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

Qphoria wrote:

peteVA wrote:Having a pissing contest

I could totally piss further

But I can piss higher!

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

Don't know who the "you guys" you are addressing are, but I don't expect that.

I'd be willing to work, if I knew what I was doing. And would be willing to work on such things as documentation if I had some idea what was going on.

Right now I am using a V3.0.17 CubeCart with well over $100 in mods, so it's not that I'm looking for custom features for nothing.

I personally would love a working multi-shop with a built-in working affiliate package so that I could provide individualized storefronts to affiliates, but still have all payments go to my account and all email address stay in my control.

Even with that goal in mind, I think the sudden sharp turn from stabilizing V1.4.1 to making it into a multi-shop and the resulting basic changes were ill-advised. I think more of what is being done is taken as a "challenge to make it work" from a coding viewpoint than it is a logical extension of what has gone before or what is truly needed by the MAJORITY of users and prospective users. More of a "It would be fun to do this" than a mundane chore of working through a defined line of improvements.

I believe this is what is turning off so many knowledgeable people who are willing and able to help. Despite the rhetoric, it remains a one-man show and that one man is liable to go off on a tangent as the mood arises, as seen by the mid-stream veering to multi-shop.

I wish I could contribute more. Or that I knew enough to fix what I want and say screw everyone else. I know that my mods are not popular. Not many here ship from Hong Kong by 5 carriers, all with different rates and zones, all of which have to be hand built by 1/2 kg from 0 to 40kg. And while many would like to have mix or match quantity discounts for entire categories, I don't see that happening any time soon. I have both with CC and would love to have them with OC, but it won't happen this week.

I paid for both of those capabilities, and would pay again. Or help in some way with testing. But I am simply a user, one of the people who generally seem to be overlooked. Overlooked for sure by the developer.

Tell me what I can do to help (except shut up) and I'll be glad to do my part. I think OC has a great potential, but, speaking of pissing, it's being pissed away by a developer wearing blinders and some knowledgeable coders bickering amongst themselves, when if they would get together they could produce a magnificent product.

This is not a pissing contest, it's people saying hey, here are some core things that are pretty much a must if you want to be a serious shopping cart, e.g. Indexes on tables, SEO URLs, Advanced Search, Centrally defined layouts, etc... of which multi-store isn't one. Most importantly you need to have a pretty stable core that doesn't undergo major changes in minor releases.

OpenCart is very well architected and all credit goes to Daniel for that. The code is lean and doesn't use bloated third party templating, email, MVC and other libraries, which makes it easy to extend (if it didn't keep changing dramatically). In fact, had it not been a one-man-band I dare say it would have gone all over the shop like osCommerce, where there are absolutely no standards in regard to code, etc.

To me, OpenCart's overall architecture (aside from a few LITTLE gripes) is far superior to either Magento (which is bloated and slow), PrestaShop (which uses lots of 3rd party libraries and has some real ugly code, especially for the admin), osCommerce (which is old and poorly written, even for those days), other osCommerce derivatives (which take something bad and attempt to improve it without much success) and many commercial carts which are also not that well architected or not open-source. OpenCart is perfect if you're a developer and want to implement features for yourself.

However, OpenCart does lack quite a few things which are pretty much a MUST in shopping carts these days, especially if you want to run a serious online store with a large range, and that's where the focus should be, not on implementing shipping or payment modules or 20% features like multi-store. Returns and Credits is one of those functional things, besides some technical things I've already mentioned.

What's important though is making sure if it does go beyond being a 1 man band, which I can only hope for, is that the architecture doesn't get screwed up and you don't have people contributing into the core who don't know what they're doing. You don't want to end up with another ugly like osCommerce.

Sorry, I meant pissing contest in the sense of who has what credentials to bring to the table.

For a "know-nothing", I feel like I can see there are some knowledgeable people, some with similar thoughts and some with some different ideas.

All I'm saying is put aside the brag and put-down and settle on the coding.

It certainly does not have to be here on the forum. As I mentioned, msn, skype, etc. Even GoToMeeting.com if that would work. I'd even spring for the fee. It would take some coordination to suit everyone's schedule, but back and forth on here, with little more than talk as the results is not accomplishing much.

Sheldmandu wrote:You don't want to end up with another ugly like osCommerce.

The only reason osCommerce ended up where it is now is because it is a one mans show.
It started of good with a qualified developers team but they all got spooked away because the head of the project wanted to run it single handed and did not care about the comments, ideas and feelings of the others.

Even now, when a lot of people want to contribute to the v3, it still sits there doing nothing while other compatitor carts take up the space osCommerce left in the 6 years it did not move forward.

It could be that you see some simularities atm. but my hope is that there are none in the end.....

peteVA wrote:Tell me what I can do to help (except shut up) and I'll be glad to do my part.

Donate? Wiki documentation? User support? Feature request white paper? Reporting specific bugs?

You and Sheldmandu want to help, that's really and truly a great thing. But it seems all you two do is talk about contributing. It's always just some vague complaint about core function, or bugs, or the niche feature Daniel wasted time developing while ignoring the niche feature you want.

i get a lot of requests about helping out with opencart but they all want to work on the core code.

just email me some docmentation or a video tutorial or something.

