From: Tom v. S. <tv...@et...> - 2003-03-31 14:52:09
|
Q: Who are the folks who maintain the current site? Feel free to vote/comment/whatever on these: * It would be cool of the wiki was done with pydiddy (on this self-hosting theme again). It needs a little work (caching and whatnot) but it seems doable enough. * Anything else people want in the website? Interface to the sandbox perhaps? * How is the current website setup as far as maintaining it goes (how are new news items added for example)? Is anyone particularly attached to the current methods? Anyone prefer a specific alternative? -T --=20 Tom von Schwerdtner <tv...@et...> The Etria Group |
From: Ian B. <ia...@co...> - 2003-03-31 16:52:57
|
On Mon, 2003-03-31 at 08:53, Tom von Schwerdtner wrote: > Q: Who are the folks who maintain the current site? Currently the website is a module in the CVS repository, and some of the developers tweak it from time to time. No one seems to have really claimed it, as in regularly updating it and taking responsibility for the content. > Feel free to vote/comment/whatever on these: > > * It would be cool of the wiki was done with pydiddy (on this > self-hosting theme again). It needs a little work (caching and whatnot) > but it seems doable enough. Sure, though I'd really like a Wiki with reST -- I've written a rough one for a recent paper. I'll put it in the sandbox soon. I like reST, though of course it's not as flexible as HTML... I'm not sure how PyDiddy's markup is in that respect. > * Anything else people want in the website? Interface to the sandbox > perhaps? We should certainly reference the sandbox, and at some point we should have an index of things in the sandbox (though we're not there now). The whole thing could probably be organized better. > * How is the current website setup as far as maintaining it goes (how > are new news items added for example)? Is anyone particularly attached > to the current methods? Anyone prefer a specific alternative? It's written in SHTML in the CVS repository. I like that it's in CVS, but there's nothing special about it otherwise. If we were to stay with a static site, or at least a static front-section, it might be better to do it with Cheetah generating static HTML. Ian |
From: Tom v. S. <tv...@et...> - 2003-03-31 17:20:00
|
On Mon, 2003-03-31 at 11:53, Ian Bicking wrote: > On Mon, 2003-03-31 at 08:53, Tom von Schwerdtner wrote: > > Feel free to vote/comment/whatever on these: > >=20 > > * It would be cool of the wiki was done with pydiddy (on this > > self-hosting theme again). It needs a little work (caching and whatnot= ) > > but it seems doable enough. >=20 > Sure, though I'd really like a Wiki with reST -- I've written a rough > one for a recent paper. I'll put it in the sandbox soon. I like reST, > though of course it's not as flexible as HTML... I'm not sure how > PyDiddy's markup is in that respect. I came to the same conclusion about using reST (now that I've investigated some more). So pydiddy is mostly out (though I might swipe parts of it). I'm playing with a "reSTKit" and a "XSLTKit" which I'm thinking of combining into a sort of wiki setup. I'd love to see what you have though (since you know webware a helluva a lot better than I do). > > * Anything else people want in the website? Interface to the sandbox > > perhaps? >=20 > We should certainly reference the sandbox, and at some point we should > have an index of things in the sandbox (though we're not there now).=20 > The whole thing could probably be organized better. I dunno what python offers in the way of CVS interfaces, but I was thinking there could be a module list and some basic info....I'll see when I get there..... > > * How is the current website setup as far as maintaining it goes (how > > are new news items added for example)? Is anyone particularly attached > > to the current methods? Anyone prefer a specific alternative? >=20 > It's written in SHTML in the CVS repository. I like that it's in CVS, > but there's nothing special about it otherwise. Ugh CVS, if this was my project I'd make you all use subversion :) > If we were to stay with a static site, or at least a static > front-section, it might be better to do it with Cheetah generating > static HTML. If were going to have a custom reST wiki sorta thing anyways, I think we might as well base the site content on reST...unless it turns out to be a hassle.....at least thats what I'm thinking right now. reST files in cvs is kinda appealing way to keep things updated. |
From: Tom v. S. <tv...@et...> - 2003-03-31 19:35:24
|
Seriously, I could stick a post-it to my cats ass and throw her into the wild and the message would get to most of you before any emails that go through SourceForge. Slowest list *ever*. -Tom, who demands all his free stuff to be top of the line. |
From: Frank B. <fb...@fo...> - 2003-04-01 10:44:42
|
Hallo, Tom von Schwerdtner hat gesagt: // Tom von Schwerdtner wrote: > * It would be cool of the wiki was done with pydiddy (on this > self-hosting theme again). It needs a little work (caching and whatnot) > but it seems doable enough. Although I wrote PyDiddy I would like a more cleaned up solution to the Wiki. Ian's Wiki is much better written, the only problem it has is that it's hard to install. PyDiddy was intended to 1) play around with porting cgi to Webware and 2) work out of the box, i.e. no SQL-db, no external dependencies outside Webware. There are several things to clean up: Saving versions takes up to much space, the crude user management (not in the tarball yet) is crude, more modular architecture etc. Nevertheless PyDiddy does JustWork(tm) for several months now in our intranet. But extending it with functionality is hard because of the inherent design problems. ciao -- Frank Barknecht _ ______footils.org__ |
