Jump to content

ploughguy

Member
  • Posts

    82
  • Joined

  • Last visited

Everything posted by ploughguy

  1. "Hi PG, Of course it works ... it's been a year and this is a highly desirable feature. The comprehensive yet entertaining documentation can be found here. Optimistic regards, Ian" Ian, why did you not offer me an answer like this? It involves much less typing, and it would have made me soooo happy. OK, OK, I will go hang out in the glum and disconsolate company of Brian, Al and Toucan. It is 9:25 am and they are probably already drowning their sorrows in the nook at the Raven and Wheelbarrow. I shall seek them there. PG
  2. Wow, here we are a year later. I am still using 4.3.4 with Estelle's mods. I have a list of 30 new requirements that have built up over the last year and I am reluctant to put more time into 4.3.4 development so I am keen to upgrade, but only if it will cause less pain than 4.3.4. Has the situation with Stock Levels for Product Options improved in version 5 to the point where it can be used? I have been updating 6.0.1 as the betas come out, and I am now at the point of trying to get the stock options to work again. It is still very mysterious. I don't know how much of it is beta bugs and how much is stuff that never worked in version 5. I gave up after the dialog above and went back to living a quiet, enjoyable life. I notice the documentation situation does not seem to have improved - the option matrix stuff is still as obtuse as ever. I am reluctant to spend too much time on this now until I hear that other poor sap has actually got it to work. Did someone mention SAP? Now that's an idea... PG
  3. bsmither said: This! Is! Sparta! a Beta! Understood! Before I switch from 434 I need to: Learn about the new user-exit technology from CC5 Get the hang of PHP Classes Learn about the new rendering engine used by the UI Work out which of my hundred-odd enhancements and add-ons are now part of the product And find or develop a solution for the remainder of them. Probably not much change out of a year for that lot. Hence the early start on V6.
  4. Problem not necessarily solved, but I managed to get it to complete by feeding it user root. After the red mist of frustration cleared, I did some more dining. I have no idea what I did to fix it, but now I have a store user configured for host: localhost and it all now works. I seriously have no idea what fixed it. The user was originally localhost - I change it to All hosts then attempted to copy it to create a separate localhost user (because "All Hosts", apparently, is not All at all - it excludes localhost.) For reasons I do not understand, the copy function in phpmyadmin did an update instead and I ended up with just a localhost user. However the installation process completed user the store-owner account and it works.
  5. Hi, I am trying to install CC691b for evaluation. I am running CC4 something in production on a hosted service and I have previously installed CC5 for evaluation successfully. Installation platform is OS X running Mavericks. MySQL 5.6.22 is installed and running. phpmyadmin can connect to MySQL. - I believe this means that the Apache, PHP and MySQL installations are acceptable. I have created a user atomacom with phpmyadmin - with a known password ( i.e. I am cutting and pasting the password to create the user and configure) - the user has access for all databases. I have created a database cc601b User atomacom has all permissions for the database but no administrator permissions. The setup dialog for CC6 runs successfully up to the database settings step, whereupon I stall with: - Incorrect database host, username and/or password. I have restarted it several time so it gets to repeat its checks - everything is green up to this point. The values I am using are: Database Settings: - Database host: localhost - Username: atomacom - Password: <paste from same source that created the account> - Database Name: cc601b - Table Prefix: cc601b Store Settings: - Default Language: English <this is not pre-initialised to its only possible value, even when page is represented for correction and the value was set previously> - Default Currency: AUD Administrator Profile - User name: admin (a user that has not been created - I believe this is a CubeCart internal user) - Password: <same as above> - Other details probably not relevant here Advance Settings; - Drop Tables Not selected (this is a new database and user) Any thoughts about what might be causing the problem?
