Before beginning to discuss the types of diskettes, hard disks, portable devices of massive storage and the others, we are going to explain some concepts that will appear in the explanation of all these devices.

As they were saying in the Godzilla publicity, "the size matters". Apart from the durability, the portability, the reliability and other more or less esoteric topics, when we look for a storage device what more matters for us generally is his capacity.

In computer science, every character (every letter, number or punctuation mark) usually occupies what is named a byte (that in turn is composed of bits, generally 8). This way, when we say that a text file occupies 4.000 bytes we mean that it contains the equivalent to 4.000 letters (between 2 and 3 text pages without format).

Of course, the byte is a very small information unit, by what his multiple ones are used: kilobyte (Kb), megabyte (MB), gigabyte (GB)... Because the computer science usually uses potency of 2 instead of potency of 10, there happens the curious circumstance of which each of these multiple is not 1.000 times major than the previous one, but 1.024 times (2 risen up to 10 = 1.024). Therefore, we have that:

The letters strip, undoubtedly! It is necessary to bear in mind that often instead of 1.024 1.000 are used, for example to do that a hard disk seems a little major than it is in fact, we say of 540 MB instead of 528 MB (taking 1 MB as 1.000 Kb, instead of 1.024 Kb).

Clear it is that not all sound letters; for example, a graphic file of 800x600 points in "real color" (up to 16,7 million colors) occupies 1,37 MB (I motivate by which there are used methods of compression like JPEG, GIF, PCX, TIFF); an operating system as Windows 95 can occupy installed more than 100 MB; 74 sound minutes with digital quality occupy 650 MB; etc., etc.

The speed of a device of storage is not the only parameter; rather it is like a car, with his speed top, average speed, acceleration from 1 to 100 and up to braking time.

The speed that usually appears in the announcements is the speed top or to blasts, which it is usually the biggest of all. For example, when it is said that a hard disk comes to 10 MB/s, it is said that theoretically, in the best conditions and during a brevísimo moment is capable of transmitting 10 megabytes per second. And nevertheless, it is possible that we never manage to come to this number.

The speed that we must be interested in is the average or supported speed; that is to say, that one that it can support of more or less constant form during valuable time spaces. For example, for a hard disk a number of 5 MB/s can be very acceptable, very far of theoretical 16,6 MB/s of the way PIO-4 or 33,3 MB/s del UltraDMA on that so much there please to comment the manufacturers, skylight.

And finally we have the average access time. It is a question of the time that on average takes the device in answering to an information request because it must begin moving his pieces, turning from the rest if it is that it turns and looking for the requested fact. In this case it measures itself in milliseconds (ms), and since it is a question of a time of waiting, lost time, the less be better. For example, a hard disk has minor 25 ms times, while a CD - ROM can overcome 150 ms. Also one speaks sometimes about the maximum time of access, which is usually like the double of the average time.

We are not going to explain here the physical theories on which each of these technologies is based, I have done it and do not believe that it was neither entertaining at all or useful for the majority; we are going to explain rather the practical peculiar characteristics of each of them.

The magnetic technology for information storage takes using for tens years, both in the digital field and in the analogical one. It consists of the application of magnetic fields to certain materials which particles react to this influence, generally being faced in a few certain positions that they preserve after stopping applying the magnetic field. These positions represent the information, good be a song of the Beatles or the bits that form an image or the last balance of the company.

Magnetic devices exist infinity; from the cassettes or the ancient tapes of music up to the modern Zip and Jaz, happening for disk drives, hard disks and similar others. They all look alike in being recording devices simultaneously that readership, in his relatively low price for MB (what stems of being very experienced technologies) and in what they are quite delicate.

They are affected by the high and low temperatures, the moisture, the blows and especially the magnetic fields; if he wants to erase safely a few diskettes, put them on a loudspeaker connected inside a car to the sun and drop them to a puddle a pair of times. And if they survive, buy actions of the company that has made them.

The optical storage technology for laser is more recent enough. His first massive commercial application was the supersuccessful CD of music, which dates back to beginning of the decade of 1.980. The technical essentials that are used are relatively simple to understand: a bundle laser is reading (or he writing) microscopic holes in the surface of a disc of plastic material, covered in turn by a transparent layer for his protection of the dust.

Really, the method is very similar to used on the ancient vinyl discs, except because the information is kept in digital format (some and zeros like vales and summits in the surface of the CD) instead of analogical and for using a laser as reader. The system has not experienced important changes up to the appearance of the DVD, which just has changed the wavelength of the laser, reduced the size of the holes and tightly the ruts so that more information fits in the same space; we go, the same method that we all use to be able to put the whole clothes in the only suitcase when we go away of trip...

The principal characteristic of the optical devices is his reliability. The magnetic fields do not affect them, scarcely neither the moisture nor the heat affect them and they can bear important blows (whenever his surface is protected). His problems take root in the relative difficulty that supposes creating recording devices at a reasonable price, a speed as so high as that of some magnetic devices and in that they need a certain care opposite to the dust and in general any imperfection in his surface, therefore it is very advisable that they have protective case. Anyway, a CD is much more probable that it survives a wash that a diskette, but better not to have to prove it.

Acronym of Small Computer Systems Interface and read "escasi", although it looks like a lie. Many people have heard to speak about these initials and in general they associate them with expensive computers or from mark and with a high yield, but not many know the porqué of the advantage of this technology opposite to others like EIDE.

The technology SCSI (or technologies, since there exists multitude of variants of the same one) offers, in effect, a valuation of very high transference of information between the computer and the device SCSI (a hard disk, for example). But although this is a very valuable quality, it is not the most important thing; the principal SCSI virtue is that the above mentioned speed stays almost constant at all times without the microprocessor realizing scarcely work.

