September 2001 - Zope

Re: [Zope] when to opt for a backend DB over ZClasses?
by Steve Spicklemire 04 Sep '01

04 Sep '01

Hi Darran, I'm cc-ing the list so other folks who look at the example might use the help. On Tuesday, September 4, 2001, at 03:25 AM, Darran Edmundson wrote: > > Hi Steve, > > Thanks for the ZPatterns tip. I've downloaded the > attendance example and, while I don't understand how > it works, I like the end result ;-) I've tried > unsuccessfully to get my head around ZPatterns on a > number of occassions, inevitably falling back on > a more familiar solution. > Well.. it's not so hard once you get the hang of it. ;-) > Any tips on coming to grips with this paradigm in > the context of your example? I'll give it a whirl! Basically, each object type gets a Specialist to manage instances of that object type (e.g., objects of type "Attendee" are managed by the "Attendees" Specialist). The Specialist delegates all responsibility for the actual storage of it's objects to a "Rack" that it contains. The objects themselves (e.g., Attendees) are sub-classed from ZPatterns:DataSkin, and are designed to further delegate all responsibility for the actual storage of their *attributes* to the Rack in which they are stored. When an object is used in an application, and the application needs a certain attribute (e.g., name) the object checks with it's rack for an "attribute provider" for that attribute. If it finds one it uses it to return the attribute to the object that needed it. If it doesn't find one, it returns NameError. > Where are your attendance > zclasses being created? The ZClasses live in the /Control_Panel/Products/AttendanceProducts folders. The ZClass *instances* are created when someone calls any of the applications' Specialists "newItem(id=...)" methods. newItem only has one argument, "id". When newItem is called, an instance of the corresponding object is created and returned to the caller. This action also fires any "WHEN OBJECT CREATED" triggers in the Rack of the Specialist (since the Specialist just forwards newItem to it's rack) > And are they persistant on > the server across multiple requests? Yes. they are persistent until someone deletes them! DataSkins have a manage_delete() method that also fire any "WHEN OBJECT DELETED" triggers in the Rack of the corresponding Specialist. The ideas of Triggers, and attribute providers are really important. Here is the SkinScript from a typical Rack (found in the Data PlugIns tab) that uses SQL: The first three are "Triggers", actions that occur when stuff happens to the object: WHEN OBJECT ADDED CALL insertSQL(mergedsectionID=self.id, targetCourseID = self.targetCourseID, section = self.section) WHEN OBJECT DELETED CALL deleteSQL(mergedsectionID=self.id) WHEN OBJECT CHANGED STORE targetCourseID, section USING updateSQL( mergedsectionID=self.id, targetCourseID = self.targetCourseID, section = self.section) Where insertSQL, updateSQL, and deleteSQL are ZSQLMethods that perform those actions. The fourth is an attribute provider, (i.e., how to I get the attributes of one of these objects when they are accessed?): WITH QUERY selectSQL(mergedsectionID=self.id) COMPUTE mergedsectionID, targetCourseID, section > I'd love to > have someone explain this to me in Socratic fashion ;-) Sounds like a novel approach. ;-) The idea is that the application uses a really simple API: newItem() to create new objects getItem() it get existing objects and the object's own manage_delete() method to destroy objects. In my own applications I use a method called getItemIds(conditions = {} ) returns a list of ids that satisfy a set of conditions passed as a dictionary. This method can call a "Select ... " query or query a catalog depending on how the objects are *really* stored. But this keeps all the data dependent stuff contained in a small number of places: The Rack, getItemsIds, and then the actual queries, db adaptor *or* a Catalog (*or* LDAP, *or* whatever else.. ). The thing I love about ZPatterns is the degree to which you can write an application *whithout* worrying about the data, but focus on the logic *first*. Most of my apps are generated first using ZODB, then, if necessary, converted to SQL using the "ZClass lever" thingy included in the example. It's nice since it generates all the SQL and skinscript for you based on the propertysheets in the ZClasses that represent the "definition" of your data. Of courses, the one that comes with the example is set up for gadfly, but "salt to taste" and all that. thanks for the questions! -steve > Cheers, > Darran. >

