[turba] turning attention to undefined indices

[Eric S. Johansson]

> Message: 2 Date: Thu, 30 Oct 2003 18:31:03 -0600 From: Eric Rostetter <eric.rostetter at physics.utexas.edu> Subject: Re: [turba] Re: turba Digest, Vol 410, Issue 1 To: turba at lists.horde.org Message-ID: <1067560263.1ebdfvzvoy0w at mail.ph.utexas.edu> Content-Type: text/plain; charset="ISO-8859-1" Quoting "Eric S. Johansson" <esj at harvee.org>:
> 
>>>> > Then why not contribute to it?
>>
>>>
>>> see previous posting where I said yes and included a sample of my notes
> 
> 
> Yep, sorry, I was on vacation for a while, just now catching up on the
> email.  So I was several days behind when I replied to that.

so I see.  Thank you.

> 
> 
>>> another option is to walk away entirely as I may be doing with horde.  I
>>> did a quick survey of a few folks in my circle and most of them agreed
>>> that using Horde is a last resort application.
> 
> 
> Without knowing *why* it is a last resort for them, we can't come to
> any useful conclusion on our end.  In many places, Horde is the first
> choice for many people, not the last.

the commentary, and maybe it is based on older versions, was they had 
serious problems with dependencies on new packages that superseded the 
standard system packages.  They also objected to the time and energy it 
took to get it running and keep running what with the formerly 
frequently appearing PHP security holes.  I picked up a sense of 
something as simple shouldn't be this hard.  fortunately, it seems like 
php has cleaned up quite a bit and is not generating quite so many 
security problems as it had in the past.

granted, this is soft information and I don't blame you if you want to 
ignore it but it is there.


>>> If IT professionals have
>>> this attitude from their experience, the project can do what it wants
>>> but it does have a problem that is driving people away.
> 
> 
> I think, without much proof though, that the user base is growing actually.

fair enough.  I think you need to try and find those who failed and ask 
them why.  also, your user base may be growing but remember that 
populations grow even if members are leaving.  They are just leaving at 
a slower rate than they are arriving.

ask yourself why did Red Hat switch from Horde to squirrel mail?  I have 
other customers that have done the same.  Most common reason is 
ease-of-use and update.  Personally, I have not used squirrel so I'm not 
sure if it's true but I'm seriously considering experimenting for my 
personal webmail.  I will let you know how it turns out ;-)


>>> adequate is the set of instructions that give a 90 percent success rate
>>> when followed by an administrator with a couple of years experience.  Do
>>> I make this all the time?  No.  But it is the goal.
> 
> 
> Sounds reasonable, but I guess it depends on the amount of experience they
> gain in those 2 years what will happen.

sorry.  Bad English on my part.  What I meant was that if you have 
administrators with a couple of years of experience, 90 percent of them 
should be successful if they follow the instructions exactly.

>>>  Yes, I
>>> could piece together.  Yes, I eventually learned that I could ignore the
>>> connect error warning but this begs the question of why should anyone
>>> have to.
> 
> 
> No, no one should have to.  But if no one reports the problem, or submits
> a fix, then it probably won't get fixed.

true.  Should I consider it reported or do I need to get a bugzilla 
account here too


>>> In order to use calendaring, you need to
>>> become a DBA.
> 
> 
> Well, you could use MCAL, but that is another issue, and no longer
> recommended anyway...
> 
> 
>>> That makes me extremely uncomfortable.  I could not make
>>> a recommendation to any customer to take on that kind of risk unless
>>> they already had a DBA on staff who was underutilized.  And even then, I
>>> would recommend they think twice about that.
> 
> 
> I could train any admin in how to do the needed DBA stuff in less than an
> hour.  So it isn't much of an issue in my opinion.  But, I really see no
> alternative.  To support the features we want to support, with any real
> performance level, you really need to use a DB for it.  Just a fact of
> life.
> 
> There are pay/rent DB systems, but I don't think an free/open program
> should force, require, or even suggest the user pay for such a service.
> It could of course support one, but not require one.

I think maybe here we need to agree to disagree.  Given a fast directory 
file system like Reiserfs or ext3 with btrees, you could build a 
database like performance using an understandable directory structure 
and dbm files.  From what I have seen of the applications, and Horde is 
just using the database as a data repository and search engine.  I know 
I may be wrong but like I said, it's what I've seen so far.


>>> I have posted the failures so many times, it got kind of redundant.  But
>>> since you asked...
>>>
>>> Notice: Undefined index: businesscategory in
>>> /var/www/html/horde/turba/templates/browse/search.inc on line 14
> 
> 
> These are not database errors.

they aren't?  Well that makes me feel like a goldplated fool.  I spend 
most of my time in Python and this kind error (KeyError) comes up on 
database problems as well so I made a bad assumption.  When I searched 
the Web for solutions to this problem, I found lots of unanswered 
requests for similar issues.

I installed the standard components without modification except for the 
configuration file as specified in the instructions.  so I'm not sure 
where to go to even look for the solution.  The log files tell me 
nothing so maybe there's some debugging a need to find and turn on.

this is also good example for why I say it's necessary to have 
checkpoints of what you should see and what you should be able to do to 
confirm correct installation.


> Using RH maybe?  

know maybe.  Red Hat 9

>If so, modify /etc/rc.d/init.d/postgresql and change the
> line (around 150 to 151 in my versions) which invokes postgresql to something
> like (sorry for the wrapping):
> 
> su -l postgres -s /bin/sh -c "/usr/bin/pg_ctl  -D $PGDATA
> -p /usr/bin/postmaster start  >> /var/log/postgresql 2>&1" < /dev/null
> 
> (which appends to a log each boot, or to start a new log each boot use)
> 
> su -l postgres -s /bin/sh -c "/usr/bin/pg_ctl  -D $PGDATA
> -p /usr/bin/postmaster start  > /var/log/postgresql 2>&1" < /dev/null
> 
> These will log everything to /var/log/postgresql instead of /dev/null which
> is the RH default log location...

I will try that but it's curious that I told the database to log to 
syslog and it does a little bit but I see nothing else.  that might be a 
clue to part of the problem.  Horde did bitch about not being able to 
connect, I turned on the TCP connection and then I get the errors I 
spoke of above.  It's almost like it is connecting but not using the 
database.

>>> I apologize for my sharpness.  It was a combination of frustration and
>>> coming down with the flu.  I do not say this as an excuse but as an
>>> explanation.
> 
> 
> Bummer.  Hope you shake that flu soon.

I have but it's one of those bugs that lasts for three days and leaves 
you feeling exhausted for three weeks.


>>> it may have been poorly said but the emperor still has no clothes.
> 
> 
> Hey, if you have a good body, show it off!

the thought of a portly gentlemen scampering merrily down the street 
wearing nothing but a crown leaves me with one thought...

ieeee!! my eyes!! they burn, they burn...  :-)

> Well, by fixing the problem.  Which I will attempt to do just as soon
> as I get time.  But right now I don't have time, and still need more input.
> 
> You might want to discuss this on the doc mailing list in the future.  It
> isn't turba specific (the documentation part that is, not your error messages).

fair enough.  I did send one set of comments to the documentation list 
and asked the moderator to approve them.  Let's solve the problem I'm 
running into and try and see if we can describe it in a way that will 
help others not have the same set of bricks.

thanks for help.  I do appreciate it

---eric

-- 
Speech recognition in use.  Incorrect endings, words, and case is
closer than it appears