By way of clarification, I was angling for a job at a local small-AM
cluster that was described as "board op with strong IT background"; when
I asked the hiring engineer for clarification on the phone; "strong
IT background" sort of evaporated.  I'm taking the interview, but...

----- Original Message -----
> From: "James L. Stewart" <[email protected]>

> Although I agree jocks shouldn't normally be consulted on the choice of
> automation systems, I think they do need to be consulted (or at least
> observed) about what they are trying to do, and how they are doing it to
> produce the product they are producing. The observer (you) then need to
> think outside-of-the-box a little and consider how all this might be
> done better.

Indeed.  Everyone needs to be pulling their ropes in the same direction...

> Compared to Simian, I find how many functions are broken out into
> separate programs in Rivendell leaves the on air part (rdairplay) to
> have a nice clean interface that leads to less on-air mistakes and
> problems with staff tampering. 

Well, that can cut both ways.  A buddy of mine ops there already, and
while I was visiting him last week, he had to make some manual changes
to the air log to make a prerecorded show and it's local breaks time out
correctly.  See more later.

>                                  The built in "music scheduler", though
> primitive, might avoid needing to purchase a 3rd party one (are they
> using one on Simian?). 

Newstalker; other station is sportstalk; I'm assuming their logs are 
hand built, though Simian's manual didn't make that sound easy; I don't 
see that there need to be as many degrees of freedom in the object tree
as Simian provides.  But then, I've never had to do it for-real.

>                         Other programs like "rdcatch" simply don't exist
> in Simain (last I checked, although you can schedule batch files from
> the main log to do things, but that is a little scary). I used to have
> "database" ("Soundhound") problems with Simian too, but things might be
> better debugged by now. Oh, and I've had to get viruses off of Simian
> systems before, of course NOT Rivendell ones. Simian systems do have
> things like "rdpanel" and "rdcartslots" and a CD ripper, but typically
> at an added cost. Finally with Rivendell, you never have to worry about
> your software getting "unregistered" for various reasons like you do
> with Simain.

Absolutely; always a big selling point.

> Related to this, I might ask the question of what "problem" are you
> trying to solve that switching from Simian to Rivendell is meant to
> fix? I've personally seen many automation systems (including BSI
> Simian) work just fine for people. So if you can identify some things
> where Simian simply can't do that Rivendell can, in areas that matter to
> the operation of the station, then there would be your selling point.
> It seems to me that they may not be using Simian to fullest extent
> anyway and perhaps it would work just fine if it was deployed better?

That's entirely possible.

The specific thing I saw that irked me was that Simian wasn't displaying 
scheduled end times for clips it knew the start and duration for, requiring
the op to calculate in his head.  Often.  I hate doing wallclock math; that's 
what frickin' computers are for.

I gather from the manual that it *knows* how to do that math, so perhaps that
column's hidden and no one realizes it.

My background is systems analysis, so figuring out problems and solutions is
in my wheelhouse; I wasn't going to try to sell Riv until I knew it was a
good answer.

> As I've been in management/technical positions over many automation
> system deployments, I'll offer my insight on comparing them to
> Rivendell. Most of the current bread of systems out there I've seen
> seem to work just fine (in many cases, "very well") when properly
> deployed. I feel Rivendell compares to most of these at least
> feature-wise. It also at least compares to some in terms of system
> stability. This is unfortunate as I'd like to say that it is superior
> to others (after all it runs on Linux), but it is not. Granted I use
> Rivendell on Debian (from Tryphon binaries, usually), so my deployment
> is not an officially supported Paravel system. Also I've done a lot of
> unusual things to and with our deployment that might be a contributing
> factor in some cases. In any case I don't consider Rivendell to be "bug
> free" or even as "bug free" as many (but not all) of the other systems
> out there. Just kind of typical, or a slightly below.

That's about the view I'd gathered from watching the mailing list.

> Next it comes down to "support". Granted this community can be very
> helpful at times, but it falls short of some of the "top notch" support
> I've seen some of the other companies provide to their customers
> (granted: at a price! $$$). It is not uncommon for me to see them
> "remote in" to their customers systems on a 24-hour basis, then diagnose
> and fix things with little end-user involvement. If you choose to use
> the Appliance CD (or a turnkey Paravel system), you can at least be
> eligible to purchase Paravel support. I do not know how this support
> compares to others that are similarly priced.

Indeed.  That would, of course, have been an item I'd have looked at; 
I hate going barefoot.

> The reason for the above paragraph is that this should be a major
> concern to an organization wanting to migrate to Rivendell. With my
> Rivendell deployment, we were originally tied in with ownership of the
> station, and thus "I wasn't going anywhere", so it was deemed safe for
> me to be the support person for our custom installation. We've since
> sold our interest in the station (we might be good at building them, but
> not so good at running them, especially when they are out of town!), so
> now the new owner has found himself more or less "married" to me which
> is beginning to worry him because he considers my rates out of his
> budget. I'm not in any way trying to rip him off, but the point is, one
> needs to worry about who is going to offer continuing support on the
> system, which ever system is chosen. All the more reason to go with a
> Paravel supported system just in case you have to buy some Paravel
> support someday. In my case I couldn't get the Appliance to easily do
> some of the things I'm doing with our system, and even if I had, it
> might have disqualified it for support anyway. That said, I see the
> possibility of the current owner of my Rivendell deployment considering
> either redeplolying on the Appliance, or worse yet, migrating to
> something like Simian or iMediatouch (or some other "lower cost" system)
> if he is serious that he thinks I am too expensive for him (I'm not
> really, he just needs to budget a few dollars for "engineering" which
> he isn't at all right now).

Not uncommon.  And yes, better is often the enemy of "Everyone knows this",
though that's probably less helpful in radio automation, there are enough 
systems that *lots* of them aren't First tier.

For the record, while I'm a newbie, I've been around Riv, and kibitzing
on the mailing list, for years; back into 0.x at least.

> I am not in any way blaming Fred at Paravel for not supporting systems
> like mine that are installed on Linux distributions foreign to him. I
> also don't blame him for any outstanding bugs in the current Rivendell
> system, as bugs exist in all systems. The difference is that many of
> the other bigger companies have much larger support staff that can be
> more responsive to their paying customers. In fact, as not even being a
> Paravel customer, Fred has personally fixed many issues I've found with
> Rivendell totally free of charge (Thank you Fred!), but he is under no
> obligation to do so, and I totally understand that.

He rightly considers them actual bugs that everyone else should get the
benefit of fixes too as well, no doubt.

> In summary: I'd consider "fixing" issues with the Simian system if you
> can (since they already bought it, unless it is an older version that
> needs upgrading for more money), 

It is 2.2.something; 2.3.0 is current.  So not too bad.

>                                   otherwise the previously recommended
> process of building up a "test" system on some old hardware (not even
> one of those expensive ASI sound cards unless they have one laying
> around from an old broken Simian system) and let them see how life could
> be on Rivendell would be a good direction to go. I would also test your
> ability to actually deploy a Rivendell system, which is NOT all that
> easy as well.

Absolutely on all points.

An excellent recap, which I wanted to reply to both to assure I wasn't going
off half-cocked, and for archive purposes on a couple points.

Cheers,
-- jra
-- 
Jay R. Ashworth                  Baylink                       [email protected]
Designer                     The Things I Think                       RFC 2100
Ashworth & Associates       http://www.bcp38.info          2000 Land Rover DII
St Petersburg FL USA      BCP38: Ask For It By Name!           +1 727 647 1274
_______________________________________________
Rivendell-dev mailing list
[email protected]
http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev

Reply via email to