Demo Discussion
Forum Config Examples Contributions Vulnerabilities
  Discussion forum about ELOG, Page 30 of 805  Not logged in ELOG logo
New entries since:Thu Jan 1 01:00:00 1970
ID Date Icondown Author Author Email Category OS ELOG Version Subject
  Draft   Thu Sep 24 00:13:31 2015 Warning Stefano colafranceschiStefano.Colafranceschi@gmail.com Linux3.1.0Elog command not working properly
  68124   Thu Sep 24 00:14:49 2015 Warning Stefano colafranceschiStefano.Colafranceschi@gmail.comQuestionLinux3.1.0Elog command not working properly
<p>Hi,</p>

<p>I have a problem running elog command from php script. Basically in the php I call system(&quot;elog -h etc etc.. And from the php command line works but not in the browser apparently without errors (php apache logfile no problem), any suggestions?</p>
  68137   Thu Oct 1 12:01:11 2015 Warning Andreas Luedekeandreas.luedeke@psi.chQuestionWindows2.7.3Re: Syntax for conditional list display

Oups: there is a flaw in this approach! You can only get different commands for Anonymous and Login Users. You cannot select different command for Admin users: they will just have the "Config" command automatically.

You could allow anonymous users to make new entries and restrict login to administrators. But that is certainly not what you want, so forget my suggestion altogether.

Andreas Luedeke wrote:
Hi Terry,
if you really want to do it, you can find a way to make it work. But it is not straight forward and not what ELOG is made for. Here some ideas how it could work:
  • ELOG allows you to have different commands for different users. You could allow "New" and "Edit" for normal users, but "Reply" only for administrators.
  • You can use "Subst on reply <attribute> = <string>" to set a specifc fields only on reply. 
    • You would hide "status" & "assignee" for new entries or edit (Show Attributes Edit = <list>), and show a "set status" field instead.
    • The line "Subst on reply status = $set status" would take care, that only an administrator can set "status" with a reply to the original entry (+ a "set assignee" field, etc.). While normal users could set "set status", it would not change the "status" field. You could provide a comment or tooltip to explain that.
    • The list view or single entry view should not show the "set status", but only the "status" attribute (List display = <list>, Show Attributes = <list>).
  • Use "Propagate attribute = status, assignee" to take care, that the original entry gets their fields set accordingly.

This is how you could do what you want. But bugzilla can do a lot more if you are looking for a bug tracking software.

Cheers, Andreas

Terry L wrote:

Hi Eddy,

Thank you for your response.  I had a feeling that was the case, but I had high hopes! :)  The reason I would like this feature is because I am trying to use ELOG as a software anomaly tracking log, where any user can create an ELOG and fill out "category", "computer" etc, but only administrators can fill out the "status" & "assignee".  If I am lucky, this will be an added feature in a future revision.

Thanks again,

-Terry

Edmund Hertle wrote:

Hey Terry,

I don't think this is possible, since in general you cannot define config settings in a way that it would only affect specific users. For example, you also cannot have different themes for different users (as stated here: https://midas.psi.ch/elogs/Forum/68112 )

Cheers

eddy

Terry L wrote:

Good Morning,

I would like to limit the ability to edit certain attributes in an ELOG logbook, depending on the user who is logged in.  I.e. if "guest" or a standard "login user" is logged in, I only want to allow the user to edit "attribute1" & "attribute2" of say, 5 attributes.  However, if an Admin user or specific user, say "Terry" or "Bob" is logged in, I want to allow ALL 5 attributes to be editable.  To expand on this, I may want to even limit the visibility of the attributes as well.

I've read though "Condition Attributes" section of the help page, but it is not clear to me how/if I can do this?  Is that section even relevant to what I want to accomplish?

Can anyone provide some guidance?

Thank you for your time & help!

Sincerely,

Terry

 

 

 

 

  68209   Fri Dec 11 14:30:10 2015 Warning Andreas Luedekeandreas.luedeke@psi.chBug reportWindows | Other3.1.1-bc8310aRe: If a Required Attribute starts with a number – No CKEditor
Hi David,
I can confirm a bug: changing the encoding does not work if you have a required attribute starting with a digit. The following minimal config reproduces the bug:
[global]
Default encoding = 1
[test]
Attributes = 7SWM 
Options 7SWM = Name
Required Attributes = 7SWM

Whatever you choose as a default encoding you'll be stuck with. If you remove the "Required Attributes" it works again.

The workaround is very simple: do not use any attributes that do start with a digit :-)

Cheers, Andreas

David Dunne wrote:
Hello, Is the below an issue or am I doing something incorrectly? Creating an attribute starting with a number and then including that attribute in the Required Attribute list prevents access to the CKEditior. This appears to be the case at least with FreeBSD (10.2 Rel) using Elog V3.1.0 and Windows 7 with Elog V3.1.1-3f311c5. The problem can be recreated using the Elog supplied sample config and adding the necessary attribute, sample config showing below. While attribute 7SWM is part of the Required Attributes the CKEditor fails to appear, remove 7SWM from the Required Attributes list and the CKEditor returns. [global] port = 8080 [demo] Theme = default Comment = General Linux Tips & Tricks Attributes = Author, Type, Category, Subject, 7SWM Options Type = Routine, Software Installation, Problem Fixed, Configuration, Other Options Category = General, Hardware, Software, Network, Other Options 7SWM = Name-1, Name-2, Name-3 Extendable Options = Category Required Attributes = Author, Type, 7SWM Page Title = ELOG - $subject Reverse sort = 1 Quick filter = Date, Type Thank you, David Dunne

 

  68245   Wed Jan 27 21:33:53 2016 Warning Devin Bougiedevin.bougie@cornell.eduBug reportLinux3.1.1attachments created using "Image" button can not be clicked-on and do not appear in the attachment table.

