strang@alberta.uucp (Lisa Strang) (09/25/89)
I am peripherally involved in the development of a medical billing application using INFORMIX on an SCO Xenix, and since it is the first time any of us are using INFORMIX - or Xenix for that matter, is there anything out in netland that documents mismatches between manuals and reality, common problems, etc? If not, and if anyone is interested, I would be willing to summarize any responses of problems/others that are mailed to me and post on the net if appropriate. Thanks, ======== Lisa Strang Edmonton, AB, Canada (strang@alberta) ========
mark@promark.UUCP (Mark J. DeFilippis) (09/27/89)
In article <1989Sep24.214421.7162@alberta.uucp>, strang@alberta.uucp (Lisa Strang) writes: > > I am peripherally involved in the development of a medical billing > application using INFORMIX on an SCO Xenix, and since it is the > first time any of us are using INFORMIX - or Xenix for that matter, > is there anything out in netland that documents mismatches between > manuals and reality, common problems, etc? You don't say if you are using ESQL/C or 4GL, or whatever, but I will make you aware of the rediculous policy INFORMIX SOFTWARE (Formerly RDS Inc) has of not providing users of their product with bug reports. Our story, and we are not alone: We started using Informix-4GL when it was in its infancy. After finding out that there were bugs that wiped out entire tables "the hard way", as well as many other bugs, and "bug like symptoms" they called features, we would call tech support and question them about it. The next day they would call back and say, "Oh yea, that's bug number XXXXX". Well thanks for the 2 days lost development time! Over the 3 years we used the 4GL to develop large scale applications our "support call file" started to become a support call book. What really bothered me was that Informix Software Inc. REQUIRES you to have maintenance to be eligible to receive software upgrades, which they charge you for. Many times these upgrades were just their bug fixes. Most of the time, support response was so slow, we found the work-around, and called Informix and provided them with the work-around when we had to call 2 days later on another bug. We were paying support $$$ to do this! It is my feeling that had we used another package at the time, we would have cut the time in half (Documented log hours for the project show 40% of logged time working on undocumented problems, bugs, features, whatever you want to call them. Oracle, and Ingress were not available under PC/UNIX, or PC/XENIX at the time. It is possible things are different now. However, I doubt it since I manage the Unix systems at Adelphi University MTH/CSC Dept, and we have a SQL and ESQL/C license with support. I have not received a bug report yet! -- Adelphi University, Garden City, NY 11530 (516) 663-1170 Department of Mathematics and Computer Science markd@adelphi.UUCP or mark@promark.UUCP UUCP: ...philabs!sbcs!bnlux0!adelphi!markd
belkin@teecs.UUCP (Hershel Belkin) (10/04/89)
I would normally e-mail this, but given the tone of a previous response I felt an obligation to publicly state my experiences with Informix (and let me assure you that I have no "connections" with Informix! :-) We have used Informix-4GL for about a year now, to re-develop an equipment maintenance and calibration recall/history application which used to run on a non-unix, higherarchical database system. Although the development cycle was not without its problems, we found Informix-4GL (and SQL) to perform very well, both in terms of system performance, and in terms of ease of development. I was most impressed with the fact that we did not need to resort to any code outside of the 4GL, even though our application is quite complex (a total of about 12,000 lines of 4GL). I don't know of many RDBM systems which provide such a comprehensive 4GL. (We selected Informix over Oracle based on cost, and depth of 4GL support). I have found that Informix technical support is quite good, but responses could be faster. As part of the support package, we receive a quarterly (I think) journal containing known bug work-arounds, technical tips, and other interesting articles. While I have no reason to doubt that there may have been serious support issues in the past, my more-recent experiences seem to indicate that Informix is cleaning up its act. We have even seen improvements within the past year. (BTW, our implementation is not PC-based, but on a HP9000 unix system) Without getting carried away here, I will just summarize by saying that while there are still a number of improvements which I would like to see (and Informix has expressed interest when I releate these suggestions...we'll see what future releases show for this :-), I am basically very happy with the product, and would probably choose it again if I had to make the choice now. -- +-----------------------------------------------+-------------------------+ | Hershel Belkin hp9000/825(HP-UX)| ...!{lsuc}!teecs!belkin | | Test Equipment Engineering Computing Services | Phone: 416 246-2647 | | Litton Systems Canada Limited (Toronto) | FAX: 416 246-5233 | +-----------------------------------------------+-------------------------+