  6. Thanks for the encouraging words, Ballesdegolf. I will keep an ear to the tracks to listen for the oncoming update. I agree with your respected pro - it needs some work, but not a great deal. Recognition of options vs attributes (and the addition of stock control for products that are options) would go a long way to making it acceptable. In fact, if it did that then it _would_ be acceptable. Provided the list of tiny buglets above were addressed and a few paragraphs of "How To Use It" appeared in the Help. We are not talking about SAP here - expectations have to be moderated based on what a small team can pull off for the money that they are charging. So - fix the bugs, resolve the product total vs option matrix total conflict and allow stock depletion by optional options. That would make me happy. As for gift certificates, I gave up on them when I realised that CC treats them as a discount voucher with a different name. That has all kinds of evil tax implications - a gift certificate is technically a deposit, not a discount. I have not checked if this has been fixed in 520. Anyhoo, I am going to murder a goat so I can examine its entrails for sign of a patch. R
  7. NeilChamp, Nice list. I am now reaching into the bottom desk drawer for the emergency bourbon, which lives beside the revolver. I am ramping up my CC5 mod skills so I can start to implement my own mods. However, I do not feel a burning urge to upgrade at the moment - this stuff is not even close to working properly. Next question - does one stick with 5.2.0 and put some time into fixing all this, assuming that CC is not going to do the whole job, or wait to see what happens next? I need to get some cornflakes to put this bourbon on. R
  8. Jim, Estelle's mod filled a functionality gap, but it was not a definitive solution. CubeCart should be aiming to implement a perfectly correct solution rather than acceptable hack, which is what Estelle's mod was. She want to create an acceptable workaround for missing functionality, and for $25, it was excellent. I think part of the problem is that there are functional dependencies that are not immediately obvious. I would like to broaden this debate to create a spec that documents the minimal reasonable functionality for product options. This will give the CC developers something to work towards that is more appropriate than a souped up version of Estelle's mod. We need some definitions: "Mandatory" means that we must have a value for this option to create an order. It may or may not mean that the customer is forced to make a choice. We can, for example, preselect a value that suits as and allow them to change it. In other circumstances, we may find it prudent to force the customer to make a conscious decision. "Finite availability" means that we need to have a physical product with the selected properties to be able to sell it. This means that this option is subject to stock control. "A Product Option" when referring to something we are selling (vs a product option in cube cart) is something we can choose to add or not, such as a foxtail for the radio aerial. It will often be another product. "A Product Attribute" is a property of the product that must be selected by the user, such as colour. If you are going to buy an RPG from my website, then you have to choose the colour - Khaki (because it is traditional), Camo (because it has street cred until your mother gets one) of hot pink (so you can find it if you drop it in the desert.) Since an RPG with no colour is probably invisible and therefore useless (how would you know you had received it?), we cannot sell the product unless you select the colour. An acceptable value here is "don't care" but we don't stock them. Note that these definitions may be inconsistently applied below. I have tried, but as the wife often points out, I am very fallible. Next, lets look at the properties of options themselves: Options may be optional or they may be mandatory Options may have finite availability or they may not Explosion for a bike store: Optional option with finite availability - handlebar streamers. Mandatory option with finite availability - frame colour. This is not an option, really. It is a product attribute, and one of the possible attribute values must be selected before I can sell the bike. Optional option with infinite availability - e.g. emailed link to online bike maintenance app - iPad, Android, or None. If the customer misses this then it does not cost either of us any actual money, and the customer can get a link from our web site later, so this is a convenience feature for the customer that causes no pain if they miss it. Mandatory option with infinite availability - warranty extension. It is important to the customer and to the vendor that a conscious decision is made about this. The vendor wants the customer to buy it because it is money for nothing, probably, and the customer will kick