Better document some elements of the ClusterFuzz UI

[alex]

oss-fuzz is going to bring a bunch of new folks to the ClusterFuzz UI. Immediately on looking at a crash I had a handful of questions:

• What does the "REDO" button do?
• Do I need to create an issue for each crash I see?
• How do I go about getting an non-reproducible hang from Fixed: NO to some other state?

[inferno-chromium]

Working on it.

[kcwu]

Please also describe what is "quarantine" and how it works. I cannot find documents about it.

[inferno-chromium]

I made some UI changes, can you take a look and provide feedback before i finalize.

What does the "REDO" button do?
-- Click the button to see its purpose.
This is for rechecking results like regression range, minimization result in case it look ambiguous.
Do I need to create an issue for each crash I see?
-- Click the button to see its purpose. So, answer is no, reproducible testcase will get auto-filed. Non-reproducible testcase are not recommended to file, unless you are planning to fix it or can see problem from stacktrace.
How do I go about getting an non-reproducible hang from Fixed: NO to some other state?
-- This was a UI bug. It should have said NA since we can't do fixed testing on one-time crashers/unreproducible testcase. Fixed now.

[alex]

These answers, plus the UI changes, are very helpful. Thanks.

[alex]

Another question, what does "known_crashes" mean? The stats page claims a number of these, but I don't understand where they come from:

[inferno-chromium]

known crashes - crashes that happened in past and already known. in this case, i see you might be confused in knowing what these looked like. we are working on an improved crash stats page (different from fuzzer stats) that should link to these.

If you are looking at the libyaml known crashes, this is https://bugs.chromium.org/p/oss-fuzz/issues/detail?id=225.

I will add these terms to Glossary soon.

[alex]

Yes, I'm looking at libyaml, which is why I'm confused. There's 1 new crash (ever), but 3 or 5 known ones! It sounds like maybe some of the "known" ones are aggregated with ones from other projects?

…

On Sun, Dec 4, 2016 at 9:04 PM, inferno-chromium ***@***.***> wrote: known crashes - crashes that happened in past and already known. in this case, i see you might be confused in knowing what these looked like. we are working on an improved crash stats page (different from fuzzer stats) that should link to these. If you are looking at the libyaml known crashes, this is https://bugs.chromium.org/p/oss-fuzz/issues/detail?id=225. I will add these terms to Glossary soon. — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#127 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AAADBFDoaNb3oS0wQ-cdCe9iVSnPnbG-ks5rE3EegaJpZM4LDALG> .

-- "I disapprove of what you say, but I will defend to the death your right to say it." -- Evelyn Beatrice Hall (summarizing Voltaire) "The people's good is the highest law." -- Cicero GPG Key fingerprint: D1B3 ADC0 E023 8CA6

[inferno-chromium]

Known crashes is a total number, not unique number of crashes. And these known ones are just 1 crash which has same crash state signature as some from other projects. This is a sanitizer check failure which is also happening with a few other projects. We will look more into resolving this confusion.

[inferno-chromium]

ClusterFuzz UI changes are done, others doc changes are tracked in #133.