Re: [cobalt-users] Chilisoft ASP

[Kris Dahl <kris@xxxxxxxxxxxxx>]

> I do agree with you that open source solutions like PHP should be better
> supported.  You should expect that.
 
I really do expect it.

> I'm surprised at your negative reaction to any ASP bundling though.  How
> is ASP different from FrontPage extensions when running on Linux?  How
> many ISPs could use linux at all for their web hosting without FPX?  Of
> course there are better solutions that are open sourced and/or free for
> both applications but Cobalt should be enabling as many customers as
> possible.  Doesn't adding ASP support onto a Linux box encourage an NT
> shop to ditch NT server and simply migrate web sites?  Once migrated to
> an OSS based platform those customers could wise up to other options
> that may be better suited to their needs?

I am not so much irritated at the fact that they are adopting it, but at the
fact that they are ignoring PHP.  We have a proven, tested, superior product
that is not in any way attached to Microsoft, and Cobalts is choosing to go
with hack of a hacked Microsoft product (which was a hack to begin with)?!?!
Its nutz.

I think that ASP on Linux doesn't help to migrate to a Linux machine as much
as it makes more people want to build ASP applications.  Basically people
are falling back on ASP purely because Microsoft is backing it--its pretty
dangerous to go that route, in my opinion.  Cobalt should announce that you
have two 'solid' (some more than others :) ) active content delivery
methods.  Put together a really cool package with PHP that should even come
installed ahead of time.  Include postgreSQL or some open source DB for the
backend and actually make an attempt before just giving into the Microsoft
way of doing things.

For the record, I don't support the use of Front Page extensions for the
same reason.  A hack of a Microsoft technology that was in and of itself is
not something I am willing to bet on.  Many of the problems with the Cobalt
products were a direct result of Front Page extensions.  I am of the opinion
that any serious ISP can get away with not offering FPX.  Basically, don't
cater to the people that want FPX, because typically they are on the low-end
of the web server market.  If you want to cater to personal web sites, etc.,
then you pretty much have to deal with FPX.  I would say that if that is
your target market, you get what you get.  Personally, I wouldn't have
anything to do with the technology because of underlying flaws in the
technolgy AND the business model.  On the other hand, the people you want to
be catering to are the people that don't need or want FPX--by and large
professional web developers.

I am pretty hard core about never going for the low-end of a market.  So if
I seem elitist, its more from experience that you can never win if you don't
aim for the high-end of the marketplace than anything.

> One of my goals is to evolve Cobalt Linux servers into a very painless
> drop-in replacement for NT boxen.  ASP support is a good way of getting
> there.  Now if only I could convolute bind enough to work with MS's
> interpretation of DNS... ;-)

I agree that a nice by-product of going with Linux-based solutions is
getting rid of NT.  By I purely consider that a by-product.  My goal is to
provide the best possible solution, no compromises.  Typically this is a
Linux solution.  Providing a 'crutch' just for the sake of ridding a network
of NT is foolhardy and really missing the point.  Getting Linux in there is
to offer superior uptime, performance, cost.  Its not really a political
thing to me.  (although I do really support a Open Source for its benefits,
not so much the political RMS standpoint).

Running a NT shop is okay as well.  Most of my former client's networks were
NT when I was a Systems Engineer.  If it is the best solution available.
While myself, see no current advantage in a Microsoft solution if you were
building one from the ground up.

So the thing that really irks me about the deal is that they are putting on
a poorer solution that isn't tested, isn't reliable, isn't backed by the
first party, isn't open source so you can fix it or continue to develop if
MS or Chilisoft drop the whole idea of ASP--by all possible counts it is
just plain inferior to PHP.  If this "solution's" only advantage is that it
helps people migrate from NT (which there is a killer app out there called
ASP2PHP try a freshmeat.net search which is a much better idea), then I
would question your motives.

The motive for Cobalt is clear: sell more boxes.  Nothing wrong with that,
and It'll probably work.  But it will be at the expense of people like
me--advocates and VARs--which are more important than a couple of former NT
customers IMHO.  I've been responsible for quite a few Cobalt machines
getting out there and I think that Cobalt will suffer more by alienating the
community that they *require* to be successful.
 
> We're not back stabbing anybody, watch for support of more tools, OSS
> and otherwise.

Excellent, I just think it is poor planning to dog a proven, excellent, OSS
solution by picking a half-ass one first.
 
> One technical correction regarding your email...  ASP is not in direct
> conflict with PHP, they can co-exist on the same server.

I didn't mean to imply this--I apologize for the confusion.  Although
*technically* this may not be true if ASP and FPX hoark your machine over
and you are having to become busy running the OS Restore CD.  That would be
an extreme case.  :)

> PC disclaimer:
> The statements above are not forward looking statements by Cobalt
> Networks but are my opinions.  Linux is a registered trademark of Linus
> Torvalds.  MS, FPX, Frontpage and all that are trademarks of Microsoft

One thing I always want to be clear about things that I post, and what I
expect from other users on this list.
    * You are free to your own opinion, regardless of who you are or who you
      work for *
Free exchange of IDEAS is the goal of this whole list.  Often time this
includes criticism.  My criticisms (while sometimes rare) are directing at
improving things.  

WHat I REALLY want is for PHP to be better supported and even encouraged on
Cobalt gear.

Final note--I just lost all of my email for the past 6 months this morning.
This has happened before and I am at the point that I don't think I trust MS
to build software that I need to rely on (even for email).  But that is
besides the point.  In there I had the results of a request I made about 2
mos ago regarding PHP usage.  I asked people to email me weather you are or
aren't using PHP.  If I remember correctly (and now don't have anything to
back it up) it was about 20 to 1 that people were using it.  I would put a
margin of error of about 100-200% (because I didn't get an actual
representation of the true marketplace--most people here tend to be saavy,
and hardly anyone emailed me saying they didn't use it).

So by my estimates roughly 33 to 95%  (closer to 33% I'd guess) of cobalt
users are using PHP on there equipment.  That represents a really decent
'market share' that PHP has gained on Cobalt gear.  This is with the package
being 'experimental'.  With no support, hardly any documentation and not
much of a mention of what it is capable on ANY Cobalt literature or on the
web site.

It just seems like a no-brainer to me that every effort should be made to
support what really could be considered a killer app for linux web servers
that even in the current unsupported, advertised, documented form is HUGELY
successful.

I doubt anyone will read all of this! :)

-k