1 0

ZODBC connection in NT
by James Johnson 04 Sep '01

04 Sep '01

Seems to me there was a post a while back about some sort of problem with an earlier version. What versions are you using? You might want to do a search of the list. I've been using mysql lately. Good luck. _________________________________________________________________ Get your FREE download of MSN Explorer at http://explorer.msn.com/intl.asp

1 0

ZPT and DreamWeaver
by alex＠quad.com.ar 04 Sep '01

04 Sep '01

I've been reading about Zope Page Templates and it looks like its targeted to DreamWeaver and GoLive (as Evan said) now, how do DreamWeaver users interact with zope? supposing you are using latest dreamweaver (4.0 + 4.0.1 patches), be it either WebDAV or FTP there's no way dreamweaver will open the extension-less zope files in its editor, unless you drag-n-drop it, and in that case, dreamweaver will refuse to show the 'design' view (rendered hml). Now, how can ZPT be targetted to an IDE that won't directly interact with zope in the first place, I just don't get it. As far as I know, Adobe has received quite a lot of complains about this behaviour (ignoring the Content-Type and just guess the file type based on the its extension name), and there is still no good signs on this changing in future releases. What do you think it's the best solution to use DW4.0.1 on Zope (for editing ZPT) ? are there any patches to zope, that will append '.html' on webdav/ftp when the user agent is dreamweaver or anything else ? I've been reading some threads on the mailing list archive, but I couldn't find any solution yet.. any suggestions would be great.. thanks in advance Alex Verstraeten (aka zxc at #zope)

4 5

RE: [Zope] Advice needed: load balancing wih ZEO and Apache on So laris.
by sean.upton＠uniontrib.com 04 Sep '01

04 Sep '01

Sounds like you need something to distribute and/or balance the load; 3 suggestions: 1 - HTTP Proxy Load Distribution - Squid does this, round-robin, put squid in front of your ZEO farm (this is what I do). I don't think that mod_proxy is capable of doing this. 2 - Layer 4 switch (Cisco LocalDirector, Intel Netstructure, etc) - not cheap, but offers some features like a choice between true balancing, outage detection, and funky routing techniques like out-of-path return for quicker network performance. 3 - LVS - Linux Virtual Server Project, attempts to do in software what # 2 does. Personally, I'm in favor of squid because it is cheap and easy, if all your machines are of equal weight in performance terms. My company uses squid in front of multiple ZEO clients, bypassing Apache altogether for access to Zope; we use pyredir as a redirector to rewrite URLs like you would with mod_rewrite; Squid is very nice as a Zope virtual host front-end, and provides ACLs for security purposes, like blocking out the ZMI from public access. Getting load balancing working in squid is simply a matter of compiling it with a flag to specify you want external DNS resolution support and using squid's built-in name resolving program dnsserver, which will look at /etc/hosts and round-robin among multiple IPs that have the same hostname. One caveat: if you put Squid (or any high-volume TCP app) on Solaris, make sure to tune your TCP connection buffers, so Solaris doesn't choke; this is one (albeit trivial) reason we use Linux 2.4 for our proxies instead of Solaris. Since squid round-robin's it doesn't deal with node reliability problems, which means you will have to rely upon another HA mechanism on the ZEO client nodes themselves, like heartbeat from the Linux-HA project, which at some point will ported to Solaris, from what I hear. If you have a decent budget, but not a lot of time, look at a L4 switch (we use for another application, and like, the Intel Netstrcuture 7140); otherwise, if you have budget constraints consider LVS or Squid, with Squid being the likely easiest path in terms of setup time. Sean -----Original Message----- From: Tony McDonald [mailto:tony.mcdonald@ncl.ac.uk] Sent: Saturday, August 25, 2001 9:14 AM To: Zope Subject: [Zope] Advice needed: load balancing wih ZEO and Apache on Solaris. Hi, I've followed the instructions at <http://www.zope.org/Members/dshaw/AdvancedSiteSetup> to set up a ZEO server and client (on the same machine). These instructions are very clear and I'd recommend them to people wanting to experiment with ZEO. We serve our content up through Apache (fast delivery of static content, CGI scripts and PHP4 served too). I'm using the VHM method of routing Apache requests through to my ZServer installation so (only a snippet shown, and this is an example); RewriteEngine on RewriteRule ^/cgi-bin - [L] RewriteRule ^/static - [L] RewriteRule ^/ltsn_images - [L] RewriteRule ^/(.*) http://localhost:18080/VirtualHostBase/http/myserver.ncl.ac.uk:80/VirtualHos tRoot/$1 [P] Following the instructions from Dave Shaws' page I have a ZEO server running on port 8080 and a client running on port 8081 One idea I've had from reading the Apache Rewrite rules page at <http://httpd.apache.org/docs/misc/rewriteguide.html> Is to use ProxyPassReverse (in some way), i.e. ProxyPassReverse / http://localhost:8080/ ProxyPassReverse / http://localhost:8081/ My question is this; What can I do to get requests from myserver.ncl.ac.uk farmed out to my ZEO server-client farm? Any pointers would be really appreciated. Cheers Tone. -- Dr Tony McDonald, Assistant Director, FMCC, http://www.fmcc.org.uk/ The Medical School, Newcastle University Tel: +44 191 243 6140 A Zope list for UK HE/FE http://www.fmcc.org.uk/mailman/listinfo/zope _______________________________________________ Zope maillist - Zope(a)zope.org http://lists.zope.org/mailman/listinfo/zope ** No cross posts or HTML encoding! ** (Related lists - http://lists.zope.org/mailman/listinfo/zope-announce http://lists.zope.org/mailman/listinfo/zope-dev )

