework is a dumb move, which will only confuse users.

I fully admit that had OSSTest been named differently then I might not
have chosen XTF as a name, but that didn't happen.  Trying to rewrite
history isn't the answer.

~Andrew

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel

^ permalink raw reply	[flat|nested] 43+ messages in thread

* Re: xenbits "official" repo for XTF (was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix)
  2016-07-08 13:06                                 ` Andrew Cooper
@ 2016-07-08 13:42                                   ` Lars Kurth
  2016-07-08 13:48                                     ` Andrew Cooper
  2016-07-12 15:13                                   ` George Dunlap
  1 sibling, 1 reply; 43+ messages in thread
From: Lars Kurth @ 2016-07-08 13:42 UTC (permalink / raw)
  To: Andrew Cooper, Ian Jackson
  Cc: Wei Liu, Lars Kurth, George Dunlap, David Vrabel, xen-devel,
	Boris Ostrovsky, Anshul Makkar, Roger Pau Monne



On 08/07/2016 14:06, "Andrew Cooper" <andrew.cooper3@citrix.com> wrote:

>On 07/07/16 18:17, Lars Kurth wrote:
>> Alright,
>>
>> it appears we are at an impasse here. Not hosting the code on xenbits as
>> suggested by David, seems to be the worst solution and will benefit
>> no-one. 
>>
>>...
>> That should address everyones concern, as far as I can tell from the the
>> e-mail thread. If anyone disagrees, please shout within the next few
>>days.
>>
>> Best Regards
>> Lars
>> P.S.: I moved fixing some of our governance issues towards the top of my
>> TODO list
>
>I have no problem with Ian's earlier suggestion:

Alright. Ian confirmed on IRC that he has no problem either. Which means
the main issue is unblocked and the repo can be created.

>
>> "CI (continuous integration)" is the keyword that many people will
>> have for osstest.
>>
>> I would suggest
>>
>>   (This is not the Xen Project's CI / Continuous Integration /
>>    automated push gate system.  For that, see
>>    <a href="wiki">osstest</a>.)
>>
>> or something.
>
>Adding something like that to the XTF documentation is perfectly fine.
>I also have no problem with the other xtf changes in descriptions/etc
>suggested.

That makes perfect sense (but also see below).

>However, OSSTest has always been known as OSSTest (including all
>references in the automated emails), and not as a xen test framework.
>Taking any steps to make OSSTest retroactively searchable as a xen test
>framework is a dumb move, which will only confuse users.
>
>I fully admit that had OSSTest been named differently then I might not
>have chosen XTF as a name, but that didn't happen.  Trying to rewrite
>history isn't the answer.

I am not trying to re-write history. Ultimately I don't mind what the
exact description of OSSTEST says, as long as it is accurate, which is why
I said "something along the lines of ...". Right now there is no wiki or
other xenbits documentation about OSSTEST which is outside the git tree
(except for a few blog posts). Which is mainly why I raised this.

Of course you are right that OSSTest is established and people who have
been around for a while know that. The key question is what newcomers who
don't know would search for: and to be honest I don't know. But it seems
reasonable to me that the "OSSTEST" page should come up when googling for
a combination/sub-set of the following keywords: "test", "xen",
"Continuous Integration", "CI", "suite", "push-gate" and maybe
"framework".  

Lars
 

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel

^ permalink raw reply	[flat|nested] 43+ messages in thread

* Re: xenbits "official" repo for XTF (was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix)
  2016-07-08 13:42                                   ` Lars Kurth
@ 2016-07-08 13:48                                     ` Andrew Cooper
  0 siblings, 0 replies; 43+ messages in thread
From: Andrew Cooper @ 2016-07-08 13:48 UTC (permalink / raw)
  To: Lars Kurth, Ian Jackson
  Cc: Wei Liu, Lars Kurth, George Dunlap, David Vrabel, xen-devel,
	Boris Ostrovsky, Anshul Makkar, Roger Pau Monne

