Archived GeneralDiscussion.

Zwiki and Python 2.2.2 --DeanGoodmanson, 2003/04/01 15:07 GMT
There seems to be a recurring thread regarding Zope 2.6: Does /this/ product work with Python 2.2?

If anyone has had success or failure with running Zope+Zwiki with Python 2.2 (instead of 2.1), please report it. If anything it will give some clues to Zope 2.7 requirements for Zwiki..

Zwiki and Python 2.2.2 --SimonMichael, 2003/04/01 16:30 GMT
Dean - A number of people have said "works fine for me". Kapil posted the best roundup of known issues on zope.org recently. If you can find it.. good luck.

questionable content --DeanGoodmanson, 2003/04/01 19:12 GMT
I removed some questionable content posted anonymously.

If the message was intended, it shouldn't have been posted anonymously and should have had a disclaimer. The graphic content should at a click-through or, if that necesary, below the fold.

rc3 released --SimonMichael, 2003/04/01 20:33 GMT
I did the cvs dance and released rc3 under the new process (fixes the missing diffform issue). Releases are now prepared on a branch, so the Zwiki trunk can be always open for business. I'll release 0.17 final a little later if all seems well.

questionable content --Simon Michael, 2003/04/01 22:55 GMT
Whoa, intense. For anyone wondering, someone posted half a dozen photos of horrifically injured and dead children.

What to do. A few thoughts.

Your thoughts ? Best wishes --SM

questionable content --SimonMichael, 2003/04/01 23:23 GMT

questionable content --DeanGoodmanson, 2003/04/01 23:42 GMT
Thanks for the disclaimer + click-through, and FrontPage cleanup. We are in our cozy comfort zones, and a shock is helpful once in awhile to the appropriate audience. I was rather bothered, just as I'm bothered by most anti-murder protests, e.g. pictures of deceased and similarly mutilated pre-born. I do not wish my children to see any such. The particular images annoyed me as they had no context (didn't wait around to read any text or check URL's.) and could have been "keep this from happening by supporting the war" to "this is happening because of the war"...but I suspected the latter due to the name. I suspect many a 3rd world orphanage to have a similar emotional/psychological effect. Even with context, certain images may evoke alternate, personal historical emotional context which could be avoided with the well placed click-through.

Peace-girl gone? I showed my wife a few days ago as she resembled a cousin. Received (by both of us) with a nod of tasteful statement, surprised there wasn't a click-through story/statement. Regarding an Iraqi child, I'm partial to global statements of belief than specific reactions.

If the vandelism keeps up please invoke IP blocking and then shields up, if it helps. Another thought on shields up: Is site-content levels (for browser & what-not controls) set via page, or server? If page, that would be a handy tool to add to an open site: All anonymous postings bump the level up. A sort of mild-shield instead of trying to track ip's or complete ShieldsUp.

A Challenge --Simon Michael, 2003/04/01 23:52 GMT

> I have posted to GeneralDiscussion before via email (in Jan). The bounce says
> I am not authorized to post, yet I am subscribed and the email is the same.

Oh ? Something's not right. If you tell me your email address I'll sort it out.


> We discussed the LatexWiki stuff briefly in Jan, but anyway if you want to
> take a look I have placed the code on my server:

Thanks.. I'll wait till it's easier. If you want to have a go, replace your render_latexwhatevertype with one of the render_stxprelink.. methods, and in the pre-render section, eg before or after t = self.stxToHtml(t), add a similar call to the latex markup processing code.

A Challenge --2003/04/02 00:02 GMT
Is someone checking the zwiki-owner address? I sent a response will full headers there. Anyway it's bob+zwiki@mcelrath.org.

And I'm sorry you guys are getting vandalized. One of the drawbacks of wiki, I suppose. Mine is supposed to be personal so I don't allow outside posting except on a couple pages. The way you're dealing seems to be reasonable (just quietly remove it & block IP's). This is not the right forum for that stuff. And I liked your peace girl, for the record. :)

fat margins --Simon Michael, 2003/04/02 00:02 GMT

> Without the dancing margins, do you still want to keep the 10% margins
> or move back back to 1 or 2%?

Though in the past I've begrudged every wasted pixel, I found the added whitespace making the wiki feel a lot more attractive. How's it for you ? Too much ?

(As poor-man's-per-user-styles, we could have dtml check for usernames in StyleSheet. Where does it all end :)

--SM torn between localized and centralized discussion, as usual

A Challenge --SimonMichael, 2003/04/02 00:06 GMT
Hi Bob - disregard, I just found your mail. Yes it does look like something fishy. If you get around to it before me, try using an address without + and see if that helps.

RFC: release branches, cvs, renaming product directory --Simon Michael, 2003/04/02 00:16 GMT

> I'm not sure what problems there would be product side.

Admin confusion caused by old ZWiki directories lying around, perhaps. Also the mental hiccup of having to adjust to a new spelling after long usage.

On the cvs side, potential problems with renaming the module but preserving history, and perhaps breaking existing checked-out copies. Just not sure.


> I see this as an opportunity to denote a major version change, and
> potentially the ability to upgrade in parrallel? Freezope won't upgrade
> from 0.9.0 due to end-user complications.

Oh ? Do you know what those are ?


> Could they then install 0.17.0 and that could be used at Zwiki and
> deprecating (without forcing immediate update/object fixing) from ZWiki
> as 0.9.0?