3 5

Disabiling tracebacks
by Serge Wroclawski 03 Sep '01

03 Sep '01

Hello, I'm sorry for such a newbiesque question, but I have not seen an answer to my question in the documentation that I read. I have a Zope server installed (by means of Debian Woody). It works, but when I purposefully fail to authenticate myself, I see a page which tells me there's been an error. If I examine the HTML source of that page, I see a Python traceback of the error. I would like to disable the Python tracebacks when I would use this software in production. Since this is a pretty obvious security issue, I'm sure that there's a simple solution, but I've yet to see it. It's been suggested that I'm running in Debug mode, but I've taken the -D out of the line which runs Zope, and re-run the daemon an the traceback persists. Thank you, - Serge Wroclawski

1 0

upload file and edit it in place?
by Milos Prudek 03 Sep '01

03 Sep '01

I need to allow users upload a text file, and edit it. Like this: <form action="&dtml-URL0;" method="post" enctype="multipart/form-data"> <textarea name="file" cols="50" rows="8"> <dtml-if file> <dtml-var file> </dtml-if> </textarea><br> <input type="file" name="file" size="25" value=""><br> <input type="submit" name="submit" size="25" value="Go!"> </form> But it does not work as desired. Although I use enctype="multipart/form-data", it still displays "[' ', <ZPublisher.HTTPRequest.FileUpload instance at 873b710>]" within textarea. How can I make it display the contents of uploaded data? -- Milos Prudek

