- User Since
- Jul 9 2019, 1:02 PM (63 w, 5 d)
Fri, Sep 25
Thu, Sep 24
Wed, Sep 23
Tue, Sep 22
Today I found out that half of the code I pulled over from github was nothing but hot garbage. It's a convoluted mess of attempts to get some "renegotiation" happen that does not have anything to do with what any of the documentation has to say about openssl.
Fri, Sep 18
Tue, Sep 15
Mon, Sep 14
Fri, Sep 11
Wed, Sep 9
Thu, Sep 3
Just like T124, it seems it makes more sense to postpone rCMS towards 2022
We're deferring this task to later. Right now CMS seems like it needs a delay into some point 2022 and right now it seems like it would introduce more work than it would solve.
Aug 28 2020
Aug 26 2020
Aug 21 2020
Aug 17 2020
Aug 5 2020
The system now produces a user report on report/user that administrators can access to see what the user has been up to.
All these changes are included in D324
Added the output of the log to the report page, which should make it easier for moderators to understand a user's behavior.
From now on we're recording whenever a bid is recorded as spam. Since this is an automated process it doesn't give much insight on what is going on with administrators, but it gives a good overview of what is going on.
Aug 4 2020
The administrative log is now able to also process bids that were posted to auctions. This should give a really good overview of whether a user is an active member of the community.
To make sure that the data can be properly used, the system also broadcasts this information to loot and cptn-hook, which allows the applications to receive the data and make it available to the network.
Jul 29 2020
Included in D324
Jul 28 2020
To make this work, there needs to be presentation for the following types so we can handle text and images:
In order for CMS to actually properly manage the interactions with users, we need to basically break every component into two aspects:
Jul 27 2020
The two actual data sets that need to be able to be imported and exported are grants and mnemonics, both identities and resources are generally implied data types and are therefore only adding bloat in a backup file.
Jul 24 2020
Now, the evolution has gotten a bit further.
Socketfire should be a combined effort between a PHP application managing all the connections from the servers, and the authentication of the requests which relays the information through a Unix socket to the CPP server actually dispatching the information to the clients that subscribe to these channels.
Jul 23 2020
Jul 17 2020
Once this ticket has been resolved I need to open a task to integrate Socketfire with YCH and other services.
Jul 15 2020
Relay and rPERM are now proper independent applications that can be managed externally - this task is therefore no longer necessary
Jul 7 2020
Every primitive should have two methods:
Jul 6 2020
Jun 30 2020
Jun 25 2020
Jun 19 2020
I decided to remove the interface to create a task completely in order to make the system more focused on being programmed.
With a simple SDK we should have an easier time just modelling an HTML template with a few basic components and then push the data to the server the way it's supposed to be .
Jun 17 2020
Jun 15 2020
Jun 12 2020
The basic model is set up. This allows our system to store all the relevant data, and generate operations that allow forwarding and rewinding a document. I've been able to divide the editor's model into three components:
Jun 9 2020
I have removed the tracking code from the repositories as a first measure, as pages get updated, the tracking code should be disappearing. Now I need to find a proper replacement for the system health monitoring.
Jun 4 2020
Jun 2 2020
Jun 1 2020
The create task page will probably never be returning in it's former shape. While it is technically possible to create a programmatic solution to the issue of creating tasks, assigning the right people, notifying them, setting the right permissions and all of this without adding extra code, the reality is that it's much more likely that a system like this will use third party applications to help itself create these data sets.
May 29 2020
At the end of the day, all content can be considered to be hierarchical and to descend from two (for the sake of simplicity) base primitive types:
May 28 2020
May 27 2020
May 26 2020
My latest big progress was that I started maintaining a document as an array of lines, and a line is a type plus an array of inlines. Every inline can have multiple properties, like bold, italics, or something similar.
This allows us to construct a document with minimal effort while maintaining a good level of compatibility.
May 20 2020
May 19 2020
May 15 2020
This can be directly taken over from CRM, which provides a module mechanism to integrate modules that it then uses to bootstrap the module.
Is implemented in the current version of SF
May 14 2020
May 13 2020
May 11 2020
May 10 2020
May 7 2020
May 5 2020
Since the window.open behavior seems discouraged by most browsers anyway I suggest we start experimenting with creating a link element that we can then click on when the download is finished and prompt the user to either save or download the file - which would be our preferred behavior anyway.
May 4 2020
D274 closes this
D283 does resolve this.