Still nada.
Anon asked about the JSON for all Q posts.
The API is still there, I just removed all the links.
http://qanon.news/Help/
Anon asked for a word count in all Q posts and I did it really quick. Just gonna drop this here.
Here's the results, sorted by occurrences.
https://pastebin.com/e1u1jxR2
You may be looking for the Delta thread.
I think I told you to come here. I did some Delta workk here
http://qanon.news/smashposts.html
That Delta is only considering the difference between a Q post and a DJT tweet. There's is nothing in there to account for DJT corrections of deltas between tweets.
The deltas you see on the smashpost page are spread out across the Q posts - since there is a different delta for each.
IE: Q posts at 12:00p
DJT tweets at 12:10p [10] delta
Qposts at 12:05p <- this would also mean the DJT tweet at 12:10p is also [5] delta.
I did it like that because I wasn't sure of the meaning of all deltas. Is a [29] valid? Only on the 5's? Good luck anon! Let us know what we can do to help.
I think most everything we've been doing here has all been resolved to either GMT or Zulu time. 8ch JSON comes in GMT/Zulu. The TrumpTwitterArchive comes in GMT/Zulu.
Correct me if I'm wrong codefags.
The deltas are what helps you find the marker.
IE: Q posts something about "win" 5 mins later DJT posts something about "Goodwin" That's a marker. (Just an example - I don't remember the deltas on the goodwin marker.)
The Delta thread where the work has been done on deltas. I'd like to see def documentation of confirmed markers.
>http://q-questions.info/research-tool.php
Looking good!
Checking the Show Delta box seemed to kill off any results for me tho. I'll try again later!
I believe you are nearly correct.
Once you have found a [marker], then the time between DJT tweets/Corrections appears to be the indicator of another marker. I don't think it goes back to a Q post delta.
Check the logic for the [5] & [1] markers.
I disregarded all negative deltas (any tweet BEFORE a Q drop). There's information there possibly - but it just introduced too much noise into the results.
Hows it looking you faggots? Things progressing as designed?
I got a nagging image issue sorted out. Now archiving Q images and reference images to my site. Just about ready to get back on the elasticsearch idea.
Superfast multitenant full text search for json. Clients in Java, C#, PHP, Pyton, Apache Groovy, Ruby etc…
I think all I need to do is write something that will input all my json into my local elasticsearch instance and then all lights are for go.
MMmmm Yes I have. I like the idea.
There are many services out there that will allow you to do this or you can create your own blockchain w/ ethereum.
Were you thinking just qposts or all qresearch?
Sounds like a good idea. Probably alot of work!
I heard that. I coasted about 2 weeks for the same reason. I've been working on tightening up the site and working on small bugs I've found.
I implemented a search for Q posts and am working on the big bread search now.
Ahh you've entered the big new world of internet interoperability! The internet is great, but it's not always the easiest to move data from platform to platform.
It's one of the reasons I stuck with straight JSON. Platform independent. Easily shared. Do you have the capability to transform into JSON/XML? What is your end goal? Share the database? Share the data? The app itself?
Big bread search update.
I may be back to Solr not being a good solution.
In trying to create a prebuilt index I've discovered that either
a) javascript just doesn't have enough memory to do it
b) javascript times out before it gets done and nothing happens.
I'm going to take a closer look at this
https://xapian.org/docs/bindings/csharp/
Moar testing today. Solr is NEVER going to work in this instance. I was hoping that I could just create an index on my dev machine and save that off and then use a worker process to add to the index. I've got one other idea to see if I can bend it to my will - but so far no workie. From what I can tell it's not possible to add to the index - it needs to be completely regenerated when you add a new document.
I don't understand how other people can add so many docs to the index and have it work. My tests were showing it to run for 12+ minutes just to generate an index and it never finished.
I'm open to new ideas if anybody has one.
The custom Google search I've got on there now does seem to work, but again it's not ideal. What I want is a list of POSTS that match and the goog search seems to find the matches, but only returns complete breads. You still have to CRTL F to find what you were looking for within the bread.
I can put together a test harness for Solr if anybody want to see if they can figure out a way to make it go.
My gut is telling me that my next best option is to move into a database in order to accomplish the bigbreadsearch. It's probably possible to do using a hosted elasticsearch solution (https://www.elastic.co/cloud @$50/mo)
On the other hand, I think that I can write an app to fill a database in a couple hours, and it would solve a few of the problems I was seeing in the other search tech. Most of the good search engines will plug into a database anyways so I think this is probably the direction I'm headed.
For elastic search?
>$50/month seems like a lot. My cost isn't nearly that much.
Derp. I clicked the wrong post.
I agree - which is why I haven't done anything on it. My hosting costs a bit more than that - ANNUALLY.
I feel like a DB is just just going to be a better solution now. I'd hoped that I'd be able to just do everything with straight JSON - but alas! You cannot.
I guess I need to find the best search engine to plug a DB into now. I'm hoping to write the code to insert my existing data into the database today, write code to insert new data into the DB tomorrow.
Yeah. It's a hosted service. It appears that deploying a custom elasticsearch is probably a large pain in the ass most folks don't want to deal with.
I have SQLServer currently set up and my host gives me a database so I'll probably go with that.
WTFERK? We already have like 3 bread searches already now? Am I totally wasting my time?
Regardless….
Interesting! Tell me more about how you are doing this. Search seems to be pretty quick. Are you using a DB backend? Straight text search? Is all this in PHP?
I've managed to import all the JSON data I have on hand. 1,569,777 posts took 25mins to import. My DB design is ultra simple. Single table that virtually matches the JSON data structure. There's no telling what the performance is going to be like just yet. Even getting a count takes 16 seconds. Ugh.
I'll run some simple tests later to see what I can figure out.