What is Informix?
Not a technical support thread. Neil Truby asked for feedback on a draft history of Informix and its products posted on his website. Mark Townsend (Oracle) challenged claims that Informix's technology was 'among the best' and that its multi-threaded architecture was revolutionary, arguing no independent proof existed and that Oracle's different architecture made such threading less necessary. Others blamed Informix's decline on weak tools, poor marketing, the Illustra acquisition and the 1997 accounting scandal. Truby conceded the piece was written from an admirer's viewpoint; the thread drifted into nostalgia about Unify and ended with no conclusion.
Auto-generated by DrWatson from the posts below — may be imperfect; read the full thread.
I'm writing a history of Informix and description of its product sets today for my website. A draft is at homepage.ntlworld.com/neil.truby. If anyone would care to review and comment, or add some text, I'd be very appreciative. thanks Neil
Neil Truby wrote: > I'm writing a history of Informix and description of its product sets today > for my website. A draft is at homepage.ntlworld.com/neil.truby. If anyone > would care to review and comment, or add some text, I'd be very > appreciative. > > thanks > Neil > > Well not to open old wounds, - but there would be many,many people that would dispute the following > Informix's database technology was widely regarded > within the industry as amongst the best available. And I don't really believe the following was ever really proved either > Informix Dynamic Server's revolutionary 'multi-threaded' architecture, ..... > was a quantum leap forward in making the most of available server resources. Part of Informix's demise was that they did actually think they had done something technically unique, and kept going on and on and on and on about it, whereas a large part of the world didn't actually think that what they had done actually made any measurable difference. And Informix could never offer any proof that it did. And I know that there will now be 100 ancedetotal (?) stories, but in the end, there never was ANY independant, verifiable, third party proof of 1) Informix technology being the best 2) Informix technolgy making the most of available server resources But I guess thats what people refer to as a lack of marketing ?
"Mark Townsend" <markbtownsend@attbi.com> wrote > Well not to open old wounds, - but there would be many,many people that > would dispute the following > > > Informix's database technology was widely regarded > > within the industry as amongst the best available. > > And I don't really believe the following was ever really proved either I am wondering how this can ever be proved by anyone. Has Oracle claims ever be proved by anyone. In ASKTOM website (official Oracle website), claims are made that query parallelism fundamentally can not scale. Have they done any study for competing databases about it. If that is true for only Oracle, then they should add a line about it, instead of generalizing that parallel query processing is unscalable. > Part of Informix's demise was that they did actually think they had done > something technically unique, and kept going on and on and on and on > about it, whereas a large part of the world didn't actually think that > what they had done actually made any measurable difference. And Informix > could never offer any proof that it did. I won't disagree with the above. In 1996 I attened an informix user conference in SFO where the user was asking why Informix development tools and end user GUI tools aren't good. Intead of replying to that, Informix was emphasizing on speed. Nothing annoys the user more than being told what they should be looking for. However Informix's demise had lot of reason, chiefly being the accounting scandal they got into in 1997. Frankly it never recovered from that. Secondly their accusition of Illustra was a disastor.
Yes, perhaps asking for comments wasn't such a good idea .....! Try to bear in mind that it's written from an Informix lover's perspective. It's not trying to put the boot in, but endeavouring to explain how a company with technology at least as good (how about if we said " ...widely regarded as amongst the best ..."?) as that of its peers failed due to its head being lodged up its arse. But perhaps your second point - that Informix Software continued to stress performance and let the tools that made Oracle dominant wither, is a decent one I hadn't considered .... We'll be writing an Oracle one written soon. Are you volunteering, Mark?! "Mark Townsend" <markbtownsend@attbi.com> wrote in message news:Om2Qa.45088$sY2.20361@rwcrnsc51.ops.asp.att.net... > Neil Truby wrote: > > I'm writing a history of Informix and description of its product sets today > > for my website. A draft is at homepage.ntlworld.com/neil.truby. If anyone > > would care to review and comment, or add some text, I'd be very > > appreciative. > > > > thanks > > Neil > > > > > > Well not to open old wounds, - but there would be many,many people that > would dispute the following > > > Informix's database technology was widely regarded > > within the industry as amongst the best available. > > And I don't really believe the following was ever really proved either > > > Informix Dynamic Server's revolutionary 'multi-threaded' architecture, ..... > > was a quantum leap forward in making the most of available server > resources. > > Part of Informix's demise was that they did actually think they had done > something technically unique, and kept going on and on and on and on > about it, whereas a large part of the world didn't actually think that > what they had done actually made any measurable difference. And Informix > could never offer any proof that it did. > > And I know that there will now be 100 ancedetotal (?) stories, but in > the end, there never was ANY independant, verifiable, third party proof of > 1) Informix technology being the best > 2) Informix technolgy making the most of available server resources > > But I guess thats what people refer to as a lack of marketing ? > >
Neil Truby wrote: > We'll be writing an Oracle one written soon. Are you volunteering, Mark?! Sure - either officially or unofficially.
