Will I "outgrow" Miva soon?

[Wendy]

Good afternoon,

I'm working on some changes to my site and implementing 5 affiliate sites which will go live one per week over the next 6 weeks. All sites are new, so not a lot of traffic at first. I will be providing a data feed to each site, and Merge Mill will be used to generate static pages for the affiliates, however, each site will be sharing my shopping cart, order status manager, etc, with proper headers for each site. (these affililate sites are NOT on my server)

My store currently has 2446 categories and 6803 products. However, before these sites go live, I will be adding another 6500 products and probably another 2000 categories.

BEFORE I invest all this time, effort and money into implementation of this project, I figured I better ask here for advice as to whether or not this is too much for a MM4 store?

What would the limit be as far as approaching the "overload" stage for Miva Merchant 4?

TIA

[dotCOM_host]

If you stay on v4, yes, you can run into serious performance issues with a large number of products and categories. On v5 with mySQL back-end - that's a completely different story. Probably the biggest store we currently host on v5/mySQL has over 2.5 million products and 36,000 categories. Page loads are in the sub 1 second range, as are search results.

So, the answer to your question is really dependant on whether you are going to stick with v4, or whether you can migrate to v5/mySQL.....

[Wendy]

hmm..........bummer!

[David Pontius]

If you have MM4, you should be okay if you have a fast dedicated server, and it's MM4 compiled. It's probably cutting it close though. I have about the same currently, over 6000 items with no issues, but am planning to migrate to MM5 next year.

Dave

[Wendy]

Thanks Dave -- I'm on a HOSTASAURUS dedicated server -- so server load won't be an issue! :)

[David Pontius]

Thanks Dave -- I'm on a HOSTASAURUS dedicated server -- so server load won't be an issue! :)

Hey, me too! The nice thing about Hosasaurus it they'll work with you. About a year ago our server load had increased, so they moved me to their newest server option, at no cost.

[Wendy]

Oh you don't have to tell me how awesome they are -- I already know! :D

Last week I emailed David because I wanted a quote on the cost to add a service to my account for one of my client's and he wrote back like 10 minutes later and was like "okay it's all set up" and I was like "how much?" (cuz I wanted to pass the cost to my client), he was like "no charge".

They're AMAZING! And so patient with all the newbie questions!

I love Hostasaurus! giggle

[Brandon MUS]

I think we could all sit here and praise Hostasaurus. They saved my butt just yesterday.....

Back on topic: Even beyond performance, I think MM5 is a much better solution. We haven't migrated yet (finishing other projects), but the reading I've done makes it more much dynamic, configurable and expandable.

[Wendy]

I'm not ready to make the MM5 jump just yet -- I have $1000's of dollars of modules in my store and I'm just not ready to "start over".

If I do, I'm not sure it's even going to be Miva -- not when there are so many other competitive products that come with stock features that you have to buy a module to do the simplest of things with Miva.

I've been using another shopping cart for new sites, since this is Miva's forum, I won't name names.

It is my intention to stay with MM4 until I suck out every penny I've invested in my store. Thank god for all the 3rd party developers because I was a complete newb when I went with Miva and without the 3rd party modules, available, as well as the original list where I learned how to customize my store, I'd still be sitting with a stock Miva store.

[William Davis]

Very interesting post. We are currently working on a new site with 20k to 25K (thousand) unique products. Have not yet decide on how many Categories, but we doubt it will be any where near of Wendy's post.

1. Should we consider M5?
2. Is it stable? (I am under the impression it took Miva 4.00, several revisions before they worked out most bugs.)
3. Are most modules available?
4. Learning curve -how bad is it? We currently have another site on M4.24c.

[dotCOM_host]

1. Should we consider M5?
2. Is it stable? (I am under the impression it took Miva 4.00, several revisions before they worked out most bugs.)
3. Are most modules available?
4. Learning curve -how bad is it? We currently have another site on M4.24c.
1. Definitely. It's much faster than v4 ever was, is running on mySQL back-end so it's more flexible and allows integration with other systems or building your own custom front-end store while accessing all MM databases natively, and has better customizability options through its StoreMorph technology.

2. Yes. Have not seen any stability problems with it - it works very well. There have been a few glitches here and there that MIVA resolved already, and the few glitches they haven't fixed yet are mostly just functionality issues, calculating taxes or shipping, for example, but overall things that won't affect your store uptime.

