In the wake of the BZ 1924808[1] discussion in Thursday's Go/No-Go meeting[2], I am proposing an addition to the Basic Release Criteria[3]. This would go into Post-Install Requirements -> Expected installed system boot behavior -> First boot utilities (appended after the existing sentence):
If a utility for creating user accounts and other configuration is configured to launch, it must be visible within 10 seconds of the first boot reaching the launch point.
Why 10 seconds? Why not? That sort of feels like the maximum length of time someone could reasonably be expected to wait. A shorter time might be better.
I don't particularly love the wording here, but I wanted to make it clear that it's not 10 seconds from power on, but 10 seconds from the time the boot up reaches the state where we expect gnome-initial-setup or its counterparts to appear.
[1] https://bugzilla.redhat.com/show_bug.cgi?id=1924908 [2] https://meetbot.fedoraproject.org/sresults/?group_id=f34-beta-go_no_go-meeti... [3] https://fedoraproject.org/wiki/Basic_Release_Criteria#Expected_installed_sys...
On 3/14/21 10:13 AM, Ben Cotton wrote:
In the wake of the BZ 1924808[1] discussion in Thursday's Go/No-Go meeting[2], I am proposing an addition to the Basic Release Criteria[3]. This would go into Post-Install Requirements -> Expected installed system boot behavior -> First boot utilities (appended after the existing sentence):
If a utility for creating user accounts and other configuration is configured to launch, it must be visible within 10 seconds of the first boot reaching the launch point.
Why 10 seconds? Why not? That sort of feels like the maximum length of time someone could reasonably be expected to wait. A shorter time might be better.
I don't particularly love the wording here, but I wanted to make it clear that it's not 10 seconds from power on, but 10 seconds from the time the boot up reaches the state where we expect gnome-initial-setup or its counterparts to appear.
[1] https://bugzilla.redhat.com/show_bug.cgi?id=1924908 [2] https://meetbot.fedoraproject.org/sresults/?group_id=f34-beta-go_no_go-meeti... [3] https://fedoraproject.org/wiki/Basic_Release_Criteria#Expected_installed_sys...
I like it.
Can we also add a "and displayed without error" clause, or maybe "completes with no visible error"? Something to explicitly capture the "sad face" bug[0].
On Mon, Mar 15, 2021 at 10:41 AM Brandon Nielsen nielsenb@jetfuse.net wrote:
Can we also add a "and displayed without error" clause, or maybe "completes with no visible error"? Something to explicitly capture the "sad face" bug[0].
I thought about that and decided to leave it off as a Basic criterion. I'd be fine with it for Final, but I was worried it would be too strict for Beta, where we expect some degree of sad face from time to time. But maybe this isn't one of those cases? If the consensus is to include a "no errors!", I won't object.
On 3/15/21 9:45 AM, Ben Cotton wrote:
On Mon, Mar 15, 2021 at 10:41 AM Brandon Nielsen nielsenb@jetfuse.net wrote:
Can we also add a "and displayed without error" clause, or maybe "completes with no visible error"? Something to explicitly capture the "sad face" bug[0].
I thought about that and decided to leave it off as a Basic criterion. I'd be fine with it for Final, but I was worried it would be too strict for Beta, where we expect some degree of sad face from time to time. But maybe this isn't one of those cases? If the consensus is to include a "no errors!", I won't object.
I expect sad face in a beta as well, but once the user gets past the install and setup phase and is actually using the OS.
My hope is an earlier criterion would encourage a bit more movement on issues that occur immediately and right in the user's face? I found it a little concerning how little chatter such an obvious (no steps to reproduce other than "install a Gnome compose") and bad looking (if ultimately harmless) bug caused.
On 3/15/21 10:45, Ben Cotton wrote:
On Mon, Mar 15, 2021 at 10:41 AM Brandon Nielsen nielsenb@jetfuse.net wrote:
Can we also add a "and displayed without error" clause, or maybe "completes with no visible error"? Something to explicitly capture the "sad face" bug[0].
I thought about that and decided to leave it off as a Basic criterion. I'd be fine with it for Final, but I was worried it would be too strict for Beta, where we expect some degree of sad face from time to time. But maybe this isn't one of those cases? If the consensus is to include a "no errors!", I won't object.
A Beta is something that we release to the general public. I don't think it's good for Fedora's reputation to have the Oops sad face screen show up during the installation / setup.
