Recently I published a new PHP MVC Framework. It is ASP.NET like,
object oriented, event driven, component based, with AJAX support,
I18N support and is highly customizable. It also has very good
documentation and plenty of examples.
Any opinions on it are welcomed.
http://qphp.net
> http://qphp.net
Not usable in Konqueror.
> Not usable in Konqueror.
To be more precise (and more constructive):
None of the http headers state that you use Xhtml.
Furthermore, the example pages say that they are xhtml transitional (in
an xml attribute), but do not even consistently quote attributes (like
href=# instead of href="#"). Firefox does try to make sense out of it,
but frankly it shouldn't.
Are those IDs in the selectbox real database IDs? I would never use a
framework that does that. Sorry.
In short, it looks very much like ASP.NET: you should trust the
framework to generate proper HTML for you (bad idea with MS-based
products) and therefore it is written mainly for IE. You even seem to be
relying on browsers sniffing the content-type, which is a really bad idea.
There are zillions of frameworks out there. Yours has probably brought
you some valuable insights, but I do not think it really adds something
to the amount we already have. But it probably is a good start. Good to
see that you thought of providing documentation.
Good luck!
Thanks for the comments. Konqueror has never been tested as it has
very small market share, less than 0.1%. Browsers that have been
tested are:
- Internet Explorer 6, 7
- Firefox 1.0, 1.5, 2.0
- Mozilla 1.0, 1.5, 2.0
- Opera 7.50, 8.0, 8.50, 9.0, 9.20
- Safari 3.0.4
- Avant Browser 10.2, 11.5
- Maxthon 1.5, 1.6, 2.0
- SeaMonkey 1.1.7
Although there are zillions of frameworks out there, there are very
few worth mentioning. I hope to qualify in Top 10, although may be too
late. ASP.NET-like frameworks are also very few, and there are pros
and cons to offer such a thing to the LAMP community. Such frameworks
really simplify the development and the maintenance, but the community
may dislike everything that comes from MS even as ideas.
Although QPHP.NET has 3 years of internal experience and is very well
tested, I published it as version 0.9 and hope by the time reaching
1.0 to become pretty good framework not only for our commercial
projects, but for the PHP community as well. Basically with this post
I am trying to figure out what features are considered valuable and to
put it in the next versions.
Full XHTML support will be added in the next version, Yahoo and Google
sites are still not XHTML based and I guess there are reasons for
that, so this feature won't bring too much value to the framework at
this time.
alexander.petrov@abv.bg schrieb:
> Thanks for the comments. Konqueror has never been tested as it has
> very small market share, less than 0.1%. Browsers that have been
> tested are:
> - Internet Explorer 6, 7
> - Firefox 1.0, 1.5, 2.0
> - Mozilla 1.0, 1.5, 2.0
> - Opera 7.50, 8.0, 8.50, 9.0, 9.20
> - Safari 3.0.4
> - Avant Browser 10.2, 11.5
> - Maxthon 1.5, 1.6, 2.0
> - SeaMonkey 1.1.7
Browser testing is something you do AFTER this
http://validator.w3.org/check?uri=http%3A%2F%2Fqphp.net%2F&c harset=%28detect+automatically%29&doctype=Inline&group=0
checked and clean, not instead.
Hello,
alexander.petrov@abv.bg wrote:
> Thanks for the comments. Konqueror has never been tested as it has
> very small market share, less than 0.1%. Browsers that have been
> tested are:
> - Internet Explorer 6, 7
> - Firefox 1.0, 1.5, 2.0
> - Mozilla 1.0, 1.5, 2.0
> - Opera 7.50, 8.0, 8.50, 9.0, 9.20
> - Safari 3.0.4
> - Avant Browser 10.2, 11.5
> - Maxthon 1.5, 1.6, 2.0
> - SeaMonkey 1.1.7
>
As stated before, first make sure your code is valid, then take into
account changes needed by the various web browsers.
>
> Although there are zillions of frameworks out there, there are very
> few worth mentioning. I hope to qualify in Top 10, although may be too
> late. ASP.NET-like frameworks are also very few, and there are pros
> and cons to offer such a thing to the LAMP community. Such frameworks
> really simplify the development and the maintenance, but the community
> may dislike everything that comes from MS even as ideas.
>
> Although QPHP.NET has 3 years of internal experience and is very well
> tested, I published it as version 0.9 and hope by the time reaching
> 1.0 to become pretty good framework not only for our commercial
> projects, but for the PHP community as well. Basically with this post
> I am trying to figure out what features are considered valuable and to
> put it in the next versions.
>
> Full XHTML support will be added in the next version, Yahoo and Google
> sites are still not XHTML based and I guess there are reasons for
> that, so this feature won't bring too much value to the framework at
> this time.
Most "large" websites - such as Google and Yahoo - don't really care
about web standards. Their purpose is to provide a nice looking and
easily maintainable website. Writing valid XHTML and CSS that works on
every browser is time consuming and that is one thing most companies
don't have or don't want to spend on a website. As long as it works,
it's OK.