No, because they both use the same "ZWiki Page" object meta_type.

For now I addressed limi's issue by configuring the zwiki cvs module to use a ZWiki directory by default; started using release branches; and left the ZWiki spelling intact for 0.17.

Zwiki 0.17 released --Simon Michael, 2003/04/02 01:58 GMT
This release goes out to the people in Iraq.

Summary:

Simpler page ids, faster performance and better memory efficiency, new general-purpose page type including tracker and fit support, more robust parenting, skin improvements, preliminary stylesheet support, code cleanups, doctest no longer used.

More information: http://zwiki.org/ReleaseNotes , http://zwiki.org/KnownIssues , http://zwiki.org

Download: http://zwiki.org/releases/ZWiki-0.17.0.tgz or http://zope.org/Members/simon/ZWiki/ZWiki-0.17.0.tgz

Best wishes, --Simon

Thanks for the work --2003/04/02 04:50 GMT
I would just like to say thanks for the good work - I just got mail-in working and I had to say this is a sweet thing you have created here. BTW - the extra whitespace looks sharp.

I did have one idea - it would be nice to have a setting available for the default mailin page that would automatically make a new default mail-in page every day (use a name + date format perhaps).

Todd Mitchell

Thanks for the work --SimonMichael, 2003/04/02 16:55 GMT
Thanks!

Maybe there should be a defaultPage method, which looks for default_page but can be overridden with a python script.

ExternalEditor --DeanGoodmanson, 2003/04/02 17:53 GMT
Version 0.7 released yesterday. http://www.zope.org/Members/Caseman/ExternalEditor

Diff problem with ZWiki-0.17.0rc2 --2003/04/02 19:30 GMT
I recently installed ZWiki on a site of mine, http://pb.cowgar.com and when I click the time for a page diff, I receive:

IOError

Sorry, a site error occurred.

Traceback (innermost last):

 Module ZPublisher.Publish, line 150, in publish_module
 Module ZPublisher.Publish, line 114, in publish
 Module Zope.App.startup, line 182, in zpublisher_exception_hook
 Module ZPublisher.Publish, line 98, in publish
 Module ZPublisher.mapply, line 88, in mapply
 Module ZPublisher.Publish, line 39, in call_object
 Module Products.ZWiki.Diff, line 156, in diff
 Module Products.ZWiki.UI, line 128, in diffform
 Module App.special_dtml, line 61, in __call__
 Module DocumentTemplate.DT_String, line 412, in __call__
 Module DocumentTemplate.DT_String, line 340, in cook
 Module DocumentTemplate.DT_String, line 332, in read
 Module DocumentTemplate.DT_String, line 519, in read_raw

Anyone know why?

Diff problem with ZWiki-0.17.0rc2 --SimonMichael, 2003/04/02 21:25 GMT
#479 zwiki last edited gives IOError, fixed in 0.17.0. Thanks.

Information Today article - Zwiki in headlines --DeanGoodmanson, 2003/04/03 00:33 GMT
aaargg! The article " Quickiwiki, Swiki, Twiki, Zwiki and the Plone Wars Wiki as a PIM and Collaborative Content Tool " gives Zwiki a whole paragraph, get's the name wrong and sometimes right, then points users at a 0.9.. version of the product. :-/

Quick rant over, more later after I read it all, maybe.

The author covered a LOT with a blogger's tenacity for hyperlinking and brevity. Good read, wish it were a wiki page.

Information Today article - Zwiki in headlines --SimonMichael, 2003/04/03 02:51 GMT
Wow! Great link, thanks.

Not much depth on Zwiki.. but we're one of those he set up, along with Swiki (good company). Also we got quoted for the definition of WikiBadge :)

edit, append methods losing page author --DeanGoodmanson, 2003/04/03 05:40 GMT
When an edit or append method is made from DTML, the parameter should include REQUEST=REQUEST to get the author noted in the RecentChanges and header.

I changed FAQ, hope you don't mind.

edit, append methods losing page author --SimonMichael, 2003/04/03 16:08 GMT
No I don't mind. That's what they're there for! Thanks!

Removed PageMaintainers badge from that page.

upgradeAll traceback --SimonMichael, 2003/04/03 16:33 GMT
Dean - saw your traceback at [DeanG"0.17.0"Upgrade]?. Any idea what happened there ? I suspect you have a page called Client-Server, another with a similar name, and they clash with the same canonical id (ClientServer).

I guess the simplest improvement here would be to just log the error and move on.

upgradeAll traeback --DeanGoodmanson, 2003/04/03 16:58 GMT

> I suspect you have a page called Client-Server, another with a similar name,
> and they clash with the same canonical id (ClientServer).

That is what happened. Removed the unneeded dashed page and the script worked.

upgradeAll traceback --SimonMichael, 2003/04/03 17:31 GMT
Thanks. I added an issue and checked in a fix, untested.

I co-opted your page for the issue - renamed it to #483 upgradeAll and upgradeId may give Copy Error, set the issue properties and reparented under IssueTracker. Nifty, huh ? Also note my favourite tweak of the moment - enter in the page management field does a reparent.

reduced zwiki.org's target cache size --SimonMichael, 2003/04/03 18:03 GMT
I'm seeing way too many proxy errors due to ZopeOutOfMemoryKiller. I've reduced the target cache size for our 2 zodb connections from 2000 to 1500 each.

