Hello,

Somehow when replying to a log entry, a log entry was created that was referring to itself. How this happened, I have no idea. The effect was that each time this enty was loaded, the elogd started to hang, going to 100% load and not responding to any http requests anymore. This problem can be reproduced by manually creating such a self-referencing log entry (see attachment). The problem entry that leads to the crash can be made by editing any elog entry and adding a line "Reply to: X" and a line "In reply to: X" where X is the MID of this entry.

1) Maybe it is possible to add a check when writing files that ensure, that is a log entry does not reference itself.
2) Maybe when loading files are preparing the thread view, elogd can detect cycles and abort.

I'm having a minor issue that were getting double entries due to the user is using the "submit" button more than once.

I seems like when there is a lag either on the machine or on the network it is possible to tap the "submit" button more than once resulting i a double or triple entry containing the same text and a almost identical timestamp.

Is there a way to aviod this?, my "solution" so far has been to select the entries and manually delete them. I'm using Elog version 3.14

Kind Regards Finn

I too have this as an occasional issue, although in my case due to a dodgy pointer.  I too manually delete the entries.

Interestingly, it gives double entries - and thus the start of a branch - even in logbooks were branches are not allowed.

I just committed some code which disables the "Submit" button after the first click and replaces the text with "Please wait...". So double submits should not be possible any more.

I'm also happy to see this change implemented as we've had to deal with the same issue at times as well. Will this change be incorporated into the latest version (314-2, aka elog-latest.exe), or will there be a new version release (that is not in Changelog yet)? If so, can you give any ETA on this new code availability?

Also I noticed that the Elog Home page still says "Current version is: 3.1.2". I assume that only means it hasn't been updated, not that it means it's the current STABLE version and subsequent releases are beta -- please correct me if I'm wrong. I just want to make sure I understand how the versions and releases work.

Endless thanks for this product and all your work Stefan.

"I committed" means that the change is sent to the GIT repository. People who compile from the source code can pull and compile immediately. Windows users have to wait until I do the next release. I'm developing on a Mac and have to boot a special (old) Windows machine to compile the .exe which each time takes me about one hour including documenation updates, changelog updates, upload of zip files etc. Since my main job is heading a research group, I only can devote this hour once in a while, depending on my work load. Sometime even the weekends are too short.

Thank your very much for your work on this - as i mentioned this is a minor issue.

Kind Regards Finn

Hello,

I am using Elog v3.1.1. When entries contain embedded images, these images are attached to the email notifications but are not properly displayed in the body of the email.
Is it expected?
For example, with an image file myimage.png the email source code contains <img alt="" height="150" src="190208_143305/myimage.png.png" width="200">
Maybe the email should make use of the Content-ID of the attached image?

Your observation is correct. The mailer inside elog should scan all email bodies and replace all <img> tags correctly. I put that on my todo list.

Stefan

Hi,
  I'm using elog 3.1.3 and I have an elogbook with an attribute with name 
Attributes = Author, CO2 Temp [deg], ...

If I add this attributes to quick list, 
Quick filter = Author, CO2 Temp [deg]

it does display on web page, but the search fails with "ERR_TOO_MANY_REDIRECTS" if I leave the default value in the corresponding quick filter box.

My understanding is that the presence of a "[]" in the default search value is the reason for failure and if I remove them the search works fine.

Is there a workaround?

I don't want to change the attribute name, since I do have quite a large number of entries in the elogbook, and I'd like to keep the attribute in the quick search if possible.

Thanks,
  Stefano

No, there is no quick workaround. The only way is to go through your logbook files manually (or with a script), and replace all occurrences of that attribute.

Stefan

I have a setting which makes ELOG a perfect solution, but there's a situation that I'm struggling to get my head around. We have 3 separate laboratories, each one containing a number of temperature chambers, which run almost constantly over a number of shifts. Each temperature chamber has it's own logbook (laptop). So far, pretty simple.
My dilemma is, our network goes down for maintenance/updates (more often than I'd like), but our operation cannot afford to stop during network interruptions.
With that said, I thought about whether I could run a "local" logbook on each laptop/chamber, and somehow mirror the local logbook to the main ELOG server.
Perhaps I'm over-thinking this...do you have any recommendations?

Sure that's easy. Install elog on each laptop separately, so they run without network. Then, set up a central elog server, and use "mirroring" as explained in the documentation at https://elog.psi.ch/elog/config.html#mirroring

So when ever the entwork comes back, you execute a manual mirror operation, and your new entries will be pushed to the central elog server.

Best,
Stefan

Thanks Stephan! I guess I was making it harder than it is.  I'm still a little fuzzy -- in this instance, am I correct in saying that each laptop would be considered a "master", and the remote (network) server considered the "slave"?  Also, I'm not sure quite sure -- which server should be assigned responsibility for performing periodic synchronization between the laptop and the central elog server?

Thanks again for all you do -- Happy Holidays!

Frank

I would call the laptops the "master" being responsible for pushing data to the central server which you can call "slave"

Stefan

Thank you again -- very much appreciated!

