Design Overhaul

This thread is getting somewhat heavy :)

Just put the PHP stuff in some well-known open source CMS and focus on the content only.

My two cents worth.

I couldn't agree more. While I'm following this thread with interest, it seems the point is getting lost in the weeds. There's no need for this site to be five years worth of *custom code*! Manuel, you've said that you're "the guy" working on it so I'd expect you would be all over the idea of using an existing system with hundreds of contributors and a world of expertise out there just waiting for you.

I've seen it before where a developer works in a "bubble" for too long and they begin to believe that their way is the only way to get things done. Heck, I've been that developer before but I was able to recognize it and do something about it. Sometimes you need to put down your tools and lift your head up from your current environment and see how others are doing things.

I'll bet there's some really cool stuff going on in there and watching your web-machine tick along every day is quite satisfying. However, I too have a burned CD somewhere with some really slick code on it that will probably never again see a production environment - we all have those. It breaks my heart to let it get stale and slide out of date, I put a lot of work into it but it's time is up and I need to move on. The development world has changed in so many years and I need to keep up.

I'm not dumping on your site here but I think that you could work smarter, not harder and leverage all the work put in by the developers, testers and 3rd party contributors of Joomla, Drupal or some other CMS that suits you.

Drop your keyboard for a couple weeks and look around, the PHP community will step up and support you. I promise.

-Mark

Rest assured that I perfectly understand the reasons why you seem frustrated for not having the contest the way you think it should be.

It seems you would like this contest be like the jobs you do for your Web design clients. You propose several mockup options to increase the chances that they like at least one of your proposals and then you implement the one they like most.

Nothing wrong with that, except that I am not your customer. If there is a customer here, it should be the thousands of users of the site that care for a design change.

This contest was mostly an experiment. No other site had done it this way before. It is totally unreasonable your claim that I am reinventing the wheel. It makes you sound as if you are trying to attack and invalidate the effort that was done.

This contest was meant to address complaints of users that write by e-mail or twitter pleading for design change. The greatest problem is that most that complained were not able to specify what was wrong.

If the current design is so bad as some claim, this contest allowed them to explain (with feasible HTML, CSS and graphics) how they would make it right, all within the constraints that are important for the site, such as those that preserve the way the site revenue is generated.

I am sure some people will still complain of whatever comes out of this edition of the contest. But only when future complaints come I will be able to evaluate what exactly they are complaining then as a whole, not just what some individuals say.

Still I appreciate the time you spent persisting in your comments. I just hope you understand that it would not be a good idea to only consider your comments, especially when you are suggesting things that for me are clearly inconvenient.

This means that I did not totally discard your suggestions. I just need to evaluate what is the outcome of this contest edition before I can decide exactly how the the next edition will be.

Maybe this edition could have been different if people like you have approached sooner and presented their feedback. Unfortunately very few people bother to comment early enough.

So, you need to be patient and wait until the plans of next edition are presented in the site blog. I hope next time everybody that is concerned really speaks up in time to influence the decisions for the best.


As for the HTML editor, the security filters do not touch the submitted HTML as long as it is valid. As far as I know, all browser editors produce valid HTML. So it does not matter if there are differences in the HTML each one produces. No user will complain because the site will not touch the HTML generated by the visual editor.

The security concerns addressed by the HTML parser and filter, are addressed for HTML that is edited manually (or submitted by a malicious robots).

From what you say it seems that you may have not noticed that there is a visual mode and a manual editing mode. That was meant to provide all the flexibility power users need.

For this contest, all users are power users, so most of them will use the HTML editing mode. The visual mode is more for editing blog posts and any other types of content for the site. That is why the toolbars make it look like a word-processing editor.


As for sub-domains, it may not be clear for you but most sections require that you are a registered user of the site because the sections were made for users of the site.

For instance, the jobs section requires that both the companies and users already be registered to the site. So, I think it would be more confusing and SEO inconvenient if they would have to go to jobs.phpclasses.org or phpclassesjobs.org, instead of just www.phpclasses.org/jobs/ .

But there are few exceptions to that. For instance the PHPClasses site has a directory of PHP user groups. The people that registered the user groups are registered users of the site. But the people looking for PHP user groups may not be.

Therefore in November I bought a better domain for putting the user groups directory, so it matches better the searches done on Google for PHP user groups. Now that I am mostly done with this design contest edition, I will be able to move the user groups top listings to that domain.


As for the time the contest takes, the main problem is that it is not the only activity in the site that requires development. Besides the usual user support and moderation of the content the site publishes, other features of the site are continuously being development.

The site design system requires and will continue to require a lot of research and development time because it is mostly a brand new development for which there is no other site from which we can borrow ideas that are already working.

The next edition will not be in 5 years. If it is not possible to allow the designers to change everything they wanted in the site, it will allow at least what could be implemented in a much shorter period of time.

Anything else will be left to future editions. One step at a time. I am just not able to predict right now the size of each step.


As for feasibility of designs, I suppose you did not pay full attention to the proposals. Some designers wanted to cheat and do things that were not asked, like redesigning the site logo, use icons of non-standard sizes or even hide them.

The rules are the same to all designers in each edition of the contest. Some proposals were rejected until the designers fixed the issues.