IssueTracker note --DeanGoodmanson, 2003/04/03 18:34 GMT
Also note my favourite tweak of the moment - enter in the page management field does a reparent.

I'm not sure what this does.

I'm looking to add a "log" field to the Issue Detail/Update page and related addition to the change properties parameter. It's much more straightforward now in 0.17.0. With that as the subject, a comment text will be added noting properties that have been changed. I'm expecting that the original values of the properties will be available as member variables? Is it this straightforward or is there a method I couldn't find to manage_GetProperty, or something to that like? (Purpose: Notify users of changes to properties and single-point of notation for why the change was made.)

escaping structured text --DeanGoodmanson, 2003/04/03 20:28 GMT
Is there a tag you can surround a block of text with to ignore STX and wikilinks on a particular section? Sometime like: Treat this next section as HTML only.

I've having trouble embedding some links, as they are triggering stx formatting. All I can do now is put them on a seperate HTML rendered page.

escaping structured text --SimonMichael, 2003/04/03 21:02 GMT
The only way to turn off STX (and everything else) for a block of text is to indent it after ::.

What kind of link triggers STX interference ?

IssueTracker note --SimonMichael, 2003/04/03 21:09 GMT
I agree with your goal, though I'm not sure how the UI should work. Merge the properties & comment form somehow ? I don't know what else to say at present.

The tweak I mentioned: in the page management form, you can type in the name of a new parent and hit enter. This used not to work.

escaping structured text --DeanGoodmanson, 2003/04/03 21:36 GMT

>>What kind of link triggers STX interference ?

These javascriptlet/favelet :

 javascript:(function(){var%20atags,i,name,a;%20anchs%20=%20document.anchors;%20for(i=0;%20i<anchs.length;%20++i)%20{%20a%20=%20anchs[i];%20name%20=%20a.name;%20a.appendChild(document.createTextNode("#"%20+%20name));%20a.style.border%20=%20"1px%20solid";%20a.href%20=%20"#"%20+%20name;%20}%20})();

Which highlights the tables on a page.

stx link(function(){var%20atags,i,name,a;%20anchs%20=%20document.anchors;%20for(i=0;%20i

url link: url link

and now for something completely different --SimonMichael, 2003/04/04 00:27 GMT
Here is a very intriguing screenshot by ToniArnold. :)

intereesting --SimonMichael, 2003/04/04 08:15 GMT
"Anti-war slogan coined, repurposed and Googlewashed... in 42 days" --via slashdot

sorry about the spam --SimonMichael, 2003/04/05 16:54 GMT
I apologize for the spam.. I have been putting this off because fixing the mail setup is tedious, complicated, error-prone and requires good concentration. Will be done soon I hope.

dtml snippet in IssueTracker --DeanGoodmanson, 2003/04/07 06:25 GMT
Would you mind explaining a bit about the code rationale for the following line in the IssueTracker and FilterIssues:

 <dtml-var "status in ['open','pending'] and '</b>' or ''">

Forgive my ignorance at this part, but I think the purpose is only to ensure open and pending statuses get displayed as bold, but there's intrinsic logic in the DTML I'm not seeing and would like to know about. (Do I need to clean my glasses, or is that a ternary operation?)

page index --goofrider, 2003/04/07 14:02 GMT
A default page index in ZWiki would be helpful

page index --DeanGoodmanson, 2003/04/07 14:07 GMT
Are you looking for something beyond AllPages and the contents ?

page index / TOC --goofrider, 2003/04/07 14:08 GMT
oh there is a TOC in ZWiki, I forgot. It's been a long time since I used plain ZWiki. ZWiki's TOC isn't available in Plone/CMF though. Any word on a port?

TOC in Plone --DeanGoodmanson, 2003/04/07 14:30 GMT
In the http://zwiki.org/plone/FrontPage example I can navigate to http://zwiki.org/plone/FrontPage/contents#FrontPage ... but that's not a definitive answer.

Perhaps the header isn't displaying the contents link? I'm not a Plone/CMF guy, so probably best to wait for the real answers. :-)

full simple minimal disappears using wikipage.zpt --2003/04/08 12:53 GMT
Is this a known issue?

Thanks
Kent

dtml snippet in IssueTracker --Simon Michael, 2003/04/08 18:03 GMT

Hi.. certainly. You'll see some eye-twisters like this, where I squeeze more functionality into a DTML expression than usual, in an attempt to reduce the overall complexity of the DTML code in the page. Does this help:

 (((status in ['open','pending']) and '</b>') or '')
 ^^^
 ||1
 |2
 3

If status is open or pending, expression 1 is true, so expression 2 returns </b>. This is "true" so expression 3 also returns </b>. If status is something else, expression 1 is false, so expression 2 returns false (the and short-circuits) and expression 3 must continue past the or and return ''. NB if we wanted to return a space instead of </b> this wouldn't work.

Well, I think it's quicker to see than to explain..

page index / TOC --Simon Michael, 2003/04/08 18:04 GMT

> oh there is a TOC in ZWiki, I forgot. It's been a long time since I used
> plain ZWiki. ZWiki's TOC isn't available in Plone/CMF though. Any word
> on a port?

Yes: in CVS and in progress.. see eg http://demo.plone.org/zwiki .

TOC in Plone --Simon Michael, 2003/04/08 18:05 GMT
NB http://zwiki.org/plone is no longer up to date with the latest skins, I'll upgrade it at some point. Actually I can probably get rid of it now.

full simple minimal disappears using wikipage.zpt --Simon Michael, 2003/04/08 18:08 GMT

> Is this a known issue?

