February 2000 - Zope

Re: Blocking downlaod of files
by Juri Jensen 26 Feb '00

26 Feb '00

Hi, >> I'm asking because I have a site with PDF-files, which I want the >>users to read through their webbrowser (via the PDF-plugin), >>but not to download. > >I'm going to guess that you're 'framing' the PDF with advertising in >some way, and don't want to lose the advertising revenue? A more user >friendly approach would be to include a low key sponsorship within the >PDF itself. > No, actually it's for a Intranet product, where I have to keep the PDF-files as safe as possible. I know I can't keep people from snopping in their cache, but if I can disable saving and printing in the PDF-file, and keep users from downloading the file directly, Im closer to what I need. but maybe it's getting too complicated for something that's not completely secure....? >HTH, > >Michael Bernstein. -- mvh./Best regards Juri Jensen ----------------------------------- Pingvin Partner - Linux løsninger email: juri(a)jones.dk -----------------------------------

1 0

Re: [Zope] Blocking download of files...
by J C Lawrence 26 Feb '00

26 Feb '00

On Sat, 26 Feb 2000 00:06:44 +0100 Juri Jensen <juri(a)jones.dk> wrote: > Hi, I'm wondering if it's possible to block the abillity to > download objects of type file (e.g. right-clicking on a file and > choose "Save Link as...")? No. Pragmatically, there is no difference between the two. > I'm asking because I have a site with PDF-files, which I want the > users to read through their webbrowser (via the PDF-plugin), but > not to download. Please realise that many users don't have a PDF viewer plug-in, and that a great many of us that do, have disabled it in favour of an external helper application. Even better, it is triviality itself to write a simple shell script that pretends to be a "helper application" that does nothing but save its data stream to a file, and thus acts identically to the "Save to file" method. -- J C Lawrence Home: claw(a)kanga.nu ----------(*) Other: coder(a)kanga.nu --=| A man is as sane as he is dangerous to his environment |=--

1 0

Assessment?
by Dave Goodrich 26 Feb '00

26 Feb '00

I'm asking the list cause we have a project to build a web app (in house) that has grown in scope, quadrupled in size. I've seen many references to SQL on this list so I just _knew_ this would be the place to ask. Currently we are looking into using PHP, WebObjects, or JSP. I'd like to get Zope in the running. PHP can handle it and I like PHP, WebObjects seems to missing available information so I don't know much. JSP, well I don't think the parties pushing for JSP can even tell me if they want to use something like jserver, JavaBeans, or what not. When I asked them if they considered Enhydra, they knew not of what I spoke... I believe Zope can do what we need. We plan to use the server to serve up approx 200 pages, pull downs and trees will be populated with query results, pages will be personalized, all the general web app stuff. The server must be able to support upto 40k requests per day. All content will come from the SQL server (we are looking at Sybase and Oracle I'm open to suggestions). SQL will be the heavy lifter using compiled in functions for queriying tables with well over 100k records. I like the SQL connectors in Zope, they are easy to use and customize. Basicly Zope must just maintain sessions, render charts from SQL results, and populate the page with data on request. My question is this, has anyone built a web app in Zope this large? Will you share your experience with

