Path: ...!2.eu.feeder.erje.net!feeder.erje.net!eternal-september.org!feeder3.eternal-september.org!news.eternal-september.org!.POSTED!not-for-mail
From: "Dmitry A. Kazakov" <mailbox@dmitry-kazakov.de>
Newsgroups: comp.unix.shell,comp.unix.programmer,comp.lang.misc
Subject: Re: Python (was Re: I did not inhale)
Date: Sun, 18 Aug 2024 12:19:10 +0200
Organization: A noiseless patient Spider
Lines: 50
Message-ID: <v9shqt$2bn73$1@dont-email.me>
References: <uu54la$3su5b$6@dont-email.me> <20240408075547.000061e8@gmail.com>
 <g52cnWOOwoz_son7nZ2dnZfqnPWdnZ2d@brightview.co.uk>
 <uvbe3m$2cun7$1@dont-email.me> <uvbfii$3mom0$1@news.xmission.com>
 <20240412094809.811@kylheku.com> <87il0mm94y.fsf@tudado.org>
 <way-20240413091747@ram.dialup.fu-berlin.de> <87il0lldf8.fsf@tudado.org>
 <choices-20240413123957@ram.dialup.fu-berlin.de>
 <v9lm2k$12qhv$1@dont-email.me> <v9m4gd$14scu$1@dont-email.me>
 <20240815182717.189@kylheku.com> <v9npls$1fjus$1@dont-email.me>
 <v9posc$1rpdj$1@dont-email.me> <v9pvoo$1sn55$1@dont-email.me>
 <v9r60h$2289h$2@dont-email.me> <v9sa91$2afht$1@dont-email.me>
 <v9sbf4$2artq$1@dont-email.me>
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit
Injection-Date: Sun, 18 Aug 2024 12:19:10 +0200 (CEST)
Injection-Info: dont-email.me; posting-host="03e87f546f2e8f955e712fe5fed38868";
	logging-data="2481379"; mail-complaints-to="abuse@eternal-september.org";	posting-account="U2FsdGVkX18mWpsLMIWhLhSrsSlzP7ZkH7orQOKUwPI="
User-Agent: Mozilla Thunderbird
Cancel-Lock: sha1:V4/STZwAJCEohUHFXoB4873NAEk=
In-Reply-To: <v9sbf4$2artq$1@dont-email.me>
Content-Language: en-US
Bytes: 3723

On 2024-08-18 10:30, Muttley@dastardlyhq.com wrote:

> The windows process API is crippled which is why threading is the main
> parallel processing method in Windows and has been since 3.0.

Threading is the main method because of its performance in a tight 
coupled application. Crossing the process borders is very expensive.

> If you think
> otherwise please give some examples. However I distincly remember trying to
> port  server to windows only to discover that sockets are some wierd
> structure in windows instead of a simple file descriptor meaning it was
> impossible to multiplex using poll() or select() on them, you had to have
> a seperate thread to manage each connection. What a fucking mess designed
> by idiots.

There is no problem using socket select with Windows. Here is an 
implementation of a server doing that:

    http://www.dmitry-kazakov.de/ada/components.htm#17.1

You can check the Windows implementation of to learn how it is done with 
Windows sockets.

>> Historically UNIX started with singe fork() call. There was nothing but
>> files, no system objects, nothing.
> 
> fork() copies the entire process space. No idea what you mean by "system
> objects". Must be some windows thing.

Objects managed by the OS.

>> On the other hand, Windows NT was developed by people influenced with
>> the VMS design. VMS had a very elaborated process communication API. So
>> Windows inherited much of that, but must importantly the concept of
>> system resources managed by the OS.
> 
> Windows NT was a slow unreliable dog.

It is not slow. Actually you can get same or better performance under 
Windows as under Linux. The problem is the default settings of the 
scheduling time interval. It is 10ms by default. Set it to 1ms and you 
will see the difference. Unfortunately 1ms is the hard limit. If you 
want it better, OK, there is VxWorks where you can go under 1ms.

-- 
Regards,
Dmitry A. Kazakov
http://www.dmitry-kazakov.de