How to Move Cheaply Out of COBOL
in [Cobol]
|
From: Howard Brazee on 7 Jun 2010 09:58 On Mon, 7 Jun 2010 10:34:17 +1200, "Pete Dashwood" <dashwood(a)removethis.enternet.co.nz> wrote: >If a man using a Bobcat to excavate his garden cuts through a main power >supply cable and plunges a few suburbs into darkness, should we blame the >diesel powered tool or the glucose powered man using it? Possibly not. What are the laws and expectations for digging in that man's property? Some places he doesn't have any rights to what is below his property, other places he has full rights. If someone is leasing a right of way for a power line, he has responsibility. .... >Software without "sanity checks" (like not allowing negative values for >something like a hole in the ground) is likely to generate tears. Software >engineers who don't understand what they are doing are also likely to cause >much wailing and gnashing of teeth. Too bad kids don't learn to use slide rules. Teaching them to estimate and make sanity checks is even more useful than understanding proportion and significant digits. -- "In no part of the constitution is more wisdom to be found, than in the clause which confides the question of war or peace to the legislature, and not to the executive department." - James Madison
From: HeyBub on 7 Jun 2010 11:28 Pete Dashwood wrote: > > Software without "sanity checks" (like not allowing negative values > for something like a hole in the ground) is likely to generate tears. > Software engineers who don't understand what they are doing are also > likely to cause much wailing and gnashing of teeth. > My point exactly. EasyTrieve allowed the unwashed to place a value of -17,000 feet in a database where the standard was the reverse. The original data edit program, designed by programmers (modest bow here), would laugh at such a value, but EasyTrieve (or similar) in the hands of a cretin would not. My point was that making databases available to the masses through easy-to-use tools invites ruination, islands disappearing, mothers smelling like elderberries, and the earth turning into one giant scab.
From: Howard Brazee on 7 Jun 2010 11:47 On Mon, 7 Jun 2010 10:28:02 -0500, "HeyBub" <heybub(a)NOSPAMgmail.com> wrote: >My point exactly. EasyTrieve allowed the unwashed to place a value >of -17,000 feet in a database where the standard was the reverse. The >original data edit program, designed by programmers (modest bow here), would >laugh at such a value, but EasyTrieve (or similar) in the hands of a cretin >would not. We are moving from an IDMS system that included some flat files and users who wrote some EasyTrieve to a PeopleSoft/Oracle system with tools such as Cognos (against the data warehouse), and SQLQuery. Getting rid of the mainframe is supposed to pay for most of this - getting rid of application programmers is what will provide the bigger cost savings. It reminds me of when they invented CoBOL and wouldn't need programmers anymore. They really, really need to know the data, and really, really need to check their results, as it is very easy to pull out of the data warehouse data that look pretty good and assume it tells what we need. I don't see that this issue was nearly as important with EasyTrieve when it pulled more simple data, from the database or from frozen files. >My point was that making databases available to the masses through >easy-to-use tools invites ruination, islands disappearing, mothers smelling >like elderberries, and the earth turning into one giant scab. But that's the next guy's problem. -- "In no part of the constitution is more wisdom to be found, than in the clause which confides the question of war or peace to the legislature, and not to the executive department." - James Madison
From: Pete Dashwood on 7 Jun 2010 21:24 HeyBub wrote: > Pete Dashwood wrote: >> >> Software without "sanity checks" (like not allowing negative values >> for something like a hole in the ground) is likely to generate tears. >> Software engineers who don't understand what they are doing are also >> likely to cause much wailing and gnashing of teeth. >> > > My point exactly. EasyTrieve allowed the unwashed to place a value > of -17,000 feet in a database where the standard was the reverse. The > original data edit program, designed by programmers (modest bow > here), would laugh at such a value, but EasyTrieve (or similar) in > the hands of a cretin would not. > > My point was that making databases available to the masses through > easy-to-use tools invites ruination, islands disappearing, mothers > smelling like elderberries, and the earth turning into one giant scab. My point would be that that can only happen if you let it. Making databases available is not the problem. Giving access to data is fine as long as you ensure that the people given the access are answerable for what they do, that you KNOW what they did (audit), and that you (and they...)can UNDO what they did as quickly and easily as they did it. It is when access is given WITHOUT these kinds of checks and balances that wickets get decidedly sticky. The same applies to ANY data repository, whether it is an ISAM/VSAM file only accessible by an application, a Relational DB accessible by SQL, or a Content Addressable store accessible from a number of cores simultaneously. The technology is irrelevant; it is how it is managed that decides whether it is safe or not. So often in this forum I have seen people posting with the idea that an indexed file accessed from COBOL is somehow "safer" than a Relational Database accessed from COBOL, because the RDB can be accessed by "anybody" who knows SQL and has access. It is nonsense. ("Anybody" could also run the COBOL app. that accesses the indexed file, or write their own app to do whatever they liked...) There is no substitute for proper management and stewardship of the data resource. Locking it up just makes it less useful. Pete.
From: Pete Dashwood on 7 Jun 2010 21:43
Howard Brazee wrote: > On Mon, 7 Jun 2010 10:28:02 -0500, "HeyBub" <heybub(a)NOSPAMgmail.com> > wrote: > >> My point exactly. EasyTrieve allowed the unwashed to place a value >> of -17,000 feet in a database where the standard was the reverse. The >> original data edit program, designed by programmers (modest bow >> here), would laugh at such a value, but EasyTrieve (or similar) in >> the hands of a cretin would not. > > We are moving from an IDMS system that included some flat files and > users who wrote some EasyTrieve to a PeopleSoft/Oracle system with > tools such as Cognos (against the data warehouse), and SQLQuery. > Getting rid of the mainframe is supposed to pay for most of this - > getting rid of application programmers is what will provide the bigger > cost savings. > > It reminds me of when they invented CoBOL and wouldn't need > programmers anymore. Yes, when you have been part of it for years and know what is going on, it is hard to accept a lesser solution. But in 5 years time, you'll be off playing golf and the new crew will be familiar with Cognos and Oracle. And some of the projected cost savings may then actually eventuate... :-) > > They really, really need to know the data, and really, really need to > check their results, as it is very easy to pull out of the data > warehouse data that look pretty good and assume it tells what we need. Then BI should be reviewing the warehouse and refining it. User interface should make it easy to flag ambiguous data for special checking. I learned this lesson in my early twenties when it became apparent to me that Management would believe anything written on green lineflo. I deliberately arranged for some outrageous figures to appear in a Management report, and sure enough, all Hell broke loose. I was carpeted and warned that any similar behaviour would result in loss of job. My defence, (to the CEO) was that if the people getting the reports knew their stuff they should have instantly realised it was a joke. (He did, and smiled when they first presented these figures to him). I kept my job and only did one or two more pranks before I moved on, and then grew up and lost my sense of humour... :-) Your point elsewhere is well taken that it is important for people to learn to make sanity checks and estimations that "keep data honest". > > I don't see that this issue was nearly as important with EasyTrieve > when it pulled more simple data, from the database or from frozen > files. I think it was just as important then as it is now, Howard. The difference is in the quality and quantity of data that is available through the warehouse, and the fact that more data available means more care is needed in interpreting it. There are more subtleties now than was previously the case, and that can be a two-edged sword. > >> My point was that making databases available to the masses through >> easy-to-use tools invites ruination, islands disappearing, mothers >> smelling like elderberries, and the earth turning into one giant >> scab. > > But that's the next guy's problem. I contend that it shouldn't be a problem at all, if data resources are administered and managed properly. Pete. -- "I used to write COBOL...now I can do anything." |