mark@promark.UUCP (Mark J. DeFilippis) (10/08/89)
In article <1680001@teecs.UUCP>, belkin@teecs.UUCP (Hershel Belkin) writes: > I would normally e-mail this, but given the tone of a previous response > I felt an obligation to publicly state my experiences with Informix (and > let me assure you that I have no "connections" with Informix! :-) > > We have used Informix-4GL for about a year now, to re-develop an Our experience with Informix-4GL was from day 1 when they put their first product out. > > Although the development cycle was not without its problems, we found > Informix-4GL (and SQL) to perform very well, both in terms of system > performance, and in terms of ease of development. I was most impressed We had to try many complex (outer join) queries, and found usually the optimizer on the backend didn't work the way it was supposed to. eg. It would select all the rows in the larger of the two tables and then do the join even though _composite_ indexes existed for one or both tables. I could go on and on. > with the fact that we did not need to resort to any code outside of We had simple needs. We needed to develop 30% of our code in C, just to do simple things like read lines in a file, or get a user id number, write a log file, the real tough stuff... > we receive a quarterly (I think) journal containing > known bug work-arounds, technical tips, and other interesting articles. That is not a bug report, it is a tech magazine that makes you feel like the big $$$ you are laying out for support is worth it. A bug report is a sequenced list of bug numbers, and a short description of what they plan on doing about it. Bugs are usually fixed when Informix comes out with a full release. We took the product through only 3 releases in 3 years. I don't know what release they are at, but it took three years to go from 1.00.00, to 1.00.05B to 1.10.00A if I remember correctly. A year is a long time to wait for a fix. Each time we had to change/remove all the bug fixes we provided work-arounds for prior to the release because they fixed some of them. At one time I had the name of every manager within Informix, and discussed several issues with Roger Sipple, the president of Informix. We were almost sued by our client, but after discussing our Informix problem, they had their legal staff draw up a letter to Informix instead. Nothing ever came of it, and the project was never completed. When it looked like it might become a legal matter, we started keeping logs of all conversations, etc. with Informix. We still have some of them, but not on disk, even compressed, they take up too much room. :-) I still have to deal with Informix at the University. We use Informix-sql and Esql/C for a Computer Science course on our Unix box. Thank God undergraduates only write 100 line programs, not much can go wrong with simple select of two rows and a simple join. Have they changed or have they just gotten bigger? -- Adelphi University, Garden City, NY 11530 (516) 663-1170 Department of Mathematics and Computer Science markd@adelphi.UUCP or mark@promark.UUCP UUCP: ...philabs!sbcs!bnlux0!adelphi!markd
cam@void.UUCP (Cam Fox) (10/10/89)
When I recently received a letter from Unify hawking their new release
of Accell (BTW.. The upgrade is free if you have a support contract,
which ain't cheap), it was accompanied by a list detailing the "bugs"
that were addressed in the new release and another much longer list
of bugs outstanding for the product. Many of the outstanding "bugs"
listed were very minor but some were what you would consider
"noticable" at the very least. This "bug" list qualifies under the last
posters qualifications, it was a multi-paged sequentialy numbered list
that contained most of what I would expect to be present in a company's
bug report database.
I was somewhat impressed to have this info available. When I worked for
a Tandy (some years ago), I was taboo to send customers a complete list of
"problem reports" for any product. My current boss mentioned that their
previous database suplier had a seriouus problem even admitting to a
demonstratable bug.
I would rather know what I'm dealing with as opposed to hitting a dead
end and THEN being told after much struggle "That is a known bug to be
fixed in "The Next Release".
Complex software is gonna have some bugs, hats off to these folks for
atleast admitting it and trying to deal with it on a "professional"
level.
Cam Fox cam@void.uucp || {texbell,texsun,attctc}!void!cam
Electro Plate Circuitry, 1430 Century Drive, Carrollton, Tx 75006
(214) 466-0818