|
A few years ago I wrote an MQ Series based application and used XML to structure the message 'packet'. The XML tags were created using the 6 character DDS field names. Response time across the network was fine. An ex-colleague of mine is now at a development site using JAVA. The developers there are doing much the same thing, except now they're using the very long and very descriptive JAVA field names as the tag identifiers. The thing runs like a 3 legged dog. The JAVA developers solution? Get a fatter pipe or a faster cpu, or just get more resources. The idea that their use of 40 character field names as tags might be causing an over head of (2 * 40) + 5 (for the brackets and slash) for a field which might in fact be 1 character long was not a consideration. We should remember that a variable name is just a code for a memory address. It's length has to simply be sufficient to distinguish it from other codes. I agree that 6 characters can be a little difficult in a large development, However the shift to 10 and 14 characters is more than suffficient. The ability to have very long field names eventually makes code unreadable (IMHO), especially with poorlu structured free-format coding. In order to maximize the 'meaningfulness' of variable names we use a mnemonics dictionary which provides a 3 character mnemonic for commonly used terms in our application (which happens to be insurance). Developers are required to create variable names using combinations of these mnemonics. This, combined with the PREFIX and QUALIFIED keywords is more than sufficient to uniquely and meaningfully define variable names. -----Original Message----- From: midrange-l-bounces@xxxxxxxxxxxx [mailto:midrange-l-bounces@xxxxxxxxxxxx] On Behalf Of James H H Lampert Sent: Wednesday, 19 October 2005 11:21 AM To: Midrange Systems Technical Discussion Subject: Re: TV Ad (I'm the man) James Rich <james@xxxxxxxxxxx> wrote: > So what if 10^36 is a large number? . . . How many >sensible english descriptive phrases can be made with 10 >characters? Not very many, and that is the problem. The >problem is not running out of unique character >combinations, it is running out of sensible descriptive >contractions. Who wants to type a 50-character-long command? We have something beyond the wildest dreams of the WinDoze, Mac, and Linux world (not to mention the DOS world): DESCRIPTIVE TEXT. We can have up to 50 characters, with no rules whatsoever, to describe what some QSYS.LIB object does. Over a quarter of a century ago, when I was in high school, we had a district-wide student timeshare system. Running on multiplexed modems, we had 5 300-Baud lines and a 150-Baud line on each of 2 phone lines. The system used an obscure OS called MUSIC (McGill University System for Interactive Computing), running on an IBM 370/135. It had six-character filenames. The first character had to be a letter, an @-sign, a $-sign, or a #-sign; the other 5 could also be digits. And everybody's files, including private ones, were in one big common library. We thought 300 baud was fast. We thought 6-character filenames were great, even with the possibility of different users' filenames colliding with each other. An operating system upgrade eventually got us longer filenames and separate private libraries (but still no hierarchical directory structure), and we thought that was downright incredible. Then came RT11, RSTS, TRSDOS, CP/M, and PC/DOS, which all used 8.3 or 8/3 format filenames, and eventually hierarchical directories. And we all thought that was great. Then came the Macintosh. Long filenames. With spaces. Fancy that. And of course, WinDoze had to be a copycat on that. I remember a company I worked for back in the late 1980s. All the DOS machines were strict command-line boxes. And the boss set them all up so that all the common commands had one- or two-character synonyms, so that nobody had to type an 8-character command. You want descriptive text to tell what an object is? Great! Put it in the descriptive text. That's what it's there for! -- JHHL
As an Amazon Associate we earn from qualifying purchases.
This mailing list archive is Copyright 1997-2025 by midrange.com and David Gibbs as a compilation work. Use of the archive is restricted to research of a business or technical nature. Any other uses are prohibited. Full details are available on our policy page. If you have questions about this, please contact [javascript protected email address].
Operating expenses for this site are earned using the Amazon Associate program and Google Adsense.