Document utilisation

[tqr]

Hi,

I'm a beginner, and i try to use document function.
I've installed openGoo on my computer.
When I have created or upload a document, where can i find it on my computer.

Somebody can help me please.
In advance thnak you.

TQR

[tqr]

I've founded. Files are uploaded in ../upload/0a.../ef../.

Ok - it's a little bit strange (or not i think)  to create a folder for each file, to save the file without its extension, to save the file with an another name.
Does someone know why  ? It would be easier to have files with their name, that way we could synchronise the upload folder on other computer.
I'm interested to have your point of view.

TQR

[ignacio]

Hi tqr,

The files are saved with those strange names so that it's easy for the application to uniquely identify files with the same name. This allows, for example, to have more than one file with the same name (or different revisions of the same file). It still could have been made different, like making a folder of what now is the filename and putting the file in that folder with its original filename. But anyway, that isn't the purpose of the file repository.

We will develop some way of file synchronization for OpenGoo, one of which is WebDav, so you will be able to mount your OpenGoo installation as a folder in your PC, maybe see the workspaces as folders, and work on the files as if you were working through OpenGoo, remembering who made the changes and saving revisions, etc.

Meanwhile you will have to get your documents by downloading them from OpenGoo, edit them with any application on your PC, and finally upload them back in.

[soomon]

but are all there folders and subfolders really needed to identify a file?
isn#t that information saved in the database?

i know that beasically it's not the sence of that filesystem that users downlaod the files with ftp or sth like that, but a filesystem that is more easy to read would be better.

maybe one can save revisions with the same filename and just append the revision to the filename?

[ignacio]

Another point in favor for using those folders is to avoid having too many files in one folder. This (I guess) could affect performance negatively.

And yet another point in favor of the folders mechanism, is that it is already implemented, whereas the other one has to be implemented . And, at least in my opinion, there are more important features to implement right now, and even bugs to fix.

Of course, if someone wants to volunteer to take on this he/she will be welcome, but not to replace the actual mechanism, but to add another one, so that we end up with 3 options: Filesystem repository, Database repository and this new Flat FileSystem repository.

[pae]

Another point in favor for using those folders is to avoid having too many files in one folder. This (I guess) could affect performance negatively.

For people contrloing the filesystem, the index option could be activated (for ext3 for exemple)

Of course, if someone wants to volunteer to take on this he/she will be welcome, but not to replace the actual mechanism, but to add another one, so that we end up with 3 options: Filesystem repository, Database repository and this new Flat FileSystem repository.

I'm agree with you. I just have a look at FileRepository.class.php.

I'll perhaps have some time (no guarantee) to implement the webdav part
Would you base it on http://pear.php.net/package/HTTP_WebDAV_Server/ or something else and where want you to place it in the source code?

[tqr]

Hi  ignacio,

Ok i understand the principle for managing files and folders.
I think that the synchronisation process you explain it's very interesting, i'm impatient.

It's trully a very good job and a very good product that offer a necessary alternative today to other existing products. I'v never find a product like this one.

Thank's for you and for all the development team, TQR

[ignacio]

Thank you for your comments.

pae, if you want contact us at contact@opengoo.org and we can coordinate better what to do with webdav. Marcos, one of the developers has been looking into it a little bit and so he can give you some directions.

Thank you very much, your collaboration would be very appreciated

[gylf]

Documents are a HUGE part of any management system for me so here are some quick thoughts:

I see nothing wrong with having duplication of files - it basically works like a backup system.  The challenge is making it streamlined in the work process so one isn't unnecessarily slowed down by having to get the file into the file system AND opengoo.

Folder <-> Opengoo "sync" is one way to accomplish this but I see that as being very challenging and time consuming from a technical standpoint.  Rather than sync folders, I would suggest scanning a root folder for all files (via cron or whatever) and automatically importing those files periodically so the user can go through the list and quickly attach them to workspaces.  I wouldn't bother preserving the folder structure exactly in opengoo, and let the user worry about how the files get onto the server (mapped drive, local upload cron, whatever).

Another possibility is just a quicker and easier way to get the document into Opengoo via the interface - right now one has to save the file locally, go to opengoo, click new upload, select the file, describe it, click "upload" ... that's a lot of clicks!  That is why I suggested a 'drag-and-drop' onto the workspaces as an alternative in another thread.  Other web apps use this (see drag-and-drop attachment feature in TiddlyWiki).

I also plan to write a little script for Ubuntu which will let me just right click a file and select "add to Opengoo" which will just directly import the file (to be classified later).  Should not be too hard to do.  Little scripts like that might help solve the problem, at least for now.

Finally, you might consider a third (or fourth) option for storage: Amazon S3.  Note that this has been implemented in the railscollab fork of projectpier: http://github.com/jamesu/railscollab/tree/master (might be based on a ruby S3 library though; not sure if an analogous one is available for php).

As always, keep up the great work!