Made of Everything You're Not

Avactis is another in a long line of e-commerce web applications written in php (similar to OpenCart and PrestaCart), this one a little different in that Avactis has different versions, each with different features, depending on how much you’re willing to spend. Avactis is a full featured product with all the bells and whistles any ambitious store would need (and then some more features stacked on top for good measure) combined with a very php like integration methodology. Unfortunately, the dated administration interface combined with a lack of a plugin architecture and theme community really holds it back from the awesome bar.

Avactis

For the uninitiated, Avactis is created and maintained by Pentasoft Corp; oddly, there’s no website for the parent company so take that for what you will. As mentioned, Avactis is based on a pay model though it’s way more upfront about it than PrestaCart and, while the free version is missing some features, the source is available for modifications. Not too bad in my opinion; at least the more cash strapped shops can still play if they want to.

The different versions of Avactis are Free, Owned ($199), Monthly Leased ($19.95 a month) and White Label ($299) each with their own features and options. Most notably the Free version doesn’t include any (useful) payment modules (Authorize.net anyone?), coupon module, data export and import, quantity discounts or search engine friendly URLs out of the box. They do offer discounts for web developers though (at least 50% and they say up to 100%), and the complete source is available without obfuscation, so at least those functions that are needed can be added which can really ease the pain of paying for the thing in my opinion.

As expected, Avactis has the ability for custom themes though I’m disappointed to see that there doesn’t appear to be any theme community in existence (compared to other cart software packages). The reasoning behind this is probably because Avactis bills itself as being “easy integration with an existing website thanks to unique Avactis tag technology”. And by unique they mean including a php file and calling php function snippets. I’m all for marketing hype but wtf is that?!? Here’s an example:

<?php include('init.php'); ?>
<?php NavigationBar(); ?>
<?php Breadcrumb(); ?>
<?php ProductList(); ?>

It’s called php and EVERY php site does this Avactis not just you. Essentially though, this isn’t a bad strategy and, in fact, is definitely a strength especially without all the hyperbole. Obviously, Avactis can stand alone and doesn’t need to be integrated into a separate site but it’s a good idea to allow easy integration into existing sites.

Avactis PHP shopping cart software

Avactis is packed with all the features any online store would want; content management system, coupon and discount mechanisms, order and customer management, packing slip builder to name a few. Each feature is also highly customizable and usually includes a plethora of options and settings; nice if you build complex product build outs or specific functionality. On the other hand though, this amount of features and customization comes at a cost in terms of ease of use, work flow and a lacking user friendly experience.

The administration interface for Avactis is a nightmare mess of pop-up windows, tabs and accordion widgets. Slick is not a word I would use to describe the experience. Functional or crappy or painful or eyebleedingworstinterfaceeverpleasekillmefortheloveofgod!; those are better words to describe it.

There’s inconsistencies all over the place; for example while editing a product the help widgets will open another pop-up window (sigh…) yet in the main menu hovering over a link will display a tool tip and in the installation process the help widgets are all inline divs.

Avactis Admin Popups

Going through the code yields such codesod qualifiers as the below:

<?php
    /**
     * Defines the possibility of uploading images by file type.
     *
     * @param $file The array consists of the $_FILES variable, for
     * the current file.
     * @return boolean
     */
    function isAllowedImageType($file)
    {
/*        $type = _ml_strtolower($file['type']);
        switch ($type)
        {
            case 'image/gif':
            case 'image/jpeg':
            case 'image/jpg':
            case 'image/jpe':
            case 'image/jfif':
            case 'image/pjpeg':
            case 'image/pjp':
            case 'image/png':
            case 'image/x-png':
                return true;
            default:
                return false;
        }
*/
        return true;
    }
?>

If that doesn’t make any sense to you suffice it to say that the above function is supposed to verify that an image’s mime type matches the list; unfortunately though, the function is, what we call “commented out” and will not be executed. All files sent to the function will validate as true so, essentially, any file type can be uploaded. While it’s entirely possible this is an old function that was replaced with something useful, and it should be noted that I never found any calls to that function (but I really didn’t look too hard), the fact that it’s still in the code-base speaks volumes, to me, about the project maintenance at the very least.

