September 1999 - Zope

RE: [Zope] Zope Performance and Large Databases
by Jay, Dylan 08 Sep '99

08 Sep '99

> -----Original Message----- > From: Michel Pelletier [mailto:michel@digicool.com] > Sent: Wednesday, September 08, 1999 06:48 > To: 'davidbro(a)namshub.org'; zope(a)zope.org > Subject: RE: [Zope] Zope Performance and Large Databases > > > > -----Original Message----- > > From: davidbro(a)namshub.org [mailto:davidbro@namshub.org] > > Sent: Tuesday, September 07, 1999 3:24 PM > > To: zope(a)zope.org > > Subject: [Zope] Zope Performance and Large Databases > > > > > But being that I have several gigs of MP3's (don't start about > > copyright stuff, these are ripped from my CD collection), we're > > talking about a large number of ZClass instances (one per > song) stored > > in the ODB. > > Be aware that this will make your Zope object database massive and > increse your memory requirements. Perhaps it would be simpler to just > serve the actualy MP3s out of a static Apache and let Zope just handle > the 'meta' information? If your running Linux, don't forget > that ext2fs > has a 2GB file size limit. I have a jukebox program written in zope that just uses an external method to scan drives for albums and load this data into a RDB. Zope does all the reports and playing an album is just a matter of redirecting to a file: url. This works very well.

1 0

Re: [Zope] The Honest Scoop on Zope
by Anthony Baxter 08 Sep '99

08 Sep '99

>>> Tim Wilson wrote > I guess VNC is a possibility, but it would probably be slower then X over > SSH and you'd want to forward that through SSH anyway. Good luck. vnc-over-ssh is probably faster than X11 over ssh, particularly if you use the -hextile option on the client. vnc's generally better in lower bandwidth situations. you could also try dxpc-over-ssh - dxpc does a bunch of caching and X11-specific compression, rather than ssh's generic compression.

1 0

Users in Ohio - surrounding states
by Ceska, Christopher 08 Sep '99

08 Sep '99

I was wondering if there would be enough interest to get a users group going. I have access to seminar facilities in the Cleveland, Ohio area.. Would anyone be interested in this? -chris

1 0

Re: [Zope] The Honest Scoop on Zope
by CHUNG-SHU YANG 08 Sep '99

08 Sep '99

I built a version of ZODBCDA3.0.3 for Linux with little modification in sql.c and used it to connect to DBMaker on NT. No problem so far. If you can find a Access ODBC driver for Linux, all you need to do should be removing the line '#include <windows.h>' in sql.c and creating a makefile. CS -----Original Message----- From: jonathon <jblake(a)stamp-coin.com> To: Phil Harris <phil(a)philh.org> Cc: zope(a)zope.org <zope(a)zope.org> Date: Wednesday, September 08, 1999 11:13 AM Subject: Re: [Zope] The Honest Scoop on Zope >On Mon, 6 Sep 1999, Phil Harris wrote: > >>link for others (note that the ODBC adapter has the possibility of running >>not ponly on Win32 but also on Unixen) > > Does that mean that there is a possibility of Zope on > a Linux platform being able to use an ACESS DATABASE > << or whatever that database program for an office suite > on a platform that wants to grow up to be an operating > system is called. >> on that non-linux OS, without using > their tools that don't work as advertized? > > xan > > jonathon > > >_______________________________________________ >Zope maillist - Zope(a)zope.org >http://www.zope.org/mailman/listinfo/zope > >(To receive general Zope announcements, see: >http://www.zope.org/mailman/listinfo/zope-announce > >For developer-specific issues, zope-dev(a)zope.org - >http://www.zope.org/mailman/listinfo/zope-dev ) >

