I have been reading up on literacy. One fascinating point is that written language (at least English) is many times richer (in vocabulary and structure) when compared to spoken language. As heard on (for example) television.
If this holds true for other languages, it may be easier to translate spoken material.
I built a "mythtv" box with a SD tuner to capture cable broadcasts. Works fine, but, of course, only SD.
I have a "HD Ready" 42" TV, with DVD. Plays DVDs great, but *I* can't get an upscaling DVD player (component in only). I won't be able to use "HDTV" boxes, and next generation HD DVDs either.
All because the anal retentives won't let me. They are afraid that I might "rip off" the signal. I am NOT buying another TV (for the next 5 to 10 years). Also, archiving or recoding 36Mbit streams does not appeal to me, anyway. Too time-consuming, and life is too short.
Why not let me use the equipment I have purchased? Must be fear -- fear that little old me is going to do some KICK-ASS ripping. Maybe I SHOULD do the ripping, after all I am already "doing the time".
How do you figure that? Let's take this a step at a time.
1 - Unless the chip implements "reversible" computations, ALL power must be dissipated as heat.
2 - N (65, 100, whatever) Watts will generate that much heat. Or other radiant energy (but your chip doesn't glow, does it?)
3 - So, your claim is that an AMD can deliver a computation at 1/4 (lets take the MOST conservative figure you quoted) the power. In other words, an AMD can deliver 100 Intel Watts of computation at 25 Watts.
Scaling this to more understandable terms: an AMD processor that dissipates around 25 watts is a K6-2 450. An Intel process that dissipates around 100 watts is a 3.4Gz Prescott.
4 - Last time I checked, that Intel processor was somewhat more that 4 times the performance of that AMD processor.
Care to back up your claim? Yes, an AMD processor may be more power efficient, but 4 to 5 TIMES more? I call bullshit.
I cannot enter such a "license". Copyright would be the closest. There is a convention that says if copying is needed to make use of a program, that copy is permitted.
For instance, a program must be copied from a disk into memory to execute it.
A "map" is only of use with a "server", so a copy is normal if using the product.
I am also allowed to rent, resell, and otherwise exploit works commercially. Again, this right is passed to me via Copyright.
I can also modify a work and resell it. Really.
I would LOVE a case like this to get into court -- one time it happened, the court ruled that an unauthorized person EXECUTING a program was a violation. I would also LOVE for shrink-wrap licenses to gain FULL power.
IP Police: STEP AWAY FROM THE KEYBOARD! YOU HAVE NOT BEEN DIRECTLY AUTHORIZED TO USE THAT SOFTWARE, AND ARE IN VIOLATION OF FEDERAL IP LAW!
But we aren't there. Not by a long shot. This (USA and Canada) is not a Facist society, and you do have rights.
Game emu is a "standard" option for MythTV. Shoutcast et al. are not there. Could be added into the framework, although I am not interested in doing that personally.
As to Guide/Recording setup. I will have to look into MCE. MythTV does "dump" live feed recording on channel change, which is the one thing I would want altered (although not sure how yet - maybe the insertion of an image? Also requires modification of the live TV buffering strategy). I'll have to play with MythTV with remote only. As it is, I have a wireless keyboard to navigate and enter data into search dialogs and such.
My hot list for MythTV:
- easy addition of hot buttons (some kind of wizard) - mods to live tv buffering - easier access (scripting) to external video software with a howto - incorporation of alternate environments (eg. games running under VMware) - possibly (if not there) dictionary smart entry via number pad (ala cellphone).
Just as a guess, from Gibsons explanation, the bug is the following sequence:
1 - The SetAbortProc is entered in the metafile. It is interpreted, but (and I have never seen the metafile code) instead of having arguments interpreted, *just* the operational code is.
2 - The metafile interpreter dispatches to the abort proc handler, which initializes the existence of the abort proc. It then attempts to scan the argument.
3 - Most likely, the argument scan routine picks up on the fact that the length is wrong, and no such argument exists. Note that the developer probably thought "Gee, this handler only takes a single arg, and the arg retriever checks it, so don't worry".
4 - The argument retriever indeed picks up on the fact that the argument is wrong, and THIS is sufficient to "abort" the metafile. Of course the existence of a handler has already been registered...