From: Ian B. <ia...@co...> - 2003-04-01 17:04:44
|
On Tue, 2003-04-01 at 04:43, Frank Barknecht wrote: > Although I wrote PyDiddy I would like a more cleaned up solution to > the Wiki. Ian's Wiki is much better written, the only problem it has > is that it's hard to install. PyDiddy was intended to 1) play around > with porting cgi to Webware and 2) work out of the box, i.e. no > SQL-db, no external dependencies outside Webware. I've made another sort of proto-Wiki, available in the Sandbox (ianbicking/Wiki/Wiki.py). There's a simple Webware UI around the UI-independant library, but there's nothing special about it. It doesn't have a lot of features, because I was originally writing it as an example application for a paper. But I like it's basic structure, and adding more features shouldn't be too hard. It uses reST and probably requires docutils from CVS. It doesn't use a database -- in the previous implementation (actually no relation to this one) the database was used exclusively for searching (links, backlinks, titles, etc). That could probably be done by keeping index files, so it could be efficient without requiring a database. Ian |
From: Tom v. S. <tv...@et...> - 2003-04-02 02:36:25
|
On Tue, 2003-04-01 at 12:05, Ian Bicking wrote: > I've made another sort of proto-Wiki, available in the Sandbox > (ianbicking/Wiki/Wiki.py). There's a simple Webware UI around the > UI-independant library, but there's nothing special about it. >=20 > It doesn't have a lot of features, because I was originally writing it > as an example application for a paper. But I like it's basic structure, > and adding more features shouldn't be too hard. >=20 > It uses reST and probably requires docutils from CVS. It doesn't use a > database -- in the previous implementation (actually no relation to this > one) the database was used exclusively for searching (links, backlinks, > titles, etc). That could probably be done by keeping index files, so it > could be efficient without requiring a database. Along the lines of using reST and whatnot... Pretty much junk at this point, but my local "sandbox" (in subversion): http://svn.etria.org/root/trunk/webware-site/server/Plugins/ I do not know the Webware internals well yet, but what I basically want is a caching setup for generated documents (such as the XML you get from reST or anything produced with XML and XSLT). My current solution seems overly complex :) I have a parent class 'Cached' which holds 'Source' objects. There are a number of different kinds of Source objects such as FileSource and DatabaseSource (in theory...right now I just have FileSource) since each type of Source has different ways of getting its mtime (a file would get it from os.stat, data from a database would prob need a mtime column...I want HTTP/FTP/Dav sources too at some point..). A class that inherits 'Cached' will hold a list of Source objects as well as a result object (which is really a Source object). Whenever a one of these documents is needed, lets say by calling Document.content(), all the source mtimes are checked against the result mtime and the cache content is returned (after being re-generated if any of the sources are newer than the current cache).=20 Make sense? It's just a slick (in theory) way of saying "gimme this document, but dont regenerate it if its in the cache". My questions: Is there a better way? Is there some internal method that will tell me where the Cache directory is? How can I get to the config options set in the Config direcotry? (Ian, you told me this on IRC at some point but it slipped my mind) Thanks, -Tom |
From: Tom v. S. <tv...@et...> - 2003-04-02 17:50:40
|
On Tue, 2003-04-01 at 12:05, Ian Bicking wrote: > I've made another sort of proto-Wiki, available in the Sandbox > (ianbicking/Wiki/Wiki.py). There's a simple Webware UI around the > UI-independant library, but there's nothing special about it. >=20 > It doesn't have a lot of features, because I was originally writing it > as an example application for a paper. But I like it's basic structure, > and adding more features shouldn't be too hard. Do you have any intention of working on this more? Also, I wanted to suck out the body of the wiki pages and wrap my own header/footer around it (so it would be in-line with the website). Know how I might achieve this? -Tom |
From: Ian B. <ia...@co...> - 2003-04-02 18:28:45
|
On Wed, 2003-04-02 at 11:50, Tom von Schwerdtner wrote: > > I've made another sort of proto-Wiki, available in the Sandbox > > (ianbicking/Wiki/Wiki.py). There's a simple Webware UI around the > > UI-independant library, but there's nothing special about it. > > Do you have any intention of working on this more? Yes, even though it's primitive I'm quite happy with the interface so far, and I like not dealing with parsing. > Also, I wanted to suck out the body of the wiki pages and wrap my own > header/footer around it (so it would be in-line with the website). Know > how I might achieve this? The WikiPage instances should return just an HTML fragment, not an entire page, so you actually have to do this. You could do this by overriding Page.writeBodyParts, or using a template in any number of ways. Ian |