- Zope-Dev - Zope lists

FCGIServer.py broken again?
by news-list.zope.dev＠innominate.de 15 Dec '99

15 Dec '99

hi, i just wanted to set up a zope server on another machine: - apache on port 80 + mod_fastcgi-snapshot - zope 2.1.1 using fcgi via socket i'm running the very same config on another machine without any problems but here i get tracebacks after the first request for a zope page: -------------------------------------------------- Traceback (innermost last): File "/usr/local/lib/Zope-2.1.1-src/z2.py", line 583, in ? asyncore.loop() File "/usr/local/lib/Zope-2.1.1-src/ZServer/medusa/asyncore.py", line 107, in loop poll_fun (timeout) File "/usr/local/lib/Zope-2.1.1-src/ZServer/medusa/asyncore.py", line 66, in poll x.handle_error() File "/usr/local/lib/Zope-2.1.1-src/ZServer/medusa/asyncore.py", line 321, in handle_error self.close() File "/usr/local/lib/zope/ZServer/FCGIServer.py", line 573, in close p.more() # free up resources held by producer File "/usr/local/lib/Zope-2.1.1-src/ZServer/Producers.py", line 108, in more getattr(self.logger, self.method)(self.bytes) File "/usr/local/lib/zope/ZServer/FCGIServer.py", line 496, in log_request time.gmtime(time.time()) TypeError: illegal argument type for built-in operation -------------------------------------------------- the code that seems to be causing the traceback is this: self.server.logger.log ( self.addr[0], '%d - - [%s] "%s %s" %d' % ( self.addr[1], time.strftime ( '%d/%b/%Y:%H:%M:%S ', time.gmtime(time.time()) ) + tz_for_log, method, path, bytes ) ) when i replace the %d for self.addr[1] to %s it seems to work but then the logging looks ugly... anybody with the same problem or maybe even a solution...?! the strange thing is that the same config is working on another machine so i guess maybe some external influence (apache, proxy) is causing the problem... thilo -- mezger(a)innominate.de innominate AG networking people fon: +49.30.308806-11 fax: -77 web: http://innominate.de pgp: /pgp/tm

1 0

Re: Threaded discussion usability (ZDiscussion inspi
by Squishdot 15 Dec '99

15 Dec '99

I'm also very interested in opinions/discussion regarding Squishdot's usability so I think its okay to post this at Squishdot's mailing list or even post this as an article at the website itself. ross boylan <rossboylan(a)stanfordalumni.org> wrote: > I posted this at a site I'm working on which uses ZDiscussions. I thought > it might be of some interest for the Zope-dev list too. I know that a > number of these ideas are in a squishdot direction, so I've added a few > thoughts on that at the bottom. > <snip> > ------------------------ sqishdot postscript > In the above, all the following are squishy: > * flat hierarchy > * expand out the postings under the heading > * reverse chrono sort > > But I think squishdot has some major usability issues itself: > * The screen is way too busy, and having multiple things going on in > different subareas of the screen is going to lose people. I say this based > partly on general principles, and partly based on having watched a couple > of people navigating around some sites. Well, its because I was literally copying Slashdot's design -- its not inherent in squishdot itself -- you could design a set of simpler screens or use the plain vanilla mode... Given the current design, could you give me more details on what and where exactly people are having problems with it? > * As far as I can tell, once stuff disappears off the bottom of squishdot, > it's just gone. And there isn't room for that much. Well, the number of articles is set via an option in the options tab And its not exactly gone -- if you search for it, its still there. The reason why things fall off the page comes from the assumption that as news gets older, it gets less important -- doesnt it? In fact, thats the reason for the previous articles box -- it should show all the article titles for the past week (any older than that would probably be not very important) and you can also control how old the articles that are listed in the previous articles box... However, if you still wanted a list of all articles, I suppose you could build a page that did that -- I just think a list of the most recent articles is the default one that most people would prefer... > * The novice view should not be the only option. Exactly what do you mean by novice view? > > Well, this seems like a natural for cross-posting to squishdot, but since > that is frowned on, I won't. > I, for one, won't frown on it :). Regards, Butch ===== Butch Landingin Squishdot maintainer http://squishdot.org squishdot(a)yahoo.com __________________________________________________ Do You Yahoo!? Thousands of Stores. Millions of Products. All in one place. Yahoo! Shopping: http://shopping.yahoo.com