5 - What is the default address for the handler? That is probably set as file relative 0 (probably for other reasons).
I don't know about MCE - requisite Windows won't install properly on our designated PVR box. However, MythTV works fine. A couple of observations. MythTV uses XML formats to build its on screen displays, and has hooks for displaying external data as well. So the artwork and menus can be changed suitably. I assume you knew this, and therefore are referring to UI functionality, and not "slickness". Because I can't review MCE, I wouldn't mind a run-down on the features that are slicker.
Linux runs on ALL those platforms (Intel, AMD, etc). And more (Alpha, IBM Mainframe, Sparc, etc.). There is really no comparision.
Consider:
Linux supports more legacy hardware with the OS core.
Consider:
Linux vendors typically support 4+ GB of object code with a typical installation.
Not that I care one way or the other about Linux/Windows comparisions, but this should give you something to think about.
The obvious conclusion to draw? That the Open Source model is SO SUPERIOR to Microsofts, that there is really no direct comparision possible. Of course, that is probably wrong. Another possibility is that Microsoft really doesn't care. Which makes sense (ob disc: I *am* a Microsoft shareholder) to me -- their goal is to maximize profit. And if the OSS model where THAT much superior, Microsoft should have adopted something like the Redhat model years ago.
Of course the problem is that Microsoft installations are more vulnerable to quiet "black hat" attacks. The general rule is to assume that if YOU found a vuln, someone else has already found it. So it really doesn't make sense to keep quiet about them, and, as a user, there seems to be some kind of speed-up on the fixes.
I don't like this a a Microsoft shareholder; more resources have to be spent on fixing 'sploits, and less on actually making money.
My experience (not that it is typical, I would *hope* it isn't):
1 - I want to make a PVR (personal video recorder). Aquire the following components: (A) AMD 1700+ mainboard, 2 PCI slots, 1 AGP slot (B) NVIDIA 5200 graphics card with s-video out, (C) Mercury TV tuner (D) 256MB RAM (E) Memorex DVD burner.
Note: Choice of components is for price. Noted that mainboard documentation states that WINDOWS XP is needed for "USB 2.0 Function". Borrow a copy of WINDOWS XP for initial installation (going to spring the $140 CDN the next day IFF it works).
Assemble hardware. Install WINDOWS XP. After initial boot, note that the optical drive it installed from IS NOT PRESENT. Of course, no network access either (given that WINDOWS XP doesn't recognize the onboard ethernet. Apparently (according to a local Windows guru) I need the drivers... PS. Installed XP a second time, to be sure that I was not hallucinating.
Installed Windows 98SE (for which I had a license). And there you go. Seems to work. Except that when the recommeded drivers are installed, the optical drives vanish yet again. Weird. And, I can't get the network going. But, able to put the contents of the motherboard CD onto the hard disk, to try XP again... Installed XP again (really, I know this is a dead horse, but I can't help myself), and loaded the drivers, and: TADA! it still doesn't work.
Give up on Windows XP. Installed Fedora Core 2. Network, USB 2.0, sound, works. Just works. Added the NVIDIA driver, and the Mercury tuner. Install MythTV - and I have a PVR. WAF (wife acceptance factor) is 6 -- because it isn't Windows, and doesn't run games on the big-screen.
Back to Windows 98. Try the Mercury tuner PVR application. Doesn't work (later, Windows guru tells me that one of the files that SHOULD have been "expanded" from.EX_ to.EXE wasn't, don't know why - but this comes back later). Increase memory from 256MB to 1.2GB to accomodate Apache and some other stuff.
Windows 98SE no longer boots. After some web searching, find out I need to reduce "physical pages" because it crashes with that much RAM. Mercury PVR application *still* won't run. Oh well. Try some of the games (freebies, mostly from cereal boxes). Sound is marginal - sometimes works, sometimes doesn't. For instance, Atari Classic Games collection - no sound, messed up main-screen. Individual games, work, but no sound. Works PERFECTLY on the kids Pentium 166 with Soundblaster, and on our Compaq Deskpro Pentium II 400 with 128MB (Windows 98SE). Why? Maybe the sound drivers?
Anyway, that's my latest Windows experience. Go figure.
