"The best of the current platform is EC. Joey still said lazily, even if he was in a meeting right now, "We can develop on it, all the conditions are met, and I think the workstation will be very profitable." ”

Nonsense, this is still hard to say? Long scolding, although EC is good, but that is only prevalent in Europe, now he is afraid of how to compete with Microsoft's indos, preferably suppressed in infancy.

Indos has been familiar with it for a long time, and he also knows that the original version 1.0 of this thing was released in November 1985, and these became exam questions that year, which is extremely impressive.

It's just that I can't figure it out for a long time, and even 386 was successfully developed in 85 years, where does the hardware resources of the PC platform come from to run this thing?

Later, by chance, I observed this system for a long time, and then I discovered the mystery, first of all, this thing is a fake graphical interface, but it is a dos plug-in, the window can not be enlarged or reduced, can not be stacked and displayed, can only show two windows on one side of the screen, no file structure.

However, because the machine used in the experiment is too high compared with 286, I really can't see the running speed of this thing, so I have been a little skeptical about the practicality of this thing for a long time, but looking at the advertisement that Ballmer made on TV at that time, a set of in1.0 only sold 99 US dollars, and it was all of a sudden.

Something is not real does not mean that no one buys, but it represents a posture and a height, I have this product, so I have given up for a long time to engage in a set of products on the EC and then wait for 386 to appear and then enter the PC graphics 'operation' system of planning, although martial arts have no first move, but the business war is like chess, and it is unimaginable to lose the first move.

"So can we understand that the graphics subsystem of the EC workstation system we are developing now is transplanted to the current AT, and all the graphics display tasks are completed by the graphics card, which liberates the CPU from heavy graphics computing, which should be supported. "I have put forward my own idea for a long time, the graphics card, there is no such concept now, the AT machine is used in the MDA display standard, which can only support the resolution of 640×480, or characters.

Joey's eyes narrowed, and he thought about it for a moment, "I don't know why you're so interested in making a graphical interface on a PC, but I have to warn you that doing so involves a lot of standard changes, is risky, and the development time is unpredictable." Moreover, the development of graphics workstations is at a critical juncture, and distraction is not the best solution. ”

Joey said very politely, because he also knows Changchang very well, since he entered the research institute in charge of the system, he and Changchang almost every day 'exchange', this godson version 3.2 is the result of their joint 'exchange'.

Unlike in the past, Godson has not moved much since it was upgraded to version 3.0 half a year ago, and version 3.2 has been successfully developed after a lot of manpower and material resources. Based on the naming mechanism, most people may think that this is only developed on the basis of version 3.0, but in fact, version 3.2 has a leap in function compared to version 3.0.

For example, a new, more robust file system and disk quota mechanism, a new signaling mechanism, a new hierarchical modular system structure and a separate I/O system, and new documentation – these are all big moves, and they are still compromised by the pressure to release (IBM claims that godson is outdated), or else be prepared to make a move on the 'exchange' intercommunication mechanism and the virtual memory management module.

All these new changes have always been only possible on mainframes and minicomputers, and it is unthinkable that the combination of Changjiu and Joey can solve this impossible task.

The new system is available in several versions, the most important being 16-bit and 32-bit.

There are considerations for making this classification, and Godon has almost become synonymous with the microcomputer 'operation' system, although the sales are large, but the relative profit is not as good as the minicomputer system, which is popular with UNIX and its variants.

Joey is familiar with this, and he himself is the founder of the most famous unix variant, BSD. Microcomputers have only been popular for a few years, and now most of them are still used by government agencies, large companies, and research institutions, and basically PDP and VAX are the main ones.

Both the PDP series and the VAX series are DEC products, which have been sold well for 20 years, and the words "" and "Wax" may still burst out of the mouths of the older generation of professors, indicating the great influence of the company.

PDP is a 16-bit series, the main thing is that it is cheap, but it is already old, and even the 32-bit VAX minicomputer that monopolized the market was 7 years old when it was 85 years old.