No it isn't.. those are in wikipage.zpt so should appear. I notice they won't appear if there's no UserOptions page, though, since they depend on that.

PrettyPrint --JordanCarswell, 2003/04/09 05:07 GMT
Simon, I am trying to provide a link on my Zwiki to a "pretty print" version of the page. I guess like Zope has on their site. I know this isn't a specific Zwiki question, but if you or anyone else could simply explain how to do this, I think it is an easy matter, but I appear to be too dense to get it. BTW- I've appreciated your pro-peace links, thanks ;-)

missing full simple minimal --2003/04/09 11:17 GMT

>I notice they won't appear if there's no UserOptions page

I'm not sure what this means, I have a UserOptions link which works, don't see a UserOptions.pt in ZWiki/skins/default

PrettyPrint --DeanGoodmanson, 2003/04/09 14:10 GMT
Could you provide a link to a Zope page that has that functionality?

PrettyPrint --JordanCarswell, 2003/04/09 14:29 GMT
Sure. If you go to http://www.zope.org, at the bottom of the page there's the "Printable Page" link.

AddSmiliesPatch code broken ? --FlorianKonnertz, 2003/04/10 05:51 GMT
The file http://zwiki.org/addSmilies.txt is displayed wrong in my browser. Other txt files are fine. Can you confirm this?

PrettyPrint - ZwikiPrinting --DeanGoodmanson, 2003/04/10 06:32 GMT
Thanks. I often overlook the obvious. http://zwiki.org/print/view_source is a method that you can add a link in the header/footer via:

 <a href="&dtml-URL0;/print">Printable Page</a>

Note: It doesn't paginate. It could use a link back to the original page. It tends to choke on pages chock full of certain dtml (e.g. DeanGSandBox, and FilterIssuse).

Wonder if ZwikiToPDF would do what you want? (I'ts Not developed, but there are products which output a PDF from HTML, You can feed it the pages URL in a similar fashion with the URL0 var.)

AddingSmiliesPatch --DeanGoodmanson, 2003/04/10 06:40 GMT
You'll have to use your browsers "view source" to get the contents of http://zwiki.org/addSmilies.txt , or right-click "save link/target as" . For some reason that text file is trying to render as HTML..or at least just tromps on white-space. (for me in IE and Phoenix, respectively).

RestructuredText for Zope 2.6 released --SergeStinckwich, 2003/04/10 13:28 GMT
This is the backport of the reStructuredText from Zope 2.7 for Zope 2.6.

PrettyPrint --PieterB, 2003/04/10 13:38 GMT
I think using /print and piping through html2fo and translating the xsl-fo to pdf using fop (see http://html2fo.sourceforge.net/ and http://xml.apache.org/fop/ )

PieterB

questions re XML import export --Craeg Strong, 2003/04/10 21:12 GMT

Hello:

I have heard rumors that ZWiki will support XML import/export. We have decided that this is the way to go for all of our Zope products-- it facilitates backups, versioning, and upgrades.

We also generally add a field on our "Add Instance" forms to add instance(s) based on an uploaded XML file. XML's hierarchical nature makes it pretty natural to handle a single instance or an entire tree this way.

It makes upgrades pretty painless:

The only requirement is that the XML import feature must know how to handle the serialization format for both the current and previous release. Even that may not be necessary given XSLT or a good set of emacs macros :-)

Is this what the zwiki mavens have in mind for this feature?

Also, can you hazard a guess as to when you might start working on it?

Thanks,

--Craeg

XML import / export --DeanGoodmanson, 2003/04/10 21:28 GMT
If you are referring to the Import/Export via the Zope Management Interface...

On my "export a page, import the page to a different folder" test I have the following caveats:

  1. You need to move the file into the server's import directory.
  2. The creation_time and Last Modified fields get changed to the import time. Only last_edit_time is preserved.

..and of course History is lost, but one should only consider History as roughly reliable (+/-) as a text editor's undo feature. (Random or Scheduled packs, renames, etc.)

Zwiki & Zope3 --DeanGoodmanson, 2003/04/11 04:35 GMT
This just in:

"Stephan Richter has ported ZWiki to Zope 3, and it now lives in the public Zope CVS repository [ http://cvs.zope.org/zopeproducts/zwiki/ ] . It's still a young version, and the skin needs work, but that's OK.

This new version supports reStructuredText, which is a real boon. Thanks, Stephan!" - http://www.pycs.net/users/0000090/ (Fred L. Drake, Jr.)

XML import / export --Craeg Strong, 2003/04/11 16:23 GMT


> If you are referring to the Import/Export via the Zope Management Interface...

Sorry, I wasn't clear. I am referring to a custom XML serialization format such as might be produced via PyXML's xml.marshal.generic.Marshaller

(cf Uche's article: http://www.xml.com/pub/a/2002/09/25/py.html?page=last )

For example:

 <ZWikiWeb>
  <ZWikiPage>
    <Metadata>
      <name>MyWikiPage</name>
      <page_type>stxprelinkdtmlhtml</page_type>
      <creator>craeg strong</creator>
      <creator_ip>123.123.123.123</creator_ip>
      <creation_time>2003-04-11 12:00:00</creation_time>
       ...
    </Metadata>
    <Content>
 <![CDATA[
   ....some stx...
    .... some dtml...
   ...some html...
 ....
 ]]>
    </Content>
   </ZWikiPage>
   ...many more pages...
 </ZWikiWeb>

What you store in the XML is presentation-free, so you are insulated against such changes release to release.