And I *still* don't have Windows XP at home -- not from lack of trying, mind you. It just won't run on what I have. Yes, the above is a VERY NEGATIVE Windows experience. Namely, IT JUST DOESN'T WORK. I guess I *could* buy a brand new box, but that just isn't going to happen.
"but I'd like to get back to coding on *at least* a recreational basis." (emphasis mine)
That would be *at least*. The course of study described would result in that. Someone could ask "I'd like to get into auto mechanics, at least on a recreational basis". And what would you suggest. How about novel writing? Oil painting? Again, emphasis on "at least".
If someone wants to "code on a recreational basis", I would suggest a different course. The "at least" suggests profesional aspirations. Which is why I suggested the given course, and attendent timelines.
Pretensious goof? I may be that. I also tried to answer the question.
That snippet is endian dependent. The argument is widened to an "int", and the address taken is for a byte pointer. Result - works on some architectures, and traps on others.
So you "coded" for 5 years, and then went "techie-upstream" for 20...
In a nutshell, you need to code for another 10 years. It takes around 15 years to build reasonable proficiency and skill.
Back to the "salt mines" for you.
1 - 20 year ago, C just started gaining commercial acceptance. Work on your C skills for a couple for years. Study old Unix source. If you can look at this code, and tell what is wrong, you are well on your way (and, yes, I know it is K&R): f(c) char c; { char *s = f2(s); }
2 - Learn Scheme. Make use of the online MIT resources. Maybe move to Common Lisp.
3 - Begin study of ANOTHER of the current crop of interpretive or "immediate" programming systems - Perl, Python, Ruby. Don't let the Scheme go, though.
4 - Tackle Smalltalk - its one of the (or the most) productive systems out there. Squeak would do.
5 - UML, XML, and the current crop of buzz.
This would give you a fair grounding. Time? 2 or 3 years C, 3 years Smalltalk, 2 years Scheme. That would get you "up to speed" as a proficient programmer in about 8 years.
Gnome is based on Win9X? That would be a BIT strange. Gnome/uses/ X11, which is in turn based on older technology, Win9X isn't even slightly in that roadmap. Gnome is/based/ on CORBA.
Please read the ellided GPL pieces below (or read a copy of the GPL). We do not mention what constitutes a derived work here, but the rights and obligations are clear. You would be wrong -- source rights extend to any 3rd party, or include full source which is then again distributable.
Ratboy
" if you distribute copies of such a program, whether gratis or for a fee, you must give the recipients all the rights that you have. You must make sure that they, too, receive or can get the source code. And you must show them these terms so they know their rights."
" You must cause any work that you distribute or publish, that in whole or in part contains or is derived from the Program or any part thereof, to be licensed as a whole at no charge to all third parties under the terms of this License."
" You may copy and distribute the Program (or a work based on it, under Section 2) in object code or executable form under the terms of Sections 1 and 2 above provided that you also do one of the following:
a) Accompany it with the complete corresponding machine-readable source code, which must be distributed under the terms of Sections 1 and 2 above on a medium customarily used for software interchange; or,
b) Accompany it with a written offer, valid for at least three years, to give any third party, for a charge no more than your cost of physically performing source distribution, a complete machine-readable copy of the corresponding source code, to be distributed under the terms of Sections 1 and 2 above on a medium customarily used for software interchange; or,
c) Accompany it with the information you received as to the offer to distribute corresponding source code. (This alternative is allowed only for noncommercial distribution and only if you received the program in object code or executable form with such an offer, in accord with Subsection b above.) "
I don't understand. AFAIK, the replacement is not networkable. I can't run GUI jobs from my mid-range, and display on the Apple -- UNLESS I use X11 -- which I then have to install. Worse, how do I run Apple stuff, and have it display on my workstation of choice.
So can you explain how X11 "sucks" as compared with the Apple drawing layer? Yes, I know that the Apple stack uses compositing, and a 3D render pipeline, &etc. But does it support multiple visuals? Does the 3D support depth queued Z shaded lines? Aside from the obvious networking issue. I can't just drop in a bunch of Apple gear unless it can interoperate with the SUN, IBM and HP gear in place.
The common language is X11, NFS, ethernet (&etc.). Did Apple open their drawing layer so that multiple vendors can implement and interoperate? If no, then why not? Are they (Apple) afraid of competition?
