Posts by Chris Gray 1 290 posts • joined 8 Jul 2009
Yeah, I know - replying to my own comment is tres tacky.
But, in a moment of weakness I went looking, and I found the 8080 assembler for the cache setup. There was another option "cache recover" - not real sure what that did, but I expect it assumed valid cache contents after a system reset/reboot.
Code is nicely commented, but would take considerable effort to really understand.
And yes, it transfered 4 bytes per bank switch.
I'm not certain, but I think my first use of CP/M was on an Exidy Sorcerer (had a Z80) with an external dual floppy box.
Later I spent an absolute fortune for a big S-100 system with dual-processor board, dual floppies, graphics card, disk controller card, system control card, SRAM card, extended DRAM card, and a partridge in a pear tree. Many of the boards were "CompuPro" from Godbout. They included the BIOS source for the hardware.
I wrote a RAM disk system using the 256K expanded memory. The CPU board had a register you could output to which controlled the upper 8 bits of the 24 bit S-100 address as output by the 8085 CPU (the floppy controller board did the full 24 bit). With care, and some "interesting" use of the DMA controller on the disk controller card, I was able to switch back and forth among the 64K "segments", saving a few bytes (4?) in the CPU registers for going between the main bank (bank 0) and a bank of my RAM disk. I created CP/M program "cache", which accepted "on", "off" and "flush". "flush" took all dirty contents from the expanded memory and wrote it to the active floppy. I was quite proud of this setup. Made programming on the system *much* faster, and with far far less "GRONK-CLUNK" going on.
I also recall a local store had a business system for sale running MP/M. It did indeed support multiple simultaneous users. Very expensive system, however.
Double sided! Ooooh - I would have been jealous. But, I did have 2 drives on mine, so it was a lot nicer for some things.
(Monstrous setup - huge S100 box with rack mount handles, dual-drive box on top, then a terminal, printer and 1200 baud modem elsewhere on the desk. Had fun playing with the graphics cards.)
Clunk, clunk, whiirrrr, clunk, clunk.
For technical reasons which are unlikely to ever go away, I want to be able to do GUI things without having any extra stuff in my process address space. The X protocol gives me that, and it sounds like Wayland has no intention of doing so. So, I'm glad things like XWayland exist and hopefully will continue to do so. My Linux is suffiently old that it still has the traditional X server, so I don't have any experience with Wayland.
And yes, I fully understand that using the X protocol makes high intensity graphical games unlikely to work. That's fine.
Apparantly most, if not all, are based on Javascript. So, I don't see them. I do see blank spaces. I have no problem at all with reasonably-sized ads without overmuch animation or glaring colours.
(I don't have any blocking with the ancient browser on my phone, and yes, El Reg does look different there. :-) )
I use Firefox with NoScript. I don't try to block ads, but I block most scripting. I find the parts of the web I care about to be quite usable. Sometimes I will temporarily (which NoScript handles nicely) allow one or two sites for scripting if I really want to see something. If that isn't enough, I simply don't care enough about that content.
I view YouTube more than I should. Ad videos can be annoying yes, but I've gotten used to them. I don't think I've ever had more than two before the main video plays. The small ads at the bottom I often don't notice (my monitor is pivoted to portrait mode).
Why don't I block ads? Simple - I believe in not trying to avoid doing my bit for how the whole web ecosystem works. Just don't try to mess up my system with scripts. Years ago I was hopeful for micropayments (yes, I know about Brave), but that's long gone.
Yikes! How often do you plug/unplug?
My ancient Samsung S4 has micro USB, and it still works just fine after 8 or so years. I don't use the phone overmuch, so only charge every second day (plus occasionally plug into my desktop to download pics/videos from the phone).
Note that the cable is always plugged in to the phone first, so that it is not live - even small sparks from the 5V will slowly corrode contacts.
Cool, thanks. I don't think I'd known about fexecve.
My point on how one should do things, and on scripts being problematic is based on the thought that anything which does something with or to a file in ways which re-lookup the file are vulnerable to timing holes - things can change between the lookups.
I also just found "flock", which allows file locking from scripts. I definitely haven't seen that before. So, it appears the downvotes are valid. Assuming of course that they are from folks who are aware of this stuff, and know that everyone who writes critical scripts uses this stuff....
Linux has far far too many scripts. They are pretty much vulnerable to timing problems by definition.
Seems to me that for anything important, you should not use:
check file properties; open file; trust file data.
but rather should use:
open file; using only the yielded fd, check file properties; using only the yielded fd, trust file data.
but, I'm not sure that all of the properties of interest are available via just the fd.
Mines a D-Link, but not one of the numbers listed.