On 08/07/16 14:42, Lars Kurth wrote:
>
> On 08/07/2016 14:06, "Andrew Cooper" <andrew.cooper3@citrix.com> wrote:
>
>> On 07/07/16 18:17, Lars Kurth wrote:
>>> Alright,
>>>
>>> it appears we are at an impasse here. Not hosting the code on xenbits as
>>> suggested by David, seems to be the worst solution and will benefit
>>> no-one. 
>>>
>>> ...
>>> That should address everyones concern, as far as I can tell from the the
>>> e-mail thread. If anyone disagrees, please shout within the next few
>>> days.
>>>
>>> Best Regards
>>> Lars
>>> P.S.: I moved fixing some of our governance issues towards the top of my
>>> TODO list
>> I have no problem with Ian's earlier suggestion:
> Alright. Ian confirmed on IRC that he has no problem either. Which means
> the main issue is unblocked and the repo can be created.
>
>>> "CI (continuous integration)" is the keyword that many people will
>>> have for osstest.
>>>
>>> I would suggest
>>>
>>>   (This is not the Xen Project's CI / Continuous Integration /
>>>    automated push gate system.  For that, see
>>>    <a href="wiki">osstest</a>.)
>>>
>>> or something.
>> Adding something like that to the XTF documentation is perfectly fine.
>> I also have no problem with the other xtf changes in descriptions/etc
>> suggested.
> That makes perfect sense (but also see below).
>
>> However, OSSTest has always been known as OSSTest (including all
>> references in the automated emails), and not as a xen test framework.
>> Taking any steps to make OSSTest retroactively searchable as a xen test
>> framework is a dumb move, which will only confuse users.
>>
>> I fully admit that had OSSTest been named differently then I might not
>> have chosen XTF as a name, but that didn't happen.  Trying to rewrite
>> history isn't the answer.
> I am not trying to re-write history. Ultimately I don't mind what the
> exact description of OSSTEST says, as long as it is accurate, which is why
> I said "something along the lines of ...". Right now there is no wiki or
> other xenbits documentation about OSSTEST which is outside the git tree
> (except for a few blog posts). Which is mainly why I raised this.
>
> Of course you are right that OSSTest is established and people who have
> been around for a while know that. The key question is what newcomers who
> don't know would search for: and to be honest I don't know. But it seems
> reasonable to me that the "OSSTEST" page should come up when googling for
> a combination/sub-set of the following keywords: "test", "xen",
> "Continuous Integration", "CI", "suite", "push-gate" and maybe
> "framework".

I think the suggested changes to XTF documentation/description should be
able to cover these cases.

I don't expect anyone to be blindly trying to find things like this; I
would expect that searches would be based on information from the
xen-devel list to start with.

~Andrew

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel

^ permalink raw reply	[flat|nested] 43+ messages in thread

* Re: xenbits "official" repo for XTF (was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix)
  2016-07-08 13:06                                 ` Andrew Cooper
  2016-07-08 13:42                                   ` Lars Kurth
@ 2016-07-12 15:13                                   ` George Dunlap
  1 sibling, 0 replies; 43+ messages in thread
From: George Dunlap @ 2016-07-12 15:13 UTC (permalink / raw)
  To: Andrew Cooper
  Cc: Lars Kurth, Wei Liu, Lars Kurth, Anshul Makkar, David Vrabel,
	xen-devel, Ian Jackson, Boris Ostrovsky, Roger Pau Monne

On Fri, Jul 8, 2016 at 2:06 PM, Andrew Cooper <andrew.cooper3@citrix.com> wrote:
> However, OSSTest has always been known as OSSTest (including all
> references in the automated emails), and not as a xen test framework.
> Taking any steps to make OSSTest retroactively searchable as a xen test
> framework is a dumb move, which will only confuse users.
>
> I fully admit that had OSSTest been named differently then I might not
> have chosen XTF as a name, but that didn't happen.  Trying to rewrite
> history isn't the answer.

I'm glad we have a solution, but it seems that we still don't all
understand one another; in which case I'm afraid something like this
may come up again.

The objection has nothing to do with OSSTest.  It has to do with the
fact that 1) contains the word "Xen", and 2) seems to describe
something much more general than what it actually is.

If you'd called it "Andy's Test Framework", or even "Cooper's
Universal Testing Solution", I don't think anyone would have had any
strong objections, because although the names very general, they're
made much more specific by adding your name to the front.

By adding "Xen" to it, you associate it with the community, and thus
the community has a right to say whether they accept the name or not.
The fact that you used the name for 2 years *privately* doesn't give
you any right to suddenly come out and use the name publicly.  You can
use whatever name you want privately, of course, but once you start to
use it publicly, the community has a right to say whether they accept
that or not.