1 0

Re: [Zope] The Honest Scoop on Zope
by Brian Salisbury 08 Sep '99

08 Sep '99

Thanks for your reply Kevin. You answered most of the other questions that didn't get answered before. The advice is helpful. As far as hosting, I've already got a dedicated server in California. So I'd rather try to make this work on it before paying someone more money to host a site of mine. But my problem is, with the dedicated option, I don't have root access. I can choose to have it co-located, which would mean I would have root access. But that means I would be managing the server myself without any assistance from them. So now I'm learning Linux and networking et al so I may be able to manage the server myself remotely. A couple more questions: Does Zope help with the organization of "html" files? Is it easier to know where stuff is, or find if you don't? Off-topic Linux question: Is there anyway to remotely administer my server with an X Window interface like KDE? Thanks again, Brian Kevin Dangoor wrote: > Hi, Brian > > I'm not sure I saw responses to everything, so I thought I'd just throw > in my couple of cents... > > -----Original Message----- > From: Brian Salisbury <brian(a)hilarious.com> > To: zope(a)zope.org <zope(a)zope.org> > Date: Sunday, September 05, 1999 11:16 PM > Subject: [Zope] The Honest Scoop on Zope > > >Is Zope scalable? Will Zope be able to support me if my site suddenly > >becomes popular and I've got to go out buy one of those huge web server > >machines? > > While this is something I have not personally run into yet, I can tell you > what people have reported in earlier messages on the list. A decent box (not > even top of the line) running linux can do 1 million relatively complex hits > per day. Zope is a good performer out of the box. Assuming you're running > one of the most popular sites on the net and you're getting more than 1M > hits per day, Digital Creations has announced something called "ZEO". This > allows you to distribute synchronized Zope DBs among multiple machines. > *That* should handle just about any performance requirement. > > >Is Zope reliable? I come from windows world, so anything that runs more > >than 30% of the time generally please me; but, what's serving my web > >documents and collaboration system has to be extremely reliable. > > My experience has been that it is. My experience has also been that if there > is any kind of reliability problem, the DC guys are there to help (free of > charge), because they want this to be a very, very stable platform. On my > site, Zope has run continuously and has only been restarted when I needed to > restart it for something. In other words, I've had no problems with it going > down. > > >Is it easy to learn? I have virtually *no* programming experience. > >Everything I've done to this point has consisted of guess-editing > >pre-cooked Perl CGI scripts. I don't know where to begin learning Zope > >and, obviously Python, since it goes with Zope. Do you remember what it > >was like to think of having to 'Program' and getting the shivers? > >That's me. I've never even looked at source code. I think I compiled > >some stuff once ;). Most of my web publishing, to this point, has > >consisted of serving mostly static web files with SSI, and CGI files for > >interactive stuff. > > Right now, Zope is not amazingly easy to learn. It has a lot of deep and > powerful concepts, and the documentation is in a state of flux because of > the big switch from Zope 1.x to Zope 2. That said, I think the learning > curve is worth working through because the platform does so much. > > Zope has a really helpful user community and quite a few very active > developers. You will start seeing a lot more prepackaged solutions. Though > I'm fairly familiar and comfortable with Python, I don't think you even need > to use python for the *vast* majority of applications because of new > capabilities in Zope 2. > > >My knowledge is getting fuzzy here. Does Zope directly interact with > >things such as a Database? I wouldn't want it having to call a process > >every time a message board action is requested. > > Zope *can* interface with an external (ie not part of Zope) database such as > Mysql, if you want it to. There is no reason it has to. (I don't use an > external database). Zope works as a long-running process, as does a database > server, so there is generally not a lot of overhead in using a database if > it makes sense for your application. > <snip>........................................</snip> > >I'd really appreciate any insight you have. If I may say so, it really > >impresses me what so many folks collaborating together can accomplish. > >You really are defining the future. > > I give a lot of credit to DC. Those guys are very sharp and have made an > excellent product. This has really encouraged all of the subsequent > development. > > Kevin

