Plz Emails Teh Skeematic



  • Lately I have been doing circuit design / assembly language programming using PIC chips. These are cheap ($US 0.50 and up), mostly 8 and 16 bit devices with very limited capabilites. The one I am using has something like 1KB of RAM (as an on-chip register file), runs at 4 mhz, and has a "RISC" instruction set with no hardware multiply or divide capabilities. 

    Initially, at least, my foray back into low-level computer engineering seemed like a pleasant escape from the world of desktop software. Standards are more exacting. Documentation is more thorough. One knows exactly what's happening when, and buzzwords will get you nowhere. After 10 years of marshalling widgets to the remote tier, I was ready for (forgive me) a paradigm shift. Rightsizing my clock oscillator down to about 4mhz seemed to be that shift.

    But now there are signs of trouble. Today, after another long and productive day of programming assembly / feeling better than people who have to use CSS, I decided to unwind reading the manufacturer's forums (microchip.com). Soon enough, I discovered that the "plz e-mail teh codez" phenomenon is by no means isolated to the world of software. Consider the following forum post, in which circuit designs are demanded from afar:

    http://forum.microchip.com/tm.aspx?m=199351&mpage=1&key=MMC%2cCard%2cQuestion&#199351

    If you don't want to click the link, the first post speaks volumes:

    Hi,
    Can anyone help me in designing post diagnostic card? i need the circuit design.

    Thanks & regards,
    Navya

    The design in question turns out to be a completely implausible USB device that transmits POST codes over Ethernet. Yeah, it would be nice if a USB device could detect POST failures and e-mail them to me. But how is any of that supposed to happen if the computer just beeps and does nothing when power is applied? Reading down just a bit, I found that this same person (who claims to be an electronics engineer) has basically spammed microchip.com as well as usb.org with the same nonsense. A thousand people (kinder than I it seems) have tried to explain the fundamentally unworkable nature of the request in a thousand ways:

    Navya, 

    The answers are going to be pretty similar to when you asked about doing the same thing via USB.
    http://forum.microchip.com/tm.aspx?m=192237&mpage=1&key=&#192237

    i.e. You would have to modify the PC BIOS to send the info via ethernet, you cannot take control of the PC before the OS loads without modifying the BIOS.
    (I'll qualify that. If the ethernet card has support for a BOOT ROM, then by placing custom code into the BOOT ROM, you might be able to access the POST code and send it. However, if you have to ask, I don't think you're capable of doing that...)

    		</span></i>   
    	  <br></p><p><i>Ric</i></p><p>Does this deter our protagonist? Not in the slightest:&nbsp;</p><p><i>
         
          Hi , <br><br> Ric, I have a doubt. I
    

    understood that we cant use usb for interfacing our device(ethernet
    controller+microcontroller) to PC.Is there any oher hardware connection
    ?(except serial ports as some laptops wont be having serial ports)we
    want a plugin type device only..ie y we went for usb plugin
    option(which will be violating specification)

    expecting ur early help,

    Navya

    I love the closing ("expecting ur early help"). A few posts later, Navya seems to be treating the thread something like a cross-examination: 

          Hi, <br> &nbsp;&nbsp;&nbsp; Can we access the PCI or ISA bus using USB device, if the system is ON? Can you please send me a suggestion? <br>  <br> thanks, <br> Navya</p>And then, the cross-examination continues with the surprise introduction of Exhibit A, block diagram of a motherboard: <br><p style="font-style: italic;">Hi,</p><p style="font-style: italic;">&nbsp;&nbsp; One more question please.. I found
    

    in net that the USB host controller IC in motherboard has connection to
    the PCI bus. USB has no seperate bus in motherboard. Can you please
    clarify it?

    Thanks,
    Navya

    This guy (or girl) really, really, really, really wants to support USB on what's essentially a boat anchor, and it's just not happening. And programmers can be stupid, but this represents an entirely different level of futility and desperation! 



  • It could just take its power from USB, and actually have a microphone to decode the boot failures.

    TRWTF is either PICs or the fact that I use them also...



  • You could also achieve the same effect by equipping a POST card with its own RJ45 jack and/or wi-fi card, or even a cellphone transmitter so it could send SMS messages. I could see something like that being mildly useful for servers and embedded systems in unstaffed or inaccessible locations.



  • @beau29 said:

    After 10 years of marshalling widgets to the remote tier, I was ready for a paradigm shift.

    Baha!

     

    I really have a hard time putting myself in the heads of the Navya's of the world. How do you get through to them, if not in their native tongue?