 -George

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel

^ permalink raw reply	[flat|nested] 43+ messages in thread

* Re: xenbits "official" repo for XTF (was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix)
  2016-07-07 17:17                               ` Lars Kurth
  2016-07-08 13:06                                 ` Andrew Cooper
@ 2016-07-15 10:21                                 ` Ian Jackson
  2016-07-15 10:49                                   ` Lars Kurth
  2016-08-08 17:04                                   ` Ian Jackson
  1 sibling, 2 replies; 43+ messages in thread
From: Ian Jackson @ 2016-07-15 10:21 UTC (permalink / raw)
  To: Lars Kurth
  Cc: Wei Liu, Lars Kurth, Andrew Cooper, George Dunlap, David Vrabel,
	xen-devel, Boris Ostrovsky, Anshul Makkar, Roger Pau Monne

Lars Kurth writes ("Re: [Xen-devel] xenbits "official" repo for XTF (was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix)"):
> Alright, [stuff]

With my xenbits admin hat on I therefore intend to:

 * Create xtf.git in the xenbits toplevel
   and give it the appropriate permissions

 * Create xtf.git/description containing

     Xen Test Framework and Suite for creating microkernel-based tests

 * Edit osstest.git/description to contain

     Xen Test Framework and Suite, used for Open Source Xen CI push gate

 * Create a directory /docs/osstest (in /var/xenbits-www/html/)
   writeable by me.

 * Move /docs/osstest-output-example to /docs/osstest/output-example
   and leave a symlink behind to keep old hyperlinks working.

With my osstest maintainer hat on I will:

 * Populate the latter with an HTML page whose <title> and <h1> are

       osstest: Xen Test Framework and Test Suite for Open Source Xen
        Continuous Integration

   and which contains a brief description and links to related
   pages/resources.

> Improvements to web searchability for "xen test framework" to ensure that
> searches for both frameworks lead somewhere sensible
> B.1) http://xenbits.xen.org/people/andrewcoop/xen-test-framework/ should
> be move under docs and re-named to "XTF: Xen Test Framework and Suite for
> creating microkernel-based tests"

Lars, you haven't specified the pathname for this.  I assume that it
would be `xtf', to correspond to the toplevel fit repo.

Andrew, do you want to edit the title in
/people/andrewcoop/xen-test-framework/ yourself and then have me use
my admin powers to move the directory into /docs/xtf/ ?  I could leave
behind a symlink to keep old hyperlinks working.

Thanks,
Ian.

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel

^ permalink raw reply	[flat|nested] 43+ messages in thread

* Re: xenbits "official" repo for XTF (was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix)
  2016-07-15 10:21                                 ` Ian Jackson
@ 2016-07-15 10:49                                   ` Lars Kurth
  2016-07-18 14:36                                     ` Ian Jackson
  2016-08-08 17:04                                   ` Ian Jackson
  1 sibling, 1 reply; 43+ messages in thread
From: Lars Kurth @ 2016-07-15 10:49 UTC (permalink / raw)
  To: Ian Jackson
  Cc: Wei Liu, Lars Kurth, Andrew Cooper, George Dunlap, David Vrabel,
	xen-devel, Boris Ostrovsky, Anshul Makkar, Roger Pau Monne



On 15/07/2016 11:21, "Ian Jackson" <ian.jackson@eu.citrix.com> wrote:

>Lars Kurth writes ("Re: [Xen-devel] xenbits "official" repo for XTF (was
>Re: [PATCH 0/2] xtf: add launcher (+1 bugfix)"):
>> Alright, [stuff]
>
>With my xenbits admin hat on I therefore intend to:
>
> * Create xtf.git in the xenbits toplevel
>   and give it the appropriate permissions
>
> * Create xtf.git/description containing
>
>     Xen Test Framework and Suite for creating microkernel-based tests
>
> * Edit osstest.git/description to contain
>
>     Xen Test Framework and Suite, used for Open Source Xen CI push gate
>
> * Create a directory /docs/osstest (in /var/xenbits-www/html/)
>   writeable by me.
>
> * Move /docs/osstest-output-example to /docs/osstest/output-example
>   and leave a symlink behind to keep old hyperlinks working.
>
>With my osstest maintainer hat on I will:
>
> * Populate the latter with an HTML page whose <title> and <h1> are
>
>       osstest: Xen Test Framework and Test Suite for Open Source Xen
>        Continuous Integration
>
>   and which contains a brief description and links to related
>   pages/resources.
>
>> Improvements to web searchability for "xen test framework" to ensure
>>that
>> searches for both frameworks lead somewhere sensible
>> B.1) http://xenbits.xen.org/people/andrewcoop/xen-test-framework/ should
>> be move under docs and re-named to "XTF: Xen Test Framework and Suite
>>for
>> creating microkernel-based tests"
>
>Lars, you haven't specified the pathname for this.  I assume that it
>would be `xtf', to correspond to the toplevel fit repo.

