Jump to content

jagorny's Content

There have been 8 items by jagorny (Search limited from 05-Aug 19)


Sort by                Order  

#159274 db_query INSERT issues...

Posted by jagorny on 03 April 2013 - 08:39 PM in Developers' Corner

Never mind. The replacement regex that detect database operations is case sensitive - changing 'INSERT into' to 'INSERT INTO' resolved the problem.



#159264 db_query INSERT issues...

Posted by jagorny on 03 April 2013 - 06:28 PM in Developers' Corner

So I'm trying to do a simple insert but for some reason I keep getting an error and all the fields in my array check out. Do I need to provide value pairs for every field?

When I try to run this simple code:

$shipment_load = array(
  "shipping_id" => $record_unser[6]["shipping"],
  "carrier" => $record_unser[6]["carrier"],
  "tracking_number" => $record_unser[6]["tracking_number"],
  "timestamp" => $srkey["timestamp"]+(60*60*24),
  "comment" => ""
);
$shipment_load_id = db_query("INSERT into ?:shipments ?e", $shipment_load);

I get the following error:

Database error: Table 'cscart.cscart_' doesn't exist (1146)
Invalid query: SHOW COLUMNS FROM cscart_

I don't understand why the table to insert is disappearing during the course of the query. I can insert non-variable values that are appropriately typed and nothing seems to work. Is the format for this different in 2.2.5?



#158617 Class includes for addons...

Posted by jagorny on 25 March 2013 - 04:57 AM in Developers' Corner

Hey folks,

So I'm working on a fairly complex Amazon integration and I am trying to establish the best place to put classes - I'd like to keep them separate in the addon because ultimately when they roll out API updates and update their models I'd like to be able to update them fairly easily.

However I am not entirely sure where/how to implement autoloading in 2.2.5 - the docs indicate that autoloading is already a feature in 3.0.x and that you simply need in include the classes you want in the 'core' subdirectory of the addon and name them class.ClassName.php - but for the 2.2.x what is the best way to set up an autoloader pull classes and where is that done from?

My gut says func.php and so I will try that tomorrow, but I hope another dev can share some experiences on pulling external class libraries into an add-on in a clean and nonobtrusive way.

Thanks in advance!



#158288 Allow users to create shipments

Posted by jagorny on 20 March 2013 - 08:20 PM in Configuration

It is important to note that once you turn this on, a completely different database storage table is used than the prior system - making tracking information inaccessible from order prior to the switch (unless you do some database voodoo to import from the orders_data table to the shipments table).



#158188 Ebay, Amazon Integration

Posted by jagorny on 19 March 2013 - 07:49 PM in Wishlist & Feature Requests

There are serious pitfalls in the process with regard to how shipping is managed and calculated, and this really prevents any sort of integration with Amazon - which rings the death knell for our use of CS-Cart - a sad day indeed.

The problem is that Amazon uses real-world shipping calculations to assess charges - these are a base fee, then per pick/pack unit, then per pound with grades. But they have different rates depending on whether your item is classified as media or non-media. And if you also participate in Fulfillment by Amazon, then you also have rate cards for the various FBA Export jurisdictions.

The problem here is the media vs non-media. If you establish rates for both media products and non-media, both sets of shipping options will appear. There is no way to have books use one set of shipping rules (for media rate) and non-books use another set of rules. Worst case scenario, you'd be forced to charge non-media rates for all customers. This is not appealing.

Second problem is that FBA/Amazon cannot export books to Canada - this is a 'Heritage Act' prohibition - but there is no way to limit the countries that a product is available to as far as I can tell. This has got to hit DVD sales also - where you would only want to make certain DVDs available to ship to certain locations, etc.

These are two essential shipping functions that CS-Cart does not provide. You should be able to determine product availability by Locations and by User Groups. This is a no-brainer to me, and functionality I assumed was tied to the Locations area.