Sure there will be bugs in a Beta. I and I think most folks expect to see a small number of SE alerts and/or Abrt notifications, but not something that Shouts "I'm Broken!"; especially during install / setup. Perhaps a middle position would be to say the sad face screen can't be displayed.
Have a Great Day!
Pat (tablepc)
On Mon, 2021-03-15 at 09:35 -0500, Brandon Nielsen wrote:
On 3/14/21 10:13 AM, Ben Cotton wrote:
In the wake of the BZ 1924808[1] discussion in Thursday's Go/No-Go meeting[2], I am proposing an addition to the Basic Release Criteria[3]. This would go into Post-Install Requirements -> Expected installed system boot behavior -> First boot utilities (appended after the existing sentence):
If a utility for creating user accounts and other configuration is configured to launch, it must be visible within 10 seconds of the first boot reaching the launch point.
Why 10 seconds? Why not? That sort of feels like the maximum length of time someone could reasonably be expected to wait. A shorter time might be better.
I don't particularly love the wording here, but I wanted to make it clear that it's not 10 seconds from power on, but 10 seconds from the time the boot up reaches the state where we expect gnome-initial-setup or its counterparts to appear.
[1] https://bugzilla.redhat.com/show_bug.cgi?id=1924908 [2] https://meetbot.fedoraproject.org/sresults/?group_id=f34-beta-go_no_go-meeti... [3] https://fedoraproject.org/wiki/Basic_Release_Criteria#Expected_installed_sys...
I like it.
Can we also add a "and displayed without error" clause, or maybe "completes with no visible error"? Something to explicitly capture the "sad face" bug[0].
We actually have this already, pretty much, for Final:
"There must be no SELinux denial notifications or crash notifications on boot of or during installation from a release-blocking live image, or at first login after a default install of a release-blocking desktop."
It was envisaged to cover desktop notifications, but I'd say it seems reasonable to count it as covering the "Oh no" screen too.
As #action-ed in yesterday's QA meeting[1], I have added the proposal as-written to the Basic release criteria[2].
If a utility for creating user accounts and other configuration is configured to launch, it must be visible within 10 seconds of the first boot reaching the launch point.
The most significant feedback was that we should also block on the sad face. As Adam noted, we can consider that under the Final criterion
There must be no SELinux denial notifications or crash notifications on boot of or during installation from a release-blocking live image, or at first login after a default install of a release-blocking desktop.
Of course, we can always adjust this criterion in the future.
[1] https://meetbot.fedoraproject.org/teams/fedora-qa/fedora-qa.2021-04-12-15.01... [2] https://fedoraproject.org/w/index.php?title=Basic_Release_Criteria&actio...
On 4/13/21 9:39 AM, Ben Cotton wrote: [Snip]
There must be no SELinux denial notifications or crash notifications on boot of or during installation from a release-blocking live image, or at first login after a default install of a release-blocking desktop.
Of course, we can always adjust this criterion in the future.
[Snip]
Does that make this[0] a blocker candidate?
Somewhat related, is there some mapping of compose test cases to release criterion? Or the inverse? Something like "Failures of test case foo, bar, and baz may constitute a violation of release criterion qux"?
On 4/14/21 2:13 PM, Brandon Nielsen wrote: [Snip]
Somewhat related, is there some mapping of compose test cases to release criterion? Or the inverse? Something like "Failures of test case foo, bar, and baz may constitute a violation of release criterion qux"? [Snip]
Nevermind, I see they're cross referenced both ways, I just somehow never noticed before. Sorry for the noise.
On Wed, Apr 14, 2021 at 02:13:31PM -0500, Brandon Nielsen wrote:
Does that make this[0] a blocker candidate? [0] - https://bugzilla.redhat.com/show_bug.cgi?id=1943683
Looks like it to me.
On Wed, Apr 14, 2021 at 1:13 PM Brandon Nielsen nielsenb@jetfuse.net wrote:
On 4/13/21 9:39 AM, Ben Cotton wrote: [Snip]
There must be no SELinux denial notifications or crash notifications on boot of or during installation from a release-blocking live image, or at first login after a default install of a release-blocking desktop.
Of course, we can always adjust this criterion in the future.
[Snip]
Does that make this[0] a blocker candidate?
Probably more likely this final release criterion:
All system services present after installation with one of the release-blocking package sets must start properly, unless they require hardware which is not present.
https://fedoraproject.org/wiki/Fedora_34_Final_Release_Criteria#System_servi...
Somewhat related, is there some mapping of compose test cases to release criterion? Or the inverse? Something like "Failures of test case foo, bar, and baz may constitute a violation of release criterion qux"?