I don't have a strong opinion. It doesn't really fit into any of the other
categories on http://xenbits.xen.org/docs, maybe with the exception of
Misc. 
Or we could create a new directory under docs that contains testing,
build, related stuff

Lars


_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel

^ permalink raw reply	[flat|nested] 43+ messages in thread

* Re: xenbits "official" repo for XTF (was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix)
  2016-07-15 10:49                                   ` Lars Kurth
@ 2016-07-18 14:36                                     ` Ian Jackson
  2016-07-18 15:19                                       ` Lars Kurth
  0 siblings, 1 reply; 43+ messages in thread
From: Ian Jackson @ 2016-07-18 14:36 UTC (permalink / raw)
  To: Lars Kurth
  Cc: Wei Liu, Lars Kurth, Andrew Cooper, George Dunlap, David Vrabel,
	xen-devel, Boris Ostrovsky, Anshul Makkar, Roger Pau Monne

Lars Kurth writes ("Re: [Xen-devel] xenbits "official" repo for XTF (was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix)"):
> On 15/07/2016 11:21, "Ian Jackson" <ian.jackson@eu.citrix.com> wrote:
> >Lars, you haven't specified the pathname for this.  I assume that it
> >would be `xtf', to correspond to the toplevel fit repo.
> 
> I don't have a strong opinion. It doesn't really fit into any of the other
> categories on http://xenbits.xen.org/docs, maybe with the exception of
> Misc. 
> Or we could create a new directory under docs that contains testing,
> build, related stuff

Can you please clarify whether you meant:

 1. "Please go ahead with the git repo changes but I object to
    your suggested docs directories"

 2. "Please go ahead with the git repo changes and the change to the
    osstest docs but I object to your suggested name for the docs
    directory for XTF."

In either of which cases I'm blocked, and that's surely a -1 and you
should specify what you would like done instead, or:

 3. "Please go ahead with the git repo changes.  I do not object
    to your proposals for the docs changes and you can go ahead
    with those unless someone else objects.

    "Furthermore, here are some observations for the future or for
    other people to consider."

In which case we are unblocked, because no-one else has objected.

Thanks,
Ian.

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel

^ permalink raw reply	[flat|nested] 43+ messages in thread

* Re: xenbits "official" repo for XTF (was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix)
  2016-07-18 14:36                                     ` Ian Jackson
@ 2016-07-18 15:19                                       ` Lars Kurth
  2016-07-18 15:30                                         ` Ian Jackson
  2016-07-18 15:31                                         ` Ian Jackson
  0 siblings, 2 replies; 43+ messages in thread
From: Lars Kurth @ 2016-07-18 15:19 UTC (permalink / raw)
  To: Ian Jackson
  Cc: Wei Liu, Lars Kurth, Andrew Cooper, George Dunlap, David Vrabel,
	xen-devel, Boris Ostrovsky, Anshul Makkar, Roger Pau Monne



On 18/07/2016 15:36, "Ian Jackson" <ian.jackson@eu.citrix.com> wrote:

>Lars Kurth writes ("Re: [Xen-devel] xenbits "official" repo for XTF (was
>Re: [PATCH 0/2] xtf: add launcher (+1 bugfix)"):
>> On 15/07/2016 11:21, "Ian Jackson" <ian.jackson@eu.citrix.com> wrote:
>> >Lars, you haven't specified the pathname for this.  I assume that it
>> >would be `xtf', to correspond to the toplevel fit repo.
>> 
>> I don't have a strong opinion. It doesn't really fit into any of the
>>other
>> categories on http://xenbits.xen.org/docs, maybe with the exception of
>> Misc. 
>> Or we could create a new directory under docs that contains testing,
>> build, related stuff
>
>Can you please clarify whether you meant:
>
> 1. "Please go ahead with the git repo changes but I object to
>    your suggested docs directories"


> 2. "Please go ahead with the git repo changes and the change to the
>    osstest docs but I object to your suggested name for the docs
>    directory for XTF."

