BasicExtensionHandler in a real filesystem?

I can't for the life of me duplicate the functionality of the BasicExtensionHandler.jsx scripting example from the Adobe Bridge CS4 SDK in a real filesystem.
Does anyone have an example of what changes to make in the SDKNode, SDKSystem, and BasicExtensionHandler javascripts to make it work with real files instead of of the leaf nodes?
Thanks!

Hello Learner,
It depends on the requirement, but in most of the cases we assign the task as general task.
To have some more information on the various attributes,
·        General task
If you define a task as a general task, all users can execute the task. This is useful if the task is used in a workflow and you only want to define the recipients in the step definition. A recipient can forward associated work items to all users.
Work items whose tasks are defined as general tasks and for which no responsible agents or default rules are defined are offered to all users of the SAP System for execution in their Business Workplaces.
·        General forwarding allowed
A work item that represents a task with this property can be forwarded by one of its recipients to all users, even if they are not possible agents of the task.
·        General forwarding not allowed
A work item that represents a task with this property can be forwarded by one of its recipients only to the possible agents of the task.
·        Forwarding not allowed
A work item that represents a task with this property cannot be forwarded by one of its recipients.
Hope this will help.
Regards,
Sam

Similar Messages

Using opendir()/readdir() with a virtual filesystem

opendir() and readdir() work as expected when running on the real filesystem (compiling as a projector). But don't seem to do anything when run on a virtual filesystem. I can open files, but can't see them -- readdir() never lists any files. Similarly, I can't seem to change the working directory to a directory in a VFS.
Is this a known limitation, or am I missing something? I can see that the VFS backend has the capbility to list files... is this not tied in to readdir()? Is there some other way to read VFS directory contents from C? (Ideally, without falling back to using interop code, though even an example of that might be nice.)
I've tried this on both crossbridge 1.0.1 and the latest 1.1.0 devbuild, and have tried to search for answers on the subject but to no avail.

Not sure then - maybe you can attach sample project with makefile and I can try poking around for a bit.
In the meantime you can also consider using alternate approach, maybe generate XML file list each time you're creating your VFS which you'd then use for walking through files in your app.

Accessing the filesystem

Is it possible to acces the real filesystem (not a vfs) through libc with abode air? Even if this do not work for a release build that would be nice. I am trying to setup a dev environement with Lua, but each time I change my assets I have to compile a vfs and my as3 project. It would be really nice to be able to add assets without compiling each time, since as3 compilation (when linking with a flascc swc) is a bit long.
Also, at the moment I am loading my lua files with an URLLoader, but then I need to execute the file as a string (doString in lua), but a lua doString is a bit harder to debug than a real file. So I'd prefer to do it with a doFile call (which need to do it through libc).

We do provide a VFS based on URLLoader; you can see an example in the filesystem sample (samples/07_FileSystem). While I haven't tried it, I believe that Air will alow you to access local files via URLLoader. That said, the current implementation of the URLLoader VFS does some processing on the files that would need to be performed each time they change, so it may not speed up your development cycle much. If that's the case, you could try writing a custom VFS by implementing the IBackingStore interface. Note that the VFS system assumes that once your program is running, all changes to the filesystem are performed either via libc or the VFS ActionScript interface. In other words, if you make any changes to the files on your native filesystem, you'll need to relaunch your SWF or Air program before you'll see the changes.

How Can I Synchronize Files From Mac to PC?

