Its a busy time for me, so progress on ZenMagick has been up and down. This week has been good – I’ve got a first working draft of the block admin UI. Only, I just realized that the current db structure doesn’t allow to manage separate settings per theme – hmmm.

Last week I had this urge to see some progress, so I worked my way through a number of small issues – nice to see the number of open tasks go down, although the big ones are still out there waiting…

I am very happy (and also rather tired) to finally having released ZenMagick 0.9.10.

The main focus for this release has been to get the new admin UI up and running. RubikIntegration have been generous with help on the design and part of the implementation and I’d like to say ‘Thank you’ for that.

Of course that is not all, but instead of repeating things over and over again, just a couple useful links:

Official Release Notes

Details about upgrading

Other than that there isn’t a lot to add, except for a link to the ZenMagick 0.9.11 roadmap. It’s already filled with all sorts of things. It will be interesting to see what we can mange to get done. If there is anything missing let me know.

mano

A mixed week with some good results, too!

Testing is progressing, unfortunately it also uncovered a few things that need fixing first :/ On the plus side I’ve upgraded useo2 to the latest version and there is even a script that can patch the code so it runs in the new admin.

Most of the required changes are using ZMDatabase rather $db and a couple other things, for example support for the category urls.

I am also very close to getting a couple sites upgrade to the latest 0.9.9+, or, better 0.9.10RC. It will be good to get some real usage to see how good the code really is :0

Apart from that there are a few improvements that now allow to cache individual template files. For example, each sidebox can now be cached or not. The current caching code itself is rather simple, but it should be easy to make it as complex as required.

This is now enabled as default, as is support for PHP short tags. That is done as a custom filter that will change all <?= or <%= to proper <?php echo syntax.

Again, this is just to get some mileage, I do not expect that a lot of people will be using this initially.

Category caching now works properly, so even a couple thousand categories won’t slow things down too much on the storefront side. The backend/admin will need a bit more time. Again, template caching will make things a lot easier in those cases.

I’ve added a new APC based cache implementation. It does cache things, clearing the cache is still a bit buggy, but if you are keen you can always use the APC admin interface to clear the cache.

The cache admin page now properly shows all used caches and groups. I hope that will make it a lot more useful Outstanding here is some sort of central caching config for each cache type. This will be possible either with smarter settings or the still unused context code.

As it turned out, .po and .pot file generation based on the current string scraper code wasn’t too hard. In fact, the hardest bit was perhaps to validate the created files.

It’s not so easy to find usable gettext binaries for Windows.. I ended up using a package from CubicWeb. It does crash, but only after having done the bits I want it to do

So, now there is a new translation project ZenMagick with a component admin UI on trasiflex. So far I’ve created a (default) en_NZ and de_DE locale. Since ZenMagick uses default strings in the code the default locale is at a 100% – yay. German is at 0% :[

If you are keen to work on a different language, let me know and I’ll add it to the repository.

I should add that the code is still using the .yaml files that are also located in the locale folders. As long as there isn’t a useful translation I can’t really see any reason to switch yet.

I’ve started a German translation of the admin UI that will be included ni the next release. It has become quite obvious that I am not a great translator. In fact, the german translation sounds a lot mor wodden and artifical to me than the english original. Go figure…

Anyway, it’s a start. Right now the strings are in a .yaml file, but in the long run I’d love to move to a web based translation tool and also move to the gettext based pomo code. If anyone is interested in helping with translations let me know.

In any case it will be good to have a second language to test switching and making sure all strings are covered.

Well, looking at the new snapshot I uploaded two days ago it should be easy for anyone interested to judge for themselfs

Having said that, since then I haven’t been idle. The catalog tree now actually reflects the category selected, editors are improved, the version check widget does some proper version comparison and a few funny things have been fixed.

Testing will start in earnest now and I expect the next snapshot really to be of RC quality.

Also, there is ongoing work on the website. I’ve started updating the wiki roadmap pages and even already created a new one for 0.9.11!

If you have any interest in a new shiny ZenMagick (and therefore Zen Cart) admin UI, now is the time to come aboard – any input is welcome and since I have no preference for what to implement next this could be your chance to decide what to implement next!

Time for another update – not the weekly one but just about the catalog manager refactoring.

The last few days I finally had a serious stab at getting this to work and I think I’ve got the basics done (screenshot at the end).
Things do not look spectacularly different from the old code, so don’t expect too much right now. However, it looks like we’ll be able to get the main layout done properly in the next week or two.

Finally, I’ve just added the task ‘Create standalone admin installer’ to my list and I’d like to mention that to hopefully get some feedback on that.

Since the new admin UI is completely separate, it is absolutely possible to run that without any other storefront/templating code. I wonder if that would be attractive as a first step to migrate to ZenMagick (should someone wish to do so). Also, no matter what, the more the merrier, so even if people would just use the admin app it would mean a lot more exposure (and finding bugs

It’s good to finally see some light at the end of the tunnel. The trip is far from being over, but at least there is now a degree of certainty about whether things will work or not.

Once the admin is stable (not complete, just everything working) it will be a lot easier to work on adding features, etc. Also, at that point there will be some more changes to the store code to catch up with all the good stuff that got created to make admin (and the MVC code in general) better.

And, once all that is done it really is time for a 1.0 release! Way to go, but it doesn’t look as impossible as, say, 6 month ago.

I just found out that, in its latest incarnation, jstree supports jquery themeroller – just what I needed! The tree is coming along nicely now, although a lot now depends on creating a proper theme…

Things have, again, been revolving around the new admin UI. I’ve made substantiel progress with the nagivation, although it is not final. Also, there is no actual design applied yet which means unless elements are created by jquery there isn’t much style at all.

On the plus side it looks like it might be possible to style most of the UI using the jquery themeroller.

As usual there have been some improvements and additions to core and store code along the way, however nothing substantial this week with perhaps one exception: I’ve added support for user based authorization. That means it is now possible to control access to admin pages by granting permission to a role (group) or individual users. (Still no UI to manage this, sorry).

As far as UI content is concerned I’ve done lots of small improvements and prepared almost all content for translation.

The biggest chunk of work this week was spend on providing a way to manage all zencart configuration options in the new admin UI. As it turned out, using widgets to render the form elements was really as easy as I had hoped! That means no need to load additional zencart code at all. For the technically inclined, the code parses the set function value and then it’s more or less just a switch to select and populate the appropriate form widget.

Finally, because I am a bit proud a nice new screenshot: