March 2000 - Zope-Dev

Initial zope distributions
by Sin Hang Kin 04 Mar '00

04 Mar '00

It will be cool to have a fresh zope installation to have the documents installed (or can be dl and imported).

1 0

Client-side Zope
by Sin Hang Kin 04 Mar '00

04 Mar '00

Continue to moving more job to the client, There will be more for Zope to perform: 1. There will be necessary to generate various section-specific temporary objects for CSS, DTD and Scripting files, framed or embeded html. 2. There will be necessary to support java or javascript access of zope in the client. In the first requirement, I would like to suggest an machenism to create an section specific temporary object (which zope create the unique name and substitute it) which dtml is able to write to it. The new object can be just a new dtml-document. As in the following <dtml-object dtdobject> <dtml-object scriptobject> <dtml-object cssobject> <script lang="javascript" src="<dtml-var scriptobject>"> .... <dtml-write target=scriptobject> function <dtml-var uniquevalidfunctionname>(){ .... } </dtml-write> I don't known about the solution of the second request. In the long run, zope might encourage separate of control and data. The current opening all tags in one dtml-method and closing them in another is very dangerous if the control get complicated. I would like to suggest DC consider a xslt like syntax which can also be used in server side xml to html conversion if the client do not support it. If zope is going seriously to the international market and xml, then unicode is important in the core level.

1 0

ZClass ID bug
by Rik Hoekstra 03 Mar '00

03 Mar '00

I have a ZClass based on CatalogAware (no problem) and on DTMLDocument/DTMLMethod (it does not matter which). A number of instances are added in a batch. All this works well, but the id representation of the resulting instances are not uniform and sometimes they're just wrong. - In a normal manage_main view of their Container, the ids will show without problem. - 'objectIds' (and 'objectItems') also returns the right values, but objectValues returns a <string> (i.e. an object) instead. This cannot be brought to show its value (at least not by me) - All the same absolute_url also returns <http://bla/bla/<string>. This makes ZClasses subclassed from DTMLDocument/DTMLMethod unusable, at least for my purposes. I solved the problem by making a ZClasses based on objectManager, which is not ideal, but workable for me. Has anyone else seen this problem? Is there a solution/patch around. i have no idea how to debug this? I will post this to the Collector as well, but I thought this would be of an interest to the list as well. Rik

2 1

Re: [Zope-dev] Using Zope in a client-server system
by Hung Jung Lu 03 Mar '00

03 Mar '00

itai tavor <itavo-(a)bigpond.net.au> wrote: >- I also need to push data to the workstations - based on events detected >at the server, I need to bring up a message on a workstation, and remove it >in response to another event. I really don't want to refresh a frame every >second for this, and I also would hate to have to write a Java applet to do >it. Bi-directional communication. Hmm... no easy way out. Java applet would be the solution, which periodically pulls from the server. But this problem seems to come up more and more often nowadays. Web browsers, being pure clients, are not good enough. What we want is some technology that implements servers on the user's end. Mini-servers, if you wish. Actually, I think this maybe a long term tendency: more and more mini-servers would be running on end-users' machines. That being said, it is actually not too hard to implement mini-servers with Python. An HTTP server that analyzes the URL (which might contain port number and form data, like http://your.machine.com:8088/?x=3&y=4) and displays a message box, probably would take about 30 to 40 lines of Python code. It'll even make a beep sound if you want. :) Quite a bargain. If you want to know more, I can say more, but it kind of depend on your configuration. Web browsers (HTTP clients) made a big impact on the way how people use computers. I believe that nowadays we also need some generic HTTP mini-servers so people can respond to outside events, there is a real need for this kind of products. The possibilities are endless. I think, if I were you, I'd devote some time into this arena. :) Applets can do the job, but why restrict yourself to the tightly-controlled, hard-to-plug-in world of browsers, when a server technology can give you a totally open blue sky? regards, Hung Jung ______________________________________________________ Get Your Private, Free Email at http://www.hotmail.com

3 2

CatalogAwareness and CopySupport, cut and paste problem
by König Alexander 03 Mar '00

03 Mar '00

I'm using a product python class that inherits from CatalogAwareness and CopySupport. The automatic indexing and unindexing on creation and deletion works fine. When a cutted object is pasted into another folder, manage_beforeDelete raises Error Type: AttributeError Error Value: aq_acquire. Debugging brought up, that absolute_url() doesn't work when the object is marked as cutted. The acquisitiontree is traversed up to th root-application object. Zope.absolute_url() tries to call self.aq_acquire('REQUEST').script. I didn't overwrite any of the involved methods and have no idea what else could have such a bad influence on the Application object. When a object is explicitly deleted, manage_deforeDelete() works well. I would be very glad if someone could help me out. Thanks, Alex

3 3

Re: [Zope-dev] Fwd: Multi-argument append() is illegal
by Tres Seaver 03 Mar '00

03 Mar '00

Martijn Pieters <mj(a)digicool.com> > > From: "Patrick Phalen" <zope(a)teleo.net> > > Fredrik Lundh has mentioned that this breaks Zope (and Medusa). > > > > If so, it's probably not too soon to begin looking into it. 1.6 is due > > this summer. > > > > <Guido's message snipped> > > Using a more advanced pattern (one that weeds out most proper appends, and > takes multiple lines into account), plus some hand weeding, brings up the > following list of 'illegal' tuple appends: > > lib\python\ExportImportXML.py(109): > self._tindex.append(self._oid, here) > lib\python\Shared\DC\ZRDB\RDB.py(199): > parsers.append(i,parser) > lib\python\Shared\DC\xml\ppml.py(278): > def __setitem__(self, k, v): self._d.append(k,v) > lib\python\ZODB\FileStorage.py(633): > self._tappend(oid, here) > lib\python\ZPublisher\cgi.py(603): > r.append(name, value) > utilities\FS.py(145): > self._tappend(oid, here) > ZServer\HTTPServer.py(285): > request.channel.queue.append(self.module_name, zrequest, zresponse) > ZServer\medusa\asyncore.py(84): > l.append (fd, flags) > > Above 8 occurences are all cases of tuples being appended to a list, using > the 'illegal' style syntax. Note that I still may have missed occurences, > and that Python 1.6 might break other things. The change from 1.5.1 to 1.5.2 > broke certain things in Zope as well, usually because of library changes. > > For those of you that grok POSIX regexp, here is what I used to search these > out: > > append[[:blank:]]*\(([^[{(,]|[[:space:]])+, > > where [:blank:] stands for whitespace (not including newlines), and > [:space:] is whitespace (including newline). This will miss aliasing of > .append (app=list.append), which I hunted out with a separate regexp, and > lines like: > > append(1 # number 1 > ,2 # number 2 > > which were again hunted (and none found) with another simple regexp. So I am > fairly sure I got em all =). > > I have posted these to the Collector. The other thing to look out for is the "extra layer of indirection", where append is invoked through another name. For example:: >>> foobar = [] >>> push = foobar.append >>> push( 'foo', 'bar' ) >>> foobar [('foo', 'bar')] This gets especially ugly if the bound method is passed as a "normal" function pointer into some completely different scope, which might invoke it with arbitrary parameters. Tres. -- ========================================================= Tres Seaver tseaver(a)palladion.com 713-523-6582 Palladion Software http://www.palladion.com

2 1

Re: [Zope-dev] problems with TM.py
by Hung Jung Lu 03 Mar '00

03 Mar '00

pavlos christoforou <pavlo-(a)gaaros.com> wrote: >TM.py defines _registered as a class variable which seems to hold the >state of an instance (whether it is registered with the transaction >manager or not). method _register() registers the instance with the TM >depending on the content of the _registered attribute. this however seems >to have the additional effect to set the instance as changed and therefore >and therefore a new version is stored in the ZODB. Additionally as >_registered is defined as a class variable it is shared among threads and >therefore one thread can interfere with the state of another. I would like to understand the source of your problem, too. However, I think there is a misunderstanding here. In TM.py, _registered is a class data attribute, which is typically a trick for "initializing" Python instance data attribute. That is, unless you explicitly do something like TM._register = 1 the class variable would never change. That is, TM._register is always None. Always always. So there should not be any problem in thread interference. If x is an instance of TM (or an instance that has TM as an interface, in the case of multiple inheritance), if you look at the dictionary of x, i.e: x.__dict__ , you will see that initially x does not have an attribute _register in its own dictionary. However, x._register is a valid reference, because it borrows from the class attribute. The statement: print x._register would print out 'None'. As soon as you assign a value to x._register = 1 you create an attribute in the dictionary of x. That is, a new item is added into x.__dict__ , which is totatally independent of TM._register. So TM._register remains as None. >I have changed all references to _registered in TM.py to _v_registered and >my problem (creating new versions for every call to FSSession) is solved, >but I am not sure what other classes use TM.py. Also I am thinking of >changing _v_registered to an instance attribute instead of a class >attribute. Any thoughts?? Is it a bug of TM.py I'd like to take a look into it, too. But first I'll have to learn what the flags _v_registered and _register mean. :( I don't even understand what is a "transaction" in the sense used by the transaction manager. If I understand correctly, the definition of "transaction" that you would like to use is one single HTTP transaction, which lasts from the beginning of a foreign HTTP client request till the end of Zope sending out the reply. The question is: is this the same definition of "transaction" as used by the transaction manager? Why would anything be stored in ZODB?!? The HTTP "transaction" should not have anything to do with ZODB database "transaction". The "versions" that are stored in ZODB, are they FSSESSION objects? I'd like to look into all this. :) regards, Hung Jung ______________________________________________________ Get Your Private, Free Email at http://www.hotmail.com

2 1

Re: [Zope-dev] Using Zope in a client-server system
by Tres Seaver 03 Mar '00

03 Mar '00

Itai Tavor <itavor(a)bigpond.net.au> wrote: > > Michel Pelletier wrote: > > > Itai Tavor wrote: > <snip> > > > >> There are a couple of features that I am not sure how to implement though: > >> > >> - I need to display constantly-updating information on the > >> workstations. I could do this by reloading the window every second, > >> but that seems ugly - I'd like to only update certain fields, which I > >> can do using layers and JavaScript, but I need to get the updated > >> data from the server. I guess that can be done with Java, but I'd > >> like to keep the system simple - a browser with HTML and JavaScript > >> only on the client side, DTML, Zope with Python Methods, and External > >> Methods, and an SQL database on the server, no Java or any other > >> complications. > > > > Ah, well, you see you've hit one of the disadvantages of using a > > stateless protocol like HTTP to solve your problem. There is no good > > general purpose solution that I know of to address this issue. > > HTTP not being bi-directional is a bigger problem than it being > stateless... I planned to use a session product that will identify > each workstation and keep a record of exactly what is happening on > it. But the biggest problem is that the browser / web server setup is > client event driven. > > >> - I also need to push data to the workstations - based on events > >> detected at the server, I need to bring up a message on a > >> workstation, and remove it in response to another event. I really > >> don't want to refresh a frame every second for this, and I also would > >> hate to have to write a Java applet to do it. > > > > This is the identical problem to the one above. > > > >> Any comments or suggestions regarding the above problems, the choice > >> of Zope for this project, or the project in general would be > >> appreciated. > > > > If you want to go far enough, you can try and use XML-RPC (also, > > stateless, but not client-bound to a certain model) and something like > > wxPython to build your own GUI that can at least mantain some sort of > > persistent or bi-directional communication between your clients and > > servers. But this is a big task. > > Programming my own GUI is already farther than I'd like to go... and > if I do, would it still be a good idea to use Zope on the server > side? If I use Zope, wouldn't I still have to use HTTP? How otherwise > can I get the client to communicate with the server? (Complete > ignorance of XML-RPC implied here.) > > > I think if you maybe rethink your problem you can compromise on > > something that does not require the server pushing data to the client. > > Yeah, without actually failing to deliver any of the features the > client is expecting... the only way I can think of is simply to > reload certain frames every second... seems like a very ugly solution. > > I need to quote on time and cost for this project, and to do that I > have to figure out the tools I'd be using. I'm sure a Zope/web > browser solution will be the quickest, but if I go that way and get > stuck, I won't be in a very enviable position. I'm sure that most > programmers would immediately start coding Java or VB at this > point... am I making a mistake trying to apply Zope to a task it's > not good for? By default, browsers aren't servers: they don't know from bind()/listen()/accept(), which is what push technology requires (unless you have a long-running, bi-directional connection between browser and server, which HTTP doesn't allow). Polling is the only "native" technique for refresh, and suffers from _horrible_ scalability problems. Options in order of simplicity: 1. Pull instead of push (ick!) 2. As Hung Jung suggests, build the "push model" stuff separately from the browser, using Python or Java to implement HTTP (and perhaps XML-RPC?) Integration with the browser is difficult: clients will have to have some mechanism for registering their mini-server with your server; if you want the broser's display to update when the push event comes in, you may be stuck. 3. Code a lightweight socket server applet in Java, persuade everyone in the world to trust you to run it in their "applet sandbox", and then "register" it with your server. 4. Write a CORBA callback object in Java, and register it with your server; the server then pushes events to your object (scalability is again an issue, as the server blocks for each client being pushed). Note that adding CORBA into Zope is a non-trivial exercise, for the moment at least; see my "notes":http://www.zope.org/Members/tseaver/CosNaming on the issues. 5. Write a CORBA event-channel applet (CosEventPushConsumer) in Java, set up a CosEvent/CosNotification channel, subscribe the applet to it, and push events to the channel from your server (decouples your server from event-delivery hassles/blockage, scales MUCH better than direct callbacks). -- ========================================================= Tres Seaver tseaver(a)palladion.com 713-523-6582 Palladion Software http://www.palladion.com