However, there are still a considerable number of PDP-11s in service, and this kind of machine generally runs unix, but you can imagine how poor the functionality is, after all, the hardware conditions are there - 64K memory addressing ability, 512k disk space, and that's it.

As one of the most widely circulated computers, the PDP series is a pity, but it is useless without the support of the 'operation' system, the latest unix and BSD both support 32-bit VAX, and no one wants to work the 16-bit system.

This has allowed a long time to see a little business opportunity, the professional computer's 'operation' system is valuable, this aspect VS is not doing well, and I don't even want to develop in this area, because in front of them lies UNIX, the most powerful 'operation' system in history.

Others don't do 16-bit systems, but VS has a unique advantage, godson monopolizes the 'operation' system of the PC, itself is a 16-bit system, this time in order to develop workstations, specially added a lot of functions.

Joey also knows what he's been thinking for a long time, making inroads into minicomputers and getting into the high-profit market.

Of course, godson version 3.2 is impossible to install directly on PDP-11, and the hardware resources are not enough, but I have been half-jokingly betting with the programmers in the group for a long time to see if I can transplant version 3.2 of godson to this kind of old machine.

The other members of the group looked at the boss as if they were looking at a monster, this kind of thing was tantamount to talking in a dream, thinking that they wanted to change their way to treat them for a long time, so they bet one after another.

There is no such fun thing without Joey, he volunteered to join this extreme challenge, so in their spare time, the two worked together seamlessly, and actually succeeded, not only running normally, but also running 'well' happily, which surprised the group members, and each lost a big meal.

No wonder these guys don't believe it, this kind of task is similar to running indosNT on 286, isn't it very prick 'excited'?

This kind of game is not simple,After the success of the long and Joey named MiNi-godson,Its functions are complete and the latest 3.2 system is no different,Once launched, it is quite popular with those PDP-11 users,This version has not been replaced and improved,Eventually accompanied by PDP-11 to the end of its life,It is said that now in some corners of the world,You can still find PDP-11 running MiNi-godson system,If you are interested, you can look for it。

The 16-bit system is an experiment, but the real highlight is the 32-bit godson 3.2, which has been tested and not made public (EC version 3.0), which has been designed for graphics workstations and minicomputers for a long time.

The reason for the delay in putting it into practice is that there is no progress on MIT's X-indo program, and secondly, it is to consider how to make this thing overcome the platform barrier and be able to support both the k32 series and Intel's yet-to-be-released 386.

Remembering that Professor Su and MIT have been working together for a long time, he even regrets that he has 'handed over' this very important work to others, and the progress is completely out of control.

So at Joey's suggestion, Chang Chang was angry and bought all the undeveloped Dongdong use rights, and recruited troops, and integrated this plan into the research and development of graphics workstations, which solved the worries.

The other is cross-platform operation, which is more of a headache. It stands to reason that Godon was designed to be more portable 'sex', but after so many version changes, there are many people who handle it, and in order to be compatible with 'sex', the system has become more complex.

It can be said that the godson on the PC is no longer the simple file system that used to be, so it has to undergo major surgery, and both Chang and Joey prefer to do it once and for all.

In order to achieve this characteristic, Joy personally 'operated' the knife and divided the core of the system into two parts, one dependent on hardware and one independent of hardware, which seemed to have returned to the time when it was "refined" and simplified. In this regard, Joey and Chang 'spent' a lot of time.

So far, the two have worked together quite happily, at least their views on the problem are a tacit mess, and they only disagree when they get the graphical interface.

As always, Joey naturally tends to make the graphical interface a program that can be opened and closed at any time, just based on a high-privilege shell on top of the core, which reduces the complexity of the system, at least not because of the instability of the graphical system and affects the operation of other programs.