If there was some better ways to "hook" into the core without editing files, and some stability in the core functions and classes that these mods would depend on so they didn't change for each point release, I think you'd find less whining and much more applaud.

Even for functions and classes that you want to change, you need to at least give some warning of deprecation so that people have time to plan their changes gradually. Or if you are going to rewrite/remove classes then there should be a bigger number change.

Even if you want to stick with your current numbering scheme, it could be something like
Framework.Major.Minor
1 = Framework change
3 = Major changes including functions/classes
0 = Minor changes, bug fixes, added new features without breaking old. no existing class/function changes

Then the cycle would look like:
1.3.0 - baseline
1.3.1 - added customer groups. No breaking of any other functions (MINOR)
1.3.2 - few bug fixes. No breaking of any other functions (MINOR)
1.3.3 - changed the template system. changed the getAddress() library function. should be 1.4.0 (MAJOR)
1.3.4 - few bug fixes. No breaking of any other functions. Should be 1.4.1 (MINOR)
1.4.0 - changed the template, changed "Router" to "Action" (unnecessarily). destroyed language class. Should be 1.5.0 (MAJOR)
1.4.1 - removed url class, removed helper image, changed the way order data was returned in the admin. Added major new shop function. Should be 1.6.0 (MAJOR)
1.4.1 through 1.4.4 - few bug fixes, no major changes or breaks. Should be 1.6.4 (MINOR)

Then we could at least count on mods working for full "MAJOR" versions. Still not ideal, but better than the current MINOR support

There is nothing wrong with people wanting to work on the core code and it should be encouraged, but you just need to make sure quality is controlled. The way to achieve that is by not allowing anyone but yourself and perhaps other 1-3 main developers to actually check into the repository, everyone else would just submit patches, which would be reviewed first. But for that to happen you need to setup and use SVN properly, not as one pleases. I would also recommend on standardising on the development environment as much as possible (e.g. Eclipse PDT + Aptana + Subclipse). This will increase productivity and make it easier for people to help each other. At least everyone in the core team should be using a proper PHP IDE with good SVN support. Then there should also be a bug tracker setup to manage to resolution process. I'd recommend using Redmine as I dare say it's the best one out there. It's not all that difficult to get a proper process up and running if you know what you're doing.

Here is my two cents. Daniel setup this up in GIT in GITHUB. That way you can control merging fixes and updates to the core of OC.

Daniel wrote:i get a lot of requests about helping out with opencart but they all want to work on the core code.

just email me some docmentation or a video tutorial or something.

Well yeah that would be grate then we would need to update the documentation or video every 2weeks when changes are made it would be hard to keep up with the changes, epically with videos.

It doesn't make much sense to use Git, because there aren't many good plugins for Git. It's far better to use SVN as it's far more mature from the tools and integration perspective and I dare say most people use Windows.

Sheldmandu wrote:It doesn't make much sense to use Git, because there aren't many good plugins for Git. It's far better to use SVN as it's far more mature from the tools and integration perspective and I dare say most people use Windows.

git works under windows very well and decentralized version system has way more pros than centralised svn

There isn't much point getting into a discussion on Git vs SVN here. SVN is more widely used and has much better tools. Even hardcore Git lovers have said that if you want something with good tools go for SVN, because of the likes of TortoiseSVN and good SVN plug-ins for just about every IDE. In this case I'd be looking for plug-ins available out of the box in Eclipse and NetBeans, and both have better SVN support than Git. The centralised vs decentralised is like the Linux vs Windows or Commercial vs OpenSource debate... Both have their pros and cons and you can find valid arguments for either. When I had to make the decision of Git vs SVN after doing significant research it all came down to availability of tools and integration and SVN was far better and it was far more widely used.

Really its all moot. The code sharing tool isn't the problem. SVN works fine. It's just that google code website itself is ugly. But Daniel isn't allowing code updates anyway and the main site isn't hosted at google code so its all a big waste of an argument.

Just another contest, really.

Mine's bigger, better, whatever than yours.

Should those who wish to be involved ever "get together" in some fashion, certainly agreeing on common tools may help things along. But until there is some form of leadership / distribution of efforts, etc. no sense talking whose favorite is better than someone else's favorite.

bmaynard wrote: The only reason on why I could see you do this is to:
1. Piss me off

This is my guess

bmaynard wrote:<snip>

What patch did you have?

Ok if you are going to want a url class then please define what attributes and methods it should have so that it is useful.

Daniel wrote:Ok if you are going to want a url class then please define what attributes and methods it should have so that it is useful.

Hi Daniel, thanks for taking the time to post, im a front end designer / developer by trade who has a very keen interest in SEO both on and off site and although I dont under stand the ins and outs of the URL class, from my limited understanding it would greatly aid the SEO of open cart and for me thats a major benefit to any ecommerce system.

If im right with this assumption I would love to hear from Sheldmandu who suggests he knows what hes doing with SEO specificaly within the opencart/ecommerce arena, I would also like to see opencart secured from any potential attack and risk, and I suspect from speaking to bmaynard that he could really give some pointers for consideration in his area of expertise.

Overall we all agree there are many users here who see open cart only from the user perspective, those who see it from a comercial prospective and those of whome see it from a code and develoment perspective, what we need to do however is take the best from all areas that make up the experience of opencart and develop a stable release and a structured road map for opencarts future.

Geoff.