That'll probably be my plan then

Might have to hack-up manage.py to work with Pony's migrations

Probably in the future we can add some Django-Pony adapter. Also, there was a projwct which automatically creates Pony entity for each Django model. Maybe it is possible to revive it https://github.com/nnseva/djony

Oh cool, for now I think I'd probably just not use Django haha. The current project is using hug.

Still don't get what hug does

It takes falcon and does magic

Okey, now what does falcon do?

HTTP APIs

Ah, I see. I use flask, so I just use json... What are the advantages?

Its faster, and you don't have to deal with globals

I like flask a lot, but I think I'm gonna switch to hug as my main for now on

Welcome :)

Hi! I saw Pony ORM recently and reading docs now :) About me - study data science on MOOC and making some scientific analytics in my lab.

Having defined a field and populated a few records -- I change my mind and decide that the field should have been unique. What must I do to change the field's schema and have that written back to sqlite? Can't find it in the doc. Thanks!

you need migrations for that, which are still a work in progress

for now, you have to alter the table manually

OK! Thanks again.

I wish there was a group like this for general Python, but I can't find one.

Kinda tempted to make one, not sure if it'd just be noise though

Kinda tempted to make one, not sure if it'd just be noise though

Yeah I'm curious if there would be enough interest

Shall I make one and we can try it?

Sure

https://telegram.me/joinchat/AaGuz0De4WkhL72Cxb5MBA

good morning ?

hi

hello

Someone uses a PonyOrm with oracle without "triggers" for insert into some tables ?

i have old project with bad arhitecture and noone tables havent triggers. i solved my problem by "before_insert" entity event

with 2 methods "check_trigger" and "create_trigger" )

So you still use triggers, but create them dynamically on first insert?

Ok, i know i can ) But , does pony to work with Oracle without triggers ?

like a Hibernate Orm in java

Currently Pony relies on triggers when working with Oracle. I think we can add another way of working with identity columns. Please open new issue for that, and provide a link to description how Hibernate does it: https://github.com/ponyorm/pony/issues

wow

Last question. :) What happend if i have tables with Data but havent triggers and i start Pony db.generate_mapping(create_tables=True) That method not erasing my Data? Just creating tgiggers if Table havent it? Maybe anybody know answer or u Alexander. But i cat test it if you dont.

. Thank you for your wonderful project! ? Happy holidays!

If you use create_tables=True option Pony should create missing tables, indexes and trigers. It should not erase any data, so you can try. But in our next release (which we should release really soon) we change how intermediary many-o-many tables are named. If you use create_table=True option, it is possible that, as part of upgrade to new Pony version, generate_mapping will automatically rename many-to-many tables if table name is not specified explicitly.

. Thank you for your wonderful project! ? Happy holidays!

Thanks! Happy holidays!

Thanks!

Welcome both

Hello. I currently work with Django and PonyORM looks very superior to the Django ORM. The syntax looks much nicer, and on migrations, one frustration with Django is that you can't rename a class easily in their migrations. Would love to see a Pony for Django (to deal with existing projects), and Pony with Flask looks like it could be even better than Django. When will pony.js be released?

Hello. I currently work with Django and PonyORM looks very superior to the Django ORM. The syntax looks much nicer, and on migrations, one frustration with Django is that you can't rename a class easily in their migrations. Would love to see a Pony for Django (to deal with existing projects), and Pony with Flask looks like it could be even better than Django. When will pony.js be released?

I recently used Pony and Flask on a project and they worked fantastically with each other

I work on oldest java project with hibernate and old spring, and recently i try fast fix some bags in DB with python and pony, and i say u - java sux if u need fast development

I fell in love with python & ponyOrm

Hello. I currently work with Django and PonyORM looks very superior to the Django ORM. The syntax looks much nicer, and on migrations, one frustration with Django is that you can't rename a class easily in their migrations. Would love to see a Pony for Django (to deal with existing projects), and Pony with Flask looks like it could be even better than Django. When will pony.js be released?

Hi Dan, We've tried pony.js in one of our projects and found that it requires some more development. We are not ready to release it yet. In the same time, we think that having a possibility to easily represent Pony ORM objects in JSON format would cover many use cases. This feature we are going to release after the migration tool.

Thank you. In your Talk Python To Me interview you also mentioned database reflection. Where does that feature stand priority-wise? If you have migrations and reflection, then I could see reworking existing Django projects with Pony/Flask. (Incidentally, Django has reflection(https://docs.djangoproject.com/en/1.10/ref/django-admin/#django-admin-inspectdb)

Dan, Migrations are almost ready, we release it in a matter of weeks. The reflection is the priority, we combine it with the https://editor.ponyorm.com, so that you can launch the reflection tool and get the database diagram and the classes.

