Re:Useful Python Libraries for Startups
by apalmer on Tuesday June 30, 2015 @02:17PM
(#9807331)
Not sure why these recommendations are start up specific. Nothing here seems particularly 'startup'. These are just good simple general purpose web related python libraries.
Re:Useful Python Libraries for Startups
by zallarak on Tuesday June 30, 2015 @02:27PM
(#9807425)
I can see that perspective. My thoughts were that each of these were not the "best" way to do things, but provided a great effort/benefit ratio. For example, if you have a very high traffic website, then you're better off using a CDN for static files instead of Whitenoise.
Re:Useful Python Libraries for Startups
by copsarebastards on Tuesday June 30, 2015 @03:08PM
(#9807702)
If the effort difference between installing nginx and using Whitenoise makes a significant difference in the success of your business, there are much larger hiccups down the road which are going to put you out of business, so you might as well give up now.
Re:Useful Python Libraries for Startups
by dimino on Tuesday June 30, 2015 @02:03PM
(#9807218)
Python-dateutil is unintuitive as all get-out in my humble opinion, so I recommend dateparser instead (recent release has many great improvements over dateutil, but check it out for yourself to decide [0]).
Some AWS specific functions, particularly for newer products or options, aren't always there immediately so you end up having to use boto in those cases. You also sometimes use provider specific options on functions, so it's not totally write-once and ignore.
Re:Useful Python Libraries for Startups
by anandpdoshi on Wednesday July 01, 2015 @12:05AM
(#9810010)
Disclaimer: I am one of the developers that built this.
Try Frappe Framework (https://frappe.io). We have received feedback that it is really good to quickly build internal projects or prototypes.
Also ERPNext (https://ERPNext.com) is an open source erp that is feature rich and has a good UI. It can help any startup to manage their business without resorting to complicated excel sheets based management.
Re:Useful Python Libraries for Startups
by zallarak on Tuesday June 30, 2015 @02:02PM
(#9807216)
We worked with a third party API for a data source which had inconsistent formats for dates. Also, we A/B test inputs and its awesome to be able to change how users input dates on the front-end without server-side code changes.
Re:Useful Python Libraries for Startups
by tuyguntn on Tuesday June 30, 2015 @03:04PM
(#9807675)
If only for testing I would agree with whitenoise, but in case of startup and you are already serving static files why do you want to replace NGINX with Whitenoise? any specific reason for Whitenoise, maybe it is better/simpler/efficient than NGINX?
Re:Useful Python Libraries for Startups
by tuyguntn on Tuesday June 30, 2015 @03:16PM
(#9807749)
If you are using python code for serving static files, probably you are not seeing lots of traffic yet, I guess you should reconsider your decision and watch "Simple made Easy"[1] talk by Rich Hickey
Re:Useful Python Libraries for Startups
by danneu on Tuesday June 30, 2015 @06:23PM
(#9808795)
Looks like Whitenoise can gzip your assets, add a hash to the filename, serve them with far-future headers, and then selectively serve the gzipped version based on Accept-Encoding headers.
Put that behind Cloudflare and your origin server is only hit when an edge location is warming its cache.
Sounds Hickey-tier simple to me, especially compared to your advice of "just use and configure Nginx".
Re:Useful Python Libraries for Startups
by JelteF on Wednesday July 01, 2015 @03:04AM
(#9810376)
When I needed a Python generation, I created abstractions between Python and LaTeX. The documentation is far from finished, but it should be fairly understandable with the examples.
Re:Useful Python Libraries for Startups
by truncate on Wednesday July 01, 2015 @04:24AM
(#9810596)
Every startup gotta deploy stuff. I remember Fabric a useful tool for quickly deploying and automating various tasks - http://www.fabfile.org/. Also not a library, but Gunicorn is pretty quick easy way to setup a quick decent HTTP server over your Python WSGI app.
Re:Useful Python Libraries for Startups
by bdarnell on Tuesday June 30, 2015 @04:29PM
(#9808223)
The phone numbers library isn't just counting digits; it has a database of which area codes exist (as well as other lesser-known rules, depending on the country). This makes it more effective at filtering out invalid/fake phone numbers, but it does mean you need to stay on top of the latest version as its database changes over time.
Re:Useful Python Libraries for Startups
by zallarak on Tuesday June 30, 2015 @12:59PM
(#9806708)
Agreed. My goal was to share what I thought were solutions to generally common problems, and were't incredibly easy to find (i.e. > 3 google searches to find). The three libraries I selected were general tasks that lots of developers end up having to re-invent (especially phone number parsing).
Re:Useful Python Libraries for Startups
by craigds on Tuesday June 30, 2015 @06:34PM
(#9808848)
Is parsing phone numbers really a common task for startups? (I've never needed to do it). Is the use-case just validating that the user didn't accidentally type their phone number incorrectly? (Seems if the user wants to give you the wrong phone number, they can just pick one out of a directory...)
Re:Useful Python Libraries for Startups
by merb on Tuesday June 30, 2015 @02:24PM
(#9807384)
Yeah but there are other Libraries around, and some people maybe like the "real" libphonenumber more.
Nothing against the article (not sure why I got down voted), I don't think it's bad to have a subjective meaning onto something, as long as I clearly state it. And the author also said it, so my comment was not against the article, that was just my subjective meaning about the article. ;)
May I ask what's the better way to parse international phone numbers?
Requests: HTTP for Humans http://docs.python-requests.org/en/latest/
I would say not much, but it's much more pythonic.
[0] https://github.com/scrapinghub/dateparser
https://boto.readthedocs.org/en/latest/
Some AWS specific functions, particularly for newer products or options, aren't always there immediately so you end up having to use boto in those cases. You also sometimes use provider specific options on functions, so it's not totally write-once and ignore.
Try Frappe Framework (https://frappe.io). We have received feedback that it is really good to quickly build internal projects or prototypes.
Also ERPNext (https://ERPNext.com) is an open source erp that is feature rich and has a good UI. It can help any startup to manage their business without resorting to complicated excel sheets based management.
Once I switched to python 3, datetime has been sufficient for my needs as timezone is also handled.
[1] http://www.infoq.com/presentations/Simple-Made-Easy
Put that behind Cloudflare and your origin server is only hit when an edge location is warming its cache.
Sounds Hickey-tier simple to me, especially compared to your advice of "just use and configure Nginx".
Docs: https://jeltef.github.io/PyLaTeX/
Code: https://github.com/JelteF/PyLaTeX
For Flask-- it removes the dependency on a global request context: https://pypi.python.org/pypi/Flask-Injector
If, however, you use Python at your startup it's more interesting, while still somewhat subjective.