This is of cardinal importance in long and complex processes in which we cannot have the computer blocked while it files the information, as for example in the edition of video, the achievement of copies of CD or in general in any operation of storage of information at big speed, proper "professional" tasks of computers of certain potency and quality as the servants of network.

The different variants of the norm are:

The types of SCSI of 8 bits allow up to 7 devices and usually use cables of 50 pines, while the SCSI of 16 bits or Wide, "breadth" in English, can have up to 15 devices and use cables of 68 pines. The denomination "SCSI-3" is used of ambiguous form, generally referring to the guy Ultra SCSI of 8 bits, although sometimes also it is used for the Ultra SCSI of 16 bits (or "UltraWide SCSI") and Ultra-2.

The checkers modern SCSI are usually compatible with the ancient norms, for example offering connectors of 50 pines along with the most modern of 68, as well as external connectors (generally very compact, of 36 pines), except in some special models that are included by devices SCSI that are designed only to control this device in particular, what lows the price of his cost.

The devices SCSI must go identified with the only number in the chain, which is selected by means of a series of jumpers or a revolving wheel in the device. At present some devices realize this task automatically if the checker supports this characteristic, what brings something more over to us to such a wished and elusive Plug and Play, "to plug and I list".

It must be born in mind that the SCSI advantages are not offered free, of course; the devices SCSI are more expensive than the equivalents with interface EIDE or parallel and also we will need a card checker SCSI to handle them, since only the most advanced basic badges and of mark they include a checker integrated SCSI. If you are thinking of buying a new computer or a motherboard, think if he would not be worth acquiring a motherboard that incorporates it therefore he could need in the future...

In the world PC the most widespread file system is the FAT16 of the versions of TWO Superiors to 3 and of the Windows 95 original, used on the diskettes and most of the hard disks. The VFAT (Virtual FAT) of Windows 95 that allows long names is only a patch on this system of file, not a system of file in himself.

Another system in rapid extension is the FAT32 of Windows 98, ME and of the version OSR-2 of Windows 95 (her "4.00.950B", as is identified herself in the system icon of the Control panel). The advantages of this system of file opposite to the previous one take root in that it is 32 bits and has a size of very small cluster, what it makes capable of admitting big hard disks and of making use of them very well to him, in addition to not needing devices as VFAT to use long names of file.

Let's go on parts; first, the clusters; there are like "drawers" in which the hard disk is divided, in which the files keep. There happens the peculiarity of which a cluster cannot be shared by two different files, by what if we have a size of cluster of 16 Kb and want to keep a file that occupies 17 Kb, will be distributed in two clusters, occupying one point and only 1 Kb of other; the rest (15 Kb) will be wasted. Yes, he has read well; we will throw 47 % of the space!!. And this it is not at all, it will see already.

The same happens if we want to store a file that only 1 byte occupies; if the cluster is 16 Kb (16.384 bytes), they will be wasted completely 16.383 bytes: 99,99 % of the space!! Since it will understand, under these circumstances it turns out to be very important to support the size of the cluster the possible minor thing to minimize the losses that cause these files, especially the very small ones. Observe the table next that relates the size of the divisions (next we will tell what they are) with the size of the cluster in FAT16 and in FAT32:

As for the size of the discs, it is not difficult to understand; if the file system gives directions of file of 16 bits, this gives us 2 risen up to 16 = 65.536 directions, which to a maximum of 32 Kb for cluster are 2.097.152 Kb, that is to say, 2 GB as maximum for FAT16. Does this want to say that we cannot use discs of any more than 2 GB? No, fortunately; but it implies that we will have to divide them in several divisions, which are each of the logical divisions (that not physical) of a disc, who manage as if there were separated hard disks (with your own unit letter and even with different types of system of file if we wish it). For example, a disc of 3,5 GB must divide at least in two divisions of 2 GB or less each one to use it with FAT16.

For FAT32 the calculation is similar, although they are not used 32 bits, but "only 28, what gives a maximum of: 2.048 GB for division!! (2 Terabytes) using clusters of 8 Kb. Undoubtedly we will not need to do any more than one division to the disc...

Observe that to support the same size of cluster of 4 Kb on a disc of 2 GB, in FAT16 we would need at least 8 divisions of as very much 255,9 MB, while in FAT32 one would be enough to us. Undoubtedly, although we could not install FAT32 it turns out to be preferable to lose something of space to have to handle a disc subdivided in units "C", "D", "E", "F"... and like that up to "J".

To end, three considerations: first, on having happened of FAT16 to FAT32, the space profit is enormous, several hundreds of MB on a disc of a pair of GB, and in my opinion this is the best (and the almost only) advantage of Windows 98 opposite to Windows 95 (not opposite to the version OSR-2, which already has support for FAT32).

Second, both systems are incompatible at level of utilities of disc. Be read Norton, the utilities of defragmentación (certainly, defragmentar is to organize a little all these pieces of file that walk dispersed in tens clusters separated on the hard disk), the compressors of disc and the others. If he installs FAT32, get rid of his ancient software for this business or the trouble can be enormous.

And third, are neither the only systems of the file, nor much less best. In case of the FAT16 I suppose that one was already waiting for it, but the fact is that the FAT32 is not a marvel either; for example, he lacks implicit safety characteristics in the system of file (like access restricted to certain users) or auto-compression of the files, characteristics that yes take more advanced systems as those of Unix and Linux, that of 32 bits of OS/2 (HPFS) and that of 32 bits of the very same Windows NT (NTFS). Why did Microsoft invent the FAT32 already having the very efficient NTFS? Mysteries of the life, friendly...