I work on oldest java project with hibernate and old spring, and recently i try fast fix some bags in DB with python and pony, and i say u - java sux if u need fast development

:( yes

I fell in love with python & ponyOrm

Me too :D

Welcome!

I recently used Pony and Flask on a project and they worked fantastically with each other

Have you ever used Django?

Have you ever used Django?

Just the tutorial a handful of times

I want to use pony, but I can't while it still doesn't have migrations.

I want to use pony, but I can't while it still doesn't have migrations.

Should be available in a few weeks

Welcome

i be suprised see telegram group in site))

how about thread safe?

should we expect asyncio/tornado support in near future?

should we expect asyncio/tornado support in near future?

in the near future we release Python 3.6 support and migrations asyncio/tornado is not easy feature, we are not planning to start working on it during the next 6 month

it's a main reason I didn't migrate on pony

In order not to block asyncio ioloop, you can run queries in separate threads. Here is a small example of PonyORM query in aiohttp handler: https://gist.github.com/grudko/4c1dcd2a2eb29bc3d98cd19bfd916cbd

Most queries that retrieve objects by primary keys can be run synchronously from asyncio/tornado, and it will be faster than using async protocol

In order not to block asyncio ioloop, you can run queries in separate threads. Here is a small example of PonyORM query in aiohttp handler: https://gist.github.com/grudko/4c1dcd2a2eb29bc3d98cd19bfd916cbd

So this part? `async with threadpool():`

Oh, I used threadpool from https://github.com/agronholm/asyncio_extras

Maybe I should learn async

The pattern @agrudko suggests may be helpful if db_session contains long queries (analytics, aggregation, etc.)

Pony ORM Release 0.7.1: Python 3.6 support and some bugfixes https://blog.ponyorm.com/2017/01/10/pony-orm-release-0-7-1/ yay ?

❤

Nice! Great job.

Pony ORM Release 0.7.1: Python 3.6 support and some bugfixes https://blog.ponyorm.com/2017/01/10/pony-orm-release-0-7-1/ yay ?

Throw an error with more clear explanation when a list comprehension is used inside a query instead of a generator expression: “Use generator expression (… for … in …) instead of list comprehension [… for … in …] inside query” yep, that was me xD

Woohoo!

7.1 doesn't come with migrations correct?

7.1 doesn't come with migrations correct?

Correct. 0.8, which we release soon, will

Correct. 0.8, which we release soon, will

Great to hear!

see yellow ribbon

see yellow ribbon

That's strange

Hello everybody. I use Pony for connect to the remote Postgres database. Periodically, the database connection is lost and Pony returns an exception: File "... / lib / python2.7 / site-packages / pony / orm / core.py", line 1566, in close try: provider.rollback (connection, cache) File "<auto generated wrapper of rollback () function>", line 2, in rollback File "... / lib / python2.7 / site-packages / pony / orm / dbapiprovider.py", line 59, in wrap_dbapi_exceptions raise InterfaceError (e) RollbackException: InterfaceError: connection already closed After that, Pony automatic reconnect, but my function call is aborted. Is it possible to test the connection and forced to reconnect? I have not found the necessary methods in Pony docs. Maybe you need to look at psycopg2 docs?

Hi Vasiliy! Pony should be able to reconnect automatically, if the connection was lost not in the middle of transaction. I think that your problem caused by the following sequense of actions: 1) An application code performs some query; 2) Due to some problem in configuration (for example, an insufficient size of some buffer) the server drops the connection, and Pony gets an exception from psycopg2; 3) Exception unrolling leads to exiting from db_session; 4) Pony tries to handle exit from db_session and attempts to peform rollback, assuming that the connection is stlii active; 5) At this moment we got another exception, "connection already closed" which masks previous exception. I think we need to improve handling of rollback so it does not mask previous exception in case the connection is closed already. Regarding your situation, I think the right way is not to attempt to reconnect, but to prevent an exception which lead to connection closing. After we fix handling of rollback, you should be able to see this original exception and will be able to fix its root

Hi. I have a question about ready to use in production. Can i safely upgrade my database schema? I saw migrations not implemtented in before.

We are working on migration tool right now. It should be released really soon

Without this tool only way it's handmade migration?

If you add a new entity or many-to-many relationship, and you call generate_mapping with create_tables=True, Pony will issue CREATE TABLE commands automatically. If you add a new attribute to previous entity, right now you need to perform ALTER TABLE command manually. Typically such commands are pretty simple, but with migration tool it will be easier to write and execute them