If I am totally off-base on this, or am missing something important, please let me know. Right now Amazon integration is a little pointless for anyone who actually is trying to use FBA.

UPDATE: It looks like the code shared that controls shipping method by product can be massaged to do what I stated above... Creating shipping methods for Books, Non-Book Media and Non-Media, and then setting up locations and charges for appropriate locations. Ideally a CS-Cart native code layer would be brought in at some point.



#150917 Olark Live Help add-on

Posted by jagorny on 11 December 2012 - 06:30 PM in Hints & Modifications

Just to clarify, the 0/20 at the top of the dashboard screen on the free account is the number of conversation you can freely initiate for the month. If you don't get a lot of live help traffic (and we generally do not) then this will probably be enough.

Concurrent conversation is a paid-only feature - each operator can have up to ten conversations open at once. However there is no monthly restriction.

With the expandability of the javascript API, it is possible to expose a great deal of information by expanding this plug in. For example, when a user is logged in you can use the API to have your registered contact name appear in your IM web visitor list. You could even extend the code to include recent order links to appear in the conversation, phone number on file, recent tracking numbers, etc. Any user information that can be mined once a user logs in can be passed to the conversation via the olark API.



#146347 TAGs what are they ?

Posted by jagorny on 04 October 2012 - 04:07 PM in General Questions

This is a frustrating topic for me personally, because tags on a store are fantastic if implemented correctly, but I have never been able to get enough interest from the developers to actually implement them in a useful way.

Right now, tags are nothing more than bookmarks for users. That's their vastly underutilized potential right now. A customer can tag stuff and go to that tag later to view what they tagged. How is this any better than a wishlist?! Oh because you can have more than one. It would make more sense to let customer create and name multiple wishlists.

Tags are supposed to serve as an alternative navigation structure - so that, for example, if I see a product has a given tag, I should be able to click on that tag and view other related products, etc. This does work, although it is laid out in a clumsy and buried way.

But from a product catalog design perspective, this system should be available across the board on the administrative backend but it is not. For example, why can't you use this as a population criteria for product blocks? Or daily deals? Or setting up a seasonal menu? Or anything else? Make tags public or private, etc... Tags in a proper content management system are much more flexible - you could allow conditional content based on tags - e.g. show or hide a block based on a product tag - why not? There are a lot of batch processes that could be simplified if tags were better implemented. You could try to set up invisible product features to accomplish this, but this is pretty clumsy and adds a lot of bloat.

I've added a few tag recommendations to user voice and gotten feedback on developers that they were good ideas, but no implementation roadmap. One day I'll end up coding add-ins myself for it I suppose...



#137030 Multi-vendor - more than one vendor per product - thoughts?

Posted by jagorny on 22 May 2012 - 05:15 PM in Wishlist & Feature Requests

Hi!

I posted this in the ideas section but wanted to drum up some support and feedback for it.

I was very excited about the multi-vendor capabilities because we presently offer an affiliate program for parish bookstores but have been looking to do something more. In particular, we wanted to provide folks with local options for purchasing items in their area where they exist.

So for example, a parish could list products in their tiny bookstore and their price and you would effectively have an amazon marketplace style selling platform. This would allow our catalog to expand in ways that we might not expect, provide analytics on products we might not otherwise see demand for, allow print on demand folks to have their work given a trial run etc.

Another option - what if you have an online store and many physical store locations? Wouldn't it be great to be able to have your customer request certain items for pick-up locally? Multi-vendor would be great for this is you could set up a vendor account for each location and then, once again, let the customer choose the vendor (in this case the location) for the product.

This choice could easily be integrated into the product options section - even allowing for price modifiers, etc. You could list it as a table with a radio button, etc.

Thoughts? Is this not how you would expect Multi-vendor to work? Don't get me wrong, if I was still running a drop-shipment enterprise the multi-vendor as it stands now is awesome - we sure could have used it back in 2006. But this seems like a really innovative application that doesn't seem like it would take a lot of development time.

Thoughts?