The disadvantage of this is that the requirements for system resources are a little higher, which is not what the current microcomputer can provide, of course, this consideration makes sense, after all, Joey's understanding is that they are making a system for a high-performance computer, not a microcomputer.

Maybe it's the powerful influence of Microsoft in my memory.,For a long time, there is such a way of thinking.,The interface of the computer as the most intuitive means of 'communication' with people,It's not unattainable.,The audience should be as large as possible.,Don't you see Mac can't do indos in the end?

For a long time, it is believed that the consumption of system resources should be reduced as much as possible, so that it is necessary to start from the bottom of the system with graphics, although the graphical interface does not necessarily have to be implemented on 286, at least when 386 comes out, it must enter the practical stage, so that the graphical interface 'operation' system can survive with an incomparably large customer base.

Neither of them is a fuel-efficient lamp, they have their own opinions, and even use various equipment to simulate the operation of the two ideas, and the result is obvious, if the hardware 'sex' can be high enough, Joey's method is once and for all, and on the current mainstream microcomputer, the long-term method is easier to achieve.

But in the end, he compromised for a long time, and he himself knew that the graphical interface was only a means and not an end, so in order not to be affected by the endless bugs in the future, he still let Joey preside over the work.

The configuration of the graphics workstation currently designed by the institute is extremely powerful, and it can support dual k32 processors to dynamically share storage while trying to reduce costs, and an independent graphics bus is also designed on the basis of the EC machine, a 2D graphics acceleration card designed and manufactured independently, a minimum of 2 trillion memory, and high-speed disks.

At that time, these hardware were fantastic configurations, and the main reason for the unprecedented unity of the hundreds of people in the institute was also based on this - to build the world's most compact and best computer.

Xu Zhixing has always been very vigilant about the psychological trends of the personnel of the research institute, guarding his territory like a mother beast, what makes him puzzled is that although the personnel of the research institute are five 'flowers' and eight 'doors', they are combined together with various thoughts, but there is no unstable military spirit, and no one is even poached by other companies, even if the treatment is higher.

There was no vague answer to this question until the end, that is, these guys are a bunch of technology maniacs, and the configuration of graphics workstations alone has greatly satisfied their psychology, and the most representative 'sexual' remarks are none other than the words of Fisnel, who claims to be the number one hacker: "What is IBM, they haven't released mainframes since five years ago, the technology can be imagined, and besides, I don't want to work in that blue 'color' dog skin, which doesn't fit my aesthetics." As for Apple, I don't like Jobs' arrogance, that's all. Only the EC platform is my choice, of course it's not strong enough, but it doesn't matter, I'll make it the first in my hands. ”

With this in mind, these guys worked together to accomplish their dreams at a rapid pace.

It is not difficult for dual processors to work in parallel, but rather the tradition of the K-series, and this technology has almost become a signature. As for the independent graphics bus, it is a long-term proposal, although it increases the cost, but the interpretation of the supremacy of 'sexuality' is unparalleled.

2 trillion memory is more troublesome, in view of the process conditions of this era, the memory chip has to be soldered on a large board, and then it can be 'plugged in' on the system motherboard, because there are too many chips needed, and it is said that it takes 0.25 square meters of circuit board double-sided soldering to accommodate so many memory modules.

As for hard drives, these guys didn't hesitate to choose the SCSI interface, the reason is that it's fast.

Sometimes I even suspect that these guys are not designing workstations, but building minicomputers, you must know that even the mainstream VAX minicomputers are only equipped with 1 trillion memory, and some of the 'sexuality' capabilities are not as good as the single-processor EC machines.

It is really with complete hardware support that Joey is able to implement various technologies on the "operation" system, including memory paging, projection, and support for various commercial "sex" software.

The above work is progressing very quickly, because they are all mature technologies, which are equivalent to building blocks. The real difficulty is the design and manufacture of the graphics acceleration card, and the corresponding graphics software is written, after all, no matter how high the machine's "performance" is, no application software is just a pile of scrap metal.