1 0

RE: [Zope-dev] Object ownership
by Toby Dickenson 15 Dec '99

15 Dec '99

> is there a method that i could use to find out an object's owner? I have this in my Intranet standard_html_header. (Its also a good example of dtml gone wild) <dtml-if get_local_roles> <dtml-in get_local_roles> <dtml-if "'Owner' in _['sequence-item'] and _['sequence-key'] != 'za'"> <dtml-if one_owner_already_mentioned> ; <dtml-else> This page is maintained by <dtml-call "REQUEST.set('one_owner_already_mentioned','yes')"> </dtml-if> <a href="mailto:<dtml-var sequence-key>"><dtml-var sequence-key></a> </dtml-if> </dtml-in> <dtml-if one_owner_already_mentioned> </dtml-if> </dtml-if>

1 0

Threaded discussion usability (ZDiscussion inspired)
by Ross Boylan 15 Dec '99

15 Dec '99

I posted this at a site I'm working on which uses ZDiscussions. I thought it might be of some interest for the Zope-dev list too. I know that a number of these ideas are in a squishdot direction, so I've added a few thoughts on that at the bottom. ------------------ original post ------------------------------------------------ Apparently it is not obvious that clicking the plus sign next to a discussion item expands to show subtopics. Kathy, who was intrepid enough to make it to the new forum, did not know that. I think this strongly suggests that the default format should simply show all posts flattened out. It might be best if they were expanded out, rather than simply showing the subject headline. Also, anything which is on the "next" screen (deja.com has a "view previous messages" button which most people didn't follow, in both senses of the word follow) tends to be out of sight and mind. I think it might be a good idea to sort in reverse chronological order, though the default ZDiscussion format is probably more natural for most people. It is very unfortunate that the current ZDiscussions (0.3) not only hides the responses to a message under the +, but gives no easy way to tell if one of the hidden items is a recent post. ------------------------ sqishdot postscript ------------------------------------------------ In the above, all the following are squishy: * flat hierarchy * expand out the postings under the heading * reverse chrono sort But I think squishdot has some major usability issues itself: * The screen is way too busy, and having multiple things going on in different subareas of the screen is going to lose people. I say this based partly on general principles, and partly based on having watched a couple of people navigating around some sites. * As far as I can tell, once stuff disappears off the bottom of squishdot, it's just gone. And there isn't room for that much. * The novice view should not be the only option. Well, this seems like a natural for cross-posting to squishdot, but since that is frowned on, I won't.

1 0

Defining permissions in inherited classes
by Itai Tavor 15 Dec '99

15 Dec '99

Hi, I've been stuck with this problems for weeks... I posted an earlier question bug got no replies, so I thought I'd try again. I have a product that is used as a mix-in class in several other products. This product adds a management interface screen and I want to define permissions to control that screen in this class. So, I do: class Publisher(RoleManager): __ac_permissions__ = ( ('Change publishing', ('manage_publish')) ) manage_publish = HTMLFile('publish', globals()) In the product that uses this class: class Something(Folder, Publisher.Publisher): __ac_permissions__ = ( ('View management screens', ('manage_tabs', 'manage_main', ....)) ..... ) manage_options = ( {'label':'Edit', 'action':'manage_main'}, {'label':'Publish', 'action':'manage_publish'} ) Now, I think this should work... that's how it's done in AccessRole.Role.RoleManager - it defines __ac_permissions__ that get added to the classes that inherit from it. But it doesn't work for me... the permissions defined in Publisher don't appear anywhere. Any idea why, or what I'm doing wrong? Thanks Itai -- Itai Tavor -- "Je sautille, donc je suis." -- itavor(a)vic.bigpond.net.au -- - Kermit the Frog -- -- "What he needs now is understanding... and a confederate victory" -- -- Dr. Jacobi, Twin Peaks --

1 0

__cmp__ in python classes?
by Benno Rice 15 Dec '99

15 Dec '99

Hi, I've got a system I'm developing that has a bunch of Python classes that I'm using. One of these classes defines a __cmp__ method, but it doesn't appear to be being called when I test for equality of two objects. I checked to see whether it was Acquisition that was making it interesting but even when I got the object directly it wasn't calling my __cmp__ method. Does Zope ignore __cmp__ methods? If so, is there any way around it? -- Benno Rice "No, no. We're *sweet* and XNFP Aries Dark Subculture- *innocent* evil bastards." friendly Internet Geek benno(a)netizen.com.au "Defend your joy"

3 4

RE: [Zope-dev] Re: Subpath traversal interface (was RE: ANN: Pyth on Methods 0.1.7 up and over to DC) and over to DC)
by Michel Pelletier 15 Dec '99