This is most valuable to folks like me that have lots of important wiki pages, don't ever customize the ZWiki default presentation, but somehow often get bitten by the presentation changes each release.

This would make it possible for me to serialize out and save all my content, throw away the old ZWiki version, upgrade to the latest and greatest, and restore all my old content.

Why use Marshal instead of the generic Python pickler? Because:

So, was this what you had in mind?

--Craeg

pageable mailout_policy --DeanGoodmanson, 2003/04/11 22:03 GMT
Perhaps I haven't visited regulations in awhile, and I thought this was in tracker but can't find it, but...

Solution for a user trying to track changes to pages they have a particular interest in:

Allow a mailout_policy to be set on a per-page basis overriding the folder value. (I don't think the other policies apply, but this sets a pattern if they need to be changed also.)

This would also allow pages dynamically edited to produce mail-outs where a comment policy resides on the wiki.

XML import / export --DeanGoodmanson, 2003/04/11 22:11 GMT
Perhaps not the response you were looking for, but:

Please describe the physical steps you might expect to accomplish: "This would make it possible for me to serialize out and save all my content, throw away the old ZWiki version, upgrade to the latest and greatest, and restore all my old content."

I think this would be useful, and potentially a foundation for building an XML-RPC interface (note: XML-RPC not the foundation for this!;-)) to edit & update pages, and the start of a ZMI replacement for bulk management of a ZWiki.

dtml pointers needed --DeanGoodmanson, 2003/04/11 22:36 GMT
1. Trying to make a searchpage search folder specific catalogs. The line in the Search page: dtml-in "aq_parent.objectValues(spec=ZWiki Page)" I have a wiki at /wiki and another at /wiki/sub . From /wiki/sub/SearchPage, how would I build in /wiki or /wiki/sub to this call?

  1. Any suggestions as when to use: hasattr vs. hasProperty ?

Re: dtml pointers needed --PieterB, 2003/04/12 21:13 GMT
Have you tried:: aq_parent.aq_parent.objectValues(spec=...)

Untested, BTW.

PieterB

PieterB

Re: dtml pointers needed --DeanGoodmanson, 2003/04/13 01:38 GMT

> Have you tried:: aq_parent.aq_parent.objectValues(spec=...)

No, but I'll give it a whirl, as that will probably cover most of the search case: Search this SubWiki and Parent wiki. Alternate cases include: Search /that/ specific wiki (on the farm(?)), and Search /these/ specific ones.

XML import / export --Craeg K Strong, 2003/04/13 05:22 GMT
DeanGoodmanson wrote:


>Perhaps not the response you were looking for, but:
>
>Please describe the physical steps you might expect to accomplish: "This would
>make it possible for me to serialize out and save all my content, throw away
>the old ZWiki version, upgrade to the latest and greatest, and restore all my
>old content."
> Not sure what you mean here. Do you mean:

"How would you enhance your marshalling code so that it could accept both the current and prior version of serialized XML?"

I imagine you could either include a xxx tag and dispatch on that, or you could look for the presence or absence of key fields. The version tag is probably the better choice, as any other way would not catch subtle changes in semantics that did not affect the syntax.


>I think this would be useful, and potentially a foundation for building an
>XML-RPC interface (note: XML-RPC not the foundation for this!;-)) to edit &
>update pages, and the start of a ZMI replacement for bulk management of a
>ZWiki.
> Yes, absolutely. But my point is that once you have done the work for XML-RPC, you have also done 99 percent of the work for what I want. All you do is add an "upload" field on a ZWikiWeb form, plus an "export" button.

If you follow Roy Fielding's theories of REST-fulness, this kind of thing is a piece of cake. Once you make sure each resource is uniquely identified by a URL, you can POST an XML representation to the resource with the semantics "replace the current value of resource with that represented by this XML" Also, we came up with a convention that appending a "/xml" to any resource and doing an HTTP GET gives you the serialized XML representation for that resource. For Zope, that is totally natural: "execute the xml() method on my object"

So for example, all I would have to do would be: perform an HTTP GET on http://acme.com/mywiki/xml and I would get the XML representation of the entire wiki. It could be produced by the marshaller for the wikiweb recursively calling the marshallers of all the contained wikipages, or you could have some other algorithm.

By making all of our Zope products adhere to these semantics, we can stitch them together really easily with the rest of the (non-Zope) world.

Here is more info on REST: see: http://internet.conveyor.com/RESTwiki/moin.cgi/

I hope I answered your question...

--Craeg

PieterB's radio weblog --PieterB, 2003/04/13 17:28 GMT
I just started my Radio Weblog . I also started writing a Review of Radio Userland 8.0.8. Are there any other using Radio for a weblog (i read DeanG's weblog).

XML continued --PieterB, 2003/04/13 17:33 GMT
Is it possible to use the PythonCommunityServer to upload reST (wrapped in XML), export it to a (Wiki)website, and have Wiki capabilities to those files? I really would like to have some kind of Wiki/Weblog/collaboration thingie to use as a single way of communicating to my wiki's/weblogs. DeanG, do you know more about the status of PyCS?

