September 1999 - Zope-Dev

manage_changeProperties and the "lines" property
by Dan Rusch 22 Sep '99

22 Sep '99

Has anyone (or does anyone "know" how to) changed an existing "lines" property of a DTML Document using the manage_changeProperties DTML Document method?

1 0

RE: [Zope-dev] minor bug in Zope content-type.
by Michel Pelletier 22 Sep '99

22 Sep '99

> -----Original Message----- > From: Terje Malmedal [mailto:tm@funcom.com] > Sent: Wednesday, September 22, 1999 5:47 AM > To: zope-dev(a)zope.org > Subject: [Zope-dev] minor bug in Zope content-type. > > > > When I send a HEAD command to a Zope server it responds with a > different Content-Type than when I give it a GET command, www.zope.org > has the same problem: Can you post this to the Collector? > [root@cornavin /root]# telnet www.zope.org 80 > Trying 209.67.167.110... > Connected to zope.org. > Escape character is '^]'. > HEAD / HTTP1.1 > > HTTP/1.1 200 OK > Date: Wed, 22 Sep 1999 09:14:46 GMT > Server: Apache/1.3.4 (Unix) > Connection: close > Ms-Author-Via: DAV > X-Powered-By: Zope (www.zope.org) Python (www.python.org) > Content-Length: 5359 > Last-Modified: Mon, 13 Sep 1999 19:51:53 GMT > Content-Type: application/octet-stream > > the GET command gives me text/html which is correct. > > I used to get application/octet-stream on GET commands as well when I > had a <!DOCTYPE > thingy in the beginning of the document, but it > seems to have been fixed when I upgraded to version 2.0.1 > > -- > - Terje > tm(a)funcom.com > > _______________________________________________ > Zope-Dev maillist - Zope-Dev(a)zope.org > http://www.zope.org/mailman/listinfo/zope-dev > > (To receive general Zope announcements, see: > http://www.zope.org/mailman/listinfo/zope-announce > > For non-developer, user-level issues, > zope(a)zope.org, http://www.zope.org/mailman/listinfo/zope ) >

2 1

Changing storage manager
by mvalente＠esoterica.pt 22 Sep '99

22 Sep '99

Hi there: In the ZODB Howto: "A Storage Manager is the database back end of Zope. The default storage manager for Zope is a ZODB3 database file object that can be found in the top level var directory of Zope (var/Data.fs). There are other storage managers for Zope, in various stages of development: o The Relational Storage Manager stores all Zope transactions in a relational database. This is still experimental." I've looked all over but I was unable to find the way to change the storage manager from the Data.fs file to a relational DB (mySQL, Sybase, etc ) or indeed to any other kind (like the Temp storage manager also mentioned in the ZODB Howto). Can anyone give me some insights or pointers on how to do this ? C U! -- Mario Valente

1 0

minor bug in Zope content-type.
by Terje Malmedal 22 Sep '99

22 Sep '99

When I send a HEAD command to a Zope server it responds with a different Content-Type than when I give it a GET command, www.zope.org has the same problem: [root@cornavin /root]# telnet www.zope.org 80 Trying 209.67.167.110... Connected to zope.org. Escape character is '^]'. HEAD / HTTP1.1 HTTP/1.1 200 OK Date: Wed, 22 Sep 1999 09:14:46 GMT Server: Apache/1.3.4 (Unix) Connection: close Ms-Author-Via: DAV X-Powered-By: Zope (www.zope.org) Python (www.python.org) Content-Length: 5359 Last-Modified: Mon, 13 Sep 1999 19:51:53 GMT Content-Type: application/octet-stream the GET command gives me text/html which is correct. I used to get application/octet-stream on GET commands as well when I had a <!DOCTYPE > thingy in the beginning of the document, but it seems to have been fixed when I upgraded to version 2.0.1 -- - Terje tm(a)funcom.com

1 0

FTP support for SQL methods
by Brian Hooper 22 Sep '99

22 Sep '99

Hi there - Should / does FTP work with SQL Methods? At least, in my environment, it doesn't, but after digging around in the code, it looks like since Shared.DC.ZRDB.DA.DA inherits from OFS.SimpleItem.Item (which defines manage_FTPstat and manage_FTPlist), it should be sufficient to just define a manage_FTPget method for DA, and that's it, right? Does anyone out there have any wisdom / insight on this? --Brian Hooper

1 0

Re: [Zope-dev] PROPOSAL: Site objects
by Adam Feuer 21 Sep '99

21 Sep '99

> Quoting Michel Pelletier <michel(a)digicool.com>: > > > o Small patch to ZServer.HTTPRequest > > > -- so it places the actual socket IP and Port numbers into the > > > request's environment (HTTP_SOCKET_IP and HTTP_SOCKET_PORT) > > > > > > -- i didn't use HTTP_HOST for mapping since it is symbolic; > > > i wanted to have a 1:1 mapping from IP address to path > > > > Hmmm.. this sounds like it wouldn't work with HTTP 1.1 'Host' header > > based virtual hosting. Where the IP addr is the same for all virtual > > hosts. > > michel, > well, yes. but the IP-based scheme seemed the hardest problem to > tackle. :-) > > it would be easy enough to change the Remapper logic to look for the > HTTP_HOST key, and if it exists, use it; othewise check the socket IP > address, and use that. i forgot to say, i'll update my code to do that! :-) cheers adam -- Adam Feuer adamf(a)pobox.com