15 Dec '99

> -----Original Message----- > From: Evan Simpson [mailto:evan@4-am.com] > Sent: Tuesday, December 14, 1999 2:59 PM > To: Michel Pelletier > Cc: 'Phillip J. Eby'; zope-dev(a)zope.org > Subject: Re: [Zope-dev] Re: Subpath traversal interface (was RE: ANN: > Python Methods 0.1.7 up and over to DC) and over to DC) > > > It's a wafer-thin cost, really! :-) it was Jim's objection, but I see the logic. > I'll be interested to see your implementation. Both > SiteRoots and Access > Rules depend abjectly on the hook. SiteRoots hook into their > containers' > traversal so that they can call REQUEST.setURL. We just do this from __bobo_traverse__ if self.base == REQUEST['SERVER_URL']. > Designating > an object as > an Access Rule causes it to be called before its container is entered, > which allows the path foolery necessary for (my version of) virtual > hosting. Most of the path foolery I think you're talking about is not needed now because absolute_url is now recursive. The Rule thing though, that might need to hook, I'd have to bug Jim about it. > When I was first designing SiteAccess, I considered hooking > __bobo_traverse__, but I was really worried about what to do > if a container > already implemented it, Then that container can't be a SiteObject. > or if a SiteRoot and an Access Rule > wanted to share > the same container. SiteRoots are containers, and we haven't implimented Access Rules yet... > It would be great if __bobo worked like > __before, but > I imagine that it's a bit late to be changing that one. How would you like the behavior to change? > I have been struggling with the new absolute_url > implementation, by the > way. It makes perfect sense, and would work fine with a > virtual hosting > solution which can insert an object into the acquisition > chain. Sadly, > mine can't. Maybe Jim can explain this better, but we find a very pleasent isomorphism between SiteObjects, Portal Objects, and virtual hosting, this is why we have a new object type that gets 'inserted' into the chain. Since a Siteobject works just like a folder, except for the URL rewriting (and that doesn't happen if the SERVER_URL doesn't match the base URL of the virtual host) then you are really only inserting a folder, but a sorta special folder that knows some tricks. -Michel

2 2

using ZClasses as DA brains?
by Anthony Baxter 15 Dec '99

15 Dec '99

Heya, Has anyone figured out the necessary magic to get the ZClass-as-brains from the current CVS to work? It seems like the 'advanced' tab needs 'manage_product_zclass_info' which is in Shared/DC/ZRDB/DA.py - but I'm stuffed if I can figure out how this is meant to work... anyone? Anthony

1 0

RE: [Zope-dev] Re: Subpath traversal interface (was RE: ANN: Pyth on Methods 0.1.7 up and over to DC) and over to DC)
by Michel Pelletier 14 Dec '99

14 Dec '99

> -----Original Message----- > From: Evan Simpson [mailto:evan@4-am.com] > On another note, the capability described above is actually > quite similar > to the __before_traverse__ behavior which SiteAccess hacks into the > traversal machinery. An Access Rule can look ahead one or more path > elements, and then easily manipulate, add, or remove them. > Also, multiple > objects can hook into the same folder's __b_t__ chain without > colliding and > with controllable invocation order, which is part of why I invented it > rather than overriding __bobo_traverse__. Any chance that DC could > incorporate it (the __before_traverse__, I mean)? It's quite > clean and > well tested at this point, and I could stop nervously examining each > release of Zope to see if it will break SiteAccess. The reason why we didn't roll it in is that we didn't want Yet Another Hook adding a per-traveral cost, and we could impliment SiteObjects without them. Whether or not we can impliment all your other features without is another question. What feature of SiteAccess uses the before hook? We were able to emulate the virtualhosting behavior with code silimar to yours (I stole some of yours actually) without havign to define the before hook. -Michel

2 1

RE: [Zope-dev] CVS
by Ken Manheimer 14 Dec '99

14 Dec '99

> I see lots of checkins that don't seem to appear when I update Zope2. > Should I be doing something different? There was a problem with the mirroring software, which lead to updates being delayed yesterday evening (over the course of around two hours) and on friday, for i'm not sure how long. (The problem was actually with the version of rsync we were running on the mirror host - after some flailing with our driver script, upgrading from rsync 2.1.1 to 2.3.1 was the solution, sigh.) > The last couple of days seems to have corrupted most of my icons. Sorry about that - check and see if you get the right thing now! Let me know if you continue to have problems. Ken Manheimer klm(a)digicool.com

3 4