Personally, I like the idea of compositing -- but why didn't Apple instead support the efforts of XFree86/Xorg?
And now on to the rest of the comment. I use X11 as a "GUI" (drawing layer). As a GUI, I use GNOME, or mwm. Every day. Yes, I may be too close to see how badly it "sucks". But it really is the only point of consistency. I deal with Solaris Sparc, Solaris Intel, HP/UX, AIX, and Linux on Intel, Sparc and PowerPC. I even use Windows Intel. X11 is available across all of these -- but the Windows platform is a sore point (cygwin allows X11 clients to run, though). OS X would be sore point, for the same reason. But the "sucky" features of X11 allow operation across almost all (except Windows) of my platforms. If Apple wants to drive a "GUI" standard, either they would have to support, or allow to be supported, the same breadth of systems, or kill off the "legacy" systems, or open up their "GUI".
Please think about the following cases (extracted from my current project):/* Semaphores and condition variables are not interchangable on Solaris. The condition variable is EVENT driven, and not based on a COUNTER. This means that the sequence:
signal condition variable
wait on condition variable will not wake up. With semaphores, this DOES work. We could "roll our own" semaphore out of condition variables, but since Solaris offers both, we just use the semaphore. */ ksema_t ellided_sema;
And consider:/* There isn't a mutex for creating and removing card handles, because the OS will do this "single-threaded". Also, we don't need one for the IOCTL entry, because any changes will be reflected in the change count, which is not updated until after the data structure has been modified.
If you are truly paranoid about this, do:
ACA_Get_ACA_Handles -> change_count_1
ACA_Get_ACA_Handles -> change_count_2
if (change_count_1 == change_count_2) then
if (change_count_1 != original_change_count)
interpret handle list */ static bool remove_card_handle(dev_state *asp) {// elided }
In both of these cases, what is being documented by the comment is something that isn't there. And, really, isn't likely to be there in other documentation either.
And, yet, I consider both of of these comments vital (as mis-understanding WILL break, or degrade the program).
They do not comment the MEANING of the data (you still don't know what ellided_sema is for, but you DO know why it is a semaphore), and certainly not the parameter of the function (although, I think it would be/hopefully/ obvious that remove_card_handle() will remove the card handle associated with, or represented by, a dev_state *. But, you know that the operation is not controlled by a mutex, or other locking device, and, hopefully, you understand why.
Good code is self-documenting? Generally, I don't think so. Code represents one possible solution, but why was that chosen? What are the tradeoffs resulting in the decisions that resulted in THIS code? The code is (in a Zen way); but the deeper question as to WHY it is generally remains unanswered.
Then again, I get well paid for asking these questions -- this is what results in optimization. Possibly my comments are more reflective of a personal thought process.
Which version of Windows detected your wireless NIC on install? (implication was that you did NOT install additional drivers to get this piece of magic to occur -- the driver was included with Windows -ver?-).
I really want *that* version of Windows -- because I tried XP and it doesn't even recognize the optical drive it was loaded from...
Linux and wired NIC issues? I would like clarification of this as well.
My MAIN point was that some of the given tasks were not IMPOSSIBLE with a CLI; which the original post had insisted was the case. Indeed (the second point), some of these tasks are EASIER with a CLI. And, with COTS GUI stuff (Common-Off-The-Shelf), the features needed may not even be implemented.
Slashdot Mirror
I have been reading up on literacy. One fascinating point is that written language (at least English) is many times richer (in vocabulary and structure) when compared to spoken language. As heard on (for example) television.
If this holds true for other languages, it may be easier to translate spoken material.
Still, I want to see the result.
Ratboy.
I built a "mythtv" box with a SD tuner to capture cable broadcasts. Works fine, but, of course, only SD.
I have a "HD Ready" 42" TV, with DVD. Plays DVDs great, but *I* can't get an upscaling DVD player (component in only). I won't be able to use "HDTV" boxes, and next generation HD DVDs either.
All because the anal retentives won't let me. They are afraid that I might "rip off" the signal. I am NOT buying another TV (for the next 5 to 10 years). Also, archiving or recoding 36Mbit streams does not appeal to me, anyway. Too time-consuming, and life is too short.
