SL4 META: list closure 2 month followup

by gwern1 min read10th May 201125 comments


Personal Blog

This is probably of interest only to the old-timers here. (Ironically, as a cross-post, this is almost by definition spam for LW, since anyone really interested in the topic would already be subscribed to SL4 and would have seen the emails.)

In 13 March, I finally got around to a long outstanding bit of cleanup: suggesting that the equally long defunct SL4 list be formally closed, things tidied up, and the lights turned out. The suggestion met with a mixed (and muted) reception.

Last night I followed up with a second email comparing SL4 activity to activity on Extropy-chat, LW, OB, and MoR; it won't surprise anyone trying to drink from the firehose here that LW is approximately 3 orders of magnitude more active than SL4 is.

25 comments, sorted by Highlighting new comments since Today at 12:58 PM
New Comment

I remember a bunch of nice things on SL4. Rolf Nelson posted his deterrence idea there first, Wei had several interesting exchanges with Eliezer, Justin Corwin reported on his AI-box experiments and gave a partial transcript... Wei's output is especially hard to find these days because it's spread thinly over everything-list, one-logic, SL4, LW, and decision-theory-workshop. Turning off posting on defunct forums is okay for me, but could you please find some way to keep the archives accessible online, so we can keep linking to individual messages like I did in this comment?

I suggest importing the SL4 archive into Google Groups, which seems likely to remain accessible for the foreseeable future. If anyone has the complete SL4 archive as a compressed Unix mailbox file (and can put it on the web and send me a URL), I'd be happy to do this. My own personal archive only goes back to 2002.

It would be great to have the Extropians list archive be available as well. It spawn off both everything-list and SL4, I think, and contained some of the earlier anthropic reasoning and Singularity discussions. I volunteer to import that into Google as well if anyone can send me a mailbox file.

For finding my non-LW posts, most (besides SL4 and Extropian posts) are archived in Google Groups. You can use the "Search author's post" boxes on these two pages to search for my pre-2002 and post-2002 posts respectively. (If SL4 and Extropians were imported into Google Groups then those search boxes would cover them as well.)

Wei, have you considered spending some time to collect your various old emails & posts? I see some stuff listed on but far less than I think there must be. (In part by collecting my old material, I've populated with a fair amount of stuff, and I think you're a much better writer & thinker than me.) Heck, just collecting all your emails on a topic and pasting them into a page would be useful.

People seem to like this idea, but nobody has sent me archives of either list (SL4 and Extropians) yet.

? I provided an SL4 archive 8 days before your comment, at Dropbox:

I wanted a raw archive in Unix mailbox format so I can feed it to Google Groups. It looks like you scraped the website, in which case I'd have to reverse the email-to-HTML process that used, which I'd like to avoid if possible. BTW, I would ask Eliezer directly for the raw archive, but I did that a few years ago and he didn't respond.

Pretty much free to ask again. Ping him as a reply to your original email?

And upload the data to github to make reliable private backup trivial. (This also works for active lists.)

Thanks for the links, they'll come in handy!

but could you please find some way to keep the archives accessible online, so we can keep linking to individual messages like I did in this comment?

From my first email:

There's no shame in shutting down the list. Let's provide a tarball of emails for archival purposes (the Internet Archive has up to February 2009:, and turn off the listserv.

Why not just leave the current archives on online indefinitely? It's stored as flat HTML and doesn't require the list to actually be active.

I've never suggested otherwise. A mailing list is not a website.

I know, I was responding to "Let's provide a tarball of emails for archival purposes (the Internet Archive has up to February 2009)". A tarball sounds like a good idea, but I was confused by the mention of the Internet Archive when the official SL4 archive is still working fine.

My point there was that will cost $8+ a year indefinitely (for the domain name alone), and my mental target is 2070; will Eliezer pay >$600 just to keep going? Probably not. I'd guess will be no more by 2030 at the latest. So it's nice to know that there's an existing backup already even if no one ever gets around to making a tarball dump of the website.

(I've learned to ask for busy people to do as little as possible; if I suggested a dump be made and only then SL4's listserv shut down, I would guarantee that the proposal go nowhere.)

[-][anonymous]10y 0

It's hard to hyperlink to a single email in the middle of a tarball.

I tend to think that this is an example of the need to beware of trivial inconveniences, the trivial inconvenience here being the need to subscribe to a mailing list.

Although I myself was perfectly willing to undergo that trivial inconvenience, I believed that many of my potential readers would not be, so I commented on Overcoming Bias instead of posting on SL4.

It would be interesting to have old-time SL4ers write an essay on lessons learned with running SL4 and what LessWrong does better or worse than it. Trivial inconvenience was more of an OB problem, I think; lack of moderation or links, and repetition, were bigger problems for me.

It'd also be a good idea to compile a categorized 'best of SL4' list of emails; for me, the worst part of SL4 was how repetitive a lot of it was (and how incredibly frustrating John Clark was with his obtuse misunderstandings of basic CS), which such a list would cut through. (As cousin_it mentions, before OB started, there was a lot of good discussion on SL4.) I might do this at some point if I ever finish a similar project of mine, reading through the archives of the Neon Genesis Evangelion ML - no promises though.

and how incredibly frustrating [name withheld to protect the guilty] was with his obtuse misunderstandings of basic CS

We have that here, too. (It's probably less frequent here.) Sometimes the misunderstandings do not get downvoted or pointed out, sometimes they do, but the negative scores and replies do not prevent the person from continuing to go on copiously and with confidence about the subject he misunderstands.

Well, inasmuch as we have that same person here, we have the same problem. (Downvoted heavily too; I haven't read the comments - about some inane looking problem about clock faces and the real number line - but given his past history I'm not inclined to be charitable.)

He's not the only one although he might have been the latest one on the subject of CS.

Are you just going to keep complaining forever until it is closed, or would you theoretically give up eventually?

From the conclusion to my second email (linked above; emphasis added):

Now. I could keep this up for the indefinite future, posting statistics about how other sites are literally 1000x more active than SL4 is. But if all the points I have presented are insufficient, then it's hard to see why simply repeating them with slight variations as the statistics change would make them sufficient. I'm done; my case has been presented. I'll stick around for a month for whatever is left to discuss, and then unsubscribe - voting with my feet, as it were.

Usually you're better than this.

EDIT: And as promised, I've unsubscribed. There were no worthwhile emails after my last email, either. Just spam of varying degrees of spamness.

So, uh, why is this being posted here?

I thought as the predecessor to the predecessor to LessWrong, there was a certain historical interest to the proposal.

For the many people here who remember SL4 when it was the best place online to observe or partake in extremely rational discussion.