Maybe next time the rules will be different to allow things that were not proposed right now. Or maybe there will be separate contests for the site logos, so designers will not need to cheat in order to propose a replacement for the current logo hiding the old logo using CSS tricks.

That is just one aspect of things that had to be constrained. Another is the use of Javascript, which is forbidden for security reasons. Some designers wanted to add distracting effects to pages using Javascript. That may help designers catch the eye of the voters but it is not feasible.

Another aspect is the use of designs that do not fit the current navigation bars in common user screen resolutions. That was somewhat tolerated in the first turn by me, even because there too many combinations of designs, resolutions and page types to test.

Still some users complained. In the second turn, that will be a different story, as only two designs will compete and those details will be reviewed before approving each design again.


As for some contest entries be hidden before voting, the problem is that if I were to split the contest in multiple stages, before moving to the next stage, authors would be able to see each other proposals and steal ideas, which is what most authors wanted to avoid.


As for imposing rules, the advantage of doing HTML/CSS/graphics contests instead of just mockups, is that you can verify many of the rules programmatically. Therefore you can tell immediately to the designers what is wrong with their proposal.

It would be very bad to let designers do a lot of work with their mockups and only after they submit them they are told what they did is not acceptable. I am sure it would upset the designers even more because they would have to trash what they did before and redo it within the constraints they only understood when they were rejected.

Still that gets more complicated because some designers are stubborn and will try to cheat, so they can workaround the rules (if they read and understood them at all).


Finally, I do not mind you continue to discuss these ideas. As you expected, I may just not have the time to reply to you quickly because there are always many things to be done.

Sorry, but the existing CMS are far to weak in features to be worth using them.

This site has too many custom features be worth the reuse of any CMS out there.

As I explained, this site has too many custom features to be worth using any CMS system. What the core of CMS systems provide is a small fraction of the custom features that this site implements.

Building upon existing CMS architectures would also not be a good idea for me, not because I want to develop everything, but because most of them rely on old PHP programming methods that are not as productive as the methods used to build this site.

Even though CMS systems have evolved over time and now some even use object oriented code, you still find them using old methods, like accessing databases writing SQL manually and managing database schemata by entering SQL files that are deeply dependent on MySQL.

If you are not an experienced developer, this may be a bit over your head, but since 2002, the PHPClasses uses an ORM (Object-Relational Mapping) generator tool named Metastorage that greatly speeds up your development, as you do not need to deal with the inefficiencies of mapping objects to tables in your head and avoids the need for writing error prone SQL by hand.

metastorage.net/

phpclasses.org/blog/post/14-2002-de ...

There is more to be said on why building on existing CMS would not be a good idea for me, but I think I already made my point clear.


Other than that, I do not work on a "bubble" as you suggest. The PHPClasses site itself was launched in 1999 precisely because I wanted to get feedback from many of the PHP components that I developed for my applications.

They are object-oriented and reusable. So, sharing them with other always helped me to get early feedback, bug reports, and improvement suggestions from people that try after they download them from here.

To what concerns this contest, I regret the fact that, despite a functional design proposal system was launched in September 2008, it was largely ignored and I did not get much feedback about it.

All was announced in the site blog. Actually it started much earlier. Here is a time line:

2002-07-19: Survey to ask users about benefits for premium subscribers that included configurable design themes
phpclasses.org/blog/post/10-Paid-si ...

2005-09-30: Explanation to why received proposals to redesign the site were not being accepted
phpclasses.org/blog/post/109-Site-d ...

2007-07-05: Premium services were finally launched with the plan to include customizable site themes later
phpclasses.org/blog/post/68-Launche ...

2008-07-31: Announcement of the intention to finally start a system that allows the users to propose new site designs
phpclasses.org/blog/post/80-9th-ann ...

2008-09-01: Announcement of the first public version of the site design system. Notice the request for feedback at the end of the article. Nobody commented.
phpclasses.org/blog/post/81-Site-de ...

2008-12-23: End of the year balance and announcement of the plans for the visual editor. Some people asked for a new design but still no feedback of the existing system.
phpclasses.org/blog/post/86-2008-ba ...

2009-06-24: Announcement of the design award sponsors. Finally some feedback started to appear, probably because people start taking it seriously after I mentioned the sponsors and the money prize.
phpclasses.org/blog/post/97-10th-an ...

2009-10-12: Design submission announcement and tutorial
phpclasses.org/blog/post/104-Submit ...

2009-12-01: Announcement of the delay of the design submission deadline at request of some designers. This was the original deadline but until then only 3 proposals were submitted, all in the previous days. It seems only then most contestants have woke up.
phpclasses.org/blog/post/109-Site-d ...

2009-12-08: Screencast on how to create and submit a design proposal
phpclasses.org/blog/post/110-Screen ...

2009-12-17: Start of the voting period
phpclasses.org/blog/post/111-Vote-o ...


I am sure there is plenty of room for improvement, specially based on the feedback from the users. But I wish people that were concerned did not wait until it was too late to start providing feedback. I hope next edition everybody keeps an eye opened and send comments and suggestions for improvements right away.

I do not always post on the blog about every little improvement. Therefore sometime ago I started posting in a special Twitter account about new developments as they are implemented, before they are even announced in the blog.

If you are interested to keep up with the site development, just follow the PHPClassesEarly account:

twitter.com/phpclassesearly