2 1

RE: [Zope] Re: export Z_DEBUG_MODE=no -- broken?
by Martijn Pieters 26 Feb '00

26 Feb '00

From: Martijn Pieters [mailto:mj@digicool.com] > Indeed, this is caused by the fact you closed STDOUT and STDERR on the > process. > > In stead of the -D, you can start Zope up with > > z2.py [options] Z_DEBUG_MODE=1 > > where [options] is your current list of options used minus > the -D flag. > Zope then will behave demonic again (run nohup) and you'll > only miss out > on extra log messages from ZServer. Harumph. I should've checked better before sending. ZDaemon will _not_ fork when it encounters the Z_DEBUG_MODE flag set. However, setting the BOBO_DEBUG_MODE flag to 1 will work. Internally, BOBO_DEBUG_MODE is the old name for Z_DEBUG_MODE, and it still works as well. However, the newer fork and watchdog daemon will not look for it. Also, the log will not be echoed to STDERR (which it would in all cases with Z_DEBUG_MODE set). Side effects of setting the (BOBO|Z)_DEBUG_MODE flag (that I could find): - When trying to access methods starting with an _ (underscore) or that have no docstring, a more descriptive error message instead of Not Found is returned. - All other Not Found errors are replaced with a debug error with the message "Cannot locate object at: [URL]" - Tracebacks in error messages will be wrapped in <PRE></PRE>, not  - The Unauthorized error will include information on whether the supplied credentials where incorrect or that there were no credentials supplied. - HTMLFile objects (used in Python products) includes the filename from which they were sourced - HTMLFile objects, when called, reload from their sourcefile if the sourcefile has changed since the last call. > > You might want to add STUPID_LOG_FILE=[path to log] as well, so that > logging messages (like that of the GUF) will be sent to the proper log > file. Note, as said above, if Z_DEBUG_MODE is set, no forking will take place, and log messages will always be echoed to STDERR, even if STUPID_LOG_FILE is set to a filename. -- Martijn Pieters, Software Engineer | Digital Creations http://www.digicool.com | Creators of Zope http://www.zope.org | mailto:mj@digicool.com ICQ: 4532236 | PGP: http://wwwkeys.nl.pgp.net:11371/pks/lookup?op=get&search=0xA8A32149 -------------------------------------------

1 0

RE: [Zope] Re: export Z_DEBUG_MODE=no -- broken?
by Martijn Pieters 26 Feb '00

26 Feb '00

From: Andrey V Khavryutchenko [mailto:akhavr@compchem.kiev.ua] > > Hi, Martijn! > > >>>>> "MP" == Martijn Pieters writes: > > MP> Whoops! Yes, if you are wondering why tracebacks still are being > MP> displayed, you are probably still running with the -D > flag in your > MP> start file. Remove it and enjoy the clarity of your > error messages. =) > > And is there way to still have tracebacks in <PRE>, but wo -D flag? > > When I start Zope with -D and leave it running, closing the control > terminal, it barfs with the following error: > > File > /usr/local/apache/zope/lib/python/Products/GenericUserFolder/User.py, > line 211, in authenticate > File /usr/local/apache/zope/lib/python/zLOG.py, line > 198, in LOG > File > /usr/local/apache/zope/lib/python/Zope/ZLogger/ZLogger.py, > line 18, in log_write > File > /usr/local/apache/zope/lib/python/Zope/ZLogger/stupidFileLogger.py, > line 40, in __call__ > File > /usr/local/apache/zope/lib/python/Zope/ZLogger/stupidFileLogger.py, > line 62, in stupid_log_write > IOError: [Errno 5] Input/output error > > Confirmed on FreeBSD and Linux. Zope 2.1.2 Indeed, this is caused by the fact you closed STDOUT and STDERR on the process. In stead of the -D, you can start Zope up with z2.py [options] Z_DEBUG_MODE=1 where [options] is your current list of options used minus the -D flag. Zope then will behave demonic again (run nohup) and you'll only miss out on extra log messages from ZServer. You might want to add STUPID_LOG_FILE=[path to log] as well, so that logging messages (like that of the GUF) will be sent to the proper log file. -- Martijn Pieters, Software Engineer | Digital Creations http://www.digicool.com | Creators of Zope http://www.zope.org | mailto:mj@digicool.com ICQ: 4532236 | PGP: http://wwwkeys.nl.pgp.net:11371/pks/lookup?op=get&search=0xA8A32149 -------------------------------------------

1 0

Creating a Folderish class in Python Product
by Shalabh Chaturvedi 26 Feb '00

26 Feb '00

Hi all, I am creating a new product with one class - called ProjectFolder - which needs to be 'folderish'. So I derive my class ProjectFolder from ObjectManger and after restarting Zope I am able to create objects of class ProjectFolder. The problem is when I add any objects under a ProjectFolder object, they get added under the ProjectFolder's parent! That is, they get added under the folder which contains the ProjectFolder object and become siblings instead of children. Any idea what is going wrong? I feel that my ProjectFolder objects somehow do not have full folderish capabilities and the calls for 'add some object(x)' to ProjectFolder are getting handled by methods acquired from the parent folder. As a result the object x gets added under the parent folder. What can I do? For creating 'folderish' classes isn't the ObjectManager base class enough? Thanks in advance, Shalabh

