[1.4.2] Removed the url class? Anymore unnecessary changes?

MWYS wrote:

dbstr wrote:

maslow788 wrote:Can someone tell me the underlying logic behind changing

Nope. Apparently the url class was stupid, because no other project uses a smiliar class.

Having one function for routing URL's thorugh is the logical option, What if someone wanted some custom functionality, perform some form of action on URL's before they are returned - Not possible anymore...

Not only that, but it's so much cleaner.

I agree, I was looking for a logical answer myself, but was told it was removed because it was "stupid".

The removal of it was more stupid, if you ask me..

Wait, what does url class mean? and whats the effect of it?

i decided to remove it because i'm after a framework where each component is needed. url class is not. i would only be keeping it around for the sake of backwards compatibility.

Also with the mulit shop its easier when sending mails out to have the web site address to:

$this->config->get('config_url') . 'index.php?route=order/invoice';

Just like:

HTTP_SERVER . 'index.php?route=order/invoice';


The only reason you are complaining is because you have to update your modules. stop complaining. opencart is a very good framework to build any application you want.

What about everyone who's using the modules that need to be updated after the drastic removal of the URL class? Users that's not fluent programmers, but have managed to dl or buy a module that may be crucial for their site.

You don't care about all theese sites that suddenly don't work after the upgrade?

If that's your attitude it will affect the trust very much to the project. How would someone dare to invest a lot of time and money if the groundwork is about to change without notice and without any real reason.

Why not let the URL class to remain at least one release and then remove it when developers have had time to produce functional modules?

carpelux wrote:What about everyone who's using the modules that need to be updated after the drastic removal of the URL class? Users that's not fluent programmers, but have managed to dl or buy a module that may be crucial for their site.

You don't care about all theese sites that suddenly don't work after the upgrade?

If that's your attitude it will affect the trust very much to the project. How would someone dare to invest a lot of time and money if the groundwork is about to change without notice and without any real reason.

I agree, but on the other hand you do not need to upgrade.

It is the same if you buy Windows 98 and Office 98 and then upgrade to Windows Vista to find that Office 98 is not working....

Why not let the URL class to remain at least one release and then remove it when developers have had time to produce functional modules?

You can put the url-class back, or wait until modules are updated.

Sure, You don't need to update and that's the route i have choosen.
But when people here have advised others not to upgrade there have been quite irritated comments from Daniel that there certainly is nothing wrong with the new version. If you then chose to rely on him You had been in deep shit.
My view is that if there is someone you should be able to trust, it's the one who leads the whole project. That is what I think is serious in this update circus.

i2Paq wrote:

rph wrote:I completely agree with you on the documentation. It's sparse. The wiki could use a little (lot of) community work on it.

Me to, but the WIKI is useless (more like userunfriendly) to add info to.
Otherwise I would have probably would have started already.
On the other hand, I do not see anybody else add info to it except Q.

I've added sections to it. I've paused because 1) a new version of OpenCart came out, and 2) regular users can't upload pictures.

The Url class or a central point through which all URLs are generated is actually extremely useful if you want to do SEO URLs they I wanted to do it by hooking into the Url class, but now that it's gone there's no way to do it and I need to change ALL the links back to using the Url class in order to make it work! Removing it was a bad idea, no argument. In osC there is tep_href_link which is the central function, in OpenCart it's now nothing.

Good to see you back, I had given you up as gone.

It seems Daniel has actually asked for some help. Maybe you will get a chance to discuss some of the suggestions you presented a couple of weeks ago.

Yeah, well we'll see if even one of my multiple suggestions gets taken onboard and implemented. You'd think you'd listen to someone who's website with 50,000+ products the DB ranks #1 on google in their country for quite a few very competitive terms, does over $1m in sales via their website and has a software engineering degree and hence knows a thing or two about SEO and performance, features and programmming.

Sheldmandu wrote:You'd think you'd listen to someone who's website with 50,000+ products the DB ranks #1 on google in their country for quite a few very competitive terms, does over $1m in sales via their website and has a software engineering degree and hence knows a thing or two about SEO and performance, features and programmming.

Why don't you try contributing code or suggest specific optimization suggestions? I'm glad you're successful and all but you keep mentioning this in your threads as if it matters to the project.

How can I contribute anything when the core keeps getting major changes between minor releases and there isn't a way to actually contribute anyway other than build and hope it gets included. For example, I've suggested a change to how layouts are handled (similar to 1.3.2, but without the problem that was the reason behind the change), I've even done the code, but we'll see if it makes it in.