Why not let me use the equipment I have purchased? Must be fear -- fear that little old me is going to do some KICK-ASS ripping. Maybe I SHOULD do the ripping, after all I am already "doing the time".
Ratboy
Are select() and poll() really that bad?
Ok, the issue is how many fds you can pass. With select() you are limited to a bitmaps worth. And performance has never been much of an issue.
Of course, poll() is a different matter -- if you are passing 100s or thousands of fds.
But, what has this got to do with the tcp connection? Not much.
So, you speed up poll() and still write small packets, and nagle won't write them out immediately... That's about the only connection here.
Ratboy.
Because they ARE registered trademarks?
Duh?
4 to 5 times more power/heat efficient?
How do you figure that? Let's take this a step at a time.
1 - Unless the chip implements "reversible" computations, ALL power must be dissipated as heat.
2 - N (65, 100, whatever) Watts will generate that much heat. Or other radiant energy (but your chip doesn't glow, does it?)
3 - So, your claim is that an AMD can deliver a computation at 1/4 (lets take the MOST conservative figure you quoted) the power. In other words, an AMD can deliver 100 Intel Watts of computation at 25 Watts.
Scaling this to more understandable terms: an AMD processor that dissipates around 25 watts is a K6-2 450. An Intel process that dissipates around 100 watts is a 3.4Gz Prescott.
4 - Last time I checked, that Intel processor was somewhat more that 4 times the performance of that AMD processor.
Care to back up your claim? Yes, an AMD processor may be more power efficient, but 4 to 5 TIMES more? I call bullshit.
Ratboy.
Fascinating.
I cannot enter such a "license". Copyright would be the closest. There is a convention that says if copying is needed to make use of a program, that copy is permitted.
For instance, a program must be copied from a disk into memory to execute it.
A "map" is only of use with a "server", so a copy is normal if using the product.
I am also allowed to rent, resell, and otherwise exploit works commercially. Again, this right is passed to me via Copyright.
I can also modify a work and resell it. Really.
I would LOVE a case like this to get into court -- one time it happened, the court ruled that an unauthorized person EXECUTING a program was a violation. I would also LOVE for shrink-wrap licenses to gain FULL power.
IP Police: STEP AWAY FROM THE KEYBOARD! YOU HAVE NOT BEEN DIRECTLY AUTHORIZED TO USE THAT SOFTWARE, AND ARE IN VIOLATION OF FEDERAL IP LAW!
But we aren't there. Not by a long shot. This (USA and Canada) is not a Facist society, and you do have rights.
Ratboy
Game emu is a "standard" option for MythTV. Shoutcast et al. are not there. Could be added into the framework, although I am not interested in doing that personally.
As to Guide/Recording setup. I will have to look into MCE. MythTV does "dump" live feed recording on channel change, which is the one thing I would want altered (although not sure how yet - maybe the insertion of an image? Also requires modification of the live TV buffering strategy). I'll have to play with MythTV with remote only. As it is, I have a wireless keyboard to navigate and enter data into search dialogs and such.
My hot list for MythTV:
- easy addition of hot buttons (some kind of wizard)
- mods to live tv buffering
- easier access (scripting) to external video software with a howto
- incorporation of alternate environments (eg. games running under VMware)
- possibly (if not there) dictionary smart entry via number pad (ala cellphone).
Ratboy
Just as a guess, from Gibsons explanation, the bug is the following sequence:
1 - The SetAbortProc is entered in the metafile. It is interpreted, but (and I have never seen the metafile code) instead of having arguments interpreted, *just* the operational code is.
2 - The metafile interpreter dispatches to the abort proc handler, which initializes the existence of the abort proc. It then attempts to scan the argument.
3 - Most likely, the argument scan routine picks up on the fact that the length is wrong, and no such argument exists. Note that the developer probably thought "Gee, this handler only takes a single arg, and the arg retriever checks it, so don't worry".
4 - The argument retriever indeed picks up on the fact that the argument is wrong, and THIS is sufficient to "abort" the metafile. Of course the existence of a handler has already been registered...
5 - What is the default address for the handler? That is probably set as file relative 0 (probably for other reasons).
Leading to the situation. No malice needed.
Ratboy.