On top of that Avactis has one of the most wasteful and useless installation processes I’ve ever seen. Initially, Avactis ships with a very minimal file set that includes a 15 mega byte (MB) file whose sole purpose is to contain all additional files in a gzipped and base64 encoded string. The sole purpose of this is to allow Avactis to programmatically write all source files to the file system during installation. For the life of me I can’t imagine what functional requirement precipitated this design decision. Considering the complexity added to the development cycle this would cause it makes no sense to me; I’m at a loss. I could be missing something though; you never know it may have a really sick benefit that I’m just not privy to.

Are any of those “issues” at all relevant? Not the function and not the installation process. Those are subjective issues that may only matter to me; it’s armchair quarterbacking at it’s finest (if I do say so myself). For everything else, well, that obviously depends on the specific needs of the project. For my needs it’s a pass simply because my clients care about the interface and Avactis looks like it hasn’t been updated since 2002 (at least).

So, while Avactis is a nice program with all of the features any store would ever need I personally feel that it’s not ready for my project.

As a continuation of my research into choosing an appropriate shopping cart application for an upcoming project I’m working on I chose to review OpenCart this time. To be honest, I hadn’t even heard of OpenCart until one of the comments from the last post turned me onto it but after reviewing it I am glad I did.

opencart-logo

As mentioned above, OpenCart is another e-commerce platform useful for setting up an online storefront. OpenCart is released under the GNU General Public License version 3 (GPLv3) which means it’s freely available for anyone to use and since OpenCart is written in php it’s right in my wheel house. The main developer of OpenCart is Daniel Kerr who, from what I can gather, is over in Great Britain but not the Australian rules footballer (in case there was some confusion from any Aussies).

Anywho, OpenCart has lot of good features available out of the box with my favorites being the Backup Manager, User Groups, really nice localization (l18n) and internationalization (i18n) options, support for multiple stores and a slick reporting overview. OpenCart also has the one feature I personally love from any and all e-commerce packages; Guest Checkouts. Personally, when I’m buying something from an online store that’s not a 500 pound gorilla like Amazon or Best Buy I hate signing up for an account. Guest checkout is the shit.

The code is really well structured and thought out; it uses a nice implementation of the MVC pattern which made things ridiculously easy to walk through and find out what was going on under the hood. OpenCart appears to be using a home grown MVC framework, which, while, in my opinion a little unnecessary,  isn’t at all a big deal. Unfortunately, the code appears to be open to Cross Site Request Forgery (CSRF) attacks and other security issues. More on this in a minute.

The OpenCart administration module is attractive and laid out logically. Everything is ready and available in such a way that most people with experience working with administration panels would feel right at home. It should be noted though that the administration panel will NOT work in Internet Explorer 6.

OpenCart Dashboard

Naturally, OpenCart also ships with the ability to have custom themes and the default theme that ships with OpenCart is quite attractive too. After reviewing the procedure for creating themes though I have to say that I’m not even a little impressed in how themes are supposed to be created; they appear, in my opinion, to be overly complicated, a little convoluted and deviates from the traditional manner used with countless other open source projects.

OpenCart Store

OpenCart has a basic module system though it isn’t what, in my experience, should be considered a module system; in OpenCart a module is more of a sidebar widget. A small distinction to be sure and not really an issue; more of an inconvenience of nomenclature than anything else but something that drives me nuts (can’t we all agree on what these terms mean already?).

There’s also a lot of manual intervention needed when configuring the system. Want to add a module? FTP the module, go to the module section of the admin, click install, click edit and fill out the form (be sure to enable the module too). Want to turn on search engine friendly URLs? Rename the file .htaccess.txt to .htaccess (through FTP or similar) then go to the administration panel, then to the settings area to enable. Very anti user friendly in my opinion.

There are other issues with OpenCart, like the flow for adding images to a product being pretty convoluted, but those are all small in the grand scheme (pretty much all my gripes can be, rightfully, dismissed as design decisions I don’t agree with). Fair enough. What isn’t acceptable is the complete lack of respect the developer has for security in OpenCart or the developers who try to help out in general. This, I’m afraid, is a deal breaker.

