Sounds nice, but I would guess it will run rather slow until caching is implemented.
Comment has been collapsed.
I've made some modifications to the tool.
It now runs as a Service, and keeps all it's data in MySql.
Subsequently, most commands take less than 1 second to execute.
In the near future, I'm planning on deploying it on a remote server (probably in AWS) for easier access.
Comment has been collapsed.
Tool location has moved to http://www.sgmt.name:8080/SGMT/
Anyone using it - please be advised you need to change your links.
Comment has been collapsed.
Update:
Added a new feature, this time designed to help group users rather than group admins:
CheckGameGiveaways - Shows the number of entries (of group users ONLY) every time a game was given away in the group. Response example: https://imgur.com/a/6ZzxQU5
The reason for this is:
When you give away a game in the group, you want to know how many entries it got last time it was given away in the group.
A SteamGifts search will give you the number of entries for each giveaway, but from all groups it was given away to.
For example: A game may appear to have 24 entries last time it was given away, but out of those 24, only 2 were from the group I'm giving it away to now. So if I create a GA only for the group - I will only get 2 entries at most (and not 24).
Comment has been collapsed.
I've taken the first step for my group to be integrated (sent an invite to sgmt.support).
Just a question: would it be possible to enable GroupUsersSummary to return in-group ratios? (total GA value given / total GA value won and total GAs sent / total GAs won)
In any case, thanks a lot for that tool. It will save me a lot of time.
Comment has been collapsed.
I've accepted your invitation, and it should sync with SteamGifts within a few days (you will see the user SGMT in your group in SteamGifts)
I don't quite follow your request. Can you please elaborate?
Are you talking about per-group or per-user?
How is it different than the data that appears right now?
Comment has been collapsed.
Thanks a lot.
Regarding my request, my group allows an in-group ratio of 0.5 for given/won GAs numbers and CV.
At the the moment, I have to do the maths manually using the SG group user page.
AFAIK, the values your tool shows are parsed from SG but the exact ratio calculation is not done (or I missed it).
Ideally, not to clutter already existing functions, you could add a new RatioChecker function. The user would select the group, then the desired ratio for GA numbers and/or CV (in my case 0.5) and the page would then display the list of users who don't meet the criteria (or would display numbers below the desired ratio in red).
I hope I made myself clearer but in any case, I'm already very grateful for the tool you've made.
Comment has been collapsed.
AFAIK, the values your tool shows are parsed from SG but the exact ratio calculation is not done (or I missed it).
Actually no. The values are taken from the tool's internal DB, which is taken by parsing giveaways data from your group.
So as far as I understand they show exactly what you need: number of GAs each user created & won in the group.
But the tool is limited to scraping up to 2 months back at any given time. So if your group is added today, the tool will ONLY get giveaway data starting January 1st. Later this data is preserved, so when new data is added you will not lose data from January.
The reason for this is to conserve memory and DB size.
As otherwise, you could add a group with 100,000 giveaways than span over a period of 5 years, and then tool would die trying to get all this data (because it will run out of memory).
If you need, as soon as user SGMT is added to your group, I can manually run it to get a longer period for your group (from what I see there is only 1000 giveaways there).
Comment has been collapsed.
Hi again. SGMT is not in the group members list yet but I noticed that my group had been added to the tool (thanks for that).
Do I have to wait longer or can you now extend how far back your tool will collect data?
Thanks.
edit: And I'm seeing a lot discrepancies in numbers in the different subtools. But may be that's expected at this point.
Comment has been collapsed.
Ok, yeah so it's not synced yet (it takes SteamGifts up to a week to sync users).
So right now the tool can't see giveaway entries/groups, so some data will be incorrect
For example:
GAs user created will be correct.
Percentage if shared GAs will not be correct.
GAs a user entered will not be correct.
Comment has been collapsed.
I'm not sure if any other group (except mine) is using the tool at the moment, but just in case, here's an update:
Due to recent limitations added on SteamGifts this tool has been working on and off for the past week.
But it's finally stabilized, and I've added various optimizations to minimize access to SteamGifts.
This can have minor impact on the accuracy of the tool (for example if within 1 hour 1 user will leave a giveaway while another will join - this will not be registered).
Comment has been collapsed.
It's been a while since the last update, but I finally got annoyed enough to do some changes in the tool.
So the current changes are as follows:
What this basically means, beyond the specific feature mentioned above (which is not very useful for 99% of all cases), is that SGMT tool can now know for every game if it was bundled, given away for free, or neither.
