> I noticed that the gcc compiler options in the tarball Makefile were not
> conducive to speed.  There, I tried changing the gcc options to:
> 
> CFLAGS = -O3 -funroll-loops -fomit-frame-pointer -W -Wall

Thank your for this hint, I changed my Makefile for the production code. However, I
could not feel any difference between the two options. The real problem is the
function getcfg(), which gets called many thousand times internally and has to parse
elogd.cfg each time. Once I implement a hash table for that function, elogd should
become faster by at least a factor of two.

> > I noticed that the gcc compiler options in the tarball Makefile were not
> > conducive to speed.  There, I tried changing the gcc options to:
> > 
> > CFLAGS = -O3 -funroll-loops -fomit-frame-pointer -W -Wall
> 
> Thank your for this hint, I changed my Makefile for the production code. However, I
> could not feel any difference between the two options. The real problem is the
> function getcfg(), which gets called many thousand times internally and has to parse
> elogd.cfg each time. Once I implement a hash table for that function, elogd should
> become faster by at least a factor of two.

Yeah - What's up with that?

I have seen this discussed before - Seems like it should be a priority to get this
fixed, as doing a hash table is straightforward, and the speed increase should be pretty
health - there are several c libraries available - check out "man 3 hsearch" for the
POSIX hash table management that already available.  Other c library searches that you
could use include bsearch (binary tree), tsearch (tree searching), btree (b+ tree). 
However, the easiest and most obvious one to use for elog appears to be a simple hash
table search (hsearch).

Is there something else  which is making this difficult to do?

> Is there something else  which is making this difficult to do?

Not really, but hsearch() & Co. are not available under Windows, so I have to extract the
source code from the GNU C libarary or so. Since the last discussion I had lots of other
topics on my to-do list, such as mirroring and cloning, but the speed issue is getting more
and more up on the priority list.

> > Is there something else  which is making this difficult to do?
> 
> Not really, but hsearch() & Co. are not available under Windows, so I have to extract the
> source code from the GNU C libarary or so. Since the last discussion I had lots of other
> topics on my to-do list, such as mirroring and cloning, but the speed issue is getting more
> and more up on the priority list.

Speaking of windows I'd like to note that when I moved my call tracking config from a slow BSD
system (PPro 200Mhz) to a faster windows system (P3 733M) I noted a huge slow down in the
interface.  Talking about perhaps 1-2 seconds before to 10-15 seconds after.  Using
sysinternals file monitor I see that elogd is hammering each log file in the directory.  Not
sure what else is going on.  309 log files - only 1.25Meg.  

Anything I can do short of pruning down the files?

[Edit:  In both cases above my default view is filtered and sorted - so that I only see things
with a specific status.  Taking away the filtering resolves this hit - but does not explain the
speed difference between platforms.]

-D

I think the text display of ASCII files, which is new in version
2.5.4, is not a good idea.  E.g. I had a large ps file attached
to one entry and it took a long time display this entry (over DSL).
Then I saw that the ps-file is displayed as text, which is not really 
useful.  

Probably it is fine to display only files ending in '.txt' per default.
In addition a file that has more than say 1000 lines should probably 
also not be displayed (as default, optional OK).  

Cheers, Heiko

[...]
> Probably it is fine to display only files ending in '.txt' per default.
> In addition a file that has more than say 1000 lines should probably 
> also not be displayed (as default, optional OK).  

No, '.txt' would definitely not be enough for me. I'm using elog to log all
administration of our network. In many cases, I simply attach a configuration
file. All those files are plain ASCII and none of them end in '.txt' - and I
would most definitely like them to be displayed inline like they are now. In
fact, this change was the main reason for me to upgrade to 2.5.4

Maybe a configuration option or a "display attachment" button would be the
best solution, then?

Cheerio,

Thomas

> [...]
> > Probably it is fine to display only files ending in '.txt' per default.
> > In addition a file that has more than say 1000 lines should probably 
> > also not be displayed (as default, optional OK).  
> 
> No, '.txt' would definitely not be enough for me. I'm using elog to log all
> administration of our network. In many cases, I simply attach a configuration
> file. All those files are plain ASCII and none of them end in '.txt' - and I
> would most definitely like them to be displayed inline like they are now. In
> fact, this change was the main reason for me to upgrade to 2.5.4
> 
> Maybe a configuration option or a "display attachment" button would be the
> best solution, then?
> 
> Cheerio,
> 
> Thomas

So to make everybody happy, it would probably be enough not to display inline any
*.ps file, is that right? Is there any other ASCII format, which should not be
displayed? PDF is binary, so it won't be displayed. What about long C files? Most
people want to see them. In the recent version there is the "Hide attachment"
link which can be clicked to not display an attachment inline. Mabe there should
be a "Hide default = 0|1" config option...

> > [...]
> > > Probably it is fine to display only files ending in '.txt' per default.
> > > In addition a file that has more than say 1000 lines should probably 
> > > also not be displayed (as default, optional OK).  
> > 
> > No, '.txt' would definitely not be enough for me. I'm using elog to log all
> > administration of our network. In many cases, I simply attach a configuration
> > file. All those files are plain ASCII and none of them end in '.txt' - and I
> > would most definitely like them to be displayed inline like they are now. In
> > fact, this change was the main reason for me to upgrade to 2.5.4
> > 
> > Maybe a configuration option or a "display attachment" button would be the
> > best solution, then?
> > 
> > Cheerio,
> > 
> > Thomas
> 
> So to make everybody happy, it would probably be enough not to display inline any
> *.ps file, is that right? 

I think there should be size limit.  Imagine a multi MB text file (whatever it is;
elogd.c is already more than 1/2 MB and is likely to increase due to your
excellent support).  A client on an ISDN line would have to wait
several minutes and during this time elogd is busy and no other client can connect
(correct?).  Of course, if somebody really wants to see this file then there is
nothing to be done, but likely someone is flipping throught the messages using
to arrows on top to find the right entry....
So a configurable size limit seems appropriate, from which on 
only 'Display attachment' is displayed.  And/Or, for files
exceeding this limit, the first N (new config option) lines could be displayed.

But this should only influence ASCII files.  E.g. the behaviour for jpeg files
should not change, which is controlled by 'Hide default'.  Maybe a 'Display/Hide
defaut extension' option, where the extensions are listed that are to be displayed
is another idea, in addition to a 'Max Display ASCII inline size' option,
which can be set to zero to disable it altogether.

In any case, whatever you think is best.

Cheers, Heiko

> Is there any other ASCII format, which should not be
> displayed? PDF is binary, so it won't be displayed. What about long C files? Most
> people want to see them. In the recent version there is the "Hide attachment"
> link which can be clicked to not display an attachment inline. Mabe there should
> be a "Hide default = 0|1" config option...