I don't know about MCE - requisite Windows won't install properly on our designated PVR box. However, MythTV works fine. A couple of observations. MythTV uses XML formats to build its on screen displays, and has hooks for displaying external data as well. So the artwork and menus can be changed suitably. I assume you knew this, and therefore are referring to UI functionality, and not "slickness". Because I can't review MCE, I wouldn't mind a run-down on the features that are slicker.
Ratboy
Consider:
Linux runs on ALL those platforms (Intel, AMD, etc). And more (Alpha, IBM Mainframe, Sparc, etc.). There is really no comparision.
Consider:
Linux supports more legacy hardware with the OS core.
Consider:
Linux vendors typically support 4+ GB of object code with a typical installation.
Not that I care one way or the other about Linux/Windows comparisions, but this should give you something to think about.
The obvious conclusion to draw? That the Open Source model is SO SUPERIOR to Microsofts, that there is really no direct comparision possible. Of course, that is probably wrong. Another possibility is that Microsoft really doesn't care. Which makes sense (ob disc: I *am* a Microsoft shareholder) to me -- their goal is to maximize profit. And if the OSS model where THAT much superior, Microsoft should have adopted something like the Redhat model years ago.
Of course the problem is that Microsoft installations are more vulnerable to quiet "black hat" attacks. The general rule is to assume that if YOU found a vuln, someone else has already found it. So it really doesn't make sense to keep quiet about them, and, as a user, there seems to be some kind of speed-up on the fixes.
I don't like this a a Microsoft shareholder; more resources have to be spent on fixing 'sploits, and less on actually making money.
Your conclusions may vary.
Ratboy.
My experience (not that it is typical, I would *hope* it isn't):
.EX_ to .EXE wasn't, don't know why - but this comes back later). Increase memory from 256MB to 1.2GB to accomodate Apache and some other stuff.
1 - I want to make a PVR (personal video recorder). Aquire the following components:
(A) AMD 1700+ mainboard, 2 PCI slots, 1 AGP slot (B) NVIDIA 5200 graphics card with s-video out, (C) Mercury TV tuner (D) 256MB RAM (E) Memorex DVD burner.
Note: Choice of components is for price. Noted that mainboard documentation states that WINDOWS XP is needed for "USB 2.0 Function". Borrow a copy of WINDOWS XP for initial installation (going to spring the $140 CDN the next day IFF it works).
Assemble hardware. Install WINDOWS XP. After initial boot, note that the optical drive it installed from IS NOT PRESENT. Of course, no network access either (given that WINDOWS XP doesn't recognize the onboard ethernet. Apparently (according to a local Windows guru) I need the drivers... PS. Installed XP a second time, to be sure that I was not hallucinating.
Installed Windows 98SE (for which I had a license). And there you go. Seems to work. Except that when the recommeded drivers are installed, the optical drives vanish yet again. Weird. And, I can't get the network going. But, able to put the contents of the motherboard CD onto the hard disk, to try XP again... Installed XP again (really, I know this is a dead horse, but I can't help myself), and loaded the drivers, and: TADA! it still doesn't work.
Give up on Windows XP. Installed Fedora Core 2. Network, USB 2.0, sound, works. Just works. Added the NVIDIA driver, and the Mercury tuner. Install MythTV - and I have a PVR. WAF (wife acceptance factor) is 6 -- because it isn't Windows, and doesn't run games on the big-screen.
Back to Windows 98. Try the Mercury tuner PVR application. Doesn't work (later, Windows guru tells me that one of the files that SHOULD have been "expanded" from
Windows 98SE no longer boots. After some web searching, find out I need to reduce "physical pages" because it crashes with that much RAM. Mercury PVR application *still* won't run. Oh well. Try some of the games (freebies, mostly from cereal boxes). Sound is marginal - sometimes works, sometimes doesn't. For instance, Atari Classic Games collection - no sound, messed up main-screen. Individual games, work, but no sound. Works PERFECTLY on the kids Pentium 166 with Soundblaster, and on our Compaq Deskpro Pentium II 400 with 128MB (Windows 98SE). Why? Maybe the sound drivers?
Anyway, that's my latest Windows experience. Go figure.