themselves (and hate the vendor) if the bike starts to rust 13 months after they buy it and they did not consciously decline the option for extra cover. Option should default to "undecided" and the customer is forced to select an option value. Optional option with finite availability I can sell you a bike with or without streamers. If I am out of stock, then I don't offer them, or I put them on backorder and send them later so I need stock tracking on these. AFAIK there is no backorder management in CC, so not offering them when out of stock is the simplest policy. In this case, the streamer is actually a separate product that I am using options to up-sell There should be an option to associate a material number and a quantity with this option (e.g. 2 x "Streamer Mk III: Handlebar, for the use of"). It should appear on the invoice as an option with a price premium (i.e. 1 x Bike - with streamers) It should appear in the sales records as the sale of the two separate underlying products (the bike, the streamers) rather than as a "bike with streamers" Orders that take this option should decrement the streamers product stock level. Streamers may, themselves, have product options such as colours. The product option drop-down for the streamers should offer an option explosion where appropriate (e.g. red solid colour, red glitter effect, red metallic, blue solid colour...) I believe this option type is essentially useless if the option is not linked to the underlying "streamers" product in this way. Mandatory option with finite availability (product attributes) This is the "SLPO" case - the "option" actually specifies a product property such as colour, of which we have a limited choice and limited availability. We have at least three selection options here - the option can have mandatory selection (where the initial value of the list box is "Please choose a colour"and the customer must make a conscious decision) - the option can have fixed pre-selection (where we preset the field to a preferred value for reasons that generally benefit the vendor in some way) - the option can have ordered pre-selection (where we preset the field to the value with a particular property e.g. the cheapest/most expensive option.) - - I modified my 4.3 store to preset "Colour" to the value that has the most stock - which generally means if they don't care about colour, they get an orange one because that is the least popular colour. It selects the stock levels from the option matrix for all values in descending order, then discards anything less than one. I added a big red bar across the top of the product page in the store that says "Be sure to select your colour" so they cannot complain if they miss it. You can see it at atoma.com.au. This has been a huge help for reducing choice depletion. I think a lot of people look at the option selection, see "Orange" and think "Yeah, why not." Product attributes generally need to be fully qualified (i.e. we must specify all possible values)- I cannot see a situation where we would have stock control on some mandatory attribute values and not others. Of course, this is a failing of my imagination, probably - perhaps someone can find an example. If product attributes must be fully qualified, then the product stock level is exactly the sum of the option matrix stock levels and it should be computed from there every time. The Product table stock value should never be used. In particular, CC should not try to maintain the product record aggregate total at all in this case. The product summary should display the total of the option matrix stock levels as a link that takes us to the product's "Options" tab. Optional option with infinite availability It is an optional option which means we don't care if the customer does or does not make a decision. I can only imagine this being used when there is no financial benefit or risk to either party. - because if there is money to be made, why would you default it to "No thanks"? - but defaulting it to "Yes Please ($100)" is just going to alienate the customers - so if there is money involved, they need to make a conscious decision - i.e. mandatory with no default. We should be able to select the default option value based on something other than alphabetical order. Mandatory option with infinite availability This is typically used for virtual products that we charge for - such as the warranty extension on the bike. We definitely need to be able to force a decision by displaying "Please choose your warranty option" to force a choice. There may be cases where we want to select a default value (but then it is not really "mandatory", so probably not.) Next steps - refine this then do a function match to CC 520.
  9. NeilChamp, Are you using an option set to assign the options? This definitely does not work. You must assign the options using the "product options" frame rather than the "options set" frame. Russ.