May 2010 was actually a pretty fortuitous time for me to research OpenCart; as soon as I began looking into the program I started seeing some discussion on OpenCart which lead me to a forum post on the official OpenCart forums. Another developer had some suggestions on how to improve the style and conventions of OpenCart and, well, Daniel really showed his ass.

Then, not a week later, I see on that there’s another war going on between Daniel and a developer who found some pretty nasty CSRF issues. Again, Daniel showed his ass (along with a good helping of ignorance mixed with arrogance this time) with nothing being resolved.

This was truly the breaking point for me. Why in the world would I ever use software written by someone who, when confronted with the issues, acts like nothing’s wrong? Nope. I have people relying on me to make choices that won’t, you know, ruin their business and OpenCart, for all it’s bells and whistles and nice code and pretty administration panel, is a horrible platform because the developer refuses to do anything about issues when presented with them.

Dreamhost is slandering me! They think I suck and have no problem telling people as much. Seriously; I write bad code by them. How do I know? Because Chris Duke over at AppModo told me so. Problem is, Dreamhost was 100% right. Yup; I fucked up.

Wp-Click-Track 0.7.1

On Monday I got an email from Chris asking for some help with one of the WordPress plugins I wrote; wp-click-track. Chris was cool enough to let me post the message he received from Dreamhost:

Hello,

I’m writing you about your database “appmodo”, please write me back as soon as possible.  There seems to be a plugin you are running that is poorly coded (missing table joins) and this 1300+ second query that it is running is causing high server load spikes. If this plugin is not fixed, or disabled, and it causes server problems, we may need to disable the wp_tracking_clicks and wp_tracking_links tables.  You can write me at justin@dreamhost.com.

The problem queries:

# Query_time: 2181  Lock_time: 1115  Rows_sent: 1  Rows_examined: 38703 use appmodo;

SELECT date_format(click_date,"%Y-%m-%d") AS first_click,
date_format(click_date,"%j") AS day_of_year FROM wp_tracking_clicks tc,
wp_tracking_links tl WHERE tc.click_id != '0'  GROUP BY first_click ORDER
BY first_click DESC LIMIT 1;

# Query_time: 1995  Lock_time: 924  Rows_sent: 1  Rows_examined: 38703 use appmodo;

SELECT date_format(click_date,"%Y-%m-%d") AS first_click,
date_format(click_date,"%j") AS day_of_year FROM wp_tracking_clicks tc,
wp_tracking_links tl WHERE tc.click_id != '0'  GROUP BY first_click ORDER
BY first_click DESC LIMIT 1;

# Query_time: 1387  Lock_time: 309  Rows_sent: 1  Rows_examined: 38703 use appmodo;

SELECT date_format(click_date,"%Y-%m-%d") AS first_click,
date_format(click_date,"%j") AS day_of_year FROM wp_tracking_clicks tc,
wp_tracking_links tl WHERE tc.click_id != '0'  GROUP BY first_click ORDER
BY first_click DESC LIMIT 1;

# Query_time: 2016  Lock_time: 930  Rows_sent: 1  Rows_examined: 38703 use appmodo;

SELECT date_format(click_date,"%Y-%m-%d") AS first_click,
date_format(click_date,"%j") AS day_of_year FROM wp_tracking_clicks tc,
wp_tracking_links tl WHERE tc.click_id != '0'  GROUP BY first_click ORDER
BY first_click DESC LIMIT 1;

Thanks!
Justin K

Justin did a nice job highlighting what the issue was; I didn’t join the tables together so it would take a really long time to complete the scans if there were a lot of rows in the tables. Another reason to always develop with large data sets; something I obviously didn’t do with wp-click-track.

Your application is useful and popular. Your users love it. Your users love you. But over the next week, something curious happens. As people use the application, it gets progressively slower and slower. Soon, the complaints start filtering in. Within a few weeks, the app is well-neigh unusable due to all the insufferable delays it subjects users to– and your users turn on you.

Anyway, there was no excuse for that; this was shoddy on my part. The only explanation I have is poor quality on my part.

