Subject:
|
Re: Musings on an open-source Brickshelf replacement
|
Newsgroups:
|
lugnet.general
|
Date:
|
Wed, 8 Jun 2005 17:46:21 GMT
|
Viewed:
|
843 times
|
| |
| |
In lugnet.general, David Eaton wrote:
|
Oh, mainly I was thinking in terms of expansion making the network stronger.
If its more generic, other communities will use it, and possibly add
features, optimization, and potentially even bandwidth.
|
Well, thats a fair point. But it still makes me nervous -- pretty soon you
could have people pulling it in several different directions, and morphing it
into something thats not really ideal for our purposes anymore.
|
|
Well yes, thats the whole idea. Id propose writing the viewer in
REALbasic, which is easy and powerful, and can compile directly for MacOS,
Windows, and Linux with a native UI on each.
|
Huh. Dont think Ive even heard of RealBasic. Some sort of cross-platform
language like Java was meant to be?
|
Heh, if you havent heard of it yet, you will (er, though I guess you just did).
Yes, its much like that, except that it doesnt suck. And instead of using a
bytecode interpreter, it compiles to native machine code for each plaftorm, just
like C (but portable).
|
MOCPages is basically the idea of a central indexer. People host their images
elsewhere, and write page content on MOCPages. So each page is intended to
be dedicated to a particular MOC, with some limited HTML formatting, probably
similar in nature to LUGNET FTX. I think each MOC has associated categories
and whatnot, and also room for web visitors to rate and comment on particular
pages/MOCs.
|
Well, that does sound similar to what weve been talking about then.
|
Its been out for a while because the site owner left on a worldwide vacation
recently, and wont be back for another month or whatnot.
|
That does illustrate the problem with any site maintained by one person, though.
And if the owner gets hit by a bus, the site and its data are probably lost for
good. What we need is something like this but maintained by a committee (about
5 people would be ideal), so that when one or several people disappear, others
can take up the slack.
|
And as far as that idea goes, Im not sure how it would be
any different making it open source, since a central index isnt really
community-run.
|
Sure it is, if its set up that way. Its not run by peers, but it could be run
by a central committee rather than by one person. Look at virtually any large
project on SourceForge, for example; thats how
they operate. (SourceForge itself operates that way, for that matter.)
|
|
|
I kinda wonder if you could combine the two.
|
Of course, itd be more work to develop, requiring both a web server and the
BrickTella app. But its certainly worth considering.
|
Yeah. Itd be a pain to write. And I dont know much about writing webservers
themselves. Im that step below where I know how to use and configure
webservers, but not write them. But provided it ever got off the ground, then
woot! Itd function pretty much the same as BrickShelf for all intents and
purposes UI-wise. The bandwidth would just be shared-- and thered be
(probably) a much lower level of attention needed on the central server side.
|
Right. Ive written web servers, which is not hard to do (e.g.
Write a Web
Server in 100 Lines), though writing an efficient web server is harder.
Integrating it with a distributed data-storage network would still be a bit of a
challenge, though. But its doable.
|
Admittedly, though, the central server would still be the limiting factor. If
it DID get bogged down, then ... ? It would be single-point-of-failure from a
web perspective, although still *usable* from a non-web-client perpsective
(although only pull requests would work if the central server went down--
nothing new would get added to the system).
|
Right. But the point is that if the central server were open-source, it
couldnt go down permanently. Even if everybody on its committee suddenly got
whisked away to Guatanamo Bay, others in the community could take the code and
data and set it back up somewhere else.
Best,
Joe
|
|
Message is in Reply To:
| | Re: Musings on an open-source Brickshelf replacement
|
| (...) Oh, mainly I was thinking in terms of expansion making the network stronger. If it's more generic, other communities will use it, and possibly add features, optimization, and potentially even bandwidth. (...) Hmm... Yeah, if users can (...) (19 years ago, 7-Jun-05, to lugnet.general)
|
14 Messages in This Thread:
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|