Imports (Was: {brace yourselves} the import is coming {Spoiler Alert: Not all of it} [i.e. blakeyrat was not utterly wrong for the first time ever] Filed under: append-only titles.)
-
I remember there used to be drawings in some of them as well... there was at least one flowchart by Swampy, and another parody one by someone else...
Tried looking on archive.org, but it's coverage seems to be limited to just the first page of each thread so I guess it's lost forever now.
It was nice that we got that warning that the old forums were going away, right?
-
@apapadimoulis any chance we can get the old forums back on read only or something for a bit so people can fix some images or whatever?
-
hopefully; I turned them off because they were crushing the cpu on the server. Despite being in read-only mode and hardly trafficked. An impressive feat if you'd ask me...
-
An impressive feat if you'd ask me...
i think CS was lonely and hoping someone would visit if they caused trouble like that!
-
I may have missed it, but are we trying the import again this weekend?
-
-
would be a good test of http://isitjustmeorservercooti.es/ if we did.... :-D
-
Why does it say that I can't troll when the forums are up?
Oh, right, I'm bad.
-
-
It's pretty hard to hit the spin button a second time! Is that intentional?
-
We have some expertise on archiving old forums now, which turns out to be.. hard.
I always try to convince customers not to do imports at all, and archive the old site in place as static HTML with our help. Alternately, just do what SitePoint did and pick the "top" topics with the most organic traffic and just migrate the best stuff. Or the last year or two. Anything to cap the amount of data converted.
-
Why does it say that I can't troll when the forums are up?
It's a bit clunky, I admit. The idea is that you went to the site because the forums don't work for you. So, the forums work, just not for you. I do admit it needs some tweaking.
It's pretty hard to hit the spin button a second time! Is that intentional?
Yes. And it doesn't actually stop the spinning... I'll fix that today.
-
Pretty much. Yes.
-
I always try to convince customers not to do imports at all, and archive the old site in place as static HTML with our help.
It's a bit late for that, no?At least until @apapadimoulis takes up Blakey's offer of hosting the static HTMLs...
-
It's a bit late for that, no?
At least until @apapadimoulis takes up Blakey's offer of hosting the static HTMLs...
Imagine all the posts we'd loose if we'd undid the import now …
-
-
Imagine all the posts we'd loose if we'd undid the import now …
what? all zero of them? i wasn't on the old forums.
-
what? all zero of them? i wasn't on the old forums.
No, I was thinking of the posts that have been created since the backup that was taken at the start of the import. I was implying that the import would be undone by rolling back to that backup.
-
-
No, I was thinking of the posts that have been created since the backup that was taken at the start of the import. I was implying that the import would be undone by rolling back to that backup.
If that was done now, it'd pretty much kill this community; I doubt many would stick around.
-
-
If that was done now, it'd pretty much kill this community; I doubt many would stick around.
The evil ideas thread is :arrows.m3u8:
-
If that was done now, it'd pretty much kill this community
yeah.... that's not a good idea.....
but hey! that means i would be setting up some sort of forum over on servercooties.com... that seems the most applicable domain i own to set up a replacement forum on.
-
Also, I have a handful of domains I'm not using, including vegetariancannibal.com and korngaveheadtogod.com
-
....why?
-
but hey! that means i would be setting up some sort of forum over on servercooties.com... that seems the most applicable domain i own to set up a replacement forum on.
I'll start on implementing a messagebus using WebSockets...
You meant using existing software, didn't you?
-
vegetariancannibal.com was going to be a WordPress blog, but I set up on Tumblr instead with a different blog name.
korngaveheadtogod.com was going to be a site of gig reviews, but it never got off the ground.If anyone can guess why
korngaveheadtogod
specifically, they'll get a cookie
-
You meant using existing software, didn't you?
yes. I'm not sure if it would be discourse or not.... i mean discourse is still very much alpha but it beats the pants off every otehr forum for UX..... most of the time anyway.
I'd have to figure that one out....
-
-
@pjh @ben_lubar I want to upgrade the docker image so you get latest image magick and latest ruby, let me know when
-
-
Test on meta or try, it appears to
-
@pjh @ben_lubar I want to upgrade the docker image so you get latest image magick and latest ruby, let me know when
If the recent update didn't do it, might as well do it when you want.
-
Actually I think it did take care of it, should have checked
-
Also @boomzilla http://what.thedailywtf.com/search work in progress totally unfinished, excerpt and proper post linking pending
-
vegetarianprogressivegrindcorecannibal.com was going to be a WordPress blog
If anyone can guess why korngaveheadtogod specifically, they'll get a cookie
Had to google it, but heh, clever.
-
Had to google it, but heh, clever.
I got precisely one link, and I am none the wiser...
-
This post is deleted!
-
Is this still imminent? Again?
Or has it finished (I'm not convinced)?
Or abandoned for the moment?
Either way, should this be is unpinned?
Paging @ben_lubar
-
Unpin it. I could theoretically do the entire import by running it again and again until it gets all the posts without Discourse throwing up, but that would make the forum members unhappy due to all the server cooties.
If Discourse ever gets to a state where it doesn't vomit all over redis every time a large number of posts are inserted, this will be revisited.
-
-
Can the import just not be rate-limited to reduce the amount Redis shits itself?
-
Maybe we should just get a bot an admin login, get it to impersonate users and make posts on their behalf.
Never mind, that would fuck up posting dates. Crap.
-
I assume there's some Discoreason this stuff can't just be inserted into the database.
I don't know what happens behind the scenes, but if it's killing Redis, surely doing (say) 100 posts then pausing for say 5 minutes (or however long it takes for the backlog to clear) then carrying on would get around it?
-
Oh hey look: I was trying to look up an old post of mine, but I can't because:
-
the old forum is still offline (and BTW its Google ranking is, AS BLAKEYRAT PREDICTED in the gutters after this amount of time-- you have to specifically put in the site URL before Google will even attempt to find shit on it)
-
Other than that dumb spectate post, none of the posts from the old forum appear to have actually been imported.
So yay! What a piece of shit! Thanks for deleting my work! Fuck you Alex, you incompetent asshole!
And whoever set the title of this thread: I was fucking right. As I knew I would be from day one.
BTW: my offer to pay for hosting of the CS site still stands. Or at least give me a copy of it so my own work isn't fucking just deleted into the ether through no fault of my own.
-
-
Can you get one of those 6¢/hr Droplets and frankencluster it with this one? Like, have this Droplet's Discourse use Redis on the other Droplet rather than the usual one, and then replicate back to this Droplet's Redis as needed? Or something?
-
If I have time at some point, I should figure out what it's spamming redis with to begin with. I remember back in July that the base import script would disable Sidekiq job creation and clean up all the imported posts in a bulk update. Not sure what happened to that.
-
Other than that dumb spectate post, none of the posts from the old forum appear to have actually been imported.
Actually, yesterday I found, by coincidence, two threads from 2006-7 that I had posted in. So there's at least three threads that made it.
-
So there's at least three threads that made it.
We're practically on the home stretch then, as far as the import's concerned...
-
Why is there still a "Site is read-only" notice floating around? Did the script crash again?