> In ASKTOM website (official Oracle > website), claims are made that query parallelism fundamentally can > not scale. Have they done any study for competing databases about > it. Well, I don't know to what you are referring to on Toms site (which isn't really official - Oracle is a big company), but of course there is an upper limit to the scalability of query parallelism (eventually you will run out of a consumable such as CPU, memory or IO bandwidth), and of course studies of different technologies and multiple ways to do different things are and have been made. It's completely fair to say the that Informix multi threaded architecture was a much more pure text book implementation than what Oracle was doing. But one reality in a world of parallel realities is that the text book answer for multi process threading was primarily useful to mitigate a set of issues introduced by using the text book answer for concurrency management - read locks. Yes - switching in and out of a blocked process IS expensive (at least at the time it was), and a threaded architecture was a much better way to time slice - but if the processes don't get blocked in the first place, then you don't have to switch as often - and at what point is a threaded architecture actually worth doing ? Pipelining query results is another one - should you try to pass results up the tree as fast as possible, or should you try to get all levels in the tree to complete in the shortest period of time - You's pays your's monies and you's takes your choice. And checkpointing - do a little a lot, with all transactions taking an overall small hit, or avoid it for as long as possible, with possible dire consequences when a checkpoint just has to be completed ? The point is that what worked for Informix wouldn't have worked for Oracle, and vice versa. Each had a different architecture, each had different set of speed of light issues, and each had their own petty foibles. Anyhow ....
Mark Townsend wrote: > Neil Truby wrote: > > I'm writing a history of Informix and description of its product sets today > > for my website. A draft is at homepage.ntlworld.com/neil.truby. If anyone > > would care to review and comment, or add some text, I'd be very > > appreciative. > > > > thanks > > Neil > > > > > > Well not to open old wounds, - but there would be many,many people that > would dispute the following > > > Informix's database technology was widely regarded > > within the industry as amongst the best available. If the point of the OP's work is marketing then it is appropriate. If the point is an objective examination then it is best to not make value claims that have no objective basis. -- Daniel Morgan http://www.outreach.washington.edu/extinfo/certprog/oad/oad_crs.asp damorgan@x.washington.edu (replace 'x' with a 'u' to reply)
"Daniel Morgan" <damorgan@exxesolutions.com> wrote in message news:3F11E0B9.97DE0E82@exxesolutions.com... > Mark Townsend wrote: > > > Neil Truby wrote: > > > I'm writing a history of Informix and description of its product sets today > > > for my website. A draft is at homepage.ntlworld.com/neil.truby. If anyone > > > would care to review and comment, or add some text, I'd be very > > > appreciative. > > > > > > thanks > > > Neil > > > > > > > > > > Well not to open old wounds, - but there would be many,many people that > > would dispute the following > > > > > Informix's database technology was widely regarded > > > within the industry as amongst the best available. > > If the point of the OP's work is marketing then it is appropriate. If the point is > an objective examination then it is best to not make value claims that have no > objective basis. What's OP? Original poster maybe? Assuming yes, the opus is to write something from an Informix admirer's point of view. And to make the point that Informix was a company with technology which could at least hold its own with the competition, but failed because of other factors. Anyway, how many commericial databases were there 10 years ago? Five? Ten tops. Then *any* of them could be "regarded ... as amongst the best available"). :-)
Neil Truby wrote: > "Daniel Morgan" <damorgan@exxesolutions.com> wrote in message > news:3F11E0B9.97DE0E82@exxesolutions.com... > > Mark Townsend wrote: > > > > > Neil Truby wrote: > > > > I'm writing a history of Informix and description of its product sets > today > > > > for my website. A draft is at homepage.ntlworld.com/neil.truby. If > anyone > > > > would care to review and comment, or add some text, I'd be very > > > > appreciative. > > > > > > > > thanks > > > > Neil > > > > > > > > > > > > > > Well not to open old wounds, - but there would be many,many people that > > > would dispute the following > > > > > > > Informix's database technology was widely regarded > > > > within the industry as amongst the best available. > > > > If the point of the OP's work is marketing then it is appropriate. If the > point is > > an objective examination then it is best to not make value claims that > have no > > objective basis. > > What's OP? Original poster maybe? > Assuming yes, the opus is to write something from an Informix admirer's > point of view. And to make the point that Informix was a company with > technology which could at least hold its own with the competition, but > failed because of other factors. > > Anyway, how many commericial databases were there 10 years ago? Five? Ten > tops. Then *any* of them could be "regarded ... as amongst the best > available"). :-) No one, not even the most rabid DB2, Oracle, Sybase, or SQL Server expert ... well maybe rabid is a bit too strong ... could doubt that Informix has decent technology. Few would argue that Windows is better than the Mac O/S. Yet the market has a way of doing what it does. And it takes no prisoners. -- Daniel Morgan http://www.outreach.washington.edu/extinfo/certprog/oad/oad_crs.asp damorgan@x.washington.edu (replace 'x' with a 'u' to reply)
"Neil Truby" <neil.truby@ardenta.com> writes: > Anyway, how many commericial databases were there 10 years ago? Five? Ten > tops. Then *any* of them could be "regarded ... as amongst the best > available"). :-) Hey, I *liked* Unify 2000 so much a decade ago that I chose to work for the company as a CPE. Then they became a "tools" company with worse marketing than Informix... -- Forte International, P.O. Box 1412, Ridgecrest, CA 93556-1412 Ronald Cole <ronald@forte-intl.com> Phone: (760) 499-9142 President, CEO Fax: (760) 499-9152 My GPG fingerprint: C3AF 4BE9 BEA6 F1C2 B084 4A88 8851 E6C8 69E3 B00B
Ronald Cole wrote: > >Hey, I *liked* Unify 2000 <snip>... > > You are SICK! Admisibly I only worked with unify three or four years ago, and ten years ago they might have been better. The way they assign an index number to each field in the database and make you reference each field by that number in your C code caused us untold hours of work to keep 40 odd databases identical so that one binary could work at any site. We wound up adding extra fields to each table called "dummy<n>" just in case we needed them for a minor change not worth the 80 odd hours required to change them all. And god help you if you added two fields in the wrong order... I still wake up screaming sometimes. Scott Burns