PyCS, PyDS was: XML continued --DeanGoodmanson, 2003/04/15 01:39 GMT
I think the functionality your looking for would be in the Python Desktop Server. From what I understand (haven't digged into the source) that's the peice the renders, builds and uploads the site. The server receives, stores and manages comments & stats.

That said: Would you like your posts built in reStructuredText? I would hope it could be built in with the current cooking code. Sending/Syncing with a Wiki would be an interesting angle. I wonder if could post the (pyds rendered) page to your wiki in HTML & wiki syntax, extract the rendered page, then plop it on your weblog. One of the major selling points of PyCS is that the content is "baked" not "fried", meaning no read-time processing is required. (Exept comment counts, as you may have heard of recent lags in Userlands servers). One of the two systems is going to take a processing hit. But how to link your wiki content to your weblog?

The current versions of a wiki/weblog I've futzed with are versions of BillSeitz included one. I got it working on 0.16.0 (and so far it sort of works on 0.17.0) for a page-a-day, page-a-month or just page-o-blogstyle views. Considering features such as comments, referer tracking (a webloggers energy bar) calendar display and categories, and simple customization left me at the foot of a steep hill so I'm back to pondering other hacks.

Look forward to reading your blog. I'll post my comments on Radio Userland soon. So far it seems rather rough around the edges. I won't pay another $40 for Radio when my license expires unless I can blog on multiple computers, and other factors justifying the price over PyDS or other blogging environment.

Another Zwiki and Zope 3 link --DeanGoodmanson, 2003/04/15 04:04 GMT
http://toulouse.amber.org/archives/2003/04/14/comparing_zwiki_code_bases.html . Nice introduction to Zope 3 through the eyes of Zwiki. I'm not holding my breath, and glad to have a better understanding where I am now.

If someone would make a page for Zwiki and Zope 3 I'd appreciate it. I'll drop this stuff there.

Another Zwiki and Zope 3 link --SimonMichael, 2003/04/15 15:40 GMT
I'll do it, if nobody beats me to it. Thanks for the link. See also this review of Zwiki for Zope 3 . We owe a big thanks to Stephan for kick-starting this.

Another Zwiki and Zope 3 link --2003/04/15 17:36 GMT
Hi all. - I read the above reviews and got a good impression of Zope3. I guess it's time to dive into.

Another remark: I'm currently occupied with work on ZopeTopicMaps. (I hope they can be useful one day for ZWiki.) So i don't have too much time to work with you here. - Cheers, Florian

ZwikiForZope3 --2003/04/16 07:27 GMT
Do we have a ZwikiForZope3 page? I'd like to make some notes about my trial. - Maybe start at ZopeRoadmap?

XML continued... --PieterB, 2003/04/16 16:55 GMT
I'm thinking of some kind of XML/rest/rdf/dublin core format that can hold both newsitems, weblog items, wiki-discussions, etc. Something like http://www.jspwiki.org/rss.rdf (Use Google cache for IE ). Or have future newsitem, wikipage, blogitem derive from some superclass item or so.

Attributes I think the wikipage should have: - date

(maybe split up action and use a item_type such as wikipage, blogitem, wikicomment, etc.)

Any idea's?

XML continued... --DeanGoodmanson, 2003/04/16 18:46 GMT
Want to itemize this in light of RSSDiscussion ?

XML continued... --PieterB, 2003/04/16 20:24 GMT

> Want to itemize this in light of RSSDiscussion ?

For example for RSSDiscussion. I'm currently working on a Java project where we are building a XML process engine. All business oriented messages will be routed through that engine. That makes it possible and quite easy to interconnect with other systems (communicating mostly in asynchronous XML btw). Furthermore, logging, backup/restore, loadbalancing, multi-channeling are much easier to create.

If I try to see such an architecture for Zwiki/Zope3, I picture it's possible as to easiliy generate RSS-feeds, easily to generate XHTML (using XSLT and docutils), easily to connect to something like userland's XMLstorage, easily to accept mailin/mailout (transform MIME-messages and process them by the engine), etc.

Something like RecentChanges would then be a view on the state of the engine.

Note that the engine can be some kind of workflow/process XML engine, but it can also be implemented in Python/Zope3 (or CMFWorkflow, or another component for Zwiki/Plone/Blogtool/discussonfora).

The other thing I'm suggesting is some kind of uniform datamodel for weblogs, discussion boards, news-items, mailinglists, etc. That would make it easier to use components. For example using hyperlinking of WikiNames can than also be used in Plone news-items or in weblogs. I think Jeffrey tries to explain the same principle in his article at http://toulouse.amber.org/archives/2003/04/14/comparing_zwiki_code_bases.html

I hope I'm not daydreaming or that nobody understands me... I hope some of these ideas will be implemented in Zwiki/Zope3, and I'm willing to help out.

PieterB

ZoPe3 wiki --FlorianKonnertz, 2003/04/17 06:35 GMT
ZoPe3 page with RemoteWikiURL to the main dev.wiki created.

Page History --DeanGoodmanson, 2003/04/17 15:22 GMT
I think I would like to add a link to a page's /manage_change_history_page so users have an easier undo capability from the DIFF screens. The "history" links over at zope appealed to me (even though I think the term "History" is misleading to those with a version control mind-set.), and the horizontal scrolling issues with /diff can irritate.

BUT..this page has the "Copy To Present" feature, which may be dangerous to: A. Catalog, B. Pre-rendering, C. Browser cache.

Insight into the functionality and social epxerience if this would be appreciated.