This particular bug affected the line chart on the admin dashboard. Ever wonder why that particular graph took forever to load? Yup, that’s why.

What’s funny, to me at least, was that I had already fixed this particular issue; I just hadn’t released it yet. This particular bug was planned to be released with 0.8 but 0.8 is taking longer to complete than I initially thought it would. So the bug fix languished and sat there. Unreleased.

So yeah, wp-click-track 0.7.1. It’s a good release to get.

I’m getting ready to put together an e-commerce site soon and, with the budget what it is (sigh), I have to use a third party package to build on instead of writing a custom platform. Not a new scenario, at a basic level, but it’s been a while since I’ve had to go with this strategy. Put plainly; I have no idea what my options are when it comes to e-commerce though I know I don’t want to go with OsCommerce or CubeCart or, and only because the client forbids it (seriously), Magento.

Should We Use PrestaShop?

PrestaShop is an open source licensed (3.0) e-commerce platform that I was recently introduced to by my friend Caroline; it’s not without it’s faults but there’s also a lot to like once you get past the boogers. PrestaShop has one of those open source business models that’s good for developers but kind of bad for the civilians. By that I mean if you know what you’re doing you can get a shop up and running with little to no cost outside of development time. If not, well you have to spend money in the PrestaStore to buy modules for payment gateways and additional functionality.

Out of the box PrestaShop includes payment modules for Bank Wire, Cash on Delivery (COD), Google Checkout, Paypal, Paypal API and, seriously, checks (I imagine because nothing screams competent company like check by mail and COD payment options). If you want to use any other payment gateway, like Authorize.net for example, you either have to buy it from the PrestaStore for 255 fucking dollars or write your own. Seriously.

The documentation is lacking to say the least. There are all sorts of references to a mythical “Developers Guide” which is coming soon with no hint as to when it’s going to be made available. Since they have an entire site dedicated to selling modules it’s hard not to think of this as intentional (which makes PrestaShop look like dicks in my opinion).

One thing that’s pretty refreshing though is their honesty about incompatibilities with various versions of php. They even know what’s wrong with each version. Right there on the download page they have a very nice warning:

* Some PHP 5 versions are bugged and prevent PrestaShop from working correctly:

* – PHP 5.2.1 (authentication is impossible)
* – PHP 5.2.6 (authentication is impossible under 64bits servers)
* – PHP 5.2.9 (image management/upload broken)
* – PHP < 5.2 (invalid date timezone)

On the one hand it’s really cool that they know there are issues with their software depending on the configuration. On the other hand though it’s a little unnerving that they know about the issues but haven’t fixed them yet. That says to me, and maybe it‘s just me, that development is a little inconsistent and poorly managed; something you DON’T want when choosing any third party solution.

The PrestaShop installation process is actually pretty nice, though  the design is guaranteed to make your eyes bleed and it’s not without it’s bugs. The installer goes through the normal and standard system checks we’ve all became used to with third party packages which is nice. The only real issue is that the installer needs the last name field to be capitalized. Weird but really a non issue in the big picture.

PrestaShop Installer Bug

As to the larger issues of bugs and management, Nick Bartlett has spent quite a bit of time documenting those. I haven’t had a chance to look into the extent of them yet, not having used PrestaShop in a production environment yet, but according to Nick’s blog, which I highly recommend checking out if you’re interested in PrestaShop, there are bugs and issues with timezones, removing orders, overzealous security hacks and the product attributes leaves much to be desired.

It should also be noted that I was never able to get the search engine friendly URLs to work. This was on Apache 2.2 CentOS 5 box. I don’t know why though I didn’t look into it too much. Could be my fault but I doubt it.

PrestaShop Administration Panel

The default theme is really gross for both the admin and client sites but it’s possible to create custom themes for the client site. You’ll have to suck it up for the admin site though.

PrestaShop Demo

Digging through the code leads to all sorts of insights. For example, the code style is a bit of a hybrid between procedural and OOP but in a very familiar way. It reminds me of a lot of projects; very logical. Interestingly, PrestaShop uses Smarty as it’s templating mechanism. For me, this is really nice; especially since I’m already familiar with Smarty.