One of my previous jobs was with a web development company and they had
their own internal developed framework created before XHTML was
available. Upgrading took too much time and internal work does not pay
the bills. Today they are still not producing XHTML valid websites.
If you want to provide a good framework you must realize that most users
will be amateur or professional web designers and developers who do want
XHTML and CSS valid websites because it makes their portfolio look so
much nicer. The are working independently and don't have to worry to
much about the time they have to complete a project.
From my experience and corresponding with other web developers I
noticed most of them use either custom build frameworks or things like
Drupal or Textpattern because they are great frameworks that provide
quality code.
- Jensen
On Sat, 19 Jan 2008 17:51:35 +0100, Jensen Somers
> Hello,
>
> alexander.petrov@abv.bg wrote:
>> Thanks for the comments. Konqueror has never been tested as it has
>> very small market share, less than 0.1%. Browsers that have been
>> tested are:
>> - Internet Explorer 6, 7
>> - Firefox 1.0, 1.5, 2.0
>> - Mozilla 1.0, 1.5, 2.0
>> - Opera 7.50, 8.0, 8.50, 9.0, 9.20
>> - Safari 3.0.4
>> - Avant Browser 10.2, 11.5
>> - Maxthon 1.5, 1.6, 2.0
>> - SeaMonkey 1.1.7
>>
>
> As stated before, first make sure your code is valid, then take into
> account changes needed by the various web browsers.
>
>> Although there are zillions of frameworks out there, there are very
>> few worth mentioning. I hope to qualify in Top 10, although may be too
>> late. ASP.NET-like frameworks are also very few, and there are pros
>> and cons to offer such a thing to the LAMP community. Such frameworks
>> really simplify the development and the maintenance, but the community
>> may dislike everything that comes from MS even as ideas.
>> Although QPHP.NET has 3 years of internal experience and is very well
>> tested, I published it as version 0.9 and hope by the time reaching
>> 1.0 to become pretty good framework not only for our commercial
>> projects, but for the PHP community as well. Basically with this post
>> I am trying to figure out what features are considered valuable and to
>> put it in the next versions.
>> Full XHTML support will be added in the next version, Yahoo and Google
>> sites are still not XHTML based and I guess there are reasons for
>> that, so this feature won't bring too much value to the framework at
>> this time.
>
> Most "large" websites - such as Google and Yahoo - don't really care
> about web standards. Their purpose is to provide a nice looking and
> easily maintainable website. Writing valid XHTML and CSS that works on
> every browser is time consuming and that is one thing most companies
> don't have or don't want to spend on a website. As long as it works,
> it's OK.
> One of my previous jobs was with a web development company and they had
> their own internal developed framework created before XHTML was
> available. Upgrading took too much time and internal work does not pay
> the bills. Today they are still not producing XHTML valid websites.
>
> If you want to provide a good framework you must realize that most users
> will be amateur or professional web designers and developers who do want
> XHTML and CSS valid websites because it makes their portfolio look so
> much nicer. The are working independently and don't have to worry to
> much about the time they have to complete a project.
> From my experience and corresponding with other web developers I
> noticed most of them use either custom build frameworks or things like
> Drupal or Textpattern because they are great frameworks that provide
> quality code.
While the discussion moves that much towards HTML it's getting offtopic:
XHTML is NOT the standard anyone must follow. HTML4.01 Strict is very OK
too, and a lot easier then _real_ XHTML (not XHTML served as tag soup).
XHTML has its specific uses and advantages, of which almost none you can
really use in production due to lack of UA support. The XHTML hype is long
over, people have calmed down, and most serious developers have gone back
to HTML4.01 Strict.
However, a flexible framework should still support XHTML (and output valid
code) if the user for some reason desires that.
--
Rik Wasmus
On Jan 19, 4:43 am, alexander.pet...@abv.bg wrote:
> Recently I published a new PHP MVC Framework. It is ASP.NET like,
> object oriented, event driven, component based, with AJAX support,
> I18N support and is highly customizable. It also has very good
> documentation and plenty of examples.
>
> Any opinions on it are welcomed.
>
> http://qphp.net
What type of feedback are you looking for - just the output or the
actual structure of the PHP programming, etc.
For example, if you're looking for the PHP stuff - I'd suggest getting
away from using global variables and check into the registry
programming pattern... thats just IMHO.
-aaron
I am looking for general feedback, someone that can compare how it
stands next to the big 3 on the market. Things as you wrote about
global variables, etc. So far it is a good feedback. I really do not
know why people dislike global vars, as they are something different
from register_globals and are not security hole, but I will change it
to register pattern as there is pretty much people against them.
news:a9157d54-6cc5-4566-a1bc-13013f115886@j78g2000hsd.google groups.com...
>I am looking for general feedback, someone that can compare how it
> stands next to the big 3 on the market. Things as you wrote about
> global variables, etc. So far it is a good feedback. I really do not
> know why people dislike global vars, as they are something different
> from register_globals and are not security hole, but I will change it
> to register pattern as there is pretty much people against them.