  10. Neilchamp, Try disabling stock management in the product's General tab. I know I had to muck around before I got it working. This is a complicated area because it is grafted on rather than designed in from the start. I use Estelle's SLPO too, and it worked very well but it was somewhat simpler. I believe this will be a better solution when it is finished. Our job is to assist with the development by: Finding and documenting each problem Finding a repeatable way of demonstrating the problem Documenting the demonstration of the problem Lobbying to get it fixed. It is very difficult for a software developer to test his own stuff. Developers tend to test the bits that caused the most trouble and assume that everything worked well. Also, when under pressure, they tend to underdesign, hoping that they will work it out as they go along. This results in nearly-working solutions that pass simple test cases. This particular problem is aggravated by the absence of documentation. It is quite possible that there are procedures where it does work perfectly, but we do not know what they are. We have to guess how to use it, so we might be doing the right actions in an unexpected order. Russ
  11. Thanks for your help, bsmither. I deleted my installation, removed the database tables then re-downloaded and re-installed. Initially, not much happened and it still looked broken. I assigned an option set first (for colour and size as above) and the Product Options table display was still out of whack so there is a bug with the assignment of option sets. I deleted the option set and assigned the attributes of the two groups independently. I still did not get an options matrix, after some unremembered mucking about, I went back to the options tab and there it was - an options matrix. And very nice it is too. I now have a product with two options with two attributes - size(A4,A5) and colour(Black,White) and the options matrix has four rows. I have assigned a stock level of 20 to each combination. This all looks very nice, but we need to get it knocked into shape. As it is, it is only half ready for prime time. Problem 1: The bug with the use of option sets (a very useful thing which I had coded for V4 myself) should be easy to fix. But there is some strange behaviour about availability of stock and this looks ugly. I consider this to be another bug, or perhaps an incomplete development. Problem 2: If I enable "Use Stock Level" for all rows of the matrix, I would expect the product's General tab stock control field (and the Product List report) to display 4 x 20 = 80, but it is still zero. Furthermore, the product in the store shows as out of stock, even though there is stock of each item. It looks like there is some kind of dodgy denormalized data storage going on here - where the master product has a stock level which is independent of the options matrix stock levels and the store owner has to keep them in synch some how. To do it properly, the General tab stock level should be the sum of the matrix values when an options matrix is defined. In my case, at least, for a product with stock control on product options, there is no version of the product that is not in the matrix. The product must have some value of colour and size, and therefore it has a stock level that should be in this table. There is no concept of a colourless or slzeless rendition of the product, and if there was then I would create "colourless" and "sizeless" as option attributes. This means that there is no scope for the overall product stock to differ from the sum of the option matrix values. The general product stock should be a virtual field that is the sum of the option matrix values if an option matrix is defined, otherwise it is the product table value. This virtual field value should display in the product overview list and the product General stock level fields. As it stands, the storekeeper has to add up the matrix values externally, then put this number into the General tab, and repeat this every time the product is restocked. Sounds like one of those nightmares where things fall apart just slightly faster that you can put them back together... Cubecart is still using ISAM files which means it is still non-transactional (and hence non-relational.) Having duplicated data values is asking for inconsistencies in the event of a crash or program error, since the database (if you can call it that) has no means of doing transactional rollback. It would be useful to have a data preening utility that checks that all the redundant values (e.g. the duplicated data in ORDER_SUMMARY and ORDER_INVENTORY tables, the duplicate stock levels across the matrix and the product data) are still consistent.
  12. I am evaluating 5.2.0 as an upgrade for 4.3.4 with Estelle's SLPO option. I am not abnormally dumb, but I cannot figure out how this works. This is what I have done: In Product Options, I have created a couple of option groups (Colour with attributes Black, White, Grey and Size with attributes A4, A5, A6, A7.) In Products, I have created a product and assigned the members of both groups to the product. Now I have three colour and four size attributes assigned under the Options tab. The Status column has a green tick for each row, and the Options Matrix column has a red cross for each row. The Options Matrix is not mentioned in the documentation under the Help link. I have not been able to find anything else about it, but I do not necessarily know all the hidey-holes where this sort of stuff is kept. I have tried various things with the Options Matrix checkboxes, (click all and save, click one from each group and save) in the hope of seeing a hint of what to do next, but I am stumped. Any hints? A stream of frustrated abuse followed by a link to some documentation is entirely acceptable. As an alternative, I tried creating an option set that contain all 7 options and assigned them to a different product. I am assuming this is a shorthand method for assigning a bunch of attributes from different groups as a collection, potentially to a large number of products. This caused a display aberation on the options page shown in the attachment. (Neither uploader is working...) The Options Matrix column displays a red X for the first attribute in a group, then the type information for the remaining items in the group. The remaining columns in these rows are slid one space to the left, so the Negative checkbox is in the Type column. It is possible that I am doing entirely inappropriate things here, but the program should be fixed to reject the operation or display it correctly.
