On 5/9/26 12:39 PM, Andrew Lunn wrote:
>>> Can we get the tool changed to add a warning, something like:
>>>
>>> WARNING: This tool uses very simple pattern matching to look for
>>> repeated words. It does not understand the complexity of English,
>>> and will often result in false positive reports. Please assume it is
>>> wrong until proven otherwise.
>>
>> There was no commit log and no cover letter AFAIK.
>> Do we know what tool was used?
>>
>> Adrien, how did you discover these repeated words?
>>
>> (If it's my script from 2021, I'll gladly update it.)
Adrien is not using my script -- they developed their own script.
> Thinking about it some more, i think the warning might actually need
> to be different.
>
> If this tool has been around since 2021, all the real problems have
> been solved, leaving only the false positives. So the warning probably
> needs to be much stronger, saying that it probably only reports false
> positives, unless the code is new.
Makes some sense.
> Maybe we also want to extend the tool to have a list all the known
> false positives?
I'm not crazy about that one. And we have no evidence that I am aware
of that my script is causing any of these patches.
--
~Randy