The reason I bring that other stuff up is because it shows I'm not some newbie with great ideas, but someone who actually knows what they're doing with a proven track record, so the suggestions I'm making are not just some ideas I came up with overnight, but things that matter.

Sheldmandu wrote:You'd think you'd listen to someone who's website with 50,000+ products the DB ranks #1 on google in their country for quite a few very competitive terms, does over $1m in sales via their website and has a software engineering degree and hence knows a thing or two about SEO and performance, features and programmming.

Yea, this would be the first reason NOT to listen to you. Try the modest approach

Hell for specific items I'm ranked #1 and I only sell 100 items and make about 100k in sales. Means jack.

Sheldmandu wrote:How can I contribute anything when the core keeps getting major changes between minor releases and there isn't a way to actually contribute anyway other than build and hope it gets included.

There's places for user contributions in the forum and main site.

Yeah, version numbering is completely screwy but I don't believe we've had a major core change since 1.4.1. I believe there was one small database/functionality change to add ssl in for multi-store. Everything else has been bug fixes.

The reason I bring that other stuff up is because it shows I'm not some newbie with great ideas, but someone who actually knows what they're doing with a proven track record, so the suggestions I'm making are not just some ideas I came up with overnight, but things that matter.

Anyone can say anything on the internet. Doesn't mean much. End of the day the only thing that matters are actual contributions.

The stuff that I'm talking about is all core changes, they're not very useful as contributions as such as they wouldn't easy for your average user to install and ever for experienced users would require quite a few changes. Either way I can't exactly do the SEO URL stuff I was going to do because the Url class has now disappeared and that was the only way to hook into it.

Can't everyone put the posturing aside and work together?

It seems this is going to take some real discussion with Daniel involved from the start. Some basic functions may need to be altered - or not. But let everyone discuss their views and come out with a long term road map of the base structure and where to go from there.

Maybe set up a msn conference, or something like that.

Having a pissing contest over who does what isn't getting anywhere, working together just might.

But it is still going to take Daniel to agree to set aside some time and participate with an open mind.

Regardless of all the little things people are asking about, the back and forth on the basics is by far the main thing that has to be decided and it seems there are half a dozen or so, maybe 10 or 12 who have some viable issues to bring to the table.

This just keeps going around and around. I'm a user, I have no idea what a URL class is, let alone most of the other stuff being talked about. But I can see some folks who certainly seem to know php and what they are talking about come and go.

Guy dropped in the other day - "Been about a year since I was here. Looks like it's still a one man show. I guess I'll come back in another year and maybe it will be straightened out." Not his exact words, but close.

I've been on a keyboard daily since the 70s. Don't know code, don't want to. But I know people and have a pretty good bullshit detector. There have been some good people come, offer some suggestions, been rebuffed and moved on. Just in the couple of months I've been active here.

There's other carts out there, someone who knows what's what can easily say "I don't need this shit" and move on.

Why not harness them and take advantage of other perspectives?

So you guys basically want a cart made-to-order but without doing any actual work?

peteVA wrote:Having a pissing contest

I could totally piss further

Let me provide some useful examples why a url class is important.

1) If you like to do url rewriting, instead of http://www.yourstore.com/index.php?rout ... uct_id=387
you can use the url class to transform it into http://www.yourstore.com/index.php/prod ... ct/102/387 or http://www.yourstore.com/index.php/winter/dress

2) A url class centralizes the route that needs to have https instead of going through a bunch of files to make modification.

3) Creating permission for each route is better control through a URL class. You don't want certain customer to access certain area, you can control the permission using the URL class instead of adding it in every page where the route exists.

4) With a URL class you can do a malform URL checking. For example, a malicious hacker can do http://www.yourstore.com/index.php?rout ... id='select * from users' ... with a URL class, you can check if there's a single quote before you allow the execution of the URL.

5) URL class can help you transform misspelled or mistyped url into something reasonable. For example, http://www.yourstore.com/index.php?route=product/roduct can be transform into http://www.yourstore.com/index.php?rout ... ct/product or http://www.yourstore.com/index.php/product/product

6) You can configure apache so that a path points to index.php so that, http://www.yourstore.com/index.php/rout ... t&path=102 can simply be http://www.yourstore.com/category/dresses

The advantage of a url class is very apparent as opencart moves forward in its development. For now, you can keep using HTTP_SERVER variable but eventually, I think for code manageability, it will change for the better.

Qphoria wrote:

peteVA wrote:Having a pissing contest

I could totally piss further

That realy pissed me off