My thought (and plan) is to always get a device with the fewest possible features (e.g. no externally visible services implemented on the router, like web serving from plugged-in storage). That should reduce the externally exposed bug footprint. And, never, ever enable remote adminstration or management. Only allow management from the local network.
And yes, I'm aware that that last rules out the semi-custom devices provided by some ISPs as "you must use this device". If you can't avoid that, then I suggest a second router between the ISP thing and your local network.
And you think the current crop of AI stuff is "intelligent"? Seriously?!
My point is that I question whether we want to emulate how humans act "intelligently", or try for something that to a greater extent we *know* makes correct choices. Humans often make correct choices, but they often make incorrect choices as well. Can computer systems with more resources do better? Probably not much if we design them to emulate how we believe humans make choices. I want them to do better.
What these "AI" things can do is what we should expect based on what they are. Feed them a mountain of stuff produced by humans and they will eventually be able to respond and act much like humans. Feed them biased stuff, and of course they will be biased.
But, is that what we want? Wouldn't we rather have machine intelligence that is factually and scientifically correct? You get that by specific algorithms and carefully selected reference data. Lots of work, but in theory the world only has to do it once.
GIGO - Garbage In, Garbage Out
I can be called lots of things ( :-) ), but in terms of what I do, I'm a compiler writer. Ignoring my first efforts in a compiler course at University and another project I was hired and paid to do, all of my compilers have been straightforward monolithic programs. No passes, no phases. Read source code and emit optimized object files. (My current project is all one program, but does create a detailed internal representation of the program.)
Makes for fast execution and minimal I/O - very important back in the days of floppy disks.
Also makes for a minimum of total code involved, hence reducing the opportunity for bugs.
Sure, if you have multiple phases/passes you can in theory manually examine the intermediate stuff, but do you really want to? It's a whole other language and set of conventions you now have to understand. Simple, specific test code can nearly always trigger any bugs you are looking for, and following the path of what's going on is far simpler in a monolithic compiler.
As for some of the other points folks have raised...
Yes, you can implement a language with more checking (array bounds, arithmetic overflows, ...) by emitting C source code. The checks are simply there in the source you emit. A good C compiler can likely optimize those into something closely resembling what a monolithic compiler would generate.
Sorry to hear that both gcc and llvm are hard to work with. I vaguely recall friends trying to send me in the direction of using one of them, but I think now I'm thankfull that I didn't.
Early compilers were often split up simply because of memory constraints. With the K & R compiler, one of the things they didn't have to deal with in the compiler proper was branch shortening. Most CPUs have branch instructions with both long and short offsets. You want to use the short-offset forms wherever possible, but its hard to know where the target of a forward branch is, because you haven't generated code for all the stuff between the branch and the target yet. The Unix PDP-11 assembler handled that for them. It also handled the details of emitting valid object files. It wasn't all about not dealing with the target instruction set.
The code generator for X86-64 in my current project is still under 8000 lines of code. There are a few things left - currently I'm doing my 'bits' types. The CPU has single bit extract/insert, but nothing more, so I'll have to special case to use those instructions. I badly need at least a move/move optimizer, and doing that for this CPU will need an abstract representation of the instructions, unfortunately - the binary format is just too complex to do directly. The AMD CPU manuals are large and complete, but as in any such technical endeavour, you have to keep in mind stuff you read a few hundred pages back when figuring out exactly what can happen in encodings, semantics, etc. Sigh.
That's likely true, but is not required. Any language (which may have been bootstrapped via C) that allows the needed lower-level operations can be used to bootstrap from. So, it ought to be possible to bootstrap your way from Modula-2, for example. Hmm. There wasn't a decent C for the IBM mainframes, so my sequence there was AlgolW => QD => QC => ALAI. Don't bother trying to look up any but the first - you won't find anything.
I've been (slowly) reading the Rust doc. I just want an understanding of what it does at the lower levels - haven't got far enough yet. What *is* "borrowing"? The language appears to not have any kind of struct with named elements - perhaps thats added with their macro stuff. They seem to think that declaring a local variable of a given type a few lines after a previous declaration with a different type, thus shadowing the old one, is a *good* thing. WTF???
I don't see any indication of intended humour, so I'll feed the troll...
APL is darn near the highest level of the programming languages from long ago. One character to multiply matrixes. (OT: I worked with Ken Iverson's nephew for a few years.)
I also wrote a RAM disk driver for CP/M. You do that by taking the assembler source to the existing disk driver for your system and modifying it as needed. I don't recall if the assembler source for the rest of the system was available from Digital Research or not.
So, a language is successful if hackers use it?
Hacking is an activity with different requirements from other programming activity, so one would expect the language preferences to differ. Hacking, I expect (not something I've ever been interested in doing) involves quite a bit of file manipulation and command execution, so more "typical" programming languages wouldn't be terribly useful. Perhaps in the past Perl rated higher?
