From NewsServer: discuss.softvelocity.com
News Group: comp.lang.clarion

>I think demand, and interest, has shown that it's necessary.<

Before you commit too heavily for something that people might back away
from, I think people need to be apprised of a few of my findings that I'm
sure will come out shortly anyway.

I've been playing detective and have not found anything that posts SQL
statements to a file to enable a roll forward even though there is a vendor
out there that say that he can do that. He relies on you running a command
line process that moves captured statements to a file. This is not
something that captures statements to an external log file when the
transaction is committed. Furthermore, it is user based, not cheap, and not
free to distribute. If it turns out that you need pay-for tools at the user
site, then it is not much, if any cheaper than ASA, and may be more costly,
all things considered.

There are more problems than hardware problems that happen in production
environments. There are the "I had a woopsy. I shouldn't have purged all
of those records. Can you get them back for me?"
- The ASA answer: "Yes but it's going to cost you. I'll think of
something."
- The Firebird answer: "Yes, but could you help me out by going around and
ask everyone if it is OK to lose everything that occurred since "sometime"
in the middle of third shift last night. Would you like me to idle our
automotive supplier manufacturing facility and its 350 users right now for
you? BTW, when you are finished and let me know what you want to do, tell
GM that they're going to get an extended lunch break tomorrow at their
Arlington Assembly plant while we airfreight parts to them."<G>

People could counter with, "Hey just run the log file maker a lot. Besides,
you could run it after the fact with the mass delete, and still get the rows
back". You'd be right, but for me this stirs thoughts of a famouse Disney
character, and breeds low confidence.

Currently, in my mind, Firebird cannot be positioned alongside of ASA, since
it would be irresponsible to deploy it in a mission-critical role, at least
until I was convinced of a solid logging solution. It appears to come in
well ahead of TPS in almost any case, and ahead of MSDE in most, but not all
situations. The ability to step between embedded demo and mission critical
client server using the same database does not seem reasonable based on what
I know so far. These reasons are the source of my reticence to commit to
Firebird after spending considerable time working with and testing it. I'm
not an expert in Firebird now by any means. However, the fact that it has
been so difficult to find a straight answer on LOG files, something required
for any mission critical app, indicates that there really is not much
experience out here.

If people are still as hot on Firebird as ever, then it may be a good
business decision. However, I do not see Firebird reducing the market for
FM3 support for ASA for the reasons cited. I simply see Firebird as an
additional revenue opportunity that will probably be a little less than
anticipated, but perhaps worthwhile nonetheless.

Regards,
Jack T.



JVZ Systems CC Customised Software - When it needs
to fit like a glove

Johan van Zyl
Owner JVZ Systems CC
PO Box 3469
Somerset West
7129
johan@... http://www.jvz.co.za tel:
fax:
mobile: +27 21 851 7205
+27 21 852 2387
082 875 4238




Signature powered by Plaxo Want a signature like this?

Add me to your address book...


---
Outgoing mail is certified Virus Free.
Checked by AVG anti-virus system (http://www.grisoft.com).
Version: 6.0.647 / Virus Database: 414 - Release Date: 2004-03-29


[Non-text portions of this message have been removed]