  13. WIth the release of 5.2.0 with Product Options, I am ready to consider upgrading from 4.3.4. My site has many important (to me) mods that I need to apply to the new release before it can go live, and I would like to do this in the cleanest possible way to reduce the cost of future upgrades. Any hints? There seem to be a couple of options for this - something called "code snippets" gets a mention in the 5.2.0 release note and there seems to be a way to integrate bigger bits of new stuff using the new template engine. I have searched the forums I have googled "code snippets" in case it was the name of a technology but it returns every bit of PHP that anyone was ever proud of, so the term is not terribly useful. The CubeCart Developer Documentation area (which I found by accident and cannot find again) sports the very promising phrase "Coming Soon" which, I am pretty sure, has been the status for some years now. I figure there should be something written down, somewhere, about each of the following: The new skin/template engine The enhancement/extension technology The mysterious "code snippets" If you know where to find documentation on any of these, then please tell me where to look. Thank you for your help.
  14. A database transaction is all of the updates required to complete a business transaction. In our case, a sale debits a gift certificate, saves one order header and hopefully many items and decreases several stock records, so a transaction may update many tables, some of them several times. If something happens to interrupt this, then there is the potential for errors to appear in the database. The purpose of a transactional update is to ensure that either all of a transaction is recorded, or none of it. According to the eloquently named Codd and Date who laid out the theory of relational databases in the 1970's, a transaction should pass the ACID test: Atomic - so the entire transaction is recorded or nothing is recorded Consistent - is complicated, and it is tricky to implement - it means that someone who starts running a long-running report sees the data as if it was a snapshot, even though it might take them a day to scan the data, and updates are being done all the time the report is running Isolation - means that transactions don't interfere with each other as they are running. Durability - means that once the data is committed, only strategic incompetence can lose it. In particular, a simple machine crash with restore can recover up to the last committed transaction, and roll back the saved bits of unfinished transactions. The big one for CubeCart is atomicity. When we make a sale, we: optionally update a gift certificate, insert an order row in the order_sum table, insert the items in the order_inv table update the stock for (hopefully) multiple items in the inventory table. There is a risk that a coding error, machine crash, DOS attack or some other nastiness intervenes between the stock update and the order save (doesn't matter which comes first.) These two are balancing updates. If this process is interrupted, then the stock-in-store will be wrong and you probably won't know until you get an order for something you no longer have in stock. Durability is another important thing for us - As I understand it, the MySQL ISAM does not maintain a redo log, so if the machine crashes, there is no way for it to clean up so when it restarts you get whatever bits of transactions made it onto the disk before the disruption. The logical grouping of related updates is not recorded in the redo logs, so there is no information describing how to clean up. We really don't know what kind of hardware our ISPs are using. I sold a three year old Dell server once to a guy who was setting up a server farm in his garage. I am pretty sure that none of his $5/month online store customers were going to ask for a site inspection. Unless you are sitting there playing with your store all day, you really do not know how many times it goes away. If we get the occasional order, we assume it is 100% available but we are buying a service from an operator we have never seen in a very competitive environment... I think some update integrity would be useful. On the whole, this is not a big deal. I run my store on CubeCart and I love it. I particularly love that I can get in there and adapt it. (Although I am less in love with this idea now that I have to upgrade to CC5...) However transactional updates would make a good product better. Convict said... "Unsure even it brings more possitives than negatives." One negative might be a higher database load - I imagine that innoDB needs more iron than ISAM - but that is provided by someone else, right? And you have to do some recoding. I have not had a look at the reorganization that when with the PHP Objects conversion but I would expect that this would make the switch to be somewhat easier. There are other benefits too - MyISAM does table-level locking so for busy stores there should be a huge performance improvement. I look forward to the day when this becomes a problem for me.
