View mode: basic / threaded / horizontal-split · Log in · Help
April 28, 2012
Designing a consistent language is *very* hard
Sorry for the noise but I think a few language designer out there 
might like this one :
http://me.veekun.com/blog/2012/04/09/php-a-fractal-of-bad-design/

I think Walter and Andrei will smile ( or cry ) reading it.
It's no trolling here, I do think this post is very valuable and 
relevant from a language design perspective.
April 28, 2012
Re: Designing a consistent language is *very* hard
Am 28.04.2012 10:18, schrieb Guillaume Chatelet:
> Sorry for the noise but I think a few language designer out there might
> like this one :
> http://me.veekun.com/blog/2012/04/09/php-a-fractal-of-bad-design/
>
> I think Walter and Andrei will smile ( or cry ) reading it.
> It's no trolling here, I do think this post is very valuable and
> relevant from a language design perspective.

Well, I cry everytime I am using it on my web site, but the effort to
port the scripts is not worth it.
April 28, 2012
Re: Designing a consistent language is *very* hard
"Guillaume Chatelet" <chatelet.guillaume@gmail.com> wrote in message 
news:vwpzirpppabcgylmvpsx@forum.dlang.org...
> Sorry for the noise but I think a few language designer out there might 
> like this one :
> http://me.veekun.com/blog/2012/04/09/php-a-fractal-of-bad-design/
>

"PHP: a fractal of bad design" is the absolute most perfect summary of PHP 
anyone could ever possibly make. It is the world's worst still-living 
non-joke language, period. Although, I genuinely cringe at calling it a 
"non-joke" language. Or "living" - PHP is like one of those botched 
homunculi from Fullmetal Alchemist, writhing around in an unholy 
half-existence. Even if you use something like Haxe-compiled-to-PHP to 
escape the language itself, you've still only escaped part of PHP's gruesome 
hell.

His sections "An Analogy" through "Core Language" are absolute spot-on 
perfection.

One of my (many) favorite quotes here is "PHP is a community of amateurs". 
So very, very true. Speaking of which, the fact that they botched the 
PHP4 -> PHP5 transition so completely, when there were obvious trivial ways 
to handle it sensibly, has always been my personal favorite evidence of 
PHP's extreme amaterishness.

I couldn't bear to get even as far as half-way through this tome, but I do 
have this to say about it: I hope I *never* end up knowing anywhere near as 
much about PHP as this guy is cursed to know.

With each passing day, I grow more and more convinced: The true measure of a 
web developer is how much they hate PHP. (Note: Not intended 
self-referentially - No narcissism implied.)
April 28, 2012
Re: Designing a consistent language is *very* hard
"Paulo Pinto" <pjmlp@progtools.org> wrote in message 
news:jngarb$14un$1@digitalmars.com...
> Am 28.04.2012 10:18, schrieb Guillaume Chatelet:
>> Sorry for the noise but I think a few language designer out there might
>> like this one :
>> http://me.veekun.com/blog/2012/04/09/php-a-fractal-of-bad-design/
>>
>> I think Walter and Andrei will smile ( or cry ) reading it.
>> It's no trolling here, I do think this post is very valuable and
>> relevant from a language design perspective.
>
> Well, I cry everytime I am using it on my web site, but the effort to
> port the scripts is not worth it.

Abandoning PHP is *always* "worth it".
April 28, 2012
Re: Designing a consistent language is *very* hard
"Nick Sabalausky" <SeeWebsiteToContactMe@semitwist.com> wrote in message 
news:jngd63$18m1$1@digitalmars.com...
>
> With each passing day, I grow more and more convinced: The true measure of 
> a web developer is how much they hate PHP. (Note: Not intended 
> self-referentially - No narcissism implied.)
>

Actually, what I really meant was: The true measure of a web developer is 
how much they hate web technologies. (PHP, of coruse, being a fairly major 
part of that.)

There does always seem to be an inverse correlation between a web 
developer's abilities and their opinion of popular web technologies.
April 28, 2012
Re: Designing a consistent language is *very* hard
Le 28/04/2012 11:27, Nick Sabalausky a écrit :
> "Guillaume Chatelet"<chatelet.guillaume@gmail.com>  wrote in message
> news:vwpzirpppabcgylmvpsx@forum.dlang.org...
>> Sorry for the noise but I think a few language designer out there might
>> like this one :
>> http://me.veekun.com/blog/2012/04/09/php-a-fractal-of-bad-design/
>>
>
> "PHP: a fractal of bad design" is the absolute most perfect summary of PHP
> anyone could ever possibly make. It is the world's worst still-living
> non-joke language, period.

Please note that Rasmus Ledorf stated many time that He « don't know how 
to design a programming language », « never intended to create a 
programing language », « don't like programing » and probably the 
funniest one « don't how to stop it » it being the success of PHP.