1 0

Fwd: Multi-argument append() is illegal
by Patrick Phalen 03 Mar '00

03 Mar '00

Fredrik Lundh has mentioned that this breaks Zope (and Medusa). If so, it's probably not too soon to begin looking into it. 1.6 is due this summer. ---------- ## Forwarded Message ## ---------- Subject: Multi-argument append() is illegal Date: Mon, 28 Feb 2000 10:41:02 -0500 From: Guido van Rossum <guido(a)python.org> I've noticed that there is some code out there that creates a list of tuples and uses code like list.append(a,b,c) to add the tuple (a,b,c) to the list. According to the documentation, this is illegal: append() only takes a single argument, and one should write list.append((a,b,c)). However, the actual append() implementation didn't mind, and implemented list.append(a,b,c) as list.append((a,b,c)). Many people are using this even though it's never been documented. I am going to rectify this in Python 1.6 -- people coming from other languages might well expect list.append(a, b, c) to mean the same as list.append(a); list.append(b); list.append(c), and it's always been my philosophy to make ambiguous syntax illegal rather than to pick one interpretation randomly. This message is simply a heads-up that you should be aware of this change (when 1.6 comes out, which should be before the summer). You can test your programs using the current CVS version (see www.python.org/download/cvs.html) You can also grep through your sources for a pattern like "\. *append *\(.*," -- which doesn't find every occurrence, but is a good starting point. If you have a smarter grep-like tool you may be able to write a tighter matching expression. Watch out for false hits though: some classes define their own multi-argument append()... --Guido van Rossum (home page: http://www.python.org/~guido/) -- http://www.python.org/mailman/listinfo/python-list -------------------------------------------------------

2 4

Using Zope in a client-server system
by Itai Tavor 03 Mar '00

03 Mar '00

Hi, I need to design sales center system containing a server and salesperson workstations. I thought it would be a good idea to build the system around Zope, for several reasons: - With Zope on the server and a standard web browser on the workstations, a lot of low level work is already done for me and I can focus on high level issues such as UI and application logic. - Easy connectivity to an SQL database. - DTML code chunks and Python Methods should be easier to manage than 1000's of lines of Java or VB code. - Easy to install and update the product and to manage and debug it through the web. - It would be easy to create new ways to view the data - for example, a WAP version to send sales statistics to managers out of the office. There are a couple of features that I am not sure how to implement though: - I need to display constantly-updating information on the workstations. I could do this by reloading the window every second, but that seems ugly - I'd like to only update certain fields, which I can do using layers and JavaScript, but I need to get the updated data from the server. I guess that can be done with Java, but I'd like to keep the system simple - a browser with HTML and JavaScript only on the client side, DTML, Zope with Python Methods, and External Methods, and an SQL database on the server, no Java or any other complications. - I also need to push data to the workstations - based on events detected at the server, I need to bring up a message on a workstation, and remove it in response to another event. I really don't want to refresh a frame every second for this, and I also would hate to have to write a Java applet to do it. Any comments or suggestions regarding the above problems, the choice of Zope for this project, or the project in general would be appreciated. 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 --

3 3