And I *still* don't have Windows XP at home -- not from lack of trying, mind you. It just won't run on what I have. Yes, the above is a VERY NEGATIVE Windows experience. Namely, IT JUST DOESN'T WORK. I guess I *could* buy a brand new box, but that just isn't going to happen.
YMMV
Ratboy
Thank you.
My set (42" Hitachi) ALSO has only component inputs. Two of them. And so I don't get to have "HD" either. The HD nazi has spoken: "No HD for you!".
Blech...
Ratboy.
Hunks eaten by this system.
should be
char *s = "address of"c; f2(s);
"but I'd like to get back to coding on *at least* a recreational basis." (emphasis mine)
That would be *at least*. The course of study described would result in that. Someone could ask "I'd like to get into auto mechanics, at least on a recreational basis". And what would you suggest. How about novel writing? Oil painting? Again, emphasis on "at least".
If someone wants to "code on a recreational basis", I would suggest a different course. The "at least" suggests profesional aspirations. Which is why I suggested the given course, and attendent timelines.
Pretensious goof? I may be that. I also tried to answer the question.
That snippet is endian dependent. The argument is widened to an "int", and the address taken is for a byte pointer. Result - works on some architectures, and traps on others.
So you "coded" for 5 years, and then went "techie-upstream" for 20...
In a nutshell, you need to code for another 10 years. It takes around 15 years to build reasonable proficiency and skill.
Back to the "salt mines" for you.
1 - 20 year ago, C just started gaining commercial acceptance. Work on your C skills for a couple for years. Study old Unix source. If you can look at this code, and tell what is wrong, you are well on your way (and, yes, I know it is K&R):
f(c) char c; { char *s = f2(s); }
2 - Learn Scheme. Make use of the online MIT resources. Maybe move to Common Lisp.
3 - Begin study of ANOTHER of the current crop of interpretive or "immediate" programming systems - Perl, Python, Ruby. Don't let the Scheme go, though.
4 - Tackle Smalltalk - its one of the (or the most) productive systems out there. Squeak would do.
5 - UML, XML, and the current crop of buzz.
This would give you a fair grounding. Time? 2 or 3 years C, 3 years Smalltalk, 2 years Scheme. That would get you "up to speed" as a proficient programmer in about 8 years.
Ratboy.
Yes, again. The attacker doesn't know which resistor is at which end. And taps the middle.
Of course, the attacker may be the receiver, in which case she KNOWS the value at one end. And that is the trivial breaking case.
Ratboy.
Subject says it all. Thanks!
Gnome is based on Win9X? That would be a BIT strange. Gnome /uses/ X11, which is in turn based on older technology, Win9X isn't even slightly in that roadmap. Gnome is /based/ on CORBA.
Ratboy
But it is normal and 100% acceptable to get all your music from filesharing. ...Oh, you live in the US. Sorry.
Ratboy.
You are wrong.
Please read the ellided GPL pieces below (or read a copy of the GPL). We do not mention what constitutes a derived work here, but the rights and obligations are clear. You would be wrong -- source rights extend to any 3rd party, or include full source which is then again distributable.
Ratboy
" if you distribute copies of such a program, whether gratis or for a fee, you must give the recipients all the rights that you have. You must make sure that they, too, receive or can get the source code. And you must show them these terms so they know their rights."
" You must cause any work that you distribute or publish, that in whole or in part contains or is derived from the Program or any part thereof, to be licensed as a whole at no charge to all third parties under the terms of this License."
" You may copy and distribute the Program (or a work based on it, under Section 2) in object code or executable form under the terms of Sections 1 and 2 above provided that you also do one of the following:
a) Accompany it with the complete corresponding machine-readable source code, which must be distributed under the terms of Sections 1 and 2 above on a medium customarily used for software interchange; or,
b) Accompany it with a written offer, valid for at least three years, to give any third party, for a charge no more than your cost of physically performing source distribution, a complete machine-readable copy of the corresponding source code, to be distributed under the terms of Sections 1 and 2 above on a medium customarily used for software interchange; or,
c) Accompany it with the information you received as to the offer to distribute corresponding source code. (This alternative is allowed only for noncommercial distribution and only if you received the program in object code or executable form with such an offer, in accord with Subsection b above.) "
And why did Apple remove X11?