  15. Does that mean it will some time in the future? The potential for database inconsistencies makes me nervous...
  16. Can anyone tell me if Cubecart 5 uses transactional updates?
  17. Problem partly solved - I believe the cause is that I have adjusted a text in this file using admin/Modules/Languages. This creates an adjustment record in the database which means the file is not in the loop. Fix is to take a note of the adjusted texts (that caused the file to be added to the DB) then delete the row from CubeCart_lang. I can't complete this, unfortunately, because cPanel (which is my interface to the hosting service) won't let me do database maintenance.
  18. Problem: On my production system, which is hosted, the Clear Cache function does not work. No errors are reported - but old texts still display. Desired response: Fix this by doing <actions unknown here> Versions Cubecart is 4.3.4 with some subsequent updates grafted in. Background I have a moderately customized CubeCart production installation running on a hosted server. My development system runs on Ubuntu "Randy Raccoon" or something similar. I have just written a GST report which I have linked into the Admin navigation tree under Statistics. In the development system, the link opens a new admin page similar to the Statistics/View Stats report page with a list of (two) report options. Clicking a report name runs it and it does its job nicely. The menu texts, links and report headers are custom texts. I have a language vector called custom.inc.php which I have put all my custom texts into so upgrades are a smidgeon simpler. It is loaded by getLang like any other language vector. After half a pleasant sunny day spend hacking code about, I copied the custom text include, the report includes and the updated navigation.inc.php file to the hosted production system, ran the Clear Cache function to refresh the texts, and refreshed my display and bingo - no change. Examining the page source shows that the links are there but they have empty texts. If I enter the report link into the url bar in Firefox, I get a page with no report links - once again, they are actually there, but the texts are empty. In Admin/Languages, I can open the custom text vector and see that the texts exist. The problem would seem to be that the cache clearing operation does not work. The hosted environment is presumably running different versions of everything, so it is not a real surprise. Has anyone else encountered and even solved this problem before? Thanks for your help. And nice wedding by the way. Well done. Pompful and ceremonious yet somehow restrained as suits times of fiscal rectitude, and as tasteful as a wedding can ever be (some hats excluded, of course - Beatrice, I'm looking at _you_...)
  19. Cubecarters, My store is using one of the standard fixed width templates which has a neat trick - it hides bulleted lists until the user mouses over the list owner... For example: * major item * another major item with detail list - detail item - detail item * major item What happens is that the list is built by a tag <ul style "display: none;"> which changes to <ul style "display: block;"> when the mouse approaches "another major item with detail list". Basically my information has been turned into an Easter Egg which is OK if you have smart customers, but how can you tell? Using firebug, I can see the html changing when I mouse over, but there is no stylesheet listed for this. I suspect it is probably javascript or some such dark magic, but I cannot find it. Can anyone tell me how to disable this horror? Thanks for your help. Russ.
  20. Sounds like it should work... What about the file ownership - can the web server get at it?
  21. In the Admin page, under group Catalog on the left, you will find Coupons and Gift Certificates. There is an Enable/Disable dropdown which lets you switch them on. A link appears in the online store page. Cubecart gift certs are treated like discount vouchers with a fixed value rather than as a deposit. This is not ideal, so check them carefully before you decide to use them.