I'm pretty sure it's all contained to the Basic, Beta, and Final release criteria pages.
Hi Chris With the beta being very very very clean, and functional, the Linux community has taken the date of 25March as the official release date.I found a design flaw, but not a malfunction with the beta.
Regards Leslie Leslie Satenstein Montréal Québec, Canada
On Wednesday, April 14, 2021, 11:01:37 p.m. EDT, Chris Murphy lists@colorremedies.com wrote:
On Wed, Apr 14, 2021 at 1:13 PM Brandon Nielsen nielsenb@jetfuse.net wrote:
On 4/13/21 9:39 AM, Ben Cotton wrote: [Snip]
There must be no SELinux denial notifications or crash notifications on boot of or during installation from a release-blocking live image, or at first login after a default install of a release-blocking desktop.
Of course, we can always adjust this criterion in the future.
[Snip]
Does that make this[0] a blocker candidate?
Probably more likely this final release criterion:
All system services present after installation with one of the release-blocking package sets must start properly, unless they require hardware which is not present.
https://fedoraproject.org/wiki/Fedora_34_Final_Release_Criteria#System_servi...
Somewhat related, is there some mapping of compose test cases to release criterion? Or the inverse? Something like "Failures of test case foo, bar, and baz may constitute a violation of release criterion qux"?
I'm pretty sure it's all contained to the Basic, Beta, and Final release criteria pages.
On Sun, Mar 14, 2021 at 11:13:45AM -0400, Ben Cotton wrote:
In the wake of the BZ 1924808[1] discussion in Thursday's Go/No-Go meeting[2], I am proposing an addition to the Basic Release Criteria[3]. This would go into Post-Install Requirements -> Expected installed system boot behavior -> First boot utilities (appended after the existing sentence):
If a utility for creating user accounts and other configuration is configured to launch, it must be visible within 10 seconds of the first boot reaching the launch point.
Why 10 seconds? Why not? That sort of feels like the maximum length of time someone could reasonably be expected to wait. A shorter time might be better.
I don't particularly love the wording here, but I wanted to make it clear that it's not 10 seconds from power on, but 10 seconds from the time the boot up reaches the state where we expect gnome-initial-setup or its counterparts to appear.
Sounds reasonable to me. +1
I mean, I'd be fine changing the time a little or wording, but I agree with the general gist of it.
kevin
On 3/15/21 11:57, Kevin Fenzi wrote:
On Sun, Mar 14, 2021 at 11:13:45AM -0400, Ben Cotton wrote:
In the wake of the BZ 1924808[1] discussion in Thursday's Go/No-Go meeting[2], I am proposing an addition to the Basic Release Criteria[3]. This would go into Post-Install Requirements -> Expected installed system boot behavior -> First boot utilities (appended after the existing sentence):
If a utility for creating user accounts and other configuration is configured to launch, it must be visible within 10 seconds of the first boot reaching the launch point.
Why 10 seconds? Why not? That sort of feels like the maximum length of time someone could reasonably be expected to wait. A shorter time might be better.
I don't particularly love the wording here, but I wanted to make it clear that it's not 10 seconds from power on, but 10 seconds from the time the boot up reaches the state where we expect gnome-initial-setup or its counterparts to appear.
Sounds reasonable to me. +1
I mean, I'd be fine changing the time a little or wording, but I agree with the general gist of it.
kevin
I agree. Speaking as someone who restarts often; especially when running tests. That extra time was becoming very annoying.
Have a Great Day!
Pat (tablepc)
On Sun, 2021-03-14 at 11:13 -0400, Ben Cotton wrote:
In the wake of the BZ 1924808[1] discussion in Thursday's Go/No-Go meeting[2], I am proposing an addition to the Basic Release Criteria[3]. This would go into Post-Install Requirements -> Expected installed system boot behavior -> First boot utilities (appended after the existing sentence):
If a utility for creating user accounts and other configuration is configured to launch, it must be visible within 10 seconds of the first boot reaching the launch point.
Why 10 seconds? Why not? That sort of feels like the maximum length of time someone could reasonably be expected to wait. A shorter time might be better.
I don't particularly love the wording here, but I wanted to make it clear that it's not 10 seconds from power on, but 10 seconds from the time the boot up reaches the state where we expect gnome-initial-setup or its counterparts to appear.
So, this kinda stalled, but is now a live issue again because we have exactly the same kind of bug again:
https://bugzilla.redhat.com/show_bug.cgi?id=1997310
so, no-one really objected to this proposal, right? Should we just go ahead and implement it, or what?