This has been confirmed usign the Demo forum on this server.  For example, please see https://midas.psi.ch/elogs/Linux+Demo/8

When an image is attached using the "Image" button in the CKEditor, that image does not appear in the attachment table for that entry.  In addition, the image can not be clicked on to view the full image.

This was not the case in v2.9.2, at least.

Is there some configuration option I could toggle to change this behaviour?  By default, we would like any attachment (whether it's made using drag and drop, the Image button, or the "Choose File" button) to appear in the attachment table *and* be clickable to view the full image.

Thanks,

Devin

  68267   Fri Feb 26 08:47:22 2016 Warning Nigel Warrwarr@ikp.uni-koeln.deBug reportLinux3.1.1-1Possible bug in elogd execute_shell

I was just playing around with gcc6's new feature for warning about misleading indentation (which can often hide real bugs) and I think it found one in elog-3.1.1-1 at src/elogd.c:22538. Here there is an if statement, which looks as though it should be inside a loop, but it isn't. The code is:

      for (i = 0; i < MAX_ATTACHMENTS; i++)
         generate_subdir_name(att_file[i], subdir, sizeof(subdir));
         if (att_file[i][0] && strlen(shell_cmd) + strlen(lbs->data_dir) + strl$
             < sizeof(shell_cmd) + 1)
{
            strcpy(p, "\"");
            strcat(p, lbs->data_dir);
            strlcat(str, subdir, sizeof(str));
            strlcpy(str, att_file[i], sizeof(str));
            str_escape(str, sizeof(str));
            strcat(p, str);
            strcat(p, "\" ");
            p += strlen(p);
         }

and the if statment is accessing the loop variable i but it is actually outside the loop. Presumably, there should be some more curly brackets here. gcc6 gave the warning:

src/elogd.c: In function ‘execute_shell’:
src/elogd.c:22538:10: warning: statement is indented as if it were guarded by... [-Wmisleading-indentation]
          if (att_file[i][0] && strlen(shell_cmd) + strlen(lbs->data_dir) + strlen(subdir) + strlen(att_file[i])
          ^~
src/elogd.c:22536:7: note: ...this ‘for’ clause, but it is not
       for (i = 0; i < MAX_ATTACHMENTS; i++)
       ^~~

  68408   Wed Aug 31 10:31:19 2016 Warning Andreas Luedekeandreas.luedeke@psi.chBug reportLinux3.1.1Wrong time in attachment "Uploaded ..." date
When an attachment is later added to an exisiting entry, we see that the string "Uploaded <date>" has the wrong
time: in our case it is one hour into the future.
I'm testing it now here in the forum.

I can reproduce it in this forum: the Uploaded time should have been 10:35, but it shows 11:35.

Cheers, Andreas
Attachment 1: gnome-fs-bookmark-missing.png
gnome-fs-bookmark-missing.png
  68460   Wed Nov 16 17:12:40 2016 Warning Grant Jeffcotegrant@jeffcote.orgRequestAll3.1.2Re: Cancel button missing when editing an entry

Andreas,

Doesn't using the Save Drafts = 0 option result in the old 'Back' button returning (replacing the Delete option) meaning you can revert back to your old (unmodified) version without saving, sure it doesn't save a draft but I'd rather be able to return back and lose the changes I didn't want then mess up an old entry, I guess the back button on the browser could also do what you want?

The issue I had was that entries were being deleted accidently by persons other than those that created them, I reduced the edit time to try and mitigate it but ended up going back to what we had in Ver 2.9.2, a 'Back' button.

Grant

Stefan Ritt wrote:

The problem is that the elog database does not allow for a copy of an entry before you make modifications (and therefore get a "draft"). This is actually very simila to other note taking programs. I use Evernote, which constantly synchronizes between devices, and there I also cannot go back to the version before I started making modification. What one would need is a version system (and Evernote has one), so you can go back to the previous verison, the pre-previous version and so on. But this requires a complete redesign of the elog database.

A quick and dirty solution would be to store the origianal entry inside the browser (using JavaScript). You could then restore the initial version with a "cancel" button. But this mechanism relies then on the browser. If you just leave the page, there is no way the browser can put back the old version.

Stefan

Andreas Luedeke wrote:

If I edit an old entry, and do some mistake while editing, then there is currently no way to savely discard the changes.

The problem is that the entry will become a draft: if I close the window without saving, then the whole entry is gone: it will be converted into a draft entry. But the draft does contain my changes, it is not the originally saved entry!

The draft mechanism should keep a copy of the entry before I opened it in edit mode, and allow to go back to that copy. The edit form has currently the buttons "Submit", "Save", "Preview" and "Delete"; it should have in addition the button "Cancel", that just closes the edit window without saving the entry and even deleting the draft that was saved while the entry was modified. This should bring you back to the previous list view.

Is anyone out there in favour of this change?

 

 

ELOG V3.1.5-3fb85fa6