[Date Prev][Date Next] [Thread Prev][Thread Next]
[Date Index] [Thread Index] [New search]

Re: Database Publishing Suggestions

I always appreciate reading such cogent analyses of the real-world
problems involved in using FrameMaker with a DBMS.  I too would be
interested in seeing actual performance benchmarks, though of course
the *best* benchmarks for your work can only be run on your own data...

On Wed, 3 Feb 1999 08:46:54 -0700 (MST), Dan Emory <danemory@primenet.com>

>At 06:22 AM 2/3/99 -0500, David Evans of Finite Matter, the developer of
>PatternStream wrote:
>>Dan, Dan, Dan,
>>I do respect your knowledge of database publishing issues, however, to
>>suggest that Unimerge is faster and better than PatternStream without ever
>>reviewing the PatternStream product demonstrates pure speculation on your
>Typically, I'm getting performance with UniMerge on a 266 MHZ CPU of 5000
>records per minute. These are complex records having upwards of 40-60
>fields, some with very long text strings. How about you volunteering some
>actual performance data on PatternStream so we can benchmark it against
>"I guess it depends on what your definition of IS, is."  The only
>>thing that YOU can say unconditionally is that Unimerge is cheaper out of
>>the box. When comparing YOUR cost (development time included) -- is
>>Unimerge really cheaper?  (I guess that is an unfair question since you
>>haven't taken the time to understand PatternStream's capabilities.)
>I do not believe PatternStream cuts development costs, particularly for the
>"high-end publishing" which you claim is the product's forte. With UniMerge,
>I can put together an application for a simple mailing list-type publishing
>effort and test it with test records containing instantiations of all the
>variabilities in record content in about 30 minutes. But such simple
>applications provides no meaningful information about capability. In a
>real-world application of some complexity, the development time is consumed
>primarily by an analysis of the customer's requirements, the content of the
>database, all the possible variations among records, and exception handling.
>Test cases must be developed for all of these issues. Then, there are all
>the issues involved in FrameMaker template design, which typically requires
>the development of more test cases. Then, a skeleton application of some
>sort must be developed to run against those test cases and analyze the
>results. After that, there is usually a back-and-forth process that goes on
>with the customer (sending samples and getting comments back) to perfect
>everything and get all the formatting and page layout issues resolved. As
>any programmer will tell you, no software product can automate this process.
>The actual final coding step is a minor part of the total programming
>effort. The same applies to high-end database publishing. I challenge you to
>prove to me that PatternStream provides any way to short-cut the development
>process described above.
>>To be clear, PatternStream will out perform any meta-tagged based system --
>>bar none. To claim expertise in a particular field, you must be
>>knowledgeable of all products and willing to try new things. I would
>>suggest that if you want to compare PatternStream to other products, try it
>>first -- contact us.
>Hmmm. Does that mean I have to be an expert in Microsoft Weird, WordPerfect,
>PageMaker, and Quark to know that FrameMaker is a better solution for
>iondustrial-strength database publishing?
>>Admittedly, PatternStream is for high-end publishing -- things that
>>Unimerge could not begin to tackle.
>And what kind of "high-end publishing" are you talking about that UniMerge
>can't tackle? Give me some examples. I can give you ample examples of
>extremely difficult high-end publishing I've done with UniMerge, and I
>suspect that some of those would be problematic with PatternStream. The fact
>is, in my 5 years of database publishing experience, I've never encountered
>an application that couldn't be cracked with UniMerge. Maybe, after
>PatternStream has been around for 5 years, your claims can be validated.
>>And be assured, PatternStream can
>>compete everywhere, from the simple to the very complex. Please understand
>>that FML is very familiar with meta-tagged systems too, and why
>>PatternStream was developed. With so many competing code based systems, how
>>can you recommend one over another -- they all do the same thing, the same
>>way, they just change the language used (meta-tags) a little bit.
>Have you, personally, ever used UniMerge to develop a high-end application?
>If not, how can you possibly know what you're talking about. The UniMerge
>command language has about 15 commands, but it's incredibly powerful. Many
>people (me included) much prefer to work with a concrete syntax rather than
>trying to do the same thing through a GUI whose underlying language base is
>hidden from the developer. When you substitue a GUI for the concrete syntax
>of a command language, you just complicate things, because the developer is
>never precisely certain of what the outcome will be, and will waste many
>hours puzzling over each discrepant behavior, and trying to figure out how
>to trick the GUI into producing the desired behavior.
>>To us, Seybold's selection as "Hot Product" makes the statement (and they
>>clearly understand what is available in the market). PatternStream doesn't
>>force anyone into a particular format -- PatternStream can do it all. Most
>>of our customers have tried the meta-tagged stuff, they are converts --
>>they understand the significance of a "LIVE" connection between the
>>database and the formatting engine.
>UniMerge has a live connction capability too. But in 5 years of database
>publishing experience, neither I nor my customers have ever found a
>compelling reason for using it. That being the case, the question for you is
>Can PatternStream handle the more common cases where a live connection is
>not only not needed, it is absolutely out of the question? UniMerge can.
>I recall a product called BrioPublish which used a GUI and a live database
>connection that would appear to be similar to PatternStream's. It took me
>less than two days of evaluation to determine that the product was useless.
>That opinion was confirmed when the product disappeared less than a year
>later. The BrioPublish "solution" simply didn't work in the real world of
>database publishing. 
>     ____________________
>     | Nullius in Verba |
>     ********************
>Dan Emory, Dan Emory & Associates
>FrameMaker/FrameMaker+SGML Document Design & Database Publishing
>Voice/Fax: 949-722-8971 E-Mail: danemory@primenet.com
>10044 Adams Ave. #208, Huntington Beach, CA 92646

-- Jeremy H. Griffith, at Omni Systems Inc.
   (jeremy@omsys.com)     http://www.omsys.com/
** To subscribe to Free Framers, email the message **
** body "subscribe framers" to majordomo@omsys.com **

** To unsubscribe, send a message to majordomo@omsys.com **
** with "unsubscribe framers" (no quotes) in the body.   **