[PLUG] Trend micro's Linux software...

[Elliott Mitchell]

>From: plug_0 at robinson-west.com
> Quoting Elliott Mitchell <ehem at m5p.com>:
> > >From: plug_0 at robinson-west.com
> > > You can get it for Enterprise Redhat too.  What Trendmicro should do in
> > > my opinion is write their own programming language and distribute a
> > > compiler for it written in assembly.  Chances are, that would protect
> > > their software.  Thing is, properly written, their Linux product should
> > > compile on any Linux system that meets standard instead of vendor
> > > specific criteria.  A better solution is getting TrendMicro to say,
> > > "here's the C/Java/C++ code to use our virus engines, you must pay 
> > > $20 a year though for access to current virus lists."
> > 
> > Develop an entire programming language of their own simply to distribute
> > one program?! Do you have any idea of the sort of complexity of
> > developing one good programming language? And the complexity of
> > developing a decent compiler? Add to this fact that writing the compiler
> > in assembly language leaves everyone out whose system has a different
> > processor. Then there is all the work to rewrite their software in this
> > new programming language. And the compiler itself has to be portable as
> > well...   ALL of these steps are *extremely* difficult!
> 
> Not impossible though, this is evidenced by the sheer number of programming
> languages that exist already.

Please read what I said.

I said it would be *extremely* difficult to do. I most certainly did not
say impossible. You're correct that there are quite a few programming
languages in existance today, however; try listing the _general_
_purpose_ programming languages in major usage today. At the scripting
level, Bourne shell, Perl, Python and BASIC (meaning VB) are the only
serious ones. Java covers the intermediate level, then C and C++ are the
only serious contenders at the low-level.

So you're suggesting that Trend Micro attempt to create a strong
challenger in a group of less than ten? In fact since they'd be doing
low-level programming they'd need to create a new language in a field
consisting of two, and they're very closely related. All of this for
*one* program? They'd need to spend a great deal more time on the
language/compiler than on writing the original program. It would cost a
fraction of the time to simply port the entire program.

Entirely possible, but it would be ludicrus to spend the required time to
do it.

> > Please think a little more and try to avoid suggesting totally ridiculous
> > ideas!
> 
> What's ridiculous is the fear of writing software for Linux in 
> the first place.  What's also ridiculous is the inability to get 

Please list any companies of significant size that are *afraid* to write
software for Linux.

> a valuable application that will work on one version of Linux to work 
> on another version.  Especially anti virus products like this 
> one are sorely needed.  Trend micro could implement a language 

The programs that truely only work on one distribution are vanishingly
rare. Certainly many programs are only /packaged/ for one distribution,
but that is not a significant barrier.

Anti-viral programs aren't needed for Linux itself, Linux is not an easy
target. A number of other companies *do* offer anti-viral software for
Linux. These are the enterprise versions though, scanning e-mail for
EvilOS virii that might pass through a server.

> like Java and not teach the world at large how to program in 
> it.  If I don't know a particular programming language, that 
> doesn't mean I can't compile a program written in it.  You 
> write a language that people don't recognize which is as 
> portable as Java, you can write multiple programs in it.  

Please state the relevance of this. In addition to the problems above, it
sounds like you're severely underrating humans. The human brain is an
incredibly powerful pattern matching machine. Even if they don't disclose
details of their new programming language, people will figure it out at
which point, what was the point of this exercise?

> Ideally, Trend would abandon protecting the program to 
> access their antivirus database and charge for accessing 
> the database instead, but I don't see this happening 
> anytime soon.

An anti-viral program isn't merely a DB access program. Their crown
jewels isn't just the patterns, the pattern matching algorithms are a
significant part too.

> > > It's surprising that Fedora isn't supported.   It seems that Trend 
> > > Micro is discriminating against up to date free verions of Linux.  
> > > Redhat Enterprise Linux is anything but free.  Strange that there's 
> > > a Suse but no Debian port.  
>  
> > Could you point to a commercial software entity doing otherwise? The
> > support the enterprise versions because no sane enterprise uses bleeding
> > edge distributions on servers (slight emphasis on /sane/). Likely someone
> > has developed a package for Debian. Most likely they are also
> > distributing a tarball, and that is usable on any system.
> 
> Does this mean that Redhat once had one free distribution for 
> everyone which has now been replaced by their more expensive than 
> Windows Enterprise alternative and the non production Fedora 
> system?  What drives the pricing of Enterprise Linux anyways?  It 
> seems the public at large is losing.  Instead of the operating 
> system should be open and free, we're seeing huge price tags on 
> Enterprise variants of Linux.  I don't think this is what Linus 
> intended for his project.  What I'm seeing and don't like 
> is that Enterprise Linux might become as expensive as commercial 
> Unix.  It doesn't look like Linux is going to stay free.  If 
> open source Linux distributions lag behind Enterprise variants, 
> what will be the advantage of Linux over Windows?  Linux is 
> popular because it's free and you can have all of the source 
> code to the system, supposedly.  Perhaps greed will turn 
> Linux into a lemon worse than Windows.  Those who think they 
> can charge a lot for a well put together Linux distribution 
> are making a huge mistake, they're going to lose the whole 
> audience.  The average Linux user isn't going to be a geek 
> anymore as adoption of Linux increases.  If the price of 
> Linux isn't contained and continues to rise, the bubble 
> will burst.  The average person will say, Windows is 
> cheaper.

Again, relevance? Notice that RH Enterprise is very definitely available
for free as White Box Linux. Notice their target market. They're charging
_companies_ lots of money to _guarentee_ their system will crash less
than once every five years. I imagine you're willing to deal with a
system that needs to be rebooted once a month, for big servers that is
too often. If you're just dealing with your desktop, then you're going to
go with cheaper options.

> As far as is Fedora Core 3 a good system.  It seems surprisingly
> stable to me.  There are a lot of updates to it.  If the updates
> continue beyond the introduction of Fedora Core 4, I don't see
> why Fedora Core 3 can't become stable.

Depends on your definition of stable. Companies tend to be buying
contracts that guarentee support for 3-5 years after the OS is released.
Fedora is ludicrously unqualified by this standard.

> Last I checked, Enterprise Redhat was in the $5k a server 
> range.  This is not what an ordinary person can afford
> for a home server.

Have you priced the Enterprise versions of other OSes? They're paying for
support, not the software itself. They're not targeting ordinary people.
If you're paying $100K/year to Oracle for software and support, $5K/year
for the OS is easy.


-- 
(\___(\___(\______          --=> 8-) EHM <=--          ______/)___/)___/)
 \BS (    |         EHeM at gremlin.m5p.com PGP 8881EF59         |    )   /
  \_CS\   |  _____  -O #include <stddisclaimer.h> O-   _____  |   /  _/
    \___\_|_/82 04 A1 3C C7 B1 37 2A*E3 6E 84 DA 97 4C 40 E6\_|_/___/