1 2

[ZGotW] Issue #6 (Open)
by Zope Guru of the Week 26 Feb '00

26 Feb '00

A new issue of "Zope Guru of the Week" is available: =================================================================== Issue #6 -- standard_manage_header =================================================================== Status: Open Zen Level: Disciple (3/8) Keywords: Management Interface Submitted by: Dylan Jay djay(a)innocent.com ------------------------------------------------------------------- When I create a ZClass with my own management methods they will by default use standard_html_header etc. This doesn't look like the standard management interface. I think we should have a standard way of making all management interfaces look the same. In fact this should be able to be changed for any subset of management interfaces so you could for instance change the way the standard zope interfaces look. *Ed. I split Dylan's proposed solutions into a separate reply.* ------------------------------------------------------------------- To reply and win undying glory and mondo guru points: http://zope.palladion.com/demos/ZGotW/6

1 0

[ZGotW] Issue #5 (Closed)
by Zope Guru of the Week 26 Feb '00

26 Feb '00

The current issue of "Zope Guru of the Week" has closed: =================================================================== Issue #5 -- Rendering DTML in Properties =================================================================== Status: Closed Zen Level: Disciple (3/8) Keywords: DTML ZClasses Submitted by: Chuck Burdick chuck.burdick(a)tradingtechnologies.com ------------------------------------------------------------------- This problem occurs when you try to make a DTMLDocument-ish ZClass. 1. Suppose you have created a ZClass that you call 'myClass'. 1. Give 'myClass' a property sheet called 'properties'. 1. Create a text property called 'content' on 'properties'. 1. Create a view to edit 'properties'. 1. Create an instance of 'myClass'. Q: If you put dtml tags in 'content', how can you render the object and interpret the tags? For example, suppose you created 'index_html' as a DTMLMethod in 'myClass' as:: <dtml-var standard_html_header> <dtml-var content> <dtml-var standard_html_footer> This will not work. It will display the text value of 'content' but the dtml tags will not have been evaluated. This is a problem I frequently encounter. I do not want to create a complicated subclass of the existing DTMLDocument classes. Instead, I want a quick-and-dirty ZClass in which DTML tags are valid. ------------------------------------------------------------------- Reviewed by: Tres Seaver tseaver(a)palladion.com ------------------------------------------------------------------- What, no valiant Zopemeister would step up to the plate? OK, here is my take on the problem. Short Answer You can't do this "quick and dirty". Long Answer Standard DTML and Python methods do not have direct access to the DTML rendering machinery, which lives in $ZOPE_HOME/lib/python/DocumentTemplate/DT_HTML.py and in $ZOPE_HOME/lib/python/DocumentTemplate/DT_String.py. A quick scan of those files convinced me completely that I wasn't going to replicate it in a Python method or two. So how do you accomplish your goal? The simplest technique would be to morph your ZClass so that it inherits from ObjectManager -- you can then stash the 'content' field's DTML into a DTML Document, say 'content_dtml', contained within the instance, and render it from your 'index_html' with the exact syntax you specified. You might even be able to use the same user interface as your first cut: in the form processing method, for instance, you could add:: <dtml-call "content_dtml.manage_upload( content )"> Another answer involves hacking the dtml-var tag to allow for a 'fmt=dtml' parameter, which would then do the recursive rendering of the property value. -------------------------------------------------------------------

1 0

More sybase weirdness
by jiva＠devware.com 26 Feb '00

26 Feb '00

So I got the SybaseDA working, but now I notice that my REC_ID's (these are autoincrementing numeric columns in my tables) seem to always show as their number and an L. So like REC_ID 10 would be 10L to the zope DA. In ISQL they come out fine. What's the deal? -- Let us never negotiate out of fear, but let us never fear to negotiate. -- John F. Kennedy

3 3

Sybase DA problems! Need help!!
by jiva＠devware.com 26 Feb '00

26 Feb '00

So, I dunno what I did, but somehow I broke my sybase DA. I can import the sy_occ module into python just fine, but when I put it into my product subdirectory, zope stops running. It doesn't give any errors or anything, it just won't start. What did I do?? Someone please help! SybaseDA version: 2.0b2 Zope version: 2.1.4 Python version 1.5.2 OS: RedHat Linux 6.1 (Fresh Install) Sybase version: 11.9.2 -- HUGH BEAUMONT died in 1982!!

3 3