  22. Ho, CubeCarters! I have a truly strange problem which I am hoping some CC-proctologist or WebServer Master might be able to help me with. Or perhaps a complete novice can point out my dumb mistake... I have created a custom language vector file en/custom.inc.php which holds texts for my personal additions. The object clearly being to keep my stuff separate from the standard product in the infantile hope that one day I will have the courage to update. I have integrated the language vector into the standard language maintenance facility in the Admin page and it has been working fine for 6 months. I have added a new and useful feature which allows a customer to get a postage quote on the cart page. Above the shipping options drop-list, there is a caption "Quote postage to", two entry fields (postcode and droplist for activated countries) and a "Quote!" link that updates the price. (The hope is to capture those who get to this point, realise they cannot get a final price without putting in personal details, get all indignant and decide to bail. My old lady is one of these...) I coded it up, sent it to my test site which is on a local linux VM, and it worked perfectly. I shipped it to the hosted production system and it bombs because the language vector is providing the default country and postcode (easiest place to put it.) First thing I did was clear the caches but no change. I put some debug in to print the language vector strings and they are blank. In the Admin page, I can see the strings have values, and when I overwrite them and update, suddenly the page works. So the update of the files was correct, the code is working correctly, I can see that the texts are in place in Admin, and if I update them in Admin, (which creates a copy in the database which overrides the file) then it all works, which means that the language caching is working correctly internally. However, before the update, the web server was not seeing the new version of the file. There were a bunch of PHP files which were updated, and their changes came through as expected. Both web servers are apache, both are running on Linux. The test system has had minimum configuration done, and I have no idea how the hosted server is configured. I would like to eliminate this behaviour because every time I upload a new feature with new strings I have conniptions because I think I have broken it until I remember about the web server and the text files. (Last time, the problem was not fixed until the hosting folks rebooted the server. Now I have worked out the update-the-language-vector trick, I can goose it myself, - - I think.) Any ideas? R
  23. It sends a confirmation message to the customer, and a confirmation message to the administrator. If you are testing and the notification address in the General Settings is the same as the email address for you test customer, then you will see two identical emails when you are testing. However, in production, customers will get one and you will get one. Now, the question is - why does the administrator get a confirmation email when an order is marked completed?
  24. Probably a bit late, but the AP shipping module uses a program on the AP website that takes the weight and dimensions, source and destination postcode, and generates a quote for standard parcel and express parcel post. It does not understand large letters (up to 20mm thick and 500g, and cheap) satchels (up to 3kg/5kg, $9.90/$18.50 anywhere in the country, but not always the best value because parcel post depends on distance and remoteness) and other useful products sold by AP. You would have to code these in yourself. AP uses a concoction called Cubic Weight, which sets a minimum of 250kg per cubic metre. Therefore, if you want to post a box of foam shipping beads, the actual weight is ignored, and a notional weight is computed by multiplying the actual volume in cubic metres by 250. For an idea of what this means, a quarter-full litre carton will weigh about 250 grams and is right on the border between cubic and actual weight. Take a sip from it, and it goes cubic. So, you have some choices: * Use the AP calculator as is, sent the dimensions small so it uses weight, and hope it averages out for you (downside, customers will buy some stuff elsewhere because their inc-freight price is cheaper.) * Closely analyze your products and develop a set of rules that you can code around the calculator. For example, I have adjusted it to get a quote for parcel post, then I compare it to a 3 or 5kg satchel, and choose the cheapest (parcel post is cheaper when the weight is low and the destination is near. The baggie looks better as you close in on 3kg and send to more exotic locations, like Darwin, which is one of the most expensive places to send to from Sydney.) However, I know 3kg of my stuff will always fit in a baggie. I know that x items of a certain product category will fit in a C4 envelope and qualify for large letter post at $1.10, provided that is all that is in the order. * Use the category-based shipping module. You can create some hidden categories (set the Hide flag in the categories table using PHPMYADMIN) that don't display in the category tree for the customers, but tell it which postage class it is in. This does not handle aggregation of products into a single shipment, so it is only useful in certain classes of store. I have to say that everything looked pretty simple with this online store caper until I started thinking about shipping. Well that, and trying to get decent product photos...
×
×
  • Create New...