[squeak-dev] AITA - returning a 408 error for a WebClient request where the socket times out?

rabbit rabbit at callistohouse.org
Sun Oct 9 04:29:32 UTC 2022


3 He is despised and rejected of men; a man of sorrows, and acquainted with grief: and we hid as it were our faces from him; he was despised, and we esteemed him not.

4 Surely he hath borne our griefs, and carried our sorrows: yet we did esteem him stricken, smitten of God, and afflicted.

5 But he was wounded for our transgressions, he was bruised for our iniquities: the chastisement of our peace was upon him; and with his stripes we are healed.



Have a good one; keep it, light.
Kindly,
rabbit
. .. … ‘…^,^


Sent from Callisto House :: decentralized mobile homeless solutions



> On Oct 8, 2022, at 23:00, rabbit <rabbit at callistohouse.org> wrote:
> 
> …
> 
> Have a good one; keep it, light.
> Kindly,
> rabbit
> . .. … ‘…^,^
> 
> 
> Sent from Callisto House :: decentralized mobile homeless solutions
> 
> 
> 
>>> On Oct 8, 2022, at 06:57, Jakob Reschke <jakres+squeak at gmail.com> wrote:
>>> 
>> 
>> Hi Tim,
>> 
>> Are you asking whether it would be good to let WebClient automatically convert socket timeouts to a WebResponse with status 408 _for everyone_?
> 
> Well that’s what I thought he said:
> 
>> WebClient code does not try to handle socket timeouts etc. If I handle NetworkError by faking a WebResponse with a 408 code
> 
> 
> So I figure a socket timeout fakes a 408 timeout response…but…
> 
> 
>> I do not think that this is appropriate, since the server has never really sent such a message (or at least it does not sound like it has in your case).
> 
> This is a good point, no actual response just a spoof.
> 
>> How much of an impact it has depends on the application using it, whether they have error handling for NetworkError or this particular HTTP status. Hence I would not pretend to the application that there was an actual response if in fact there was none.
> 
> I like your point, no response but NetworkError ought to signal an exception, not spoof a response. Andso the Socket interface throws exceptions that must be handled by the Client using said socket. If no response yet poorly handled client-side self-generated Exception, guessing what the real issue was with no information. It just kicked rocks. Problematic when the Client really needs to talk. For a long time sockets had issues! Self-repair for resilient code helps with the presence or delivery of better details! Silent communications requires ESP, another joke, who’s laughing?
> 
> Especially if the Client was doing something wrong. The Client should be told or who knows what the Client may assume. It can exacerbate delusions & psychosis due to this info-absent trauma? Quiet termination requires advanced Qigong!
> 
> This was the selected policy, consequences be damned. Not very friendly or collaborative, challenging to a successful project application the Client may be working for.
> 
> Huh. Somebody is celebrating outside! Bull City Dreams.
> 
> 
>> In your application, you may of course choose to convert the NetworkError to a response, if you think that makes sense for the purpose of this application.
> 
> Yeah, but I kinda like the exception handler mentioned on another thread. I believe Dave said the handler should be this but I do not like the silent consumption of the NetworkError, 
> 
> Andsoi more generally I’m gonna do.do the following defaultAction. Because I come correct and keep it real, and deliver on
> my promises and always speak the truth, don’t you ever doubt that. My words hold pure power in their sincerity. The straight path. I said I was going to build a corner stone and that’s what I continue to do for what? 20 years now? Better late than never.
> 
> NetworkError<*rabbitzutility*>>>#defaultAction
> 
>     [^socket read]
>         on: NetworkError 
>         do: [:ex | ex sender notify: ‘Kick rocks for:’, 
>             ex description, ‘ on my end.’,
>            ‘ Why won’t you respond? ‘,
>            ‘A twisted sort of system-wide security policy?.‘,
>            ‘Fuck, you provide zero information. ‘,
>            ‘It’s bad you, know!! I had no warnings,’,
>            ‘I don’t do subtle innuendo, does not compute;’,
>             ‘Thinking everyone was good with ‘,
>            ‘my decision. I was under the‘,
>            ‘impression we had the time to do real engineering. ‘,
>            ‘Sockets not simple. Very hard to‘,
>            ‘understand. Due to my past trauma’,
>            ‘as a 100% SC disabled veteran, i’,
>            ‘Tend to catastrophize my delusions of persecution: thus C-PTSD. Not a good time rolling!’,
>            ‘And my psychosis's understanding of myself. ‘
>            ‘Confusion reigned for years. Everyone hated me and wanted me to go away. Now I have been transforming by the Holy Spirit and I just don’t fucking care anymore. Neither does she. I know myself now. And I’m not gonna put up with anymore bullshit from whomever is throwing it around. Mutual respect is a requirement! A head’s up would have been steller! Who failed the mission? You could have said something #notify: Nope. Look what you did. Sigh.’
>https://youtu.be/hHMtEt8YLo4’].
> 
> 
>> 
>> Kind regards,
>> Jakob
> 
> Der Anführer kontrolliert alles,
> rabbit 🐇🐇🐇🐇🐇🐇
> 
>> 
>>> Am Fr., 7. Okt. 2022 um 23:01 Uhr schrieb tim Rowledge <tim at rowledge.org>:
>>> So far as I can see the WebClient code does not try to handle socket timeouts etc. If I handle NetworkError by faking a WebResponse with a 408 code, Am I The Asshole? Will it cause the entire intertoobs to catch fire?
>>> 
>>> tim
>>> --
>>> tim Rowledge; tim at rowledge.org; http://www.rowledge.org/tim
>>> No one is listening until you make a mistake
>>> 
>>> 
>>> 
>> 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.squeakfoundation.org/pipermail/squeak-dev/attachments/20221009/27456409/attachment.html>


More information about the Squeak-dev mailing list