Definitely go ahead with the git repo.

I don't object to a new xtf directory in the docs root per se. I mainly
object to how the documents are presented in
http://xenbits.xen.org/docs/unstable

As far as I understand, the docs directory structure maps onto headline
categories in http://xenbits.xen.org/docs/unstable. Is this correct? If
not, feel free to go ahead and create the directories, as presumably we
can easily change headline categories that appear in
http://xenbits.xen.org/docs/unstable later, without having to move files
around. 
 

I would only object to creating a new docs/xtf directory,
- iff directory structure is exposed in the web index in
http://xenbits.xen.org/docs/unstable are linked, and
- iff there is only one document in this directory now, and
- iff this will likely be the case in future.
And I would have the same concern for osstest. If there are several
documents already in each, then go ahead and create them.

If there is a link between directory structure and web index, my preferred
alternative to creating the two separate directories would be to group all
the test related documents into a docs/testing directory.

It would be helpful to users and newcomers if all test related documents
could be found under a Testing headline in
http://xenbits.xen.org/docs/unstable/

Regards
Lars

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel

^ permalink raw reply	[flat|nested] 43+ messages in thread

* Re: xenbits "official" repo for XTF (was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix)
  2016-07-18 15:19                                       ` Lars Kurth
@ 2016-07-18 15:30                                         ` Ian Jackson
  2016-07-18 15:36                                           ` Lars Kurth
  2016-07-18 15:31                                         ` Ian Jackson
  1 sibling, 1 reply; 43+ messages in thread
From: Ian Jackson @ 2016-07-18 15:30 UTC (permalink / raw)
  To: Lars Kurth
  Cc: Wei Liu, Lars Kurth, Andrew Cooper, George Dunlap, David Vrabel,
	xen-devel, Boris Ostrovsky, Anshul Makkar, Roger Pau Monne

Lars Kurth writes ("Re: [Xen-devel] xenbits "official" repo for XTF (was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix)"):
> I don't object to a new xtf directory in the docs root per se. I mainly
> object to how the documents are presented in
> http://xenbits.xen.org/docs/unstable

The documents under there are generated from xen.git#master.

> As far as I understand, the docs directory structure maps onto headline
> categories in http://xenbits.xen.org/docs/unstable. Is this correct?

What ?  No.

docs/ is just the HTTP namespace section for documents hosted on
xenbits.  Its subdirectories currently include:

  - one directory for each existing stable and unstable Xen branch,
    generated from the in-tree documentation source.

  - some example osstest output, referred to from the osstest.git
    in-tree README

  - a subdirectory `xcp' which is presumably now obsolete and retained
    for historical reasons.

It appears that it doesn't have an index.html, so we get a webserver
directory listing.

> I would only object to creating a new docs/xtf directory,
> - iff directory structure is exposed in the web index in
> http://xenbits.xen.org/docs/unstable are linked, and

I don't understand what you are asking.

> - iff there is only one document in this directory now, and
> - iff this will likely be the case in future.

I'm proposing that the initial contents of the xtf/ directory would be
the initial contents of the /people/andrewcoop/xen-test-framework/
directory, only with the titles changed as discussed.

So there would be one `document', which is really an HTML page, called
index.html, so it would replace the webserver directory listing.

> And I would have the same concern for osstest. If there are several
> documents already in each, then go ahead and create them.

Is your concern that it is a bad idea to create a directory rather
than a file ?

I think this is wrongheaded.

It is better to create a directory now, with an index.html.  Otherwise
it will become difficult to turn it into a directory later.

> If there is a link between directory structure and web index, my preferred
> alternative to creating the two separate directories would be to group all
> the test related documents into a docs/testing directory.

I think you are confused.

> It would be helpful to users and newcomers if all test related documents
> could be found under a Testing headline in
> http://xenbits.xen.org/docs/unstable/

This is certainly not possible because the url you quote is generated
from xen.git.