You don't know why global variables are bad? LMGDAO
"asdf"
>
>
>>I am looking for general feedback, someone that can compare how it
>> stands next to the big 3 on the market. Things as you wrote about
>> global variables, etc. So far it is a good feedback. I really do not
>> know why people dislike global vars, as they are something different
>> from register_globals and are not security hole, but I will change it
>> to register pattern as there is pretty much people against them.
>
>You don't know why global variables are bad? LMGDAO
That's a bit harsh. The dangers of global variables are not intuitively
obvious; I have no doubt that you were not aware of the dangers until
someone told you.
--
Tim Roberts, timr@probo.com
Providenza & Boekelheide, Inc.
Well, I think it is more prejudice than real danger. If you read more
carefully my last post I am barely mentioning 2 types of "global"
variables.
1. This is the danger one. It can be used if register_globals = On. Of
course I will never turn it On. The security hole is well known.
2. If we define "global ConfigSettings" this way we can have it
accessible from everywhere and it is very convenient way to use it
without passing it as parameter to every function. I do not see any
danger here. I have read articles like this one
http://blog.case.edu/gps10/2006/07/22/why_global_variables_i n_php_is_bad_programming_practice
and reasons like:
- The downside to using global configuration variables is that they
can be modified from anywhere in the program
- All it takes is one unknowing programmer...
- The reason is that dependence on global object instances limits the
flexibility of your program
are not serious for me. I already did a workbench using global and
registry pattern approach and the later is a bit slower, so I would be
thankful if one can provide me with any info that brings the real
downside.
alexander.petrov@abv.bg wrote:
> Well, I think it is more prejudice than real danger. If you read more
> carefully my last post I am barely mentioning 2 types of "global"
> variables.
>
You've never had to try to figure out where in 500K LOC a global
variable got changed... Yes, it is a real danger.
> 1. This is the danger one. It can be used if register_globals = On. Of
> course I will never turn it On. The security hole is well known.
Which has nothing to do with global variables, despite the name.
> 2. If we define "global ConfigSettings" this way we can have it
> accessible from everywhere and it is very convenient way to use it
> without passing it as parameter to every function. I do not see any
> danger here. I have read articles like this one
> http://blog.case.edu/gps10/2006/07/22/why_global_variables_i n_php_is_bad_programming_practice
>
And it can be set to an invalid value from anywhere. It's a huge problem.
> and reasons like:
> - The downside to using global configuration variables is that they
> can be modified from anywhere in the program
> - All it takes is one unknowing programmer...
> - The reason is that dependence on global object instances limits the
> flexibility of your program
>
> are not serious for me. I already did a workbench using global and
> registry pattern approach and the later is a bit slower, so I would be
> thankful if one can provide me with any info that brings the real
> downside.
>
Not until you have that bug. Or someone else has to work on your code.
--
==================
Remove the "x" from my email address
Jerry Stuckle
JDS Computer Training Corp.
jstucklex@attglobal.net
==================
On Jan 19, 2:43 am, alexander.pet...@abv.bg wrote:
>
> Recently I published a new PHP MVC Framework.
It would be great if you could explain why you think your framework is
better than what's already out there. So far, the only selling point
I can see is that it's more understandable to ASP.Net types, but are
there enough of those among PHP developers to justify the exercise?
As to the supposed "elegance" of Java and C# and "bizzare" (your
spelling) nature of Perl, this is largely a matter of opinion. One
only needs to try to retrieve a remote Web page (which in .Net
requires half-page of code spawning three objects) to wonder where the
bizarre truly lies...
> Any opinions on it are welcomed.
Applications written with your framework won't scale beyond a single
DB server, because your code as written is unaware of the possibility
of DB server replication. Your TApplication class has only one Db
property, while scalable applications need two DB server connections;
reading can be done from any server in a replicated setup, while all
writing must be done by the master server. Granted, the vast majority
of applications out there will never scale beyond a single DB server,
but you wanted an opinion, so heree it is... :)
Cheers,
NC
@NC
QPHP was built 3 years ago when there was no such amount of existing
frameworks. A couple of successful commercial projects has been
created on top of it. I did create it as architecture and help for the
implementation from other guys. Now from the point of project manager
I can tell you what is the most valuable thing of a framework - it
creates a frame and brings discipline to the development process. The
simpler it is, the better it is. Everything else are just pluggable
libraries.
Now from a private framework I am bringing it to a public one. Whole
new different story. I am collecting feedback from a couple of sources
and can tell that it is a hard thing to satisfy everyone. Everyone
suggests, well, you know, it is good, but it does not have ... what my/
our personal problems are!
I have been in the software industry for 10 years and can see how
people associate themselves with different ideas. The linux folks will
never share the values of the MS folks and vice versa. An ASP.NET/VB/
Delphi guy will be horrified to see how in most of the popular
frameworks people still parse raw request to see if a button has been
clicked and a linux guy will be horrified to hear something working on
Win/IIS platform. I did check the forum PHP software a month ago and
was scared how we still have "nice" mixture of SQL statements and HTML
tags in the most popular solutions.
So comparing the architecture style there are not too much solutions
on the market, beside Prado I even do not know the rest if at all.
And if you want second or third DB connection it is a simple thing,
just ask in the forum. Retrieving remote page is a very simple thing
also, there is an example.