I have a G4 laptop and a PC Desktop (HP Pavilion d4100e, AMD X2 processor, NTFS file system), and I need to be able to work on both. To work on both, I need to synchronize the file systems so that my 30k files on the Mac appear on the PC and I can progressively synchronize as I make changes on one machine or the other.
I've tried a multitude of solutions, but have thus far been unable to set things up so that I can synchronize from one machine to the other. I learned that the copying process from Mac to PC would be interrupted if any of the files had a number of special characters that could occur on the Mac but not the PC, specifically: / \ ? " * . I more recently learned that a space at the end of a Mac file name would, once the file was sent over to the PC, be impossible to open and impossible to change the file name. I fixed all these problems using the mv_sed unix program. Nevertheless, a simple folder copy in the finder would quickly lock up--it would continue forever but no data would be transmitted. I then used the Synk program to synchronize to the mounted PC drive (mounted w/ 'smb').
I got all my files over to the PC, but when I try to synchronize any changes I made since the initial transfer, Synk locks up as well--it keeps running forever w/o any transmission of data. At this point, I realized that I could not list the contents of some of the folders on the PC from the Mac (using Terminal, smb connection). After some experimentation, it became clear that there wasn't anything wrong w/ specific file names, but that it wasn't possible to list the contents of folders that had a large number of files. I fixed that, but I ran into a multitude of additional problems being able to list folders. I don't have the time to figure out why. My guess as to why Synk locks up is that it can't list folders for synchronization.
At this point, I tried to make the file system on the PC more UNIX like, in the hopes that this would make synchronization possible. I installed cygwin on the PC, and I tried to use rsync to synchronize the files. Using cygwin, I seem to be able to list any folder from my Mac. So far, so good. But rsync is unable to transfer many files to the PC. It seems to try to create a temporary copy and then it can't read its own temp file, apparently because the file was never created. It also goofs up at points and puts a copy of a folder inside itself. Also for some reason it copies files over even when they already exist.
So far, no good solution. Is it possible to synchronize a real filesystem from Mac to PC? How? I'm guessing Apple doesn't have much incentive to insure this is possible.
Cheers,
Peter

Hi Peter!
Not sure if this meets your needs and I've never used it but a quick search on the InterWeb reveals this http://www.powerfolder.com/.
Hope this helps! bill
1 GHz Powerbook G4 Mac OS X (10.4.7)

[vmware] Best way to share files between host and client

Good afternoon!
I have successfully installed VMware server on my machine and set up a virtual machine, powering Windows XP for study issues (emulation via wine does not do the trick for me). The virtual machine runs fast and stable but does not meet my requirements unless I have two issues faced:
First:
How to access a USB pendrive monunted by the host system properly within the client?
Second:
What is the best way to share file system space between host and client? Should I set up a Samba share which is accessible for both host and client or is it save to grant my VM access to my physical hard disc as a whole (using it as a partition)?
Thanks for answers in advance, I have still quite a lot to learn, when it comes to VMs.
Regards,
~cg

Do not setup your VMware to be able to access your physical disk as a whole. This would have you mounting the filesystem twice, corrupting it in the process.
You /can/ setup VMware with Sharing Folders that lets you share your filesystem (either / or just /home or something else) and that is what I would recommend.
As for accessing the pen drive from both operating systems at once, I'd say use the same Sharing Folders method. I always share /mnt so that I can access my cdrom/USB/etc from the vmware OS.
The sharing folders is basically the same as a samba share, but uses vmware's own method (HGFS). Samba is more secure (if you are worried about your VMware OS breaking from it's shell and attacking your real filesystem, which is unlikely). You need the vmware tools installed to make use of it.

JSF output truncated when integrating with OpenCms

Hi all!
I'm quite new to JSF and am trying to integrate it with OpenCms, a JSP based content management system. In OpenCms all files reside in a virtual filesystem (VFS) and are accessed through the opencms-servlet which is mapped to "/opencms/*".
While trying to integrate JSF with OpenCms I added another prefix servlet mapping for the faces servlet: "/faces/*". When I now try to access a JSF-page I have to use a URL like this:
http://myserver:8080/mywebapp/faces/opencms/page1.jsp
For my understanding the faces servlet now tries to read a file /opencms/page1.jsp from the real filesystem, which does not exist. Is this correct? How could I get faces-servlet to obtain page1.jsp from the VFS? However, I do not get an exception or a message telling me that the resource page1.jsp is missing.
Instead the faces-servlet seems to do it's work and forwards to the opencms-servlet which in turn forwards to the JSP-servlet and processes the JSP. I'm sure the whole JSP is processed as I added log-statements from the beginning to the end, of which all are executed. However, the output of the JSP is truncated whenever there exists at least one JSF tag in the page.
Any ideas what's going wrong here?
Is my understanding of the executing order correct?
many thanks in advance for your help,
- martin