XML continued... --DeanGoodmanson, 2003/04/17 15:32 GMT

>> I hope I'm not daydreaming or that nobody understands me... I hope some of these ideas will be implemented in Zwiki/Zope3, and I'm willing to help out.

I think we get it whether it's dreaming or not. :-)

Can you define some baby steps to starting a foundation for these features? ..something that could be hacked in an afternoon,perhaps. (Your top 2 favorite Wiki XML-RPC methods?)

Are there other baby steps (I prefer the term bite sized features, but..) can be done in the current Zope2 Zwiki to establish a code and feature precedent, and test-bed ?

For other dreaming, please add your 2 bits at PluginAPIDiscussion

I have an idea for a page add-on which only requires a list of Zwiki pages. I'm pretty sure I can do it with a list property and changes to product code, but I'd much rather do it through a plug-in API..

Search tips --DeanGoodmanson, 2003/04/17 21:14 GMT
After months of using a Zwiki I finally learned that you can use wildcards (*) in the SearchPage and Tracker. (Zope TextIndex catalog searches)

  1. Are there other search mechanisms (regexp,etc.) that work?
  2. What's a good way to promote these uses? TrackerSearchTips has some good info. Should similar info be put right on the SearchPage?

- --PieterB, 2003/04/18 08:57 GMT
Did a bit of ZopeRoadmap updating. Created ZopeProductsRoadmap and added the link to the Plone Q1 status

BTW Next Zope3 milestone is expected today.

editConflictDialog --WimB, 2003/04/18 13:13 GMT
When editing my personal page, I get a NameError Module Products.ZWiki.UI, line 203, in editConflictDialog

editConflictDialog --WimB, 2003/04/18 13:17 GMT
But now it's ok????

ZwikiScrumDiscussion --PieterB, 2003/04/18 17:18 GMT
ZwikiScrumDiscussion is getting crowded. Please join.

- --2003/04/20 16:06 GMT
Is that possible to add a spelling checker on edit page? As a foreigner and lover to zwiki as compared to other wikis, sincerely like to see it has this feature. (MoinMoin has it already.)

hello --simon, 2003/04/21 16:43 GMT
Hi all.. I have been unable to do zwiki work for a while due to 1. taxes 2. burnout 3. client deadlines. Hope to be back in action soon. Top of my list is reducing the noise on our list (spam blocking, etc).

re: hello --DeanGoodmanson, 2003/04/21 17:37 GMT
Need inspiration? SpamBabe.jpg

Most of the junk I've seen has been easily pluckable from the IssueTracker.

re: hello --simon, 2003/04/21 19:29 GMT
Very cute!! :)

Mailout zwiki.org changed? --FlorianKonnertz, 2003/04/23 11:48 GMT
Since 04-18 i haven't got any emails. So I checked my subscriptions and the mailinglist subcriber list: There are only two subscribers (including me), why? Was there a change in mailout on did i miss anything else?

P.S. Thanks for the nice foto, Dean! :)

Mailout zwiki.org changed? --SimonMichael, 2003/04/23 15:15 GMT
Doh!! Looks like I lost the list subscribers. I did re-create the list recently to get it listed at http://zwiki.org/mailman/listinfo and to fix the the testhostname links in the archives (didn't work). Shoot. How many were there ?

Mailout zwiki.org changed? --SimonMichael, 2003/04/23 15:40 GMT
Ok I'm trying to back out of this. Pardon any mailman chaos.

Mailout zwiki.org changed? --SimonMichael, 2003/04/23 15:45 GMT
Ok, the list should be back to the way it was before the 18th. Messages since the 18th were not archived, hopefully this one will be. What a pain.

Mailout zwiki.org changed? --SimonMichael, 2003/04/23 16:00 GMT
You need to stop the qrunner before changing Mailman's db around. Try one more.

5-day list disruption --SimonMichael, 2003/04/23 16:06 GMT
Ok, just to recap for the archives:

Florian: Since 04-18 i haven't got any emails. So I checked my subscriptions and the mailinglist subcriber list: There are only two subscribers (including me), why? Was there a change in mailout on did i miss anything else?

Simon: Doh!! Looks like I lost the list subscribers. I did re-create the list recently to get it listed at http://zwiki.org/mailman/listinfo and to fix the the testhostname links in the archives (didn't work). Shoot.

Simon: Ok, the list should be back to the way it was before the 18th. Messages since the 18th were not archived.

Now, I'll have to find a better way to fix testhostname and not being listed in the overview. Maybe someone can show me how to do it via python.

5-day list disruption --SimonMichael, 2003/04/23 16:14 GMT
I'll come back to this later, but it looks to me as if everything is working as before, except: the threaded archive view is not being updated.

I don't know enough about mailman to handle these things gracefully, so my next move may be to throw away the archive and start it afresh. Let me know if you have any insight.

5-day list disruption --SimonMichael, 2003/04/23 16:54 GMT
Aha, now it's working. Weird.

Mailman --PieterB, 2003/04/23 20:15 GMT
you can easily recreate a mailman admin from the unix mbox using pipermail. I don't know the exact syntax by heart. Let me know if you need Mailman assistence (or use mailman-lists).

structured text now starts at H2 not H3 --Simon Michael, 2003/04/25 14:35 GMT
limi requested this for plone and it makes sense. I'm not sure what migration/confusion issues may arise, let's find out. The change takes effect when a page is saved or you call /clearCache.

test 2 --Simon Michael, 2003/04/26 17:20 GMT
test 2

spam fixed, 500 issues, topics --simon, 2003/04/26 18:26 GMT
The serious #450 zwiki.org open to spam is resolved. Also, we just passed 500 issues ! Whee!

