nopCommerce 3.60 roadmap and estimated release date. Let's discuss.

This topic was automatically closed 365 days after the last reply. New replies are no longer allowed.
9 years ago
I agree with the comments that the focus should be put for a while on performance and architecture.
My recommendations are:
1.- Use unit of work pattern to make all saves transactional. Wrapping all saves in a single transaction would also greatly increase performance.
2.- Convert all DB operations, services and controllers that use them to async. It's relatively easy (although a bit time consuming) and it would greatly reduce server load, allowing many more requests per server
3.- The already mentioned server farm support. It doesn't have to take advantage of all cloud services facilities initially, but at least it should support multiple instances.

In terms of architecture, I would suggest:
1.- Split domain in multiple bounded contexts. Entities like customer and product have many concerns mixed up, making them very heavy and complex. The domain could be separated into Catalogue, Sales, Reporting, Security bounded Contexts having lighter versions of the same entities on each context, but with only the right concerns on each. This would make future development and maintenance easier and also developing plugins easier.
2.- Load plugins in a separate app domain each. This would make the system more reliable, as a failing plugin wouldn't bring the whole site down. Also, it's currently impossible/dangerous to use commercial/community plugins on a customized nop if you've had to upgrade some nuget packages.
9 years ago
a.m. wrote:

I've prepared a draft of the roadmap for the next version of nopCommerce (3.60). Here are some of the most interesting ones:

- Web farm support (postponed)
- Ability to assign catalog price per currency (postponed)
- New modern public store theme (most probably)
- Move to ASP.NET vNext (if released in the near time)
- And a lot of other enhancements



Looks great! Way to go!!
9 years ago
My Xmas wishes for nopCommerce 3.60 by priority:

- Performance please!!!
    - OutputCache + Donut caching?
    - Differentiate read and write database connections
          - Read from slaves
          - Disable change tracking in EF for read connections
    - Try to avoid DB queries in For loops (We have already implemented this in some places and it works great!).
    - Lucene based search and faceted filtering?  
    - ...
- Webfarm support (including Azure)
- Change release and code maintenance strategy so it is easy to
    - Accept minor modifications and bug fixes to already released major versions (3.60, 3.60.1, etc...)
    - Feature branches
    - Create Spikes (vNext, EF7) without affecting normal development
- More general extensibility and customization points:
   - Add events when something happens in domain (not in db like now).
   - Pass more information to view models so you can do more things in Views (ex: Pass price as numeric value not only the formatted one).
- Razor based email templates (so customer's can add custom logic).
- New admin dashboard:
    - Bootstrap, .less, and font awesome based
    - New and improved design
       - Move from horizontal tabs to vertical "tabs"
       - Bigger fonts
       - May be we could get some inspiration with Umbraco dashboard (they have done a nice job there).
       - Reduce header size
    - Remove dependency from Kendo Grid (is it illegal to use it now?)
    - More extensibility admin points
       - Add new menu items to existing menus
       - Custom columns in existing grids
- Move to Github (Even Microsoft is doing it!)
- Entity Framework 7 (if released in the near time)
- Spec tests and more unit tests for the web.
- Create something like a "steering committee" so that responsibility does not fall solely on Andrey and his co-workers and the project is more open to the community.

Great job!
9 years ago
@rchamorro that would be awesome list.
9 years ago
In my view the major item that is missing from the roadmap is performance. I know a lot has happened in this regard in 3.4.  However when I loaded 100.000 products / 500 categories / bunch of attributes into database for performance test .. I realized I would still have to go through and manually optimize NOP so that users don't see a 7 second wait time every time they want to see a product page  on 8 CPU / 8 GB RAM server..

Filip
9 years ago
anyone has big database available for download to test ?
9 years ago
Request to show product images on the order detail page's "Products" tab for easier to know which actual products are ordered.
9 years ago
Hi,

I agree with the speakers above that we need to improve the vertical performance (that is to be able to handle more concurrent users). Web farm/garden and Azure instances is a very important step to take in that direction!

Anyone tried Varnish cache with nopCommerce? (https://www.varnish-cache.org/)

/Johan
9 years ago
I agree that a new modern theme is needed: a more visual theme with default icons that can be managed that are used to conserve horizontal screen space for categories, manufacturers, main menu, footer, and header links. For example, the header links could be replaced with icons like a Profile Silhouette, Shopping Cart, Email Envelope, and Wishlist Icon. In addition, instead of the popup for unread messages a call out (red overlay on icons that display a count) would be less intrusive and still indicate to users they have unread messages in their inbox.

Another UI improvement would be to put Product Reviews on the same screen as the product, either a tab or preferably near the bottom so customers can read them and get additional information and support for making a purchase.
9 years ago
I am somewhat new to a serious treatment of nopCommerce (3.50). I was surprise to discover that search only works against product names and does not incorporate category names, tags, descriptions, product specification (despite documentation), and product attributes (which seems like a fundamental oversight). Since all these aspect describe the product, one might presume they would be legitimate collective search targets. Consequently, provided I am not missing something, getting "search" working right would seem to be a high priority.
This topic was automatically closed 365 days after the last reply. New replies are no longer allowed.