Ian.

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel

^ permalink raw reply	[flat|nested] 43+ messages in thread

* Re: xenbits "official" repo for XTF (was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix)
  2016-07-18 15:19                                       ` Lars Kurth
  2016-07-18 15:30                                         ` Ian Jackson
@ 2016-07-18 15:31                                         ` Ian Jackson
  1 sibling, 0 replies; 43+ messages in thread
From: Ian Jackson @ 2016-07-18 15:31 UTC (permalink / raw)
  To: Lars Kurth
  Cc: Wei Liu, Lars Kurth, Andrew Cooper, George Dunlap, David Vrabel,
	xen-devel, Boris Ostrovsky, Anshul Makkar, Roger Pau Monne

Lars Kurth writes ("Re: [Xen-devel] xenbits "official" repo for XTF (was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix)"):
> If there is a link between directory structure and web index, my preferred
> alternative to creating the two separate directories would be to group all
> the test related documents into a docs/testing directory.

Also, I am annoyed that just when I thought we had agreed a
compromise, you have opened another, irrelevant, can of worms.

Ian.


_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel

^ permalink raw reply	[flat|nested] 43+ messages in thread

* Re: xenbits "official" repo for XTF (was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix)
  2016-07-18 15:30                                         ` Ian Jackson
@ 2016-07-18 15:36                                           ` Lars Kurth
  2016-07-18 15:58                                             ` Ian Jackson
  0 siblings, 1 reply; 43+ messages in thread
From: Lars Kurth @ 2016-07-18 15:36 UTC (permalink / raw)
  To: Ian Jackson
  Cc: Wei Liu, Lars Kurth, Andrew Cooper, George Dunlap, David Vrabel,
	xen-devel, Boris Ostrovsky, Anshul Makkar, Roger Pau Monne



On 18/07/2016 16:30, "Ian Jackson" <ian.jackson@eu.citrix.com> wrote:

>Lars Kurth writes ("Re: [Xen-devel] xenbits "official" repo for XTF (was
>Re: [PATCH 0/2] xtf: add launcher (+1 bugfix)"):
>> I don't object to a new xtf directory in the docs root per se. I mainly
>> object to how the documents are presented in
>> http://xenbits.xen.org/docs/unstable
>
>The documents under there are generated from xen.git#master.
>
>> As far as I understand, the docs directory structure maps onto headline
>> categories in http://xenbits.xen.org/docs/unstable. Is this correct?
>
>What ?  No.
>
>docs/ is just the HTTP namespace section for documents hosted on
>xenbits.  Its subdirectories currently include:
>
>  - one directory for each existing stable and unstable Xen branch,
>    generated from the in-tree documentation source.
>
>  - some example osstest output, referred to from the osstest.git
>    in-tree README
>
>  - a subdirectory `xcp' which is presumably now obsolete and retained
>    for historical reasons.
>
>It appears that it doesn't have an index.html, so we get a webserver
>directory listing.
>
>> I would only object to creating a new docs/xtf directory,
>> - iff directory structure is exposed in the web index in
>> http://xenbits.xen.org/docs/unstable are linked, and
>
>I don't understand what you are asking.
>
>> - iff there is only one document in this directory now, and
>> - iff this will likely be the case in future.
>
>I'm proposing that the initial contents of the xtf/ directory would be
>the initial contents of the /people/andrewcoop/xen-test-framework/
>directory, only with the titles changed as discussed.
>
>So there would be one `document', which is really an HTML page, called
>index.html, so it would replace the webserver directory listing.

>
>> And I would have the same concern for osstest. If there are several
>> documents already in each, then go ahead and create them.
>
>Is your concern that it is a bad idea to create a directory rather
>than a file ?
>
>I think this is wrongheaded.
>
>It is better to create a directory now, with an index.html.  Otherwise
>it will become difficult to turn it into a directory later.

In that case, forget my objection and go ahead and create the directories.
I clearly don't understand how the docs toolchain works and was assuming
that the docs directory structure gets exposed in the web server index,
without being able to specify an extra indirection. Given that is not the
case, the directory structure does not matter.