I set up an initial mailman topic or two, but they don't work like I hoped. Is there a way to set it up so eg people can choose not to receive delete/rename/property change notifications ?

Search tips --Simon Michael, 2003/04/26 20:51 GMT
zwiki-wiki@zwiki.org (DeanGoodmanson) writes:
> After months of using a Zwiki I finally learned that you can use
> wildcards (*) in the SearchPage and Tracker. (Zope TextIndex catalog
> searches)

I thought SearchPage was a simple string match (no regexps). Tracker is a catalog search, so catalog syntax works there.

editConflictDialog --Simon Michael, 2003/04/26 22:26 GMT
Hi Wim, welcome back.


> When editing my personal page, I get a NameError *Module
> Products.ZWiki.UI, line 203, in editConflictDialog*

Thanks - missing imports fixed in UI.py for 0.18rc2.

Search Tips --DeanGoodmanson, 2003/04/27 02:59 GMT

>>I thought SearchPage was a simple string match (no regexps).

That seems to be the case. :-/

diff feedback --DeanGoodmanson, 2003/04/30 15:36 GMT
I would prefer the diff screen buttons be disabled instead of removed when not available.

I keep clicking "previous", then looking at the content, click again intending "previous", but by this time "next" has filled in under my mouse and I click the wrong thing.

calling ZWiki.Utils method from DTML --BillSeitz, 2003/04/30 18:09 GMT
I'm back on the idea of expanding WikiWords all over the place. I wrote a function Utils.expandWikiword(s), which I can call find from in ZWikiPage.py. But I'd also like to call it from within standard_wiki_header. Maybe this is more of a stupid Zope question, but can you tell how to do this? (In a previous incarnation I had a method within the ZWikiPage.py class, and could call that OK...)

expanding wiki words --DeanGoodmanson, 2003/04/30 18:25 GMT
I can't give you a good answer, probably not if I had more information, but I'm curious anyway...

Would you mind describing what expanding WikiWords entails? Is it an extension to the existing "wikilink" function?

While I've got your attention..whats your opinion on fuzzy WikiNaming? Such that ZWiki and ZwiKi would both resolve to ZWiki. Or just aggressively promote FreeformLinks? Would it lead to further ambiguity over names such as "reStructuredText" ?

python code for expandWikiname() --BillSeitz, 2003/04/30 19:44 GMT
Here's my function - the goal is to "expand" the name by inserting a space before each capital letter :

 def expandWikiword(word):
    import string
    #if all caps or caps/nums, then return same
    if (string.upper(word) == word):
        return word
    else:
        #what about z20... case?
        #else insert space before each cap, then left trim
        import re
        eWord = re.sub('([A-Z])', r' \1', word) #insert space before every cap letter
        return string.lstrip(eWord) #strip out leading space

also changing create-page "?" to square brackets --BillSeitz, 2003/04/30 19:49 GMT
so that those expanded words make an enclosed phrase (and because I've always hated the question-mark because it confuses non-wiki people) :

 if getattr(folder,'relative_urls',None):
                return '[%s<a class="new" href="%s/editform?page=%s" title="create this page">]</a>' % \
                       (expandWikiword(morig), quote(self.id()), quote(m))

here's the normal way it's called --BillSeitz, 2003/04/30 19:51 GMT
from within ZWikiPage.py :

 return '<a href="%s" title="%s">%s</a>' % \
                       (quote(m),linktitle,expandWikiword(morig))

So want to call this from within standard_wiki_header --BillSeitz, 2003/04/30 19:59 GMT
something like this :

 <TITLE><dtml-var expr="expandWikiword(getId())">

But that doesn't work. It generates a "NameError: global name expandWikiword is not defined"

If I try :

 <TITLE><dtml-var expr="Utils.expandWikiword(getId())">

Then I get "NameError: global name Utils is not defined"

If I try :

 <TITLE><dtml-var expr="ZWiki.Utils.expandWikiword(getId())">

Then I get :

   File "D:\Zope\WebSeitz\lib\python\AccessControl\ZopeGuards.py", line 47, in guarded_getattr
    try: v = getattr(inst, name)
AttributeError: Utils

So how do I call this sucker?

Solution for now --BillSeitz, 2003/04/30 20:08 GMT
I put back the stub of the function I used to have within ZWikiPage.py :

 def title_exploded(self):
        """
        created by Seitz so that standard_wiki_header can show WikiName of 
        page split into separate words for search engine purposes
        """
        from Utils import expandWikiword
        return expandWikiword(self.getId())

Then I call that from within standard_wiki_header :

 <TITLE><dtml-var expr="title_exploded()">

of course I might hate this feature once I roll it out --BillSeitz, 2003/04/30 20:15 GMT
I know I'll get plenty of false-splits like Z Wiki and Py Thon, etc.

But I'm going to take it live anyway and think about how I feel about it.

This is all part of my WebSeitzWiki:SpreadingWikiweblog master plan.

calling title_exploded --DeanGoodmanson, 2003/04/30 23:02 GMT
Do you have all of the Zope security decorations/attributes/whatever above the function to expose it similar to the other public API's in the Zwiki code? (See folder(), etc.)