On Tue, Aug 24, 2021 at 4:00 PM Adam Williamson adamwill@fedoraproject.org wrote:
On Sun, 2021-03-14 at 11:13 -0400, Ben Cotton wrote:
In the wake of the BZ 1924808[1] discussion in Thursday's Go/No-Go meeting[2], I am proposing an addition to the Basic Release Criteria[3]. This would go into Post-Install Requirements -> Expected installed system boot behavior -> First boot utilities (appended after the existing sentence):
If a utility for creating user accounts and other configuration is configured to launch, it must be visible within 10 seconds of the first boot reaching the launch point.
Why 10 seconds? Why not? That sort of feels like the maximum length of time someone could reasonably be expected to wait. A shorter time might be better.
I don't particularly love the wording here, but I wanted to make it clear that it's not 10 seconds from power on, but 10 seconds from the time the boot up reaches the state where we expect gnome-initial-setup or its counterparts to appear.
So, this kinda stalled, but is now a live issue again because we have exactly the same kind of bug again:
https://bugzilla.redhat.com/show_bug.cgi?id=1997310
so, no-one really objected to this proposal, right? Should we just go ahead and implement it, or what?
Let's go ahead and implement it. Only QA types would consider waiting 2 minutes for this just to (a) time it (b) see what happens, if anything (c) try to find logs and/or some sort of work around. All other mortals will properly hit the power button, give it all a second go - which of course will just hit the same problem, i.e. broken.
On Wed, Aug 25, 2021 at 12:00 AM Adam Williamson adamwill@fedoraproject.org wrote:
On Sun, 2021-03-14 at 11:13 -0400, Ben Cotton wrote:
In the wake of the BZ 1924808[1] discussion in Thursday's Go/No-Go meeting[2], I am proposing an addition to the Basic Release Criteria[3]. This would go into Post-Install Requirements -> Expected installed system boot behavior -> First boot utilities (appended after the existing sentence):
If a utility for creating user accounts and other configuration is
configured to launch, it must be visible within 10 seconds of the first boot reaching the launch point.
I'm not exactly clear on what "the launch point" is, i.e. when I should start counting.
My guess is that we start counting when the wall paper is up and the top bar is displayed.
Have a Great Day!
Pat (tablepc)
On 8/25/21 06:43, Kamil Paral wrote:
On Wed, Aug 25, 2021 at 12:00 AM Adam Williamson adamwill@fedoraproject.org wrote:
On Sun, 2021-03-14 at 11:13 -0400, Ben Cotton wrote:
In the wake of the BZ 1924808[1] discussion in Thursday's Go/No-Go meeting[2], I am proposing an addition to the Basic Release Criteria[3]. This would go into Post-Install Requirements -> Expected installed system boot behavior -> First boot utilities (appended after the existing sentence):
If a utility for creating user accounts and other configuration is
configured to launch, it must be visible within 10 seconds of the first boot reaching the launch point.
I'm not exactly clear on what "the launch point" is, i.e. when I should start counting.
test mailing list -- test@lists.fedoraproject.org To unsubscribe send an email to test-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/test@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
On Wed, Aug 25, 2021 at 12:44 PM Kamil Paral kparal@redhat.com wrote:
On Wed, Aug 25, 2021 at 12:00 AM Adam Williamson < adamwill@fedoraproject.org> wrote:
On Sun, 2021-03-14 at 11:13 -0400, Ben Cotton wrote:
In the wake of the BZ 1924808[1] discussion in Thursday's Go/No-Go meeting[2], I am proposing an addition to the Basic Release Criteria[3]. This would go into Post-Install Requirements -> Expected installed system boot behavior -> First boot utilities (appended after the existing sentence):
If a utility for creating user accounts and other configuration is
configured to launch, it must be visible within 10 seconds of the first boot reaching the launch point.
I'm not exactly clear on what "the launch point" is, i.e. when I should start counting.
Could we just say something along the lines that "Applications and the shell must start reasonably fast and not e.g. run into a 60 second timeout when starting." and leave it up for interpretation?
On Wed, Aug 25, 2021 at 6:44 AM Kamil Paral kparal@redhat.com wrote:
If a utility for creating user accounts and other configuration is configured to launch, it must be visible within 10 seconds of the first boot reaching the launch point.
I'm not exactly clear on what "the launch point" is, i.e. when I should start counting.
