LW Migration Announcement
post by Vaniver · 2018-03-22T02:18:19.892Z · LW · GW · 14 commentsContents
14 comments
The votes are in, and of the eligible 376 voters, 102 voted to Migrate and 15 voted to archive. With 87% in favor, we’re going ahead with the transition, which will begin tomorrow (3/22) at 6pm Pacific time. Trike will take a snapshot of the database, transfer it to us, and we’ll begin the database import of new material since the last import. At 7pm, the DNS server will point to our server instead of Trike’s, and so people visiting lesswrong.com will see the new site. The import will run in parallel with the site operating, so there won’t be any downtime of the new site.
Some things to note:
- Old links will continue to work, redirecting to the right new URL. We’ll be watching the analytics to notice any failures, and you can also alert us through Intercom (which we’ll shut off a few days into the transition, as problems get fixed).
- Lesserwrong links will redirect to the same page at lesswrong.
- We’ll add old karma scores to new karma scores, and may adjust the vote weight algorithm accordingly.
- Please report issues here.
Also, for those of you near Berkeley, we’ll be throwing a launch party on April 7th at 7pm (LW, Facebook).
14 comments
Comments sorted by top scores.
comment by bogus · 2018-03-22T06:19:21.993Z · LW(p) · GW(p)
Great news, overall. However, do notice that, by all indications, this migration will in fact break links, for the case of "links to comments on a deleted LW1 post". As I mentioned on LW1 itself, given such a LW1 "permlink", you can freely explore "parent" comments and replies. Lesser Wrong does not support individual "permlink" pages, so it simply links to the individual comment as part of an "all comments" listings, which breaks in the "deleted page" case (It also impacts the case of a page with a large number of comments--if only from a sheer usability POV--because a user is now effectively forced to load all comments instead of just requesting what she's actually interested in-- and possibly even with "what she's interested in" being buried in a deep subthread somewhere, beyond the "load more comments" feature and nowhere to be found in what actually was loaded!). I mention this in a simple comment here rather than opening a formal "issue report", because I do not regard this as a true "bug" with the site that can be easily fixed or that even has truly significant impact. However, since (due to the vicissitudes of early LW1 history, with at least one major contributor having wiped his whole presence from the site!) there is actually quite a bit of early content that's impacted by this, it may be something to think about at a later time, for a more complete restoration of early LW history.
Replies from: habryka4, Raemon↑ comment by habryka (habryka4) · 2018-03-22T06:24:20.397Z · LW(p) · GW(p)
Oh, I wasn't aware of that functionality. Thanks a lot for pointing this out, and we will get around to fixing this.
Replies from: Kaj_Sotala↑ comment by Kaj_Sotala · 2018-03-23T14:18:39.502Z · LW(p) · GW(p)
There's also an argument for not fixing it, in that if an author deletes his whole posting history then he might want all traces of it removed, and making the comments (which contain information about the post) inaccessible respects that wish more than keeping them available.
EDIT: though that said, people were promised that existing links would continue to work when they voted about the migration, so sticking to that promise is a more important priority and we should actually fix this.
Replies from: totallybogus↑ comment by totallybogus · 2018-03-23T14:30:50.191Z · LW(p) · GW(p)
This seems quite right for the new feature of "private blogs" here on LW2, and for "drafts", which IIRC behaved as you describe, even on LW1. But ordinary posts on LW1 were unambiguously public (both in the Main and in the later-added Discussion section)--roughly equivalent in visibility to "Featured" posts here on LW2; and once you choose to make a post potentially eligible for such status, it would be quite unfair if you could wipe both its text and its attached comments with the click of a button. (Note that a 'deleted' post did become anonymous, in addition to having its text wiped out; people might have been able to figure out the original author if she happened to be mentioned in the comments but this only occurred rarely, so there was quite a bit of reputational protection. And really, much of the value in large comment threads is actually in side-discussions that are only tangentially related to the 'OP', if at all.)
↑ comment by Raemon · 2018-03-22T23:27:18.961Z · LW(p) · GW(p)
Do you have an example of such a link? (Mostly want to make sure I understand what syntax we're referring to here)
Replies from: sdr↑ comment by sdr · 2018-03-23T02:09:19.961Z · LW(p) · GW(p)
Not grandparent, but browsing through my private notebook for potentially breaking links, eg:
http://lesswrong.com/r/discussion/lw/deg/less_wrong_product_service_recommendations/6yry <- which is one specific advice (and a good one at that) vs https://www.lesserwrong.com/r/discussion/lw/deg/less_wrong_product_service_recommendations/6yry <- which is 404. This actually do have a high impact both on other sites linking specific comment threads, and by extension, on SEO in general (linked page with content changed to empty).
( Relatedly, https://www.lesserwrong.com/non-existing-page returns HTTP 200 instead of 404, which is more wrong, than http://lesswrong.com/not-existing-page )
Replies from: Radmincomment by bogus · 2018-03-22T07:57:23.214Z · LW(p) · GW(p)
There's currently a lot of useful content at the LessWrong (LW1) Wiki, wiki.lesswrong.com. What will happen to it? It seems that at a very minimum, you should request a wikidump as well from Trike; it would even be relatively easy to make a public dump available, using the tools that MediaWiki makes available for this purpose.
Replies from: habryka4, Vaniver↑ comment by habryka (habryka4) · 2018-03-22T08:01:59.642Z · LW(p) · GW(p)
The wiki will stay as it is for now, and stay available as it always has.
comment by bogus · 2018-03-22T16:32:39.554Z · LW(p) · GW(p)
Hi, it currently appears to me that the LW2 functionality around 'recovery email' addresses is subtly broken, and if I am correct, this will impact LW1 users who did not set a recovery email until after the first database import to LesserWrong, as well as any users wishing to change their associated 'recovery' email in the future. Please see this subthread (or GreaterWrong link) for details about the issue. I'm not confident that opening a formal "issue" ticket on GitHub would be appropriate here, because all I have is circumstantial evidence, and I am quite unsure how things were expected to work in the first place. However, I do think some cursory attention from the devs would very much be appropriate, if only to ascertain whether there's an actual issue here.
Replies from: habryka4↑ comment by habryka (habryka4) · 2018-03-22T19:27:17.379Z · LW(p) · GW(p)
Thanks for letting us know! I figured out what the bug was.
It seemed that for users who were imported from LW1 and who didn't have an email set at the time of the import, didn't get their email properly updated when edited. I fixed this bug locally and will push that before we move to LW1.