2 1

Re: postgres logging Re: [Zope] GUF & SQL login problems
by Lee Harr 03 Sep '01

03 Sep '01

>Okay, I changed postmaster.opts to include d=2, and I verified using >"pg_ctl status" that it applies. Now where is the log file? The queries >do not go to syslog... > >There is a file that looks like a log: /var/lib/pgsql/data/pg_log. It is >growing, but when I start psql and do "SELECT * from pg_log", I get >"ERROR: pg_log cannot be accessed by users"... > Actually, I use the -l switch to pg_ctl to specify the location of the log file. As long as the file you are talking about is not actually pg_xlog that might be it. You might try more /var/lib/pgsql/data/pg_log Any file in that directory will not be a database or table. Or look in the .../pgsql directory and see if it is not there. It looks like there is also a way in postgresql.conf to send logging to syslog. >From the postmaster man page, it says that postmaster runs in the foreground by default and prints log messages to stdout. So the easiest thing to do may be just to start it by hand in the foreground and watch the log on the console, or redirect it to a file using postmaster [other switches here] -o -d 2 > logfile _________________________________________________________________ Get your FREE download of MSN Explorer at http://explorer.msn.com/intl.asp

1 0

Re: [Zope] ZSQL method and database connection acquisition.
by Dieter Maurer 03 Sep '01

03 Sep '01

jon(a)totient.demon.co.uk writes: > any chance you could be more explicit. You read the ZClass chapter in the Zope book? > This has confused me enough already > > it's "Use Database Methods" permission? > > do you add this on the Product Folder, the Zclass thing or the > instance? or on the zsqlmethod. The ZClass allows you to remap the permissions protecting the ZClass' methods. Say, such a method is protected by "Use database methods". But the ZClass uses the method to update a shopping card and wants the permission to be "Update shopping card". It can use a permission mapping to rebind the "Use database methods" as necessary. The context of the ZInstance can then associate roles with the mapping target as required. > should it be set to 'view' or 'use database methods'? Whatever is more appropriate for the use of your Z SQL method.... E.g. when the SQL is just a search (no modification) quite restrictive (can read only non-sensible information), then "View" may be sufficient. If, however, it can modify data, "Use database methods" would be more appropriate. > does the open database connections come into use at all? Only, when you want to test your Z SQL methods inside the ZClass. > this is driving me crazy. it should be so simple. It is not too difficult. Look at the chapter on ZClasses in the Zope Book. When I last read it, it was quite understandable. > (I've never liked the zope permission thing... and the language is terrible > 'permission that requires this permission') Seems to be a documentation bug: means "operation requires a permission..." or "permission is mapped to permission..." Dieter

1 0

RE: [Zope] how to capture the referring URL??
by Oliver Bleutgen 03 Sep '01

03 Sep '01

>> Don't use referring url ;-) >> Seriously, it depends of your problem, but it may be solved by using >> some kind of user sessions... >> >> If you could explain more your problem, maybe someone could find another >> solution? > Here is a shot at what I am working on. > This is an "Email to a Friend" thing. > Thanks to anyone that can get me in the right direction. And explain why > I > shouldnt use HTTP_REFERER. Using http_referer for more than statistcal reasons is a generally problem because of two reasons: 1. security-wise: Never trust the client! The browser constructs the referer information and it can send for that whatever it wants to. So it makes no sense to try to prevent malicious users from doing something based on the referer header of their browser. Sites like geocities etc. try to hinder direct linking to binary files on their site by requiring a local (for them) referer. This is why various download utilities allow to set a custom referer header. 2. it may not function Apart from security considerations, which might not be important in your case, various proxies strip referer information for privacy reason, so that they don't show up in logfiles etc.. So point 2 seems more important in your case. But you seem to be in control of the page which link to your "Email to friend" form. If these pages are also in zope, you could easily pass the "referer" information to that form, i.e. <a href="link_to_email_to_friend_form?url=<dtml-var URL0 url_quote>">email this page to a friend</a> To sum up, using the referer header is not adviced for security uses only, for others it's just not optimal, IMO. cheers, oliver

1 0

how to capture the referring URL??
by Trevor Toenjes 03 Sep '01

03 Sep '01

another newbie Q... I am trying to figure out how to capture the PREVIOUS URL in a variable. It looks like I need to do something with PARENTS and absolute_url. What is the proper syntax? Thanks, -Trevor Toenjes

4 6