Sai Emrys wrote:
> On Fri, Jun 27, 2008 at 11:22 AM, Nomad of Norad -- David C Hall <
> [log in to unmask]> wrote:
> 
>> What about mirroring of an existing site onto the LCS site?  At that, I've
>> noticed that some wikis contain direct copies of articles from wikipedia
>> proper, which apparently they update automatically as that article changes
>> at wikipedia.  It might also be worth creating, as well, a Wayback sort of
>> thing, independent of archive.org, as a hedge in case archive.org (God
>> forbid!) ever goes byebye,
>>
> 
> That depends.
> 
> Running a spider to get content without the cooperation of the existing host
> is a pretty major undertaking.
> 
> If however the existing host / user cooperates, it's not hard at all; you
> either save the site as a zip/tar/whatatever and unpack it on our servers,
> or just use rsync over ssh w/ a cron job (or any of a large number of other
> strategies) to make a real mirror.

Would that rsync be done from the host-server side or from the LCS side? 
I had envisioned this as something an LCS user would select from a menu 
at the LCS site, fill in a field pointing to the page or pages at his 
site that he wanted preserved/mirrored, and set an interval for it to do 
this (say, once a week, once a month, etc, depending on how active he 
thinks his site is)...

As far as crawling a site...  so far as I can tell, Wayback subcontracts 
their spidering out to some other company whose name escapes me at the 
moment...

Actually, they also suggest a provided browser-toolbar thingy that 
itself automatically preserves everything you websurf to that is at your 
site, quietly in the background, at your discretion.  (I haven't tried 
it, so I don't know what sorts of parameters it sets to how and when the 
stuff gets copied and why.)

On the other hand, your probably right, that doing a Wayback-style 
preserve-all-versions-of-this-page thing is probably not something to 
tackle right now... maybe if/when LCS becomes big enough to support such...

> I suppose somewhere along the line it'd be useful to develop some kind of
>> open-source dictionary-builder app, something that would do for dictionaries
>> and language-guides what wikipedia has done for encyclopedia media, and
>> make/offer it as the standardised program at LCS.
>>
> 
> I know for sure that someone (Henrik?) posted about exactly this not too
> long ago.
> 
> I don't know whether that dictionary protocol has a good frontend interface,
> and I don't think it can *build* one so much as *store* it, but still...
> this I think is something that is vastly more an issue of community uptake
> than of availability.
> 
> If simple pragmatics like storage space, accessible servers, etc are an
> issue, we can probably fix that. If it's a matter of deciding on a standard
> to use, getting people to adopt it, developing pretty front ends, etc though
> - that's up to y'all.
> 
> 
>> Another thing just crossed my mind:  There have been radio ads lately for a
>> program that automatically duplicates the contents of your computer onto a
>> personally-accessible site online, so you can restore data after an HDD
>> crash or something.  It occurs to me that LCS could provide a
>> personally-accessible mirror of the contents of a user's selected local
>> folders, maybe even something that provides so many levels deep of a backup,
>> so you could go back to a folder corresponding to a week ago, that is a
>> separate image than that which was changed yesterday, and allowing the user
>> to set an arbitrary number of back copies (i.e. 7 folder-copies back) and
>> dumping the oldest copy/ies as the newest copy comes in.  Perhaps
>> maintaining it, or the oldest ones, as a zip file or set of zip files.
>>
> 
> There are many applications that will allow you to do this relatively easily
> using SCP or SSH. Our host is compatible with this type of solution.
> 
> I can't recommend a Windows-based program though, as I almost exclusively
> use Kubuntu (albeit considering moving to OSX).

True, but of course we DO want something that a Windoze user can 
configure really easily, from a menu pane or something, rather than 
something he has to dig down under the hood and configure by editing a 
text file or issuing switches from a command-line interface or something...

Making it open-source just means it can be ported easily (or 
not-so-easily, depending on what authoring system or programming 
language its written in).  I.e. you could have a Windoze version, a BeOS 
version (if someone was willing), an AmigaOS version, and so on....

> 
>> We could provide some sort of open-source mirror-the-folder(s) program that
>> is provided with the LCS membership, with the included benefit that  if you
>> die, you can specify (at your discretion) that those folders and zips be
>> opened to the public for perusal or otherwise be preserved for posterity.
>>
> 
> So long as the legal ramifications are considered, this could happen. (E.g.
> one would tell us ahead of time, and ensure that we're contacted in the
> event it's necessary; preferably one would already have put up a usable
> website that's just not publicly accessible, so we can just flip a switch
> and make it public. We don't really have the resources to do website
> creation & maintenance.
> 
> The idea being here that the folder you keep your works-in-progress in,
>> where your conlang scratchpad files are, and so on, would be preserved so
>> that you, and/or future generations, can see the bits and pieces of your
>> language as it developed, and/or the bits and pieces that didn't quite make
>> it to the dictionary-site yet... so that that stuff won't die when you do.
> 
> 
> I'd suggest that this sort of thing could be well done using one of the
> blog-like CMSs.
> 
> E.g. in my case, I use LiveJournal extensively. Many of my entries are
> locked. I have instructions left, in a form verifiable to be mine and
> duplicated offline, for some parts of it to be made public on my death. For
> the server admins, this would be as simple as a single SQL query to just
> make those items public.

True, but what I had envisioned was something that the end-user could 
set from the LCS website, check or uncheck boxes in his configurization 
screen on-site.

> However, if I may be so bold: why the worry about deferring this until after
> you're dead?
> 
> Why not live openly *now*, while you're still around to collaborate and
> explain and present?

Just bracketing in case some user wants to keep certain things a secret 
for the time being, but wants them preserved for posterity regardless.

F'rex, in my own conworld, there is a deep secret that right now, if I 
placed it up on the site, would constitute a Major Spoiler, but it is 
something intended to be revealed later on into the planned story-arc.

JNS, the creator of Babylon5, kept a lot of the story-arc details and 
stuff locked away in a safe when he was writing and producing the 
series.  But I'm sure, if he died, he'd want that stuff preserved for 
posterity, and so that maybe someone else could pick up where he left 
off in the B5 story.

Mind you, that's story, not conlang, but since so much of conlanging is 
also related to conworlding and sometimes story-arcing...

> 
> - Sai


-- 
       Nomad of Norad (David C. Hall)  ---  *TeamAmiga*
[log in to unmask] --- http://www.joshua-wopr.com/
  - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
For a dementedly wacky sci-fi continue-the-story project,
join my WebBBS. http://www.joshua-wopr.com/phpBB/index.php
----------------------------------------------------------
This is the Emergency Backup .Signature File, it kicks in
if the Regular Backup .Signature File fails to load.