Hi!
Working with JSF 1.1 works, but now I want to switch to JSF 1.2, which does not work :-(
I have a simple demo application (webapp 2.4 and webapp 2.5) without OpenCms which works. OpenCms 7.0.2 without JSF works as a 2.4 and as a 2.5 webapp. As soon as I want to access a JSF-page with an URL like http://localhost:8080/opencms-7.0.24/faces/opencms/mypage.jsp I get only truncated output like the following:
<form id="j_id_jsp_2033902881_1" name="j_id_
{code}
Any Ideas what could cause the truncation?
There are now log messages and no exceptions in any logfile :-(
tia,
- martin

What, exactly, is a webdav realm?

The web technologies admin document's glossary defines a WebDAV realm as "A region of a website, usually a folder or directory, that’s defined to provide access for WebDAV users and groups."
This strikes me as a bit unclear, as a "region of a website" can be many things. Do they mean a location on a *web server* that holds files for a particular website hosted on the server?
And in the context of the small lab network I administer, would I use one realm for every web service offered by the server? Or would I use different realms for each blog, wiki, group, etc?
Thanks,
Jon

Hi Jon,
generally, I think a Realm refers to an area within a website, as defined by some base portion of its URL, where access authorization is restricted. Usually, the simplest method to do this on Apache is with the .htaccess files. You define a Realm as the URL at which this authorization is enforced.
At its most basic level, a "region of a website" might just be a directory, such as http://example.com/test-dir/. The web directory test-dir may map directly to a real filesystem location, such as /Library/WebServer/Document/test-dir/.
Or a "region of a website" may instead map to a specific environmental context within an application. The URL components following the domain name may be a series of values that represent a certain state in the application. This can either be done by simply appending a query string of name=value pairs to the URL (like the URL for this post- there is no folder named ?threadID=2547419 on the disscussions.apple.com server). Or it can be more semantic, SEO friendly URLs like http://example.com/articles/2010/08/What-is-a-realm
A specific website have have many Realms defined within itself. It is up to you, the admin, to decide when and where you need to create a Realm to restrict access or activity.
WebDav is almost always configured inside a Realm because of the extra abilities the clients accessing the website may have: i.e. DELETE or PUT methods.
If your "small lab network" is only accessed by the people in your small lab, and not be anonymous people out on the internet, then you could probably get by without anything too complicated. Maybe one Realm for the entire website.
But if your wiki is used by some people, your blog is used by a different group and you don't what to share data or content between them, then yes, create a separate Realm for each area of content:
create a Wiki-realm for /wiki
create a blog-realm for /blog
This will allow you to manage access to each content area separately. If you have Open Directory set up, you can configure your realms to use those user accounts so you don't have to create new username/passwords for each Realm.
The documentation to do this is all pretty readily available from Apple.
Does this answer your question?
Cheers,
Matt

Recovering partitions

Well, I screwed up one of the HDDs that I have installed, causing the computer to not be able to recognize the filesystem type. The HDD was on sdb and had two partitions, one NTFS for Windows, and another Fat32 that I was storing media on. I did some reading around and found testdisk and I was able to get back into the Windows partition, but I've no idea how to find the media one. Does anyone have any ideas? This media one is the most important one for me to recover.
I tried photorec too, which comes with testdisk, but it was only pulling stuff from the windows partition, not the other one.

