free implementation? factorial?
in [Cobol]
|
Prev: Data Representation in COBOL
Next: xml acucobol
From: Richard on 13 Jan 2006 14:48 > Certainly some of the other classics, such as tree traversal, aren't > an easy fit for traditional COBOL. It's not hard to implement binary > trees in COBOL, with or without pointers, but it's atypical. What is usually done using trees in C is done using ISAM files in Cobol, which may in fact be indexed by btrees, or by arrays and SEARCH ALL (which may be a binary chop), or with relative files using the record numbers as pointers. The advantage of doing things in files is that these can be shared and can be permanent which is important for business. If you want to introduce Cobol do not teach recursion and btrees but teach the difference between an invoice and a credit note.
From: Howard Brazee on 13 Jan 2006 14:53 On 13 Jan 2006 11:48:16 -0800, "Richard" <riplin(a)Azonic.co.nz> wrote: >What is usually done using trees in C is done using ISAM files in >Cobol, which may in fact be indexed by btrees, or by arrays and SEARCH >ALL (which may be a binary chop), or with relative files using the >record numbers as pointers. Kind of like teaching the difference between sorting an internal table with a version of CoBOL that doesn't have a command for it, and a version of CoBOL that does have a command for it.
From: Michael Wojcik on 15 Jan 2006 11:33 In article <1137181696.726963.129770(a)z14g2000cwz.googlegroups.com>, "Richard" <riplin(a)Azonic.co.nz> writes: > [I wrote:] > > Certainly some of the other classics, such as tree traversal, aren't > > an easy fit for traditional COBOL. It's not hard to implement binary > > trees in COBOL, with or without pointers, but it's atypical. > > What is usually done using trees in C is done using ISAM files in > Cobol, which may in fact be indexed by btrees, or by arrays and SEARCH > ALL (which may be a binary chop), or with relative files using the > record numbers as pointers. Agreed - that's why I wrote that tree data structures are "atypical" for COBOL. Turing-completeness and the like aside, it's clear that programming languages favor certain approaches over others, simply in the syntatic sugar and semantic assistance they provide. The Algol family of languages - such as C, Pascal, and their derivatives - accomodate tree structures with little programmer inconvenience. LISP and its derivatives encourage them; they're the basic data structure provided by the language. On the other hand, there are languages like APL, where the multidimen- sional array is the basic data structure; the implementation may use trees under the covers, but from the programmer's point of view they're not a natural fit. And there's COBOL, where the basic data structures are hierarchical records, arrays, and files. > The advantage of doing things in files is that these can be shared and > can be permanent which is important for business. Obviously there are various tradeoffs between in-memory and on-disk storage in most implementations, though unified virtual memory management and hierarchical storage management are steadily removing many of the distinguishing characteristics. > If you want to > introduce Cobol do not teach recursion and btrees but teach the > difference between an invoice and a credit note. Agreed, but as I noted in my response to Louis, introducing recursion in COBOL is a different subject from introducing COBOL using recursion. (And I certainly wouldn't try to teach B-trees, or B+trees, in COBOL; they're not a natural fit and they're outdated.) -- Michael Wojcik michael.wojcik(a)microfocus.com If Mokona means for us to eat this, I, a gentle person, will become angry! -- Umi (CLAMP & unknown translator), _Magic Knight Rayearth_
From: Rick Smith on 15 Jan 2006 12:45 "Michael Wojcik" <mwojcik(a)newsguy.com> wrote in message news:dqdtg901874(a)news1.newsguy.com... [snip] > (And I certainly wouldn't try to teach B-trees, or B+trees, in COBOL; > they're not a natural fit and they're outdated.) If those trees are *out*, what's *in*?
From: Howard Brazee on 17 Jan 2006 10:05
On Sun, 15 Jan 2006 12:45:43 -0500, "Rick Smith" <ricksmith(a)mfi.net> wrote: >> (And I certainly wouldn't try to teach B-trees, or B+trees, in COBOL; >> they're not a natural fit and they're outdated.) > >If those trees are *out*, what's *in*? I suppose, external databases. |