Re: [9front] Change of Hashsize in acid (slow startup on netsurf)

[Stanley Lieber <sl@stanleylieber.com>]

On January 5, 2021 4:10:19 PM EST, hiro <23hiro@gmail.com> wrote:
>i did not manage to quote the relevant part:
>
>Authentication-Results: mx.google.com;
>       dkim=neutral (body hash did not verify) header.i=@gmail.com
>header.s=20161025 header.b=frcOH6pW;
>       spf=pass (google.com: domain of 9front-bounces@1ess.inri.net
>designates 216.126.196.35 as permitted sender)
>smtp.mailfrom=9front-bounces@1ess.inri.net;
>       dmarc=fail (p=NONE sp=QUARANTINE dis=NONE) header.from=gmail.com
>Received: from mail-vk1-f180.google.com ([209.85.221.180]) by 1ess; Tue Jan
>
>the old server, ewsd seems to have done the same, but somehow gmail
>didn't take it serious (perhaps people have trained gmail well enough
>by manually pressing NOT SPAM).
>
>On 1/5/21, hiro <23hiro@gmail.com> wrote:
>> the 9front mailing list seems to be making a weird subdomain
>> quarantine decision:
>>
>> Authentication-Results: mx.google.com;
>>        dkim=neutral (body hash did not verify) header.i=@gmail.com
>> header.s=20161025 header.b=frcOH6pW;
>>        spf=pass (google.com: domain of 9front-bounces@1ess.inri.net
>> designates 216.126.196.35 as permitted sender)
>> smtp.mailfrom=9front-bounces@1ess.inri.net;
>>        dmarc=fail (p=NONE sp=QUARANTINE dis=NONE) header.from=gmail.com
>>
>> but the from header does not specify a subdomine, so quarantine should not
>> apply
>>
>> ---------- Forwarded message ----------
>> From: Jonas Amoson <jonas.amoson@gmail.com>
>> Date: Tue, 5 Jan 2021 14:10:09 +0200
>> Subject: [9front] Change of Hashsize in acid (slow startup on netsurf)
>> To: 9front@9front.org
>>
>> It is slow to debug netsurf with acid, as it takes a long time to load
>> all the symbols from the binary. It is the process of checking if a
>> symbol has been encountered before, the function unique(), that makes
>> the loading slow.
>>
>> Changing Hashsize from 128 to 32768 in acid takes down the loading
>> time from 4 minutes to 8 seconds for me (45 times faster). Philippe
>> reported a similar speed increase (30x from 15 min to 30 sec).
>>
>> I don't know if it is desirable to change the hashsize in the 9front
>> repo, as it probably is only netsurf that has so many symbols to load,
>> that it matters. But it might be that it doesn't harm anything either.
>>
>> /Jonas
>>
>

i failed to get the ptr record updated until earlier today. hopefully gmail will stop complaining once it propagates.

sl