Glibly, it's 10 seconds before the point where you go "huh, this should have launched by now." :-)
It's admittedly a bit of a fuzzy description, but I think that's okay as a starting point. It's not like we expect testers to sit there with a stopwatch. This will be, IMO, a subjective-ish opinion. The 10 seconds is there more as an indicator of the magnitude of acceptable delay, not as a firm cutoff point.
On Wed, Aug 25, 2021 at 2:30 PM Ben Cotton bcotton@redhat.com wrote:
On Wed, Aug 25, 2021 at 6:44 AM Kamil Paral kparal@redhat.com wrote:
If a utility for creating user accounts and other configuration is
configured to launch, it must be visible within 10 seconds of the first boot reaching the launch point.
I'm not exactly clear on what "the launch point" is, i.e. when I should
start counting.
Glibly, it's 10 seconds before the point where you go "huh, this should have launched by now." :-)
Well then, can we have a phrasing which will not make people ask what the "launch point" is? If we can't define it, don't put it there, because it will only confuse readers.
One option was proposed by Kalev - e.g. "it must start reasonably fast and not wait for an unexpected/undesired timeout".
Or if we want to have some exact numbers, what about "it must not add more than 10 seconds to a regular system boot speed". Of course with the slight problem that you don't know what a regular system boot speed is, if this is your first installed system and you're looking at the first system boot. But for a regular tester this shouldn't be a problem, and if you really want, you can always measure and compare it with a second+ boot.
On Wed, Aug 25, 2021 at 11:03 AM Kamil Paral kparal@redhat.com wrote:
Or if we want to have some exact numbers, what about "it must not add more than 10 seconds to a regular system boot speed".
I have no objections to that.
On Wed, 2021-08-25 at 11:33 -0400, Ben Cotton wrote:
On Wed, Aug 25, 2021 at 11:03 AM Kamil Paral kparal@redhat.com wrote:
Or if we want to have some exact numbers, what about "it must not add more than 10 seconds to a regular system boot speed".
I have no objections to that.
It seems Ben went ahead and merged this into the page. Personally I think specifying "10 seconds" is a bit much - 10 seconds on what system? If it takes 11 on my ten-year old laptop, are we really going to block a Beta release on that? - but at least we have something. I'll maybe propose an adjustment next week. In the mean time, I'll propose the bug we have in F35 Beta as a blocker for review at the meeting tomorrow. Thanks!
On Thu, Sep 16, 2021 at 7:17 PM Adam Williamson adamwill@fedoraproject.org wrote:
On Wed, 2021-08-25 at 11:33 -0400, Ben Cotton wrote:
On Wed, Aug 25, 2021 at 11:03 AM Kamil Paral kparal@redhat.com wrote:
Or if we want to have some exact numbers, what about "it must not add
more than 10 seconds to a regular system boot speed".
I have no objections to that.
It seems Ben went ahead and merged this into the page.
A long time ago, actually: https://lists.fedoraproject.org/archives/list/test@lists.fedoraproject.org/m...
But I also missed his announcement, and when this discussion was renewed, I thought the criterion still wasn't finalized and in effect. It seems I wasn't the only one :o)
On Fri, Sep 17, 2021 at 3:35 AM Kamil Paral kparal@redhat.com wrote:
But I also missed his announcement, and when this discussion was renewed, I thought the criterion still wasn't finalized and in effect. It seems I wasn't the only one :o)
I also forgot that I had done it when the discussion picked back up. But yeah, I'm definitely flexible on the timing. I think the reason we went with a number was to avoid getting bogged down in what "reasonable" meant. But the existing criterion has its own ambiguity, so if we can make it better*, we should.
* But what does "better" mean?! :-D
What if we replaced the "10 seconds" with something like "reasonably fast". With different machines people could have different expectations, so somebody could consider 10 seconds a long time, while others could take 20 for normal. This could make more room for usual experience and expectations.
On Fri, Sep 17, 2021 at 1:13 PM Ben Cotton bcotton@redhat.com wrote:
On Fri, Sep 17, 2021 at 3:35 AM Kamil Paral kparal@redhat.com wrote:
But I also missed his announcement, and when this discussion was
renewed, I thought the criterion still wasn't finalized and in effect. It seems I wasn't the only one :o)
I also forgot that I had done it when the discussion picked back up. But yeah, I'm definitely flexible on the timing. I think the reason we went with a number was to avoid getting bogged down in what "reasonable" meant. But the existing criterion has its own ambiguity, so if we can make it better*, we should.
