From: Per Jessen on
Rene Veerman wrote:

> funny how i've been topposting for over a year here and the complaint=
s
> start when i tell some people not to butt into my business and choice=

> of tools.

Rene, the only reason I mentioned it was because your language was
becoming abusive and annoying. If it hadn't, I wouldn't have mentioned=

your equally annoying top-posting. I've been around for long enough to=

know that most top-posters will eventually see the light.=20


--=20
Per Jessen, Z=C3=BCrich (9.8=C2=B0C)

From: Rene Veerman on
exactly. the knock-on problems you mentioned are well solved and well
documented.
realtime programmers using threads have to get their heads around it
on their first realtime project.

i don't like doing my code in c(++), or worse; having to interface
between c(++) and php.
i chose php because my code can stay close to simple english that way.
what you're suggesting is highly intrusive in my work-style, one that
you're not affected by at all.
in fact if you make things more difficult for me, i have less time to
release opensource code of my own.

On Wed, Mar 24, 2010 at 10:59 AM, Per Jessen <per(a)computer.org> wrote:
> Tommy Pham wrote:
>
>> When you do use AJAX, there is a slight difference in your app design
>> then when you don't use AJAX.  That's the way I see threads.
>
> A threaded design makes for a lot more than a slight difference IMHO.
> Once you've said "threading", the next words in rapid succession are:
> mutexes, semaphores, locking, spin-locks, signals, race conditions,
> atomic updates, cache coherency, asynchronous IO etcetera.  They are
> all perfectly well-known but complex concepts, and I would always
> choose C and/or assembler to work with those.
>
>
>
> --
> Per Jessen, Zürich (9.5°C)
>
>
> --
> PHP General Mailing List (http://www.php.net/)
> To unsubscribe, visit: http://www.php.net/unsub.php
>
>
From: Per Jessen on
Rene Veerman wrote:

> again:
> a) you're determining the contents of my toolset, without it affectin=
g
> you at all. the way you want it php will degrade into a toy language.=


Rene, it seems to me that you and I are advocating two opposite
positions on the topic of threading in PHP, so aren't we both trying to=

determine the contents of each others toolset?=20

> b) i will aim for all possible decreases in development time and
> operating costs during, not only in the grow phase but also in hard
> economic times. any business person knows why.

Given that the lifetime effort (=3Dcost) of any software project is
divided into 25% development and 75% maintenance, you really ought to
focus on the latter. If you want more performance at a minimal cost,
surely you should opt to write in a compiled language where you'll get
far more bang for the buck. =20


--=20
Per Jessen, Z=C3=BCrich (9.8=C2=B0C)

From: Rene Veerman on
On Wed, Mar 24, 2010 at 11:13 AM, Per Jessen <per(a)computer.org> wrote:
> Rene Veerman wrote:
>
>> again:
>> a) you're determining the contents of my toolset, without it affecting
>> you at all. the way you want it php will degrade into a toy language.
>
> Rene, it seems to me that you and I are advocating two opposite
> positions on the topic of threading in PHP, so aren't we both trying to
> determine the contents of each others toolset?
>

Per: that's EXACTLY the point.
You are determining it for me, i'm not for you.
Simply because you dont have to use the language features you atm
think you don't need in php.


>> b) i will aim for all possible decreases in development time and
>> operating costs during, not only in the grow phase but also in hard
>> economic times. any business person knows why.
>
> Given that the lifetime effort (=cost) of any software project is
> divided into 25% development and 75% maintenance, you really ought to
> focus on the latter.  If you want more performance at a minimal cost,
> surely you should opt to write in a compiled language where you'll get
> far more bang for the buck.
>
zend? facebook compiler?

i'm staying with php coz the trend is towards jit compiling bro.
From: Stuart Dallas on
Heh, you guys are funny!

On 24 Mar 2010, at 08:58, Rene Veerman wrote:

> On Wed, Mar 24, 2010 at 10:47 AM, Per Jessen <per(a)computer.org> wrote:
>> Rene Veerman wrote:
>>
>>> popular : facebook youtube etc
>>>
>>
>> Rene, I must be missing something here. That sort of size implies
>> millions in advertising revenue, so why are we discussing how much
>> performance we can squeeze out of a single box? I mean, I'm all for
>> efficient use of system resources, but if I have a semi-scalable
>> application, it's a lot easier just getting another box than trying to
>> change the implementation language. OTOH, if my design is not
>> scalable, it's probably also easier to redo it than trying to change
>> the implementation language.
>
> again:
> a) you're determining the contents of my toolset, without it affecting
> you at all. the way you want it php will degrade into a toy language.

And how exactly are you defining a toy language? If you want features like threading, why not switch to a language that already supports it?

> b) i will aim for all possible decreases in development time and
> operating costs during, not only in the grow phase but also in hard
> economic times. any business person knows why.

Yup, this is very good practice, but deciding that one particular tool is the only option is a fatal business decision. Use the right tool for the job!

What you're trying to do here is akin to taking a hammer and whittling a screwdriver in to the handle. It's ridiculously inefficient, and imo, pretty stupid.

>>> and you're still trying to impose a toolset on me.
>>
>> I didn't think I was - you're the one who seem to be fixed on PHP as the
>> only solution, and advocating that it be enhanced to suit your
>> purposes.
>
> no, php is just my toolset of choice, and i think it should grow with
> the times and support threading and shared memory.
> maybe even a few cool features to enable use-as-a-cloud.

PHP is a hammer, and a bloody good one at that, but you seem to want it to be a tool shed. Accept that it's a hammer, go visit a DIY store, find the right tool for the job and get on with your life!

The fact is that even if we all agree that PHP needs threading, and one or more people start working on putting it into the core, it will likely be many months before you see any sight of a working version, and even longer before you see a stable release.

-Stuart

--
http://stut.net/