Back on the article, the author is wrong when stating that it isn't 
possible to run several different versions of PHP on the same server. 
This is possible, I did it several times and for big websites.
April 28, 2012
Re: Designing a consistent language is *very* hard
On Saturday, 28 April 2012 at 09:41:06 UTC, Nick Sabalausky wrote:
> "Paulo Pinto" <pjmlp@progtools.org> wrote in message
> news:jngarb$14un$1@digitalmars.com...
>> Am 28.04.2012 10:18, schrieb Guillaume Chatelet:
>>> Sorry for the noise but I think a few language designer out 
>>> there might
>>> like this one :
>>> http://me.veekun.com/blog/2012/04/09/php-a-fractal-of-bad-design/
>>>
>>> I think Walter and Andrei will smile ( or cry ) reading it.
>>> It's no trolling here, I do think this post is very valuable 
>>> and
>>> relevant from a language design perspective.
>>
>> Well, I cry everytime I am using it on my web site, but the 
>> effort to
>> port the scripts is not worth it.
>
> Abandoning PHP is *always* "worth it".

Rewriting seldom brings anything new. It almost like starting
from zero, without the benefit from all the bug fixes the
software has had along the years.

The hours wasted porting to a new language, could be used adding
features to existing code base, or refactoring.

I know how I am speaking about, many of the big bucks we get on
consultancy jobs, are from companies that decide to keep rewriting
their software on the flavour of the month framework, just because
"everyone else is doing it".

So I rather keep my "almost clean" PHP scripts running, as getting
to the trouble of switching ISP just to be able to rewrite the 
site
in a more sane language.

--
Paulo
April 28, 2012
Re: Designing a consistent language is *very* hard
On Saturday, 28 April 2012 at 11:42:19 UTC, Paulo Pinto wrote:
> Rewriting seldom brings anything new. It almost like starting
> from zero, without the benefit from all the bug fixes the
> software has had along the years.
>
> The hours wasted porting to a new language, could be used adding
> features to existing code base, or refactoring.

 There are times where a full re-write is better than keeping the 
old code. I'm rewriting my code for SmartMerger (TES3 project) 
from C to D, will probably be renamed to SmartMergeD). I'm 
getting stuck more and more adding and rewriting portions of code 
that are basically simple containers for different types; like 
associative arrays and array searching of different types.

 It is quite refreshing to see how compact and simpler the code 
is; And this is the first project I'm fully embracing OO. I can't 
wait to see how D will look as it matures and the compiler gets 
to a fully stable condition.
April 28, 2012
Re: Designing a consistent language is *very* hard
On Saturday, 28 April 2012 at 12:32:17 UTC, Era Scarecrow wrote:
> On Saturday, 28 April 2012 at 11:42:19 UTC, Paulo Pinto wrote:
>> Rewriting seldom brings anything new. It almost like starting
>> from zero, without the benefit from all the bug fixes the
>> software has had along the years.
>>
>> The hours wasted porting to a new language, could be used 
>> adding
>> features to existing code base, or refactoring.
>
>  There are times where a full re-write is better than keeping 
> the old code. I'm rewriting my code for SmartMerger (TES3 
> project) from C to D, will probably be renamed to SmartMergeD). 
> I'm getting stuck more and more adding and rewriting portions 
> of code that are basically simple containers for different 
> types; like associative arrays and array searching of different 
> types.
>
>  It is quite refreshing to see how compact and simpler the code 
> is; And this is the first project I'm fully embracing OO. I 
> can't wait to see how D will look as it matures and the 
> compiler gets to a fully stable condition.

My statement is based in a few Fortune 500 projects, where I took 
part
in the migration project. Some of those took several years to 
complete,
and the only impact on the business was dropping the revenue.

But I do concede, that there are use cases where it makes sense.

--
Paulo
April 28, 2012
Re: Designing a consistent language is *very* hard
On Sat, Apr 28, 2012 at 10:18:33AM +0200, Guillaume Chatelet wrote:
> Sorry for the noise but I think a few language designer out there
> might like this one :
> http://me.veekun.com/blog/2012/04/09/php-a-fractal-of-bad-design/
> 
> I think Walter and Andrei will smile ( or cry ) reading it.
> It's no trolling here, I do think this post is very valuable and
> relevant from a language design perspective.

I've written some simple apps in PHP... never learned enough of it to
run into all the problems listed on that page. But I *did* run into
stupidities involving ===. IMNSHO, *any* language that has === is
fundamentally, irreparably dainbramaged, and needs to be scrapped and
redesigned from scratch. This includes that hellspawn of evil incarnate
known as Javascript.

After reading this article, I'm seriously considering adopting D as my
language of choice for CGI. ;-)


T

-- 
"Computer Science is no more about computers than astronomy is about
telescopes." -- E.W. Dijkstra
« First   ‹ Prev
1 2 3
Top | Discussion index | About this forum | D home