Did you read the article here on El Reg indicating that the fire has been put out, the plant folks had plenty of warning, there never was any serious issue, and there has been no release of radiation? The Russians probably want to shut the plant down to cause problems for Ukraine (apparantly it supplies 1/4 of their power). The simplest way to do that is to cause enough fuss that the Ukraine authorities shut it down themselves.
To quote the El Reg article:
"In other words, the cores Super-Earths need to be cooled to much lower temperatures before they solidify. Their larger-sized cores also mean they lose heat at a slower rate than Earth’s too."
But, the article also says that the higher pressure in super earths makes their iron melting point *higher*, meaning less cooling to reach solidifacation. The first sentence in the above has it backwards. Or am I missing something?
ELF is "Executable and Loadable Format" - its the file format of object files and libraries used on Linux and some other systems. It isn't an ABI. It isn't even an API.
Now, its certainly possible that that latest BeOS stuff does use a system call ABI much like Linux uses (which is distantly based on the SVR4 stuff, I believe), but my gut feel is that that would mean a number of interfaces had to change significantly (callbacks, auxilliary code, etc. done via expected methods and class members).
Er, why do you say that?
I just looked at the "installed files" for Wine here on my Ubuntu Mate system. Two binaries of about a Meg each ("wine64-preloader" and "wine64"). Plus dozens of shared libraries and a whack more things in a "fakedlls" directory, some named <xxx>.exe .
This *is* a quite old distribution.
I read about BeOS when it was first coming out. Sounded all good, until I got to the part where its OS ABI (Applications Binary Interface - a difference beast from an API) was based on the C++ object representations of the specific C++ compiler they used for it. Makes it painful for native code written in any other programming language.
Has that changed at all since then?
"Client SOC Architecture". I wonder what his NDAs are like? Could he be targetted at doing something at Intel that is quite similar to what he did at Apple? I doubt he would be terribly interested in diving into X86-64 stuff (no-one wants to do that!), unless it is at an architecture-independent level. Intel have announced they will be doing SOC's based on ARM and RISC-V, so those, or general stuff, is my guess. (See "Integrated Device Manufacturing", IDM 2.0)
Go to the Lego website and click on "Pick a Brick".
Prices there are high, it doesn't include all in-stock parts, and it doesn't include out-of-stock parts, but it satisifies many needs if you don't want to use Bricklink or BrickOwl.
And of course, if you are missing a part or two from a recent set, just ask nicely and they will send you replacements.
Yep, take a look at bricklink.com
Now owned by Lego, but still quite independent. Hundreds of Lego repackers all over the world. Some specialize, others have huge stocks, etc.
A friend of mine is one such seller. There was a recent Walmart sale of "sets" which are just lots of bricks of various sizes, shapes and colour. He got 50, and they are now here for me to sort the 75,000 parts, for his Bricklink store. It's going to take a fair while - my guess is one of the 11 different bags of pieces per week, roughly. He would do it *much* faster, but I've got computer work to do, and he's got lots of regular sets to sort.
Thanks - I tried. In my upper-right corner I just get a link to Git for the doc sources. If I try the browser's print function, I only get the one page, as expected.
So, I went and got my old Kindle (the original "Kindle Keyboard"). The web browser is under "Experimental", but it had bookmarks for Wikipedia, Google, etc. None work. The only pages I could load with it are some very old straight HTTP pages. Errors it shows suggest that it can't do HTTPS. Oh well, thanks anyway.
Sorry, I forgot that I had gone there. I don't live in a browser - my eyes weird out with a white background (like I see now). I want something I can put on my Kindle and read in an easy chair. I guess I *could* try the Kindle's ancient web browser, but I'm not holding my breath that that would work. I was looking for a PDF (likely a .docx would be *huge*).
I'm a language developer and compiler writer (both professional and as a hobby). Strange, I know. Anyway, I'd like to get good documentation of the Rust language (and maybe standard libraries, and some tutorial stuff). From what I've seen, the official way to do that seems to be to install the whole Rust distribution. I won't do that, except perhaps in a virtual machine, since my current language/compiler project needs my current setup. I found one free book for my Amazon Kindle, but it is very bad (not proof-read, mostly cut-n-paste, too much on crates, etc.) Anyone know where to find proper docs? Or, can they be legally extracted from the Rust distribution?
Given that the vortices now appear to be quite deep, do we really know that they aren't exactly like the ones that form in our bathtubs as they drain? Would a relatively small loss of atmosphere be noticeable? Has anyone tracked Jupiter's mass over time? Do we really know that there aren't a bunch of locked-open stargates down there?
The Register - Independent news and views for the tech community. Part of Situation Publishing
Biting the hand that feeds IT © 1998–2022
