That's also why I want to add a mongoDB backend. (as well)
I am also thinking of moving the Views completely to javascript (ejs would probably be my choice then).
That would result in a client <--JSON --> Server <--- JSON ---> Database
szenario. Which looks pretty clean, at least to me.
No problem, I just took the standard Bootstrap layout.
But right, the red info box is a little heavy weight. But I like to point out that this is not meant to be production ready, yet.
Anyway I am not a Designer.
I like my blog template, though ;)
NDB is a good hint.
But the problem remains that I want PoW to be able to support SQL DBs also. I think SQLAlchemy is a really mature framework for that area. So I need to generate the models in two ways..
I think I need to separate this as an extension since I do not want to have two PoWs, where one is for GAE and one for SQL.
I think flask is ment to be really micro.
Misses ( and wants to miss IMO ) the generative approach.
Also misses ( and wants to ) the db layer etc
So these approaches are really different.
Code: i will cleanup and DRY the generatots. But generated code is the one i was working on and that the user will work on and read. This is ways better.
Doesn't want to be better but more generative, so quicker first app results. Idea is to add many rails-alike helpers like pagination, observation, etc...
Easing small apps life .;)
Really works fine from really many different devices for me... Damn i really dont wanr to ask if you can replicate thsi .. 'cause you can.
But i just cant advice something other than to try a different browser...
Pep8 => yes, need to review / rewrite
Dry => again yes, working currently on it for generate_*
did you look at some genrated code, think this is cleaner, though. The generators are the oldest ..
Intersting that you chose a different framework for almost any field. So there is room ;)
I am thinking about a google app engine port but sqlalchemy is a showstopper here .. Need the models to talk GQL ... So i stopped that port for now. Much to do in the base..
1 point by pythononwheels 7 minutes ago | link | edit | delete
Hey oinksoft. Dont just punch on top... To your points: 1. The shebang: yes, ouch, gets removed. 2. Pow_console: it is a remark that i too, code from there. I always comment the source if code is not mine.
3. The mentioned code snippets from the homepage are default twitter-bootstrap sources..
Criticism is ok but this is bashing...
So, no, its no joke ;)
One important remark: if you check it out from github, be sure to get the Beta1 branch.
The master branch is really long behind the current development. I also stated that in the screencast.
And as i think about it, some of the comments might refer to the master branch (like the one saying there is no directory structure) which is not the case in Beta1...
I wrote the PoW framework. And want to give some short feedback on the comments here.
Yes, i need to get things cleaned up.
Shebangs are old and not used. So can and will be dropped. Good point i really forgot they were still there.
Yes, I am rewriting generate_* to encapsulate specific actions in functions... Can be much cleaner..
Typos on page where really bad , sorry.
And yes, the videos where really too long. But were online in just for 2 days. I replaced them with versions of 6minutes and (i think) 14 minutes for the weblog.
Some other remarks are just not true:(or outdated)
1. how do i install ? => 1. Watch 6 min screencast (thats exactly why i made them )but its just a simple unzip and pip the needed libs.
2. Directories hardcoded => no, are configurable.
3. Base in just one directory. => no. Structured stubs and generates a very structured mvc dir-tree.
4. Server errors on first 2 days => yes, was due to serving with simple_server (which is really not for prod ) now running apache mod_wsgi. Stable since 2 weeks... ;).
Again: lots to do in code refinement of course. Please keep in mind its pre-beta. The focus of the last weeks was in the generated code, more than the generators. Can you also give some feedback on that ? To my mind much better an OO than the flat generators which where really more form follows function. (which is no excuse for having much potential in there)
Thanks for feedbacks. Working dayly in improving the thing ;).
But my main point was to have a fluent workflow.
That's also why I want to add a mongoDB backend. (as well) I am also thinking of moving the Views completely to javascript (ejs would probably be my choice then).
That would result in a client <--JSON --> Server <--- JSON ---> Database szenario. Which looks pretty clean, at least to me.