I don't understand. AFAIK, the replacement is not networkable. I can't run GUI jobs from my mid-range, and display on the Apple -- UNLESS I use X11 -- which I then have to install. Worse, how do I run Apple stuff, and have it display on my workstation of choice.
So can you explain how X11 "sucks" as compared with the Apple drawing layer? Yes, I know that the Apple stack uses compositing, and a 3D render pipeline, &etc. But does it support multiple visuals? Does the 3D support depth queued Z shaded lines? Aside from the obvious networking issue. I can't just drop in a bunch of Apple gear unless it can interoperate with the SUN, IBM and HP gear in place.
The common language is X11, NFS, ethernet (&etc.). Did Apple open their drawing layer so that multiple vendors can implement and interoperate? If no, then why not? Are they (Apple) afraid of competition?
Personally, I like the idea of compositing -- but why didn't Apple instead support the efforts of XFree86/Xorg?
And now on to the rest of the comment. I use X11 as a "GUI" (drawing layer). As a GUI, I use GNOME, or mwm. Every day. Yes, I may be too close to see how badly it "sucks". But it really is the only point of consistency. I deal with Solaris Sparc, Solaris Intel, HP/UX, AIX, and Linux on Intel, Sparc and PowerPC. I even use Windows Intel. X11 is available across all of these -- but the Windows platform is a sore point (cygwin allows X11 clients to run, though). OS X would be sore point, for the same reason. But the "sucky" features of X11 allow operation across almost all (except Windows) of my platforms. If Apple wants to drive a "GUI" standard, either they would have to support, or allow to be supported, the same breadth of systems, or kill off the "legacy" systems, or open up their "GUI".
Ratboy.
Please think about the following cases (extracted from my current project): /* Semaphores and condition variables are not interchangable on Solaris. The condition variable is EVENT driven, and not based on a COUNTER. This means that the sequence:
/* There isn't a mutex for creating and removing card handles, because the OS will do this "single-threaded". Also, we don't need one for the IOCTL entry, because any changes will be reflected in the change count, which is not updated until after the data structure has been modified.
// elided
/hopefully/ obvious that remove_card_handle() will remove the card handle associated with, or represented by, a dev_state *. But, you know that the operation is not controlled by a mutex, or other locking device, and, hopefully, you understand why.
signal condition variable
wait on condition variable
will not wake up. With semaphores, this DOES work. We could "roll our own" semaphore out of condition variables, but since Solaris offers both, we just use the semaphore.
*/
ksema_t ellided_sema;
And consider:
If you are truly paranoid about this, do:
ACA_Get_ACA_Handles -> change_count_1
ACA_Get_ACA_Handles -> change_count_2
if (change_count_1 == change_count_2) then
if (change_count_1 != original_change_count)
interpret handle list
*/
static bool remove_card_handle(dev_state *asp)
{
}
In both of these cases, what is being documented by the comment is something that isn't there. And, really, isn't likely to be there in other documentation either.
And, yet, I consider both of of these comments vital (as mis-understanding WILL break, or degrade the program).
They do not comment the MEANING of the data (you still don't know what ellided_sema is for, but you DO know why it is a semaphore), and certainly not the parameter of the function (although, I think it would be
Good code is self-documenting? Generally, I don't think so. Code represents one possible solution, but why was that chosen? What are the tradeoffs resulting in the decisions that resulted in THIS code? The code is (in a Zen way); but the deeper question as to WHY it is generally remains unanswered.
Then again, I get well paid for asking these questions -- this is what results in optimization. Possibly my comments are more reflective of a personal thought process.
But I am very curious about this.
Ratboy.
Which version of Windows detected your wireless NIC on install? (implication was that you did NOT install additional drivers to get this piece of magic to occur -- the driver was included with Windows -ver?-).
I really want *that* version of Windows -- because I tried XP and it doesn't even recognize the optical drive it was loaded from...
Linux and wired NIC issues? I would like clarification of this as well.
Ratboy.
Sorry to reply to myself. But, I must correct.
My MAIN point was that some of the given tasks were not IMPOSSIBLE with a CLI; which the original post had insisted was the case. Indeed (the second point), some of these tasks are EASIER with a CLI. And, with COTS GUI stuff (Common-Off-The-Shelf), the features needed may not even be implemented.
Ratboy.