2 1

ZClasses question
by Doug McNaught 08 Sep '99

08 Sep '99

OK, I've read the ZClasses tutorial, and done some playing around. I have a test class i'm trying to create, but I'm not sure if what I'm trying to do is possible. The idea is to have a "Random Image" object. It acts like a Folder in that you can drop images into it, but when rendered it picks a random object from its contents and returns the HTML for it. So if I have an RI called "foo", I can put images into it via cut/paste, upload, etc (just like a folder) but if I refer to <dtml-var foo> in a document or method, I get the HREF for a random image in the RI. I can do this with a regular folder by creating an index_html (which could just as well be called 'pluck_random_item') in the image folder containing <dtml-var expr="_.whrandom.choice(objectValues())"> and then using the following in my documents: <dtml-with imagefolder> <dtml-var index_html> </dtml-with> but I would really like to have the cool object behavior described above. I've been thrashing about trying different base classes and suchlike but haven't found anything that works. I though that creating an 'index_html' (which is basically a 'render' method) in a folder would have that method called automatically upon <dtml-var foldername> but it looks like that's not the case. Is what I want possible? If so, what kind of ZClass do I need to make? -Doug -- Doug McNaught doug(a)mcnaught.org http://www.mcnaught.org/~doug

1 0

Zope+Apache config
by Tim Wilson 08 Sep '99

08 Sep '99

Hi everyone, I'm configuring Zope to work with Apache. A quick question: I compiled Zope 2.0 with # python w_pcgi.py Then I started up Zope by typing #./start -p '' & No problems. My site was available at http://foo.bar:8080/ I followed the directions in the Zope and Apache howto and added some lines to my Apache config file. I made sure the final Rewrite rule pointed to the Zope.cgi file in my Zope directory. After restarting Apache, my non-Zope pages are still available just as they should be. Unfortunately, http://foo.bar/atlas/ (which, of course, refers to a Zope-generated page) returns an error which says: Forbidden You don't have permission to access /atlas/ on this server. Any ideas? If I understand this correctly, the Zope.cgi file should have all of the information necessary to access my Zope pages, e.g., ZServer is running on port 8080. -Tim -- Timothy Wilson | "The faster you | Check out: Henry Sibley H.S. | go, the shorter | http://slashdot.org/ W. St. Paul, MN, USA | you are." | http://linux.com/ wilson(a)chem.umn.edu | -Einstein | http://www.mn-linux.org/

5 9

The Honest Scoop on Zope
by Brian Salisbury 08 Sep '99

08 Sep '99

Hello List, I've been kicking around the idea behind Zope for quite a while. There are a some questions I have about it. These are things I consider important to publishing on the web. Is Zope scalable? Will Zope be able to support me if my site suddenly becomes popular and I've got to go out buy one of those huge web server machines? Is Zope reliable? I come from windows world, so anything that runs more than 30% of the time generally please me; but, what's serving my web documents and collaboration system has to be extremely reliable. Is it easy to learn? I have virtually *no* programming experience. Everything I've done to this point has consisted of guess-editing pre-cooked Perl CGI scripts. I don't know where to begin learning Zope and, obviously Python, since it goes with Zope. Do you remember what it was like to think of having to 'Program' and getting the shivers? That's me. I've never even looked at source code. I think I compiled some stuff once ;). Most of my web publishing, to this point, has consisted of serving mostly static web files with SSI, and CGI files for interactive stuff. My knowledge is getting fuzzy here. Does Zope directly interact with things such as a Database? I wouldn't want it having to call a process every time a message board action is requested. How endless are the possibilities? In the Web Server/RDBMS world, one just has to program a script to call the DB to do whatever they need it to do. I wouldn't want to have a great idea for my new site, then find out Zope, which I invested all of my time in money in, isn't able to accomplish x-task. Is everything in Zope stored in a (I'm guessing) Object Database? I installed Zope on Windows 98 to poke around. (I tried to install it on my dedicated server, of which I don't have root access, with no success). It doesn't look like anything is stored in a physical file system. I see two potential problems with this. First, I would think it hard to transfer an existing site to Zope. Plus, wouldn't it be very difficult to convert your Zope site to X-Platform should I have the need to at some point in the future? I don't want to paint myself into a corner. Let say I wanted to have someone manage the technical aspects of my Server(s). How hard it would be to have someone maintain Linux, its connection, and Zope? Could I have someone I could call on to write in custom scripts to accomplish what I need in Zope? What kind of expenses would something like this incur in a Linux OS environment? I would like to know how everything works, but I'm afraid I can't do it all by myself. On the same token, I'm not a Fortune 500 company who has money burn. These are generally my biggest concerns. Right now, I'm trying learn to Linux, Networking, and how I can more efficiently run a web site. I'm considering buying a PC that strictly runs Linux so I can have it there full time for experimental/learing purposes. Now pile on top of this the facts that I have returned to school with a full-time schedule, and I'm still maintaining my current site. That means I don't have hours everyday to play around with Linux and Zope. I'd really appreciate any insight you have. If I may say so, it really impresses me what so many folks collaborating together can accomplish. You really are defining the future. Thanks, Brian Salisbury

