Hi Stefan,

I've been slowly moving threads, and twice now so far (and reproducably) had elog crash.

In each case, it is trying to move a thread with more than 24 entries; it copies the first 24 entries, then
crashes with "Segmentation Fault".  It does not erase the lock file /var/run/elog.pid

I have got around this by manually copying the entries beyond no 24, then deleting the thread entry by entry.

I am aware that I have an old and limited machine (586, 256MB RAM, running Slack 10), and at first I was
"content" to write it off as that; but when it crashed for the second time at exactly the same entry (the
twenty-forth) even though the size of the entries would have been significantly different, I wondered if there
was some factor within  elog that could affect this.

I've not tried it with Copy to:, but imagine it will also be affected as the only difference with this and Move
to: is the deletion of the thread after all the entries had been copied.

> Hi Stefan,
> 
> I've been slowly moving threads, and twice now so far (and reproducably) had elog crash.
> 
> In each case, it is trying to move a thread with more than 24 entries; it copies the first 24 entries, then
> crashes with "Segmentation Fault".  It does not erase the lock file /var/run/elog.pid
> 
> I have got around this by manually copying the entries beyond no 24, then deleting the thread entry by entry.
> 
> I am aware that I have an old and limited machine (586, 256MB RAM, running Slack 10), and at first I was
> "content" to write it off as that; but when it crashed for the second time at exactly the same entry (the
> twenty-forth) even though the size of the entries would have been significantly different, I wondered if there
> was some factor within  elog that could affect this.
> 
> I've not tried it with Copy to:, but imagine it will also be affected as the only difference with this and Move
> to: is the deletion of the thread after all the entries had been copied.

This rings a bell: it's probably related to some internal stack overflow, since the entries are copied 
recursively. I have an idea on how to fix that, but I need time for that.

> > Hi Stefan,
> > 
> > I've been slowly moving threads, and twice now so far (and reproducably) had elog crash.
> > 
> > In each case, it is trying to move a thread with more than 24 entries; it copies the first 24 entries, then
> > crashes with "Segmentation Fault".  It does not erase the lock file /var/run/elog.pid
> > 
> > I have got around this by manually copying the entries beyond no 24, then deleting the thread entry by entry.
> > 
> > I am aware that I have an old and limited machine (586, 256MB RAM, running Slack 10), and at first I was
> > "content" to write it off as that; but when it crashed for the second time at exactly the same entry (the
> > twenty-forth) even though the size of the entries would have been significantly different, I wondered if there
> > was some factor within  elog that could affect this.
> > 
> > I've not tried it with Copy to:, but imagine it will also be affected as the only difference with this and Move
> > to: is the deletion of the thread after all the entries had been copied.
> 
> This rings a bell: it's probably related to some internal stack overflow, since the entries are copied 
> recursively. I have an idea on how to fix that, but I need time for that.
Thanks Stefan,  I'll be keeping an eye out on any annoucement about this one!

> Hi Stefan,
> 
> I've been slowly moving threads, and twice now so far (and reproducably) had elog crash.
> 
> In each case, it is trying to move a thread with more than 24 entries; it copies the first 24 entries, then
> crashes with "Segmentation Fault".  It does not erase the lock file /var/run/elog.pid
> 
> I have got around this by manually copying the entries beyond no 24, then deleting the thread entry by entry.
> 
> I am aware that I have an old and limited machine (586, 256MB RAM, running Slack 10), and at first I was
> "content" to write it off as that; but when it crashed for the second time at exactly the same entry (the
> twenty-forth) even though the size of the entries would have been significantly different, I wondered if there
> was some factor within  elog that could affect this.
> 
> I've not tried it with Copy to:, but imagine it will also be affected as the only difference with this and Move
> to: is the deletion of the thread after all the entries had been copied.

I reworked the internal memory allocation, since there was a stack overflow going over 24 entries. It should be now 
much better. Give a try to revision 2226.

Hi,

A user tried to import a CSV file, which caused e-log to add a field called "date" to the list of attributes (and then crash). This caused the log-book to be blocked until someone (guess who) would go edit the elogd.cfg file and then trigger a reload.

1. suggestion : E-log should not crash in this case

2. suggestion: E-log should not allow invalid attributes to be added via CSV Import, which causes the log-book to be blocked.

For the time being, I will just  "Deny import" (by the way, the doc says it is "Deny CSV import", but I think the syntax is "Deny import". Not really important.

I think this should be quite easy to reproduce.

Thanks a lot

Soren

If the CSV file contains a "date" column, elogd tries to interprete the date to the internal format. Now a date can be written in a huge number of variations, and I'm sure I did not cover all. So please send me your CSV file and I will fix the crash. 

This must be related to the unknown cookie "style". There was some work recently on elog to make it robust against unknonw cookies. The latest modification was on version 2192, and you probably have an older version. I made a release elog276-2.exe for you. Try it and let me know. You can also just delete your cookies in Firefor and it should be fine, but it's better not being able to crash elog. 

It's here. 

Ah, the GET /nor/elog.rdf gave me somt hint: Your Firefox browser checks the RSS feed of elog regularly. If the logbook is empty, the routing producing the elog.rdf file went into some kind of infinite loop, but this only happens on an empty logbook. If fixed this and updated the 276-2 file just now, so try again.