- But what does "better" mean?! :-D
-- Ben Cotton He / Him / His Fedora Program Manager Red Hat TZ=America/Indiana/Indianapolis _______________________________________________ test mailing list -- test@lists.fedoraproject.org To unsubscribe send an email to test-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/test@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
On 9/17/21 07:29, Lukas Ruzicka wrote:
What if we replaced the "10 seconds" with something like "reasonably fast". With different machines people could have different expectations, so somebody could consider 10 seconds a long time, while others could take 20 for normal. This could make more room for usual experience and expectations.
On Fri, Sep 17, 2021 at 1:13 PM Ben Cotton bcotton@redhat.com wrote:
Providing for the varied experience and expectations also provides room for more bug reports. Some of those reports will likely be for times we either can't or won't fix. Personally I like having a number. I start noticing a restart taking longer than expected when there are no signs of progress and around 10 seconds have passed. I start thinking "this is taking too long at around 20 seconds, and by 30 seconds I start thinking this needs a bug report.
Have a Great Day!
Pat (tablepc)
On Fri, Sep 17, 2021 at 3:35 AM Kamil Paral kparal@redhat.com wrote:
But I also missed his announcement, and when this discussion was
renewed, I thought the criterion still wasn't finalized and in effect. It seems I wasn't the only one :o)
I also forgot that I had done it when the discussion picked back up. But yeah, I'm definitely flexible on the timing. I think the reason we went with a number was to avoid getting bogged down in what "reasonable" meant. But the existing criterion has its own ambiguity, so if we can make it better*, we should.
- But what does "better" mean?! :-D
-- Ben Cotton He / Him / His Fedora Program Manager Red Hat TZ=America/Indiana/Indianapolis _______________________________________________ test mailing list -- test@lists.fedoraproject.org To unsubscribe send an email to test-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/test@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
test mailing list -- test@lists.fedoraproject.org To unsubscribe send an email to test-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/test@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
On Fri, 2021-09-17 at 09:34 +0200, Kamil Paral wrote:
On Thu, Sep 16, 2021 at 7:17 PM Adam Williamson adamwill@fedoraproject.org wrote:
On Wed, 2021-08-25 at 11:33 -0400, Ben Cotton wrote:
On Wed, Aug 25, 2021 at 11:03 AM Kamil Paral kparal@redhat.com wrote:
Or if we want to have some exact numbers, what about "it must not add
more than 10 seconds to a regular system boot speed".
I have no objections to that.
It seems Ben went ahead and merged this into the page.
A long time ago, actually: https://lists.fedoraproject.org/archives/list/test@lists.fedoraproject.org/m...
But I also missed his announcement, and when this discussion was renewed, I thought the criterion still wasn't finalized and in effect. It seems I wasn't the only one :o)
I wrote that mail a week or more ago, but my mail client kindly sat on it and never sent it out till now...
On Wed, 2021-08-25 at 17:01 +0200, Kamil Paral wrote:
On Wed, Aug 25, 2021 at 2:30 PM Ben Cotton bcotton@redhat.com wrote:
On Wed, Aug 25, 2021 at 6:44 AM Kamil Paral kparal@redhat.com wrote:
If a utility for creating user accounts and other configuration is
configured to launch, it must be visible within 10 seconds of the first boot reaching the launch point.
I'm not exactly clear on what "the launch point" is, i.e. when I should
start counting.
Glibly, it's 10 seconds before the point where you go "huh, this should have launched by now." :-)
Well then, can we have a phrasing which will not make people ask what the "launch point" is? If we can't define it, don't put it there, because it will only confuse readers.
One option was proposed by Kalev - e.g. "it must start reasonably fast and not wait for an unexpected/undesired timeout".
Or if we want to have some exact numbers, what about "it must not add more than 10 seconds to a regular system boot speed". Of course with the slight problem that you don't know what a regular system boot speed is, if this is your first installed system and you're looking at the first system boot. But for a regular tester this shouldn't be a problem, and if you really want, you can always measure and compare it with a second+ boot.
Yeah, I would be in favour of making the language as vague as we want the evaluation to be. If it specifies an exact time, we should be able to measure it with a stopwatch. If we can't, let's just say "within a reasonable time" or something vague like that.
-
Adam Williamson -
Ben Cotton -
Brandon Nielsen -
Chris Murphy -
Kalev Lember -
Kamil Paral -
Kevin Fenzi -
Leslie S Satenstein -
Lukas Ruzicka -
Matthew Miller -
pmkellly＠frontier.com