I've got things working - sort of.  Ran into one strange problem that has me scratching my head.  I have two different laptops, each running a local instance of their own logbook.  Both are functional, but for some strange reason, one looks great, and the other is missing its graphic format.  I've attached a screen capture of that logbook, and a copy of the config file.  Do you see something that I've done wrong?

Thanks,

Frank

Sorry -- dumb mistake.  I moved the "theme" files to the resource folder.  Works like a champ...life is good!

When there are attachments in an entry, logbook is adding additional "MIME-Version" header to e-mail messages.

Spam filter in our university system is mostly giving warnings:

• X-Amavis-Alert: BAD HEADER SECTION, Duplicate header field: "MIME-Version"

But sometimes it is not redirecting further the message.

• Remote Server returned '< #5.6.0 smtp; 554 5.6.0 Bounce, id=27666-07 - BAD HEADER>'

In code I noticed there is always "MIME-Version" header added to the message and additionaly it is added again when a file is attached. I think it is not neccessary to add again this header.

I just checked the current installatin, and just found one MIME-Version in the header, even with attachments:

MIME-Version: 1.0
--------------717E3059287C82F12BC14941
Content-Type: multipart/related;
  boundary="------------4543EB0E5AAC14B461D70512"

No idea where your second header comes from, but sometimes I have seen some SPAM filters adding things to headers.

Stefan

Hello,

Is it possible to remove the "active" checkbox a user sees when they click on "config"?  Alternatively, is it possible to have a new user arrive directly at the logbook they chose, rather than starting out at the config screen?

We are running elog-3.1.1 on SL6 with webserver authentication.  For the most part this works great, but somehow a few users become inactive after they login for the first time.  They claim they didn't un-check "active," but I can't think of any other way this would happen.

Many thanks,
Devin

Hi,
  I reopen this old thread since I'm having a similar problem.

I'm running ELOG V3.1.1-0767eb0 on ubuntu with webserver authentication, and I'd like to remove 'active' checkbox as well.

The problem that we found is that if any user click on the [Set None] button at the bottom of the Config page in order to unsubscribe from all mail notification, *all* the checkbox are toggled, not only those for the email notification. And this includes 
also the "active user" as well.
Moreover, if an user click directly on the "active user" checkbox, a warning dialog box appears, asking for confirmation, but this is not the case is [Set none] is pressed: so, in that case, a user deactivate itself without warning.

Looking at the web page as well as at the code, I did not find a possibility, but maybe someone knows better.

Thanks in advance,
  Stefano

> Hello,
> 
> Is it possible to remove the "active" checkbox a user sees when they click on "config"?  Alternatively, is it possible to have a new user arrive directly at the logbook they chose, rather than starting out at the config screen?
> 
> We are running elog-3.1.1 on SL6 with webserver authentication.  For the most part this works great, but somehow a few users become inactive after they login for the first time.  They claim they didn't un-check "active," but I can't think of any other 
way this would happen.
> 
> Many thanks,
> Devin

> Hi,
>   I reopen this old thread since I'm having a similar problem.
> 
> I'm running ELOG V3.1.1-0767eb0 on ubuntu with webserver authentication, and I'd like to remove 'active' checkbox as well.
> 
> The problem that we found is that if any user click on the [Set None] button at the bottom of the Config page in order to unsubscribe from all mail notification,
> *all* the checkbox are toggled, not only those for the email notification. And this includes also the "active user" as well.

I can confirm that this bug is present even in this forum and in my current version (ELOG V3.1.3-aded4ae).
And I am positive that this is a bug: it was certainly not the intended behaviour.

Hi,
  a possible solution is to change the current SetAll() and SetNone() javascript functions to something like:


  <!--
function SetNone()
{
  for(var i=0,els=document.querySelectorAll('[name^="sub_lb"]'); i<els.length; i++)
  {
    els[i].checked = false;
  }
}
function SetAll()
{
  for(var i=0,els=document.querySelectorAll('[name^="sub_lb"]'); i<els.length; i++)
  {
    els[i].checked = true;
  }
}
//-->

It works on Chromium 58.x and Firefox 53.x, I don't know about other browsers.

The javascript snippet is generated in function elogd.c: void show_config_page(LOGBOOK * lbs) (line 13960-984 on my version of the source code).

Best,
  Stefano

Hi,
  sorry to bother again on this issue, but it is causing me some trouble, as many of my users are using the "SetNone" button in their config page, and doing so they deactivate their account, which I have to 
re-activate manually.

The warning about "Are you sure you want to deactivate your own account ..." is actually triggered only when someone toggle the "active" radio button, but not when the "SetNone" function is used. In that 
case no warning is present, and the user just deactivate his/her account. I'm sure that this is not the intended behavior.

I attach a modified version of elogd.c (based on git c1866af75bb2fe62aa868c4cf562f43a08e4b499 ) with the modified javascript I proposed earlier in this thread.

best,
  Stefano

Thanks for the fix, I merged it into the current development branch.

Stefan

I changed the code ot include the CSS file directly in the email body. So no need any more for external CSS files. I dropped the "Email CSS URL" option as well.

Stefan