And new features can be developed based on that (if anyone asks me to).
Examples:
Comment has been collapsed.
Update:
It's been a while since my last update, and this time the update was forced upon me.
It's all started 7 years ago, when I first created the tool...
initially the tool was deployed on a single smallest AWS server, which hosted both the tool and the DB, and was free.
However as more data and more groups were collected, I was forced to move to a bigger server (that cost money).
I moved to the cheapest spot instance I could find.
However I soon found out the hard way, that spot instances don't last.
And at the first spot instance replacement, I lost my entire DB contents, and needed to rebuild it from scratch.
That's when I decided to move my DB to a manager MySql.
While the service itself remained on a spot instance.
Until that single spot instance was not enough to load the data of a group, so I had to use 2 spot instances.
One for scraping & collecting the data.
And the other for UI and serving data to users.
So my expenses grew from a couple of bucks a month.
Which I could cover by AWS usage credits.
To 25$ a month for a long while
to $40+ a month in recent months.
While the spot instances became less and less reliable, and forced me to redeploy my code after each replacement (and also made the tool unavailable).
So this week, after redeploying the service every day of the week, I finally broke down and decided to migrate to AWS lambda.
lambda is not trivial, and it's been a while since I worked with Python (I used to work with it all the time back when I first wrote the tool. But did not touch it at all in recent years).
And my task was double: Migrate from a deployed service to a lambda, and migrate from Python 2 to Python 3.
Fortunately, with the current state of AI and chatbots, something that would have taken me weeks if not months, in the past, I was now able to accomplish in a couple of days.
So I'm happy to announce that SGMT (or at least the user-facing functionality) is now running on an AWS lambda.
Which both reduces my monthly AWS bill.
And reduces the failures and downtime of the SGMT tool.
One thing that does change, is the address under which the tool can be found.
I've updated the addresses both here, and in the QGG group recruitment page.
However, if anyone is using the links directly - they need to remove ":8080", replace http with https, and remove the "/" at the end of the address (if there is one).
Comment has been collapsed.
I started with Amazon as it was free at first (which is probably the reason why they offer it for free at first).
And migrating to a different hosting service now, would take a considerable effort from me.
AWS lambda will allow me to reduce 2/3 of what I'm paying for services now ($12)
(The part that is updating the data still needs to run on a virtual server).
So the most expensive part right now is the Database (roughly $15 per month).
So I'm looking into migrating it into a cheaper AWS data storing service.
ChatGPT has been a lifesaver in this regard - doing all the research into options and prices for me.
Comment has been collapsed.
AI Coding Assistant is really ground-breaking. I learned to code last year and have integrated them into my workflow from the beginning, and I'm always fascinated by the things I can accomplish by mixing some AI use with my own knowledge
Comment has been collapsed.
2,886 Comments - Last post 2 minutes ago by DontSh00t
313 Comments - Last post 18 minutes ago by MarshallBanana0815
1,679 Comments - Last post 21 minutes ago by MarshallBanana0815
15 Comments - Last post 26 minutes ago by ShannonA81
114 Comments - Last post 45 minutes ago by CapnJ
409 Comments - Last post 1 hour ago by rhylek
101 Comments - Last post 1 hour ago by Bin246
134 Comments - Last post 6 minutes ago by CptWest
163 Comments - Last post 9 minutes ago by fernandopa
235 Comments - Last post 12 minutes ago by mourinhos86
208 Comments - Last post 16 minutes ago by hieeeen
2,819 Comments - Last post 30 minutes ago by AndyFrost
8,813 Comments - Last post 33 minutes ago by MarshallBanana0815
391 Comments - Last post 35 minutes ago by igel2005
Hey everyone,
I've been working for the last few months on a group management tool, that will greatly reduce the effort needed to manage a SG group, by automating as much of it as possible.
Tool current features:
The tool is built to run as a Service, and uses MySql DB to store it's data.
Run time of each command (except UserCheckRules) is less than 1 second.
I'm currently using the tool to manage my own group:https://www.steamgifts.com/discussion/woEqN/
And it's working out well. I have a good overview of what's happening in the group, and a good level of control.
And it takes seconds to check rules compliance, etc.
Source code for original (Python 2) service can be found here
Source code for the new (Python 3) AWS lambda can be found here
Tool itself can be found here
Comment has been collapsed.