Thanks for your help

With a logbook defined by 

 [test]
 Theme = default
 Comment = Test of auto-increment
 Attributes = Batch
 Subst Batch = %Y%m%d-##

the auto-incrementing of the Batch attribute within dates works when the logbook contains only entries for
today's date but otherwise will give a batch number of "01" for each entry.  

Changing line 8714 of elogd.c as follows, from:

      /* if date part changed, start over with index */
      if (strlen(attrib[index]) > 0 && strncmp(attrib[index], retstr, loc) != 0)
         old_index = 0;   
      else
         /* retrieve old index */
      if (atoi(attrib[index] + loc) > old_index)
         old_index = atoi(attrib[index] + loc);
to
      /* if date part changed, start over with index */
      if (strlen(attrib[index]) > 0 && strncmp(attrib[index], retstr, loc) != 0)
         break;            /* <------------- */
      else
         /* retrieve old index */
      if (atoi(attrib[index] + loc) > old_index)
         old_index = atoi(attrib[index] + loc);

appears to fix this bug when I test it.  This code is inside a loop stepping back through all log entries, and
the variable old_index is already set to zero before the loop. The existing code resets old_index whenever the
prefix of the attribute string (containing a date) does not match the "current" value of that prefix as found in
retstr (set using strftime).  So, if there are any records for dates other than today then old_index is reset
and attribute values will be set with the counter = (old_index+1) = 1.  The modified version stops comparisons
once a different date is seen, but assumes that records are date ordered.  An alternative patch:

      /* if date part matches */
      if (strlen(attrib[index]) > 0 && strncmp(attrib[index], retstr, loc) == 0)
          /* retrieve old index */
         if (atoi(attrib[index] + loc) > old_index)
            old_index = atoi(attrib[index] + loc);
  
does not make this assumption and also appears to work OK when I test it.

> An alternative patch:
> 
>       /* if date part matches */
>       if (strlen(attrib[index]) > 0 && strncmp(attrib[index], retstr, loc) == 0)
>           /* retrieve old index */
>          if (atoi(attrib[index] + loc) > old_index)
>             old_index = atoi(attrib[index] + loc);
>   
> does not make this assumption and also appears to work OK when I test it.

Yes, this is the right one, it searches all entries to have the same date than the current one, and then looks for the 
largest index which will be stored in old_index and later incremented by one. I committed your patch to the 
repository. Thanks a lot.

We are currently having problems with Elog, I never had a problem installing it before but with my last 3 install I can't get it to work. We are running under Windows XP, the installation works perfectly but when I try to use the shortcut pointing to the localhost 8080, I get an error page from Internet Explorer telling me that the server cannot be found. I've rebootied the computer to make sure the server was running but still the same problem. Is there a basic requirement I should look at (Dotnet versionX or specific DLL required)?

No, there is no Dotnet or DLL required. Just start the ELOG server manually by clickin on Start/Programs/ELOG Server/Start ELOG server manually and watch if you can spot any error. One person reported that downloading again the package solved a similar problem.

 Thanks for the feedback, I found the problem, someone installed the AVG toolbar in Internet Explorer in the computers, and for some reason the AVG Error Redirection option prevents Elog from opening in Internet Explorer.

A few days after I create entries in my logbooks, they are being deleted. Has this happened to anyone before?

Never as far as I know. Try to do logging on your elog server ("Log file = ..."), then you will see all write/delete accesses. So if someone deletes your entries, you will see a trace. 

Hello

I have a config file  :

Attributes = Author, Author Email, Category, Customer, server,HPSD_ID, Cust_ID ,Subject ,  Change Window
; which fields will be shown !!
List display = ID,Author, Category, Customer, server, HPSD_ID, Cust_ID,Subject ,  Change Window

Type Change Window  = datetime
# Type Change Window End = datetime
Options Category = Network , Server ,
Options Customer = Arag{A}, Frieslandbank{B}, Super de Boer{C}, NRE{D}

{A} Options server = arag-lin1,arag-lin2,arag-lin3,arag-lin4, arag-lin5, arag-lin6, arag-lin7
{B} Options server = rc1upas02, rc1upas22,rc1upas03,rc1upas04,rc1uptsm01,rc1upas07,rc1upas08
{C} Options server = xia61031,xia61032 ,xia61033,xia61034,xia61035,xia61036,xia61037,xia61038,xia61039,xia61040,xia61032 ,xia61033,xia61034,xia61035,xia61036,xia61037,xia61038,xia61039,xia61040
{D} Options server = server27,server32,server34 ,server37,server38,server40
Extendable options = Category , Customer,server
Required Attributes = server , customer, Subject, Change window Begin , Change Window End
 

This works OK, However when I change the {A} Options  to {A} Moptions in the above code i get an error which tells me that a required field is not selected. So, with the pull down it works, with Moptions ( select boxes )  it will say that a required field is not selected.  Any hints ?

regards Arno

Thanks for reporting this bug. I fixed it in SVN revisoin 2213. 

Greetings,

I was wondering, is it possible to wrap lines in the config file ? 

I have to add a dropdown lost which is kinda long and typing everything on one line will make ik kinda unreadable.
Somehow wrapping the line so each entry will get on a separate line would make it much better readable. (which
makes less errors).

W.