h a l f b a k e r yNot from concentrate.
add, search, annotate, link, view, overview, recent, by name, random
news, help, about, links, report a problem
browse anonymously,
or get an account
and write.
register,
|
|
|
| |
//Giant Radioactive Mutant Koala// |
|
| |
Ye Gods, we hope not .... |
|
| |
For most users, the professionally maintained server farm
where their accounts reside is has better redundancy, and
more consistent, reliable backups than their local machine.
(Present company excepted, of course.) |
|
| |
// the professionally maintained server farm
Do you know something I don't? |
|
| |
I think this is a pretty good idea, actually. Although implementing the inevitable next request - carefully re-inserting all this stuff into the halfbakery a deletion took it out of - will be painful to implement. The next will be a full REST interface, and then people will have their own skins for it, and uh. It'll still be mostly around 100 people talking to each other. |
|
| |
So, what would this snapshot actually contain? |
|
| |
- all your ideas and all annotations (yours or someone else's) on your ideas; number of +/- votes; dates
- all your annotations and links, with the title of the idea they were on
- all your votes
- some sort of signature over that state so you can "prove" that you didn't make it up. |
|
| |
I guess having some sort of timestamp and allowing you to request an incremental backup relative to the last timestamp you've got would be good, too. |
|
| |
//Do you know something I don't?// Probably, but none of it
concerns the Halfbakery's infrastructure. |
|
| |
I had rather exaggerated ideas about the "experimental
university research network." |
|
| |
[jutta], I seem to remember reading something about there being restrictions on what kinds of commercial activity you could get into here - but I've always wondered if users could sponsor full-site backups; and in return receive a disk containing the data (with a viewer of some sort, and the password & vote tables either missing or encrypted). |
|
| |
It wouldn't be a sale, exactly; it'd be a method and system for storing a redundant, widely distributed off-site backup, with each recipient/safeholder agreeing to return the disk if the need were to arise. |
|
| |
I imagine that even the biggest user contributions would be pretty tiny compared to the kind of files people deal with every day nowadays. What about the illustrations? Also, what about ideas one has annotated? |
|
| |
I don't think that this would be workable as a 'backup' device, only as a snapshot of history. The reason being that reloading the backup (or the various individual backups, as we seem to be envisioning here) will run into all kinds of problems with annos that have subsequently been deleted, users that have subsequently been deleted etc. |
|
| |
"Snapshot" is a better term. Title changed. |
|
| |
No, that's too easy. We need one dedicated person, living on a pension, to save the days traffic by hand in a word document. On a 486. While tippling. |
|
| |
(Ah... sent the email before noticing this...) |
|
| |
Well the easiest way to do this, is really to just do a
whole site backup. Zipped into an archive, timestamped,
and hashed. |
|
| |
You post the archive as a torrent (or download link), and
post up a PGP signed copy of the hashlist for each
archive. |
|
| |
This keeps complexity down to a minimum, and
encourage backup of other people's ideas as well. |
|
| |
Actually the easiest way is to make posts offline (which usually works out better anyways, especially if you're in a prolific period), save that in a directory, or append to a log file, and do the same for your own brilliantest annotations as well; that way you have your own backup for your stuff, secure, should you need to recreate. |
|
| |
Or use the Wayback Machine. That way you've other people's annotations as well (for the purposes of an accidental account deletion, I think we can assume the permission of annotators to recreate their annotations). |
|
| |
Or whatever. Currently not manically churning out posts day after day, I'd probably use the WM to get titles and rewrite some of the ones I really like. |
|
| |
//It'll still be mostly around 100 people talking to each other. |
|
| |
… Ten thousand people, maybe more,
People talking without speaking,
People hearing without listening,
People writing songs that voices never share,
And no one dared
Disturb the sound of silence…
|
|
| |
Type your account name into the halfbakery search page.
If there's a space in your name, surround with quote
marks. Save the resulting page.
Then, either write and use a script to scrape all those
pages, or just pull them off the wayback archive if you
ever need them. (You can just look at your user page on
wayback to see your old ideas, but you won't be able to
find all your comments that way.) |
|
| |
I think this is pretty clean, provided you have a name
which isn't also a word. But you'll also get pages where
you haven't commented, but in which someone refers to
you. |
|
| |
Scraping them while logged in would also capture which
way you'd voted. You won't get votes on ideas you didn't
comment on, though. |
|
| |
The script method should be pretty straight-forward, and
is no hassle for jutta. Assuming of course that it's polite
and doesn't hammer the server with hundreds of requests
at once. |
|
| |
You can bash (pun intended) something together in Linux shell script
to do this on your local servers
in less time than it takes to drink half a dozen bottles of Piper
Heidseck ... |
|
| |