PHP A COMPUTER LANGUAGE AGAINST EVERYONE!
Report by Lorenzo Romano January 2021 ©
In reality, the PHP language for computers is not exactly the latest arrival because it was introduced in
1990s and immediately found expansion with the installation of relatively fast telephone channels (9800
bit / s against the 300 bit / s of the previous) but its best performance is found on the sections in
Its architecture is also not new but derives from a paradigm used by none other than
IBM for its 1960s printers.
In fact, IBM downloaded the cpu from the unnecessary time lost for formatting the documents,
delegating to the print server the set of data and commands for the line jump, the font,
possibly the color in graphic programs, etc.
It should be borne in mind that most users refer to shared systems in the report
server – client, for which data processing on a centralized system becomes essential!
PHP expands this concept because it processes messages “in house” (ie on the server) and transforms them into
“Pure” HTML to be sent to printers and / or video on the client side. In this way the PC or the
mobile phone or tablet, it will only have to have an excellent Internet browse.
This has considerable advantages because, for example, while PC memory is almost always
expandable to Gigabyte level with minimal expense, that of other devices is not because
sealed in the factory and if the user needs to install more programs or some
additional program that cannot work in the expansion sim (among other things, generally little
reliable), is forced against his will to discard his current mobile phone to buy one
other with more central memory and obviously higher cost, to the great delight of the producers!
Those who have had experience with the first programming languages such as Fortran, Assembler,
Cobol up to PL1 because Dennis Ritchie’s first C language entered the scene soon after that
together with Ken Thompson (**) with whom he perfected the Bell Unix system passing from
meticulous Assembler to the formidable C, he will find many common things there.
Here, I say this to introduce a fantastic property of PHP: said language is a kind of
fruit salad of various flavors because it can easily mix with other languages, among which most
important are the multidimensional databases: the majestic Oracle and the very light MySQL, I don’t have
experience of interfacing with IBM DB2 whose installation is also accessible on a PC, because I have it
used only on mainframes of classes 370 – 9000.
This great flexibility of PHP is achieved with modest definitions entering and exiting one
subroutine or from a plug-in; for example if the main program is an HTML construct (and in
generally it is), PHP processing is accessed simply by entering parameters and structures
in paragraph limiters <? ?> This also works for access to DBs, just write in the
paragraph the connection string and the affected database can be anywhere.
If you have noticed, in the brief excursus of these lines, we have practically tracked a series of
characteristics that make PHP a language of a very high specialized level without however
remain complicated for those who have to write source programs, perhaps they are a bit tedious
(characteristic of Cobol and Assembler), or as complicated as the routines written in C
case in point in C ++ Object Oriented. I don’t like this last feature at all because it pays off
the corrective and / or evolutionary maintenance of a program is complicated if done after some time
and / or by other programmers. PHP allows you to build self-documented structures, extremely
linear, easy to expand and maintain as in the case of system upgrades
operating and / or communication software. So why complicate your life by getting entangled in
complex routines unless strictly necessary?
The next episode I will talk about security hoping to have thematic contributions from readers!
(**) It is said that Thompson, like all high-level engineers, enjoyed some free time after the delivery of a project and coincidentally he wrote a program for playing chess based on a Bell 32A prototype PC. There Unix was born, the first operating system with a real-time interface. Previously, including Word processors, they processed in batches while with Unix they spoke directly with the CPU thanks to the “round robin” module that assigned priorities to all the procedures pertaining to the kernel in fixed or dynamic polling.
IBM, UNIVAC, RCA, Honeywell and others understood with decades of delay that the realization of RT (real time) programs are much more advantageous and economical than batch processing, not so much for the power of the CPUs used, but for the response times. obtainable in the face of an urgent request for information. In fact, many giants left the scene …
An interesting memory: I developed a conversion program between a database operating on a batch procedure resident on mainframe – with unacceptable times for this reason – and the same one converted into Cobol on VAX 6000, the first between loading and processing of about 1, 5 million lines of General Ledger took almost 5 minutes (the scheduling was sequential with hundreds of other uninterruptible programs), while the little VAX with only 1 Gbyte of central memory and roughly the same load of programs, took a little less 35 seconds. This was possible thanks to the different memory management function capable of mapping the entire Gigabite in the few milliseconds in which the “round robin” now assigned the disk, now the memory to the program concerned!