Also, as previously mentioned, there’s a module framework though, also as previously mentioned, the documentation doesn’t exist. So, in theory, it’s possible to extend on PrestaShop though I haven’t had a chance to dig into how that’s done exactly. Expect that to happen in the future though.

All and all, even with the bugs and “issues”, and keeping in mind that I haven’t actually used PrestaShop, I’m not against it. Yes, it sucks that the module everyone and their mother is going to need (Authorize.net) is a paid module but it’s still better than OsCommerce.

In my pursuit for financial independence I’ve been taking on random freelancing gigs from some really smart and interesting clients. One of the more respected clients I work with has been using Expression Engine for their main platform for years, and while I was initially skeptical, I’m beginning to believe there is potential for Expression Engine to be a useful tool too. There’s just one little thing; it’s possible to create a debugging nightmare pretty easily.

Expression Engine Escaping Madness

Expression Engine is built by the same company who put together pMachine, one of my favorite blogging software from back in the day, so I had some pretty high hopes for it. Then I started reading some off the cuff comments about Expression Engine, especially in comparison to my mortal enemy Dolphin CMS, and I started getting a little nervous. Then, when I started seeing how the flow worked, my head almost exploded.

See, all the style and creative stuff is stored in the database. Because Expression Engine has it’s own meta templating language (similar to Smarty in syntax and style; to me anyway) all the templates are available and ready for anyone to make modifications to. On top of that, Expression Engine allows for the inclusion of custom php inside of the stored template files which gets executed with the dreaded and evil “eval()” tag.

Confused? Me too. To help clear things up here’s a snippet of Expression Engine templating code:

{assign_variable:my_weblog="default_site"}
{assign_variable:my_template_group="site"}
{embed="global/header"}
Page Content Here.

It should be pretty obvious what the code above is doing, but because I get a lot of shit for not being verbose (I’m looking at you Reddit), here’s what’s happening:

1. A variable called “my_weblog” is being created with the variable “default_site”.
2. A variable called “my_template_group” is being created with the variable “site”.
3. The header template file is being included.

Not so bad right? I didn’t think so either but there’s also the inclusion of raw php. The below is perfectly valid to do in Expression Engine (assuming the “Allow PHP in Tempaltes” setting is enabled):

<?php
$my_weblog = 'default_site';
$my_template_group = 'site';
include 'global/header.php';
?>

The above is a translation of the Expression Engine code by the way (if you hadn’t picked up on that). This, too, isn’t bad per se, but it does break a few very important rules which I’ll get into in a moment. Annoying and sort of dangerous? Absolutely. But I can see where the appeal lies in allowing this sort of functionality (and, yes, even if you have to use eval() to do so).

That being said, my head almost exploded when I saw how the logic was laid out when mixing both the Expression Engine template tags with php functionality. Keep in mind that Expression Engine has a setting that allows you to set when in the processing flow you want the php to be executed. If that sounds confusing just know that in the below example the Expression Engine stuff is executed before the php code.

Here’s what I mean:

{exp:query sql="SELECT name FROM exp_freeform_entries WHERE entry_id = '1'"}
<?php 
$name = '{name}';
?>

The above simply grabs the name from the table and then sets it up for use by php. Once again, perfectly valid usage it would seem, though the more astute people will immediately see the issue.

Since Expression Engine executes the template tags first this is kind of a snap. The thing is though there’s no escaping going on there. The above will work great when the value of name is something like Eric or John but what if the value is “Eric O’Reily”?

Yeah; it’s gonna break with a parse error. But worst of all when it does break the error message you’re going to get is going to reference the call to eval() and not the actual template file. This is going to make debugging a bit of a bitch. On top of that, there’s no native method to escape anything within Expression Engine itself. So adding the usual call to addslashes() isn’t possible.

So, while Expression Engine is pretty snazzy and nice it isn’t without it’s pitfalls. Mind you, the escaping issue isn’t impossible to avoid; it’s more a question of design than anything. It is something that needs to be watched out for because, yeah, doesn’t seem there’s going to be a change anytime soon.