1 0

RE: [Zope-dev] PROPOSAL: Site objects
by Michel Pelletier 21 Sep '99

21 Sep '99

> -----Original Message----- > From: Adam Feuer [mailto:adamf@pobox.com] > Sent: Tuesday, September 21, 1999 3:52 PM > To: Evan Simpson; Jim Fulton > Cc: zope-dev(a)zope.org > Subject: Re: [Zope-dev] PROPOSAL: Site objects > > o Small patch to ZServer.HTTPRequest > -- so it places the actual socket IP and Port numbers into the > request's environment (HTTP_SOCKET_IP and HTTP_SOCKET_PORT) > > -- i didn't use HTTP_HOST for mapping since it is symbolic; > i wanted to have a 1:1 mapping from IP address to path Hmmm.. this sounds like it wouldn't work with HTTP 1.1 'Host' header based virtual hosting. Where the IP addr is the same for all virtual hosts. -Michel

1 0

RE: [Zope-dev] using persistent storage inside Zope
by Michel Pelletier 21 Sep '99

21 Sep '99

> -----Original Message----- > From: Adam Feuer [mailto:adamf@pobox.com] > Sent: Monday, September 20, 1999 12:54 PM > To: zope-dev(a)zope.org > Subject: [Zope-dev] using persistent storage inside Zope > > > zope folks, > if this has already been covered somewhere, let me know-- i just > couldn't find information about it on the Zope site! > > i'm building a Zope product that needs to store some persistent > information in the Zope object database. the information needs to be > accessible to all instances of that product. I would suggest not making this data a class attribute, and instead make it an instance of an object that provides a service to your instances via a protocol you define yourself. For example, a user folder provides an authentication server to other zope objects. > my approach was: write a class that derives from > Persistence.Persistent that will store the data and mediate access to > it. then, in the product's __init__.py file, instantiate this class > and put the instance into the Zope object database, then call > get_transaction().commit() to commit it. > theoretically, once my instance is in the Zope object database, > other objects can access it-- adding or subtracting data. and the > instance will handle committing its own changes, so it won't forget > anything between restarts of Zope. Your sorta halfway there, but I would make your 'storage' a full fledged Zope object, not some sort of product class attribute, if this is what you mean. Maybe I'm the one confused. > my question: how do actually do this?? :-) i found a kludgy way to > store a class instance in the Zope object database from the product's > __init__.py file, but i can't figure out how to connect to this object > from a random instance of my product. > > amy i going about this all wrong? is there a better way of storing > and accessing persistent objects from inside of Zope? or if my > approach is correct, how do actually implement it? > > here's what i am doing in the product's __init__.py file: > > # get the ZODB root-- kludge! > root=context._ProductContext__app._p_jar.root() > # now make the object instance if we need to > if root.has_key('FooObject'): > print "FooObject already in DB" > else: > print "made new FooObject." > root['FooObject'] = Foo.Foo() > get_transaction().commit() > > > the problem seems to be, i don't know how to get access to the Zope > object database's root object from inside an instance of my product. > what i want to do inside the product base class, during an add or > delete of an instance of the product, is something like: > > # somehow get the root of the Zope ZODB :-) > root=Somehow.get.ZODB.root() self._p_jar.root() > root['FooObject'].remove_item(someitem) > get_transaction().commit() All persistence objects have an _p_jar attribute which represents the current database 'connection'. _p_jar.root() will give you the root object. But what it looks like your doing is stuffing something in the object database that Zope is not aware about? Why not just make it a Zope object? This smells like a hack. -Michel

3 2

Length of Search Results/ZServer Crawling
by Jason Spisak 21 Sep '99

21 Sep '99

Zopsters, Does anyone know why this code:    Outside the Catalog, renders the number of all the objects in the Catalog and not the Amount of MyClasse instances in the Catalog? I'd like to get the number of objects of a give type with some property values that are cataloged, but I'm unable to. Also, ZServer has been slowing to a crawl and taking 2-3 minutes to give back responses at times. Very vague, I know, but I don't see activity on all of the threads, so I don't see how more threads would solve the problem. Right now I am running what looks like the default 6 threads. All with 13% of CPU percentage. I've got about 20 people hammering this thing non-stop all day long. The spead is good when I restart Zope, but then it deteriorates the longer the process is leftr running. -- Jason Spisak webmaster(a)hiretechs.com

3 2

Re: [Zope-dev] what you see and what dont is all dom
by cwitty＠newtonlabs.com 21 Sep '99

21 Sep '99

Richard Wackerbarth <rkw(a)dataplex.net> writes: > On Mon, 20 Sep 1999, Martijn Pieters wrote: > > In Zope, the inputs to a rendered document depend on the acquisition path > > taken, and there is an unlimited number of combinations. I don't think that > > a dataflow scheme is therefore applicable to Zope. > > I tend to agree. Cacheing of targets, which is implicit in dataflow, only works > when there are a limited number of targets and those targets are likely to be > used in the future. Well, there are certainly sites where very few HTTP requests would be repeated, and caching would be a waste of time; but I'll bet there are sites that would be willing to trade off some of the power of Zope (by avoiding some of the more interesting uses of acquisition) for significantly better performance. Carl Witty

1 0