3 2

Re: [Zope] Editing DTML Methods with Emacs ?
by Doug McNaught 08 Sep '99

08 Sep '99

Michel Pelletier <michel(a)digicool.com> writes: > The problem is XEmacs. It first deletes the file and saves a backup > '~', and then an original. Because it deletes the origninal object > instead of overwriting it, Zope assumes that the *new* file is to be a > document. I bet a gazzillion dollars that there is some lisp variable > somewhere in EFS to disable this behavior. The variable is 'efs-make-backup-files', which is a list of host types (efs has a concept of host type and will do different things when talking to VMS, say, than when talking to Unix) on which to make backup files when saving. If you set it to 'nil' efs will never make a backup file on a remote machine. If you want to make backup files on 'real' ftp servers but not on Zope's, this could be a problem, as 'efs-process-host-type' is "unix" when connected to a Zope server, just as it is with normal systems. I don't know whether there might be a way to make Zope recognizable to efs. I did do a test with setting 'efs-make-backup-files' to nil and doing a few efs edits through localhost (standard ftp server). The inode number of the file in question did not change, indicating that efs just did a 'put' when saving the file, rather than deleting and recreating. Thus from Michel's comments, Zope should do the Right Thing and preserve the document type if you nil out the above variable. Hope this helps. -Doug -- Doug McNaught doug(a)mcnaught.org http://www.mcnaught.org/~doug

1 0

CRM and Zope
by Bjorn Stabell 08 Sep '99

08 Sep '99

Hi there, We're currently evaluating customer relationship management (CRM) solutions for future projects. The main contenders are BroadVision and Vignette so far, but it's hard to navigate the forest of CRM solutions. Looking at Zope I'm amazed at the flexibility already existing in the product, and knowing the power of open source and a committed community of developers often associated with beautiful systems like Zope (software is like art and so it can be beautiful), it's only going to get better. I am, however, concerned with Zope's maturity and comprehensiveness compared to current state-of-the-art commercial solutions. I'd like to hear other people's opinions as well. In particular I'd like to know about: 1) Work-flow. Zope doesn't seem to have any inherent work-flow capabilities, so I guess that'll have to be developed? Is anyone working on tools and toolkits for this? 2) Personalization. Again, this can undoubtably be developed in Zope, but is anyone working on toolkits to lift the personalization features to a higher level? Something similar to Vignette's self-learning ability (and collaborative filtering like NetPerceptions), and having a business layer that allows business types to do targeting is also useful. 3) Content management. With big websites you need content management systems that'll help you keep the content consistent and up-to-date. Again, this will have to be developed in Zope. 4) Scalability. How scalable is Zope when it comes to handling many requests, and how does it perform when the database grows? Are there any inherent limitations? 5) Replication, distribution. Can the Zope database be replicated and the copies kept synchronized? If not, does anyone have an idea how much work it'll be to develop this in Zope (it should be possible to do this at a pretty low level, transparently, just above the object database layer). The good news is that Zope is already so flexible that I think it won't take much work, relatively speaking, to develop modules that'll lift Zope up to the same level as commercial solutions. Any efforts going on already? Bye, -- Bjorn Stabell <bjorn.stabell(a)ogilvy.com> CTO OgilvyInteractive, Beijing

2 1