Everything
php
Learn Community Releases The PHP Project
Home » Project » Newsgroups » php.internals » Re: BUG #18630 »

Re: BUG #18630

php.internals

Uwe Schindler

22 years ago
You can also check vor availability of this single function with GetProcAddress(GetModuleHandle("kernel32.dll"),"funcname") and use the pointer from there if not NULL. If this is the only unknown function in Win95/NT4 then we should not boot out all people because of a single function. That construct would do the same. At 21:48 30.10.2003, you wrote:
>At 09:10 PM 10/30/2003 +0100, Marcus Börger wrote: >> >>"Andi Gutmans" <andi@zend.com> wrote in message >> >>news:5.1.0.14.2.20031029233328.032153e8@127.0.0.1... >> >>> Argh, this only works on Windows 2000 and later or Windows 98 and later. >> >>> I guess I'll need to #ifdef it somehow. Anyway, I'd still like you >> to check >> >>> it as I assume you're not using Windows 95 or NT 4 :) >> >>Even microsoft has stopped support for those, so why should we care? > > >Someone claimed that in the far east Win95 and NT 4 are still popular. >I would like to have a discussion here and decide on this issue. We are >only talking about the support of our binary distribution. I think I can >make the source compatible using #ifdef. > >Andi > >-- >PHP Internals - PHP Runtime Development Mailing List >To unsubscribe, visit: http://www.php.net/unsub.php
----- Uwe Schindler thetaphi@php.net - http://www.php.net NSAPI SAPI developer Erlangen, Germany

Andi Gutmans

22 years ago
At 11:01 PM 10/30/2003 +0100, Uwe Schindler wrote:
>You can also check vor availability of this single function with >GetProcAddress(GetModuleHandle("kernel32.dll"),"funcname") and use the >pointer from there if not NULL. If this is the only unknown function in >Win95/NT4 then we should not boot out all people because of a single >function. That construct would do the same.
Yeah, I also thought of that but I don't really want to do run-time checks. It is a possibility though. Andi

Ilia A.

22 years ago
On October 30, 2003 05:48 pm, Andi Gutmans wrote:
> Yeah, I also thought of that but I don't really want to do run-time checks. > It is a possibility though.
If it is done once during startup, the 'harm' should be minimal and in vast majority of cases performance on win32 is not really that much of an issue. In persistent environment it won't even be felt. On the other hand do we really wish to continue supporting an operating system that has been abandoned by it's developer? Ilia

Andi Gutmans

22 years ago
At 11:07 PM 10/30/2003 -0400, Ilia Alshanetsky wrote:
>On October 30, 2003 05:48 pm, Andi Gutmans wrote: > > Yeah, I also thought of that but I don't really want to do run-time checks. > > It is a possibility though. > >If it is done once during startup, the 'harm' should be minimal and in vast >majority of cases performance on win32 is not really that much of an issue. >In persistent environment it won't even be felt.
Yeah, we'd obviously do it once and save a flag. I just meant it's ugly but I can live with it.
>On the other hand do we really wish to continue supporting an operating >system >that has been abandoned by it's developer?
Its customers haven't abandoned it because its developers want muchos $$$ for upgrades :) Andi

Uwe Schindler

22 years ago
At 10:27 31.10.2003, Andi Gutmans wrote:
>At 11:07 PM 10/30/2003 -0400, Ilia Alshanetsky wrote: >>On October 30, 2003 05:48 pm, Andi Gutmans wrote: >> > Yeah, I also thought of that but I don't really want to do run-time >> checks. >> > It is a possibility though. >> >>If it is done once during startup, the 'harm' should be minimal and in vast >>majority of cases performance on win32 is not really that much of an issue. >>In persistent environment it won't even be felt. > >Yeah, we'd obviously do it once and save a flag. I just meant it's ugly >but I can live with it.
Some runtime checks for "undocumented" functions in SunONE webserver are also in /sapi/nsapi.c at module startup. The function pointer is then in a global variable: http://lxr.php.net/source/php-src/sapi/nsapi/nsapi.c#218
>>On the other hand do we really wish to continue supporting an operating >>system >>that has been abandoned by it's developer? > >Its customers haven't abandoned it because its developers want muchos $$$ >for upgrades :) > >Andi >
----- Uwe Schindler thetaphi@php.net - http://www.php.net NSAPI SAPI developer Erlangen, Germany

Matthias Nothhaft

22 years ago
Hi List, I noticed that file_exists(""); // or null instead of "" under my Windows machine with php 4.2.2 returns true!? My Debian machine with php 4.3.3 works fine and returns false! Am I alone with that problem or is it a bug? Regards, Matthias

Thomas Seifert

22 years ago
On Fri, 31 Oct 2003 23:35:08 +0100 php@mahono.de (Matthias Nothhaft) wrote:
> Hi List, > > I noticed that > > file_exists(""); // or null instead of "" > > under my Windows machine > with php 4.2.2 returns true!? > > My Debian machine with php 4.3.3 works fine > and returns false! > > Am I alone with that problem or is it a bug? > > Regards, > Matthias
Hmm, I would say "no file" will always exist. Therefore I would call that behaviour undefined. thomas

Steph

22 years ago
I just started a post-grad course. The majority of the network I deal with here day-in-day-out is NT4.0 and yeah that makes me cringe too. But it has to be asked, if this is the case in a UK uni network then why assume anything? The guys here are also committed to using opensource, and PHP is taught on my course... *nix runs the stuff behind it all, in every case. (The usual teccie/management divide.) So right, in the Far East NT 4 is 'still popular'? :I get real, we're 'advanced'!