You are viewing a single comment's thread from:
RE: Proposals: STEEM citizenship comes with both rights and responsibilities
Thanks for the mention!
A couple more have been added since you wrote this so there might be even more to get excited about.
I just hope that some of use receive enough votes to put our words back into action!
Awesome. It's great that something new and exciting is happening.
The developer portal... I probably shouldn't be bringing this topic at all but the closest single word to describe it... probably "rubbish".
The last pull on github is about six years ago, lol
It all looks abandoned and that isn't good for attracting new developers and investors...
I spent almost two days to try to install steem-python on Windows 11 and then gave up after trying for more than a dozen times. Fixing one install error, another one popping instantly. I want to use the most up-to-date python distributions, can't risk my pc being compromised. Moreover, I have the feeling that even I manage to install it, then some stuff won't work due to API methods changes. No wonder some witnesses have their own libraries, developed from scratch, non-open-source.
I think this all has to be addressed if we're looking in for a brighter future :)
Maybe remlaps recent post could help you
https://steemit.moecki.online/hive-151113/@remlaps/getting-steem-python-to-run
Thank you, that definitely looks promising, I'll give it a try, although meanwhile I was having fun with json-ing the APIs directly. :)
@the-gorilla, I couldn't find in Condenser a method that returns a list of the active API full nodes?
Beem has NodeList().get_steem_nodes() and I have no idea where that info comes from...
Code:
from beem import Steem
from beem.nodelist import NodeList
nodelist = NodeList()
nodelist.update_nodes()
nodes = nodelist.get_steem_nodes()
print(nodes)
Output:
['https://api.justyy.com', 'https://cn.steems.top', 'https://api.steemit.com', 'https://api.steem.buzz', 'https://api.steem.buzz', 'https://steem.61bts.com']
The NodeList class has some standard nodes and updates itself depending on availability. There is an account @fullnodeupdate for this purpose. The results of the check are stored there for everyone to access. Unfortunately, the node list is no longer maintained by beem (on Steem?). The last account update was in November 2022, so the current new nodes (e.g. mine) are not even included there.
If you would like to know which current nodes are available with which response times, you can use these pages:
https://steemyy.com/node-status.php
https://ecosynthesizer.com/steem/nodes
You can add the best nodes for you to the Steem object as a list.
Thank you! When I was searching online for the most up-to-date list of the full nodes, I found a few resources that are different from those you shared :)))
Justyy's page says these response times are from the browser but that didn't made any sense to me, as logically that would be the response times from where that page is hosted, so I pinged some of the nodes from my command prompt and got completely different response times.
That whole matter came to me as I am a sucker for efficiency and I'd like to use the fastest node that's as close as possible to me :)))
The most logical thing would be the full-nodes list to be available through a Condenser method but it seems that isn't available, one of the downsides of decentralization, I suppose :)
Not necessarily. For example, if the queries are executed on the client side. This is also the case on Justyy's page. I can also confirm this, because in my case the scripts are blocked by default and I can see on Justyy's site that the information is only available when I allow the requests to the respective servers. The speed is therefore only measured in the browser.
I can't tell you how the times are measured. However, if you measure the speed with Ping, the data will of course differ.
I think you're right and it seems to me those response times on the page are bigger because they include the api method processing time as well ;) Ping will always be faster unless it's a method that does nothing :)
Ah, I can't help you much with Python. Anything I do with Steemit is either with Condenser (on the main site) or via PHP / JavaScript and I predominantly use steemchiller's API (for reading data anyway)
I'm sure that I saw a superior Python library but can't remember where. I don't think it's been checked for security vulnerabilities but I think it's had a lot of work. @moecki might remember where it is.
Well, I have working knowledge of both JS & PHP but Python... is just more fun :D
Perhaps you mean the Beem library? It's also quite old now and what I don't like about it, it works both on STEEM and HIVE, so potentially unsafe too. Probably I am just too cautious. :D
Condenser - I suppose you're the perfect person to ask to confirm there are already many differences between the Dev Portal docs and the up-to-date methods? :)
Thinking more of it - a new proposal for maintaining the Dev Portal makes sense and while I would normally hesitate to do real development work, I would love to spend some time working on updating the docs.
I’d like to think that’s something that would get support.
I’ve not looked - if I need to do something, I’ll see how it’s done elsewhere or find an equivalent. Some of the communication with Hivemind is something I’ll try to understand better this year but other than the /tags fix, I’ve not needed much knowledge of that yet. Although my introduction of avatar and cover images in communities which crashed the site might have benefited from this capability!
For my own projects, I use steemchiller’s API for reading data and steemjs for broadcasting.
I see some differences that I don't understand yet.
For instance, a json request to condenser_api.get_account_count results at the moment:
1930142 (when I use https://steemd.steemworld.org )
and
3655647 (when I use https://api.steemit.com )
More confusion in. :D
2M accounts sounds more realistic :)
Good question that I can't answer! It could be related to the Hivemind database that each API is querying. api.steemit.com is guaranteed to contain everything but it's more likely to be a slower API. So 3.6m is probably true.
I don't know... I mean about that number. I guess I have to write a small script that loops all the APIs and calls that same method. Some other basic methods as well, to compare the output and the execution times, perhaps that would be useful.
By the way, another major block of info that the Dev Portal lacks, is Glossary with all the building blocks of STEEM such as, what's Condenser, what's Hivemind. There is a glossary right now but it's missing a lot...
I need that information. I currently just run condenser on my localhost. I initially did this through Docker but it was faaaaar too slow so I eventually managed a workaround.
I'd like to know how all of the parts link together. I'd like to fix the search box at the top of the screen but that's another repository too - and I also think it's another link into Hivemind. So much to learn and understand and so little documentation to help us!