Lars

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel

^ permalink raw reply	[flat|nested] 43+ messages in thread

* Re: xenbits "official" repo for XTF (was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix)
  2016-07-18 15:36                                           ` Lars Kurth
@ 2016-07-18 15:58                                             ` Ian Jackson
  2016-07-18 16:06                                               ` Lars Kurth
  0 siblings, 1 reply; 43+ messages in thread
From: Ian Jackson @ 2016-07-18 15:58 UTC (permalink / raw)
  To: Lars Kurth
  Cc: Wei Liu, Lars Kurth, Andrew Cooper, George Dunlap, David Vrabel,
	xen-devel, Boris Ostrovsky, Anshul Makkar, Roger Pau Monne

Lars Kurth writes ("Re: [Xen-devel] xenbits "official" repo for XTF (was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix)"):
> In that case, forget my objection and go ahead and create the directories.
> I clearly don't understand how the docs toolchain works and was assuming
> that the docs directory structure gets exposed in the web server index,
> without being able to specify an extra indirection. Given that is not the
> case, the directory structure does not matter.

There are four information structures.

Note that `directory structure' and `link structure between html
pages' can be different for the same collection of documents.  This is
because there is no necessary correspondence between the directory
structure of html file storage (normally visibile in urls) and the
logical structure of the document forest (represented in
table-of-contents pages, cross-page hyperlinks, and so on).

From least to most visible:

1. The xen.git source tree docs/ subdirectory has a directory structure
in the source code which is used when organising and editing the docs.

2a. When xen.git is built, it generates docs/html/ which contains a
directory structure.  The directory structure is controlled by the
individual Makefiles in the rest of docs/.

2b. When xen.git is built, the directory structure in docs/html/ is
converted into an index.html which is put at the top of docs/html/, by
a script `gen-html-index' which simply peers at the titles of the
files.  The same script is used to generate an index.html for each
subdirectory.  So in practice this is very similar to 2a.

3. The webtree http://xenbits.xen.org/docs/ is primarily a set of
files and directories which are simply served up by apache, and there
is no index.html at the top level, so the top level is a webserver
directory listing.  There are various subdirectories, which are
directly exposed, including

   - /docs/unstable/           xen.git#master's version of 2a/2b
   - /docs/unstable-staging    xen.git#stagin's version of 2a/2b
     [etc. for the stable branches, autogenerated]
   - /docs/osstest-sample-output/   rename proposed


The proposal is to add

   - /docs/osstest/
   - /docs/xtf/

to 3.  1 and 2a and 2b will remain unchanged.

Ian.

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel

^ permalink raw reply	[flat|nested] 43+ messages in thread

* Re: xenbits "official" repo for XTF (was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix)
  2016-07-18 15:58                                             ` Ian Jackson
@ 2016-07-18 16:06                                               ` Lars Kurth
  0 siblings, 0 replies; 43+ messages in thread
From: Lars Kurth @ 2016-07-18 16:06 UTC (permalink / raw)
  To: Ian Jackson
  Cc: Wei Liu, Lars Kurth, Andrew Cooper, George Dunlap, David Vrabel,
	xen-devel, Boris Ostrovsky, Anshul Makkar, Roger Pau Monne



On 18/07/2016 16:58, "Ian Jackson" <ian.jackson@eu.citrix.com> wrote:

>Lars Kurth writes ("Re: [Xen-devel] xenbits "official" repo for XTF (was
>Re: [PATCH 0/2] xtf: add launcher (+1 bugfix)"):
>> In that case, forget my objection and go ahead and create the
>>directories.
>> I clearly don't understand how the docs toolchain works and was assuming
>> that the docs directory structure gets exposed in the web server index,
>> without being able to specify an extra indirection. Given that is not
>>the
>> case, the directory structure does not matter.
>
>There are four information structures.
>
>Note that `directory structure' and `link structure between html
>pages' can be different for the same collection of documents.  This is
>because there is no necessary correspondence between the directory
>structure of html file storage (normally visibile in urls) and the
>logical structure of the document forest (represented in
>table-of-contents pages, cross-page hyperlinks, and so on).

Thanks for clarifying

>
>The proposal is to add
>
>   - /docs/osstest/
>   - /docs/xtf/

Go ahead and create them. We can figure out how to present the docs to the
user later. 

I just noticed that docs/INDEX controls what is shown on the generated
html pages. Sorry for opening this can of worms unnecessarily.