Amarildo wrote:I always save a copy of the MBR in case something goes wrong. Once deleted the partition table and couldn't get my partitions back, but I had a copy of it. It's also useful if you thing some malware is in the MBR (specially since you use Windows too).
This is good advice; however, most newer computers use the GUID Partition Table (GPT), not MBR, so backing up the MBR won't be of much use with them. To create a backup of a working GPT, I recommend using gdisk; its "b" option will back up the GPT data to a file. You can then restore the data using the "l" option on the recovery & transformation menu.
Of course, this doesn't help Somnus. I'm posting it just in the hope that somebody will follow this advice and be saved Somnus's pain in the future.
As to Somnus's problem, if you've recovered one partition, it's conceivable that you'll be able to recover the other just by creating a new partition in a suitable spot. If the recovered partition is the first one on the disk, and if they were both primary partitions, use fdisk (or gdisk if the disk was GPT) to create a new partition beginning on the first sector after the end of the existing partition. If the second partition was a logical partition, then I recommend creating a new primary partition with a gap of 2048 sectors. (I don't recommend creating a new logical partition because that requires writing data to the disk in an area that might hold filesystem data if you guess wrong about the partition sizes.) If the recovered partition is at the end of the disk, create your new primary partition beginning with sector 2048.
Once you create the partition, you can try mounting it. If that fails, though, do not use "fsck" or other recovery tools on the partition -- at least, not in a mode that changes data in the partition. A failure to mount the partition could indicate filesystem damage, but it could also be that you guessed wrong about where the partition begins. If you guessed wrong and use "fsck" to "fix" the data, you'll end up doing more damage. You could use "fsck" in a mode that scans but does not write data to see if it seems to think there's a real filesystem in the partition. Read the "fsck" man page (or, better, the "dosfsck" man page, since it's a FAT partition you're trying to recover) to learn about its options. I don't recall what they are, offhand.
This approach of guessing about where the partition begins is an act of desperation, but it sounds like you're at that point....

Leopard & Single User Startup

Since upgrading my iMac to Leopard,it will not boot in single user mode. Is this a previously known bug? Also wondering if the fsck code has changed with Leopard.

As far as I know, it is still Command-S early at boot to get single user mode.
What does "fsck code" means ? I don't think the syntax of fsck changed, but I'm wondering why you would execute fsck, unless you have a real filesystem issue that Disk Utility does not repair (which is using fsck under the hood)

Linking Somewhere to get a Package Compiled

Hey There,
I'm trying to create a package for util-vserver.. I'm compiling it with dietlibc support but it's installed in /opt/diet.. Since util-vserver's configure searches diet in /usr/lib/dietlibc/ I should create that as a soft link in order to get util-vserver compiled..
Is it possible to warn (and abort installation) the user about creating a symlink if the link doesn't exist ?

I bet there's a switch (to configure) where you can specify where the diet libraries and compiler are. Creating symlinks in the real filesystem in a PKGBUILD is a *bad* idea.
EDIT: if you 'return 1' in a PKGBUILD, makepkg fails.
EDIT2: Another solution is to run a little sed script on Makefile to change the directories before running make.
Last edited by bender02 (2008-06-08 17:33:11)

OCFS2 and file locking (why not?)

Hi,
I have three machine connected to a SAN Storage. I'm trying to build a Mailcluster with them, storing the mailboxes on the shared storage and using DNS round robin to connect to the mailservers.
For the shared storage I setup OCFS2 (1.2.6) with Debian Etch Kernel 2.6.22.6. The installation process was very easy and the storage is running, all machines can read and write the shared storage.
BUT: File locking doesn't seem to work. I've read that OCFS2 is posix file locking compatible and that it uses a DLM (which is running on my machines). But I've tested flock and fcntl with a small C programm. When I'm starting this both machines get the lock on the same file. I think this will lead to data inconsistency, won't it?
OCFS2 is a "real" filesystem, isn't it? Can I use it for my scenario? Did I miss something?
Thanks!