3. Unfortunately, not yet. There are quite a few modules available, and some are really obsolete since you can do a lot of the UI work in MM5 out of the box, without any additional modules. However, there are still a few popular v4 modules that are not available for v5 - for example, soft doods delivery (downloadable products). Have a look at your store, see what features and functionality you will need, and see if you can accomplish all that with MM5 and 3rd party modules. If modules don't exist yet - either stick with v4 or contact developers to see what it would take to upgrade their modules. If you pay them, chances are they may upgrade a module they have been delaying on under time and financial constraints required to upgrade their modules to v5.

4. There is a bit of a learning curve, but if you don't faint looking at HTML code, chances are you will figure out StoreMorph fairly quickly. If you are completely alergic to any sort of "coding", it may not be for you, but then again if you are using v4 and have seen any of the template modules available for it, the migration to v5 is not all that difficult. Yes, you'll need to spend a few afternoons testing things, trying different things, etc, until you get a hang of it, but it's pretty straightforward overall. And of course, there are plenty of people on this forum who can help you with StoreMorph coding issues if you do get stuck.

[William Davis]

4. There is a bit of a learning curve, but if you don't faint looking at HTML code, chances are you will figure out StoreMorph fairly quickly. If you are completely alergic to any sort of "coding", it may not be for you, but then again if you are using v4 and have seen any of the template modules available for it, the migration to v5 is not all that difficult. Yes, you'll need to spend a few afternoons testing things, trying different things, etc, until you get a hang of it, but it's pretty straightforward overall. And of course, there are plenty of people on this forum who can help you with StoreMorph coding issues if you do get stuck.

"StoreMorph"? Whats that? -the only time we need to work with HTML is when we need to change something in VK templates modules. Is M5 more code dependent -we have two peoplen who work on the site that don't know one line of code.

[dotCOM_host]

If you know how to edit VC templates (not just settings for the VC template modules) then working with StoreMorph is very similar. The 'tokens' may be different and there are more options available to you, but the concept is about the same. You may want to read up on StoreMorph on MIVA's web site and see if it's something you are comfortable with. Otherwise, I would suggest outsourcing your store development to a qualified MIVA developer - let me know if you need any suggestions.

[Wendy]

I'm willing to bet there are MM4 stores with hundreds of thousands of products and categories that may have some performance problems if they are not running static pages.

However, if my own site is using Merchant Optimizer & the affiliate site's shoppers aren't even on my server -- and only hitting my miva store and server for the checkout pages I'm wondering how much of a problem it will really be and if it would really be worth trashing my entire store that I've spent $1,000's on not only in terms of money, but sweat equity?

I guess I'm stubborn, but I just don't think I want to take that MM5 jump until it has the same and complete functionality as my current store.

[William Davis]

Remik, I have spoken to five different Miva developers that are active participants in this forum. None have even given me the courtesy of following-up after our initial conversation when they said they would. Not even to give me a quote or to tell me they are too busy. Please provide me the names -I hope it's none of the ones I have spoken with. I would never mention names because it not professional -not even off list so please don't ask, I would just never refer them.

For the record it’s not Leslie, we love her!

[Vic - WolfPaw Computers]

Personally I think the two are nothing a like. StoreMorph is very limited in what 'tokens' can be used and where without the need for custom components. At least with the Viking Templates and OpenTokens, there are published resources for what tokens you can use and in what portions of the templates.

I think StoreMorph is crippled by the lack of examples and a complete listing of tokens and variables that can be used.

Now Weiland's Toolkit (http://www.emporiumplus.com/merchant2/merchant.mvc?Screen=PROD&Store_Code=wcw&Product_Code=1AA00223) breaks down some of the barriers, but I still think StoreMorph comes up way short to replace the functionaliy lost from v4/OUI.

[Bruce - PhosphorMedia]

True, for now. However, the long term benefit is that the "token" langauge is MIVA's and they can move it forward, while ensuring backwards compatibility and without "fracturing" the language and syntax. One of the drawbacks for the average end user was decerning the difference between template tokens and OpenUI (as you know from the numerous corrections you've made to posts like "Why doesn't this token script work?".

Technically, I think the basic concept of SMT is better than OpenUI...although, it also suffers from the lack of "put it there" ability. Meaning, a developer can't really make a module and offer the dead simple "Where do you want it" selection. That being said, there are always a trade-offs.