Lars

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel

^ permalink raw reply	[flat|nested] 43+ messages in thread

* Re: xenbits "official" repo for XTF (was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix)
  2016-07-15 10:21                                 ` Ian Jackson
  2016-07-15 10:49                                   ` Lars Kurth
@ 2016-08-08 17:04                                   ` Ian Jackson
  1 sibling, 0 replies; 43+ messages in thread
From: Ian Jackson @ 2016-08-08 17:04 UTC (permalink / raw)
  To: Lars Kurth, Andrew Cooper, Lars Kurth, Konrad Rzeszutek Wilk,
	David Vrabel, George Dunlap, Boris Ostrovsky, Wei Liu, xen-devel,
	Anshul Makkar, Roger Pau Monne

Ian Jackson writes ("Re: [Xen-devel] xenbits "official" repo for XTF (was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix)"):
> With my xenbits admin hat on I therefore intend to:
> 
>  * Create xtf.git in the xenbits toplevel
>    and give it the appropriate permissions

I created a group `xtf-committers' and put Andrew in it.

>  * Create xtf.git/description containing
> 
>      Xen Test Framework and Suite for creating microkernel-based tests
> 
>  * Edit osstest.git/description to contain
> 
>      Xen Test Framework and Suite, used for Open Source Xen CI push gate

I have done the above.

xtf.git is currently just what you get from `git init --bare'.

Ian.

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel

^ permalink raw reply	[flat|nested] 43+ messages in thread

end of thread, other threads:[~2016-08-08 17:04 UTC | newest]

Thread overview: 43+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2016-05-19 13:46 [PATCH 0/2] xtf: add launcher (+1 bugfix) Roger Pau Monne
2016-05-19 13:46 ` [PATCH 1/2] xtf: remove setting ROOT path in common.mk Roger Pau Monne
2016-05-19 13:52   ` Andrew Cooper
2016-05-19 13:46 ` [PATCH 2/2] xtf: add a launcher script Roger Pau Monne
2016-06-09 15:04 ` xenbits "official" repo for XTF (was Re: [PATCH 0/2] xtf: add launcher (+1 bugfix) Ian Jackson
2016-06-13  8:59   ` George Dunlap
2016-06-13 10:10   ` Andrew Cooper
2016-06-13 14:03     ` George Dunlap
2016-06-13 14:11       ` Ian Jackson
2016-06-14 17:27         ` Andrew Cooper
2016-06-15  8:31           ` George Dunlap
2016-06-20 17:03             ` Ian Jackson
2016-06-20 17:10               ` Konrad Rzeszutek Wilk
2016-06-21 10:17               ` Wei Liu
2016-06-21 10:36               ` David Vrabel
2016-06-21 10:42                 ` Ian Jackson
2016-07-01 19:04                 ` Ian Jackson
2016-07-06 11:19                   ` Lars Kurth
2016-07-06 14:22                     ` Konrad Rzeszutek Wilk
2016-07-07 11:10                       ` Lars Kurth
2016-07-07 11:26                         ` Andrew Cooper
2016-07-07 13:59                           ` Ian Jackson
2016-07-07 14:02                             ` David Vrabel
2016-07-07 14:58                               ` Ian Jackson
2016-07-07 15:00                                 ` Ian Jackson
2016-07-07 14:11                             ` Andrew Cooper
2016-07-07 17:17                               ` Lars Kurth
2016-07-08 13:06                                 ` Andrew Cooper
2016-07-08 13:42                                   ` Lars Kurth
2016-07-08 13:48                                     ` Andrew Cooper
2016-07-12 15:13                                   ` George Dunlap
2016-07-15 10:21                                 ` Ian Jackson
2016-07-15 10:49                                   ` Lars Kurth
2016-07-18 14:36                                     ` Ian Jackson
2016-07-18 15:19                                       ` Lars Kurth
2016-07-18 15:30                                         ` Ian Jackson
2016-07-18 15:36                                           ` Lars Kurth
2016-07-18 15:58                                             ` Ian Jackson
2016-07-18 16:06                                               ` Lars Kurth
2016-07-18 15:31                                         ` Ian Jackson
2016-08-08 17:04                                   ` Ian Jackson
2016-07-07 15:22                           ` Konrad Rzeszutek Wilk
2016-06-14 17:01   ` Konrad Rzeszutek Wilk

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).