Worth reading the oss mails to find the answer - see for example
http://oss.oracle.com/pipermail/ocfs2-devel/2004-July/000529.html

Is it possible to install Arch on a usbstick ?

Hello there,
i would like to install arch on an usbstick like slax, puppy or dsl. Is it possible or does even exists a stick version of arch ?
Best regards
FightRight

FightRight wrote:
Thank you for this information. But that means in my case, that i have to format the stick with the unionfs or aurfs and setting up the fstab with the right options that you have describe.
Did i understand you correct ?
aufs/unionfs is not a file system that you can use to format a physical medium such as a usb-stick. It is a sort of pseudo-file-system that makes a read only real filesystem appear to be writable, but the writing is done not to the original file-system, but to another one, which could be a disk, memory or whatever. To use such a file-system quite a few changes are necessary to the way a Linux system is booted.
You can find some more details at aufs.sourceforge.net and some description of how it is used in larch in the documentation on the larch web-site (see below).

Find results in warning - problem?

Hi,
A find results in the following warning:
WARNING: Hard link count is wrong for /proc: this may be a bug in your filesystem driver. Automatically turning on find's -noleaf option. Earlier results may have failed to include directories that should have been searched.
Is this something to worry about?
Cor

wanderingidea wrote:
Hi,
A find results in the following warning:
WARNING: Hard link count is wrong for /proc: this may be a bug in your filesystem driver. Automatically turning on find's -noleaf option. Earlier results may have failed to include directories that should have been searched.
Is this something to worry about?
Cor
I get this warning too - I've not noticed it before. I guess it's from a recent findutils update.
% ls -lid /proc
1 dr-xr-xr-x 107 root root 0 2005-02-08 08:22 /proc
The number in the first column is the inode number. The third column is the number of the hard links to the inode. The slightly weird thing is I can't find any other hard links to it:
% sudo find /proc -inum 1 -xdev -ls 2> /dev/null
1 0 dr-xr-xr-x 85 root root 0 Feb 8 08:22 /proc
I don't think it's something to worry about - /proc isn't a real filesystem. I guess the links are something to do with processes - the number of hard links to /proc seems to constantly change.
- olly

Regular Preventive Maintenance schedule for xsan volume ??

hi
Is there is any weekly , monthly regular Preventive Maintenance schedule which we should perform for xsan volume like defragment , some commands etc
What are the steps for regular Maintenance for xsan ,servers which apple recommends ?
thanks

Hi,
I do not do such maintainance. However every night a cvfsck -nv is run. This is a readonly file system check. If in the result of this a line exist which says:
Filesystem would have been modified
Then a real filesystem check is needed.
A defrag is only performed on files, not on empty space. It takes very very long, and I do not expect much of it. So I do not do this.
Hope this helps,
Regards
Donald

Voting and clusterfile disks on raw blocks?

Is this possible?
Is there any way to get some document which describe when to use hardware clusterware and when to use Oracle one?
Our config:
Oracle10gR2 Itanium (RAC; ASM, DBS), Linux RHEL 5.3, HP EVA storage.
THX
Edited by: Funky on Jul 18, 2009 8:06 PM

Yes, it is possible (and done a lot) to put vote and registry on raw (block) devices. I don't know a document that discusses oracle vs 3rd party clusterware but my advice would be to use whatever feels comfortable for you. When your organization already uses some other clusterware in other projects and there is knowledge within your company, you should consider using that. When you don't, use oracle clusterware.
To me, the pure oracle approach is favourable because I want to have components from as few vendors as possible because I like to avoid the complexity that comes with all these different softwares and vendors.
Also, I am happy with datafiles on ASM and vote/registry on raw devices while other dbas feel more comfortable with 'real' filesystems where they can easily use operating system commands to copy/move files around.
Bjoern

BasicExtensionHandler in a real filesystem?

Similar Messages

Maybe you are looking for