  • @dhromed said:

    I really have a hard time putting myself in the heads of the Navya's of the world. How do you get through to them, if not in their native tongue?

     

    There is something in Linguistics called the Saapir-Whorf Hypothesis (sp.?) which asserts that people's entire way of thinking is shaped by their language. To some extent, we think in our language(s) even when we're not speaking or hearing them. Saapir-Whorf notices the big differences amongst languages and tries to draw parallels in attitude and behavior.

    I'm not sure I completely believe Saapir-Whorf. I feel capable of "thinking" things I can't neatly express in English, and in amateur hands Saapir-Whorf can seem like a thinly disguised form of ethnic stereotyping. But there's probably some truth to it. I mean, reading those posts, I can't help but think Navya's native language must be much more suited to the practice of law than the practice of engineering. 

     



  • @jpa said:

    It could just take its power from USB, and actually have a microphone to decode the boot failures.

    TRWTF is either PICs or the fact that I use them also...

     

     

    Taking this a step further, couldn't one cut the wires off of the PC Speaker (or unplug it) and wire it into one of the PIC's digital inputs, thus allowing the PIC to "hear" the failure beeps? The PIC could power on with the motherboard and count the beeps. The e-mail could probably be generated if, say, two or three beeps occur instead of the expected single beep. This is, of course, somewhat motherboard-specific (or at least BIOS-specific), but I can believe that perhaps a few DIP switches could be used to switch amongst Phoenx mode, AMI mode, etc. and support just about everything. 

     I don't know if this would meet Navya's spec (USB is not involved), but it seems like a workable design. 




  • @beau29 said:

    @dhromed said:

    I really have a hard time putting myself in the heads of the Navya's of the world. How do you get through to them, if not in their native tongue?

     

    There is something in Linguistics called the Saapir-Whorf Hypothesis (sp.?) which asserts that people's entire way of thinking is shaped by their language. To some extent, we think in our language(s) even when we're not speaking or hearing them. Saapir-Whorf notices the big differences amongst languages and tries to draw parallels in attitude and behavior.

    I'm not sure I completely believe Saapir-Whorf. I feel capable of "thinking" things I can't neatly express in English, and in amateur hands Saapir-Whorf can seem like a thinly disguised form of ethnic stereotyping. But there's probably some truth to it. I mean, reading those posts, I can't help but think Navya's native language must be much more suited to the practice of law than the practice of engineering. 

    I'd say your skepticism is justified. If you go and peruse Not Always Right, you'll see plenty of instances of English-speaking customers of all types going through approximately:

    Customer: I want [request which is impossible for whatever reason]
    Business: Sorry, we can't do that because [explanation, possibly involving the fact that the customer is in entirely the wrong type of store]
    Customer: Okay, just give me [same request possibly with minor changes which do not alter the obvious impossibility of the base request]
    Business: We still can't do that because [variation on the explanation]
    [Repeat some number of times until either customer walks out in a huff, business kicks them out because there are other people waiting, or business manages to come up with a red herring of some kind, such as referring them to another business which also will not be able to comply with the request but which at least will get the customer out of the way for now.]

    I think it's just a particular type of stupidity, combined with arrogance of the "the world revolves around me, so naturally anything I want must be possible" type. Language has little to do with it, beyond making it harder for people to understand "no", not that they'd listen to it anyway.



  • @The Vicar said:

    Customer: I want [request which is impossible for whatever reason]

     

    Sometimes "whatever reason" ends up being plain physics, or more specifically hardware limitations. I mean, a CGA display adapter can only display 4 colors at 320x200 resolution... asking one's supposedly brilliant grandkid for a second opinion is not going to change that fact. I'm not sure what's to blame here.... self-centerdness? the schadenfreude of seeing a supposed expert proven wrong? a belief that the proverbial "squeaky wheel" gets the metaphorical "grease?"

    The Real WTF™ is that JPA and I have basically articulated a potential solution to the problem here, in a quasi-recreational forum for programmers. I have always said that the world should listen more closely to what its CS graduates are saying...



  • Now I come to think about it, it's kind of a WTF that PC hardware manufacturers still haven't come up with something more user-friendly than either beep codes or the two-digit figure on a POST card readout after all these years; I mean, how hard could it be to offer a POST card with an LCD screen that prints a scrolling line of text describing the error state, or a MIDI-synthesised voice reading out the error instead of a beep code? (Admittedly the latter might add to the price of the board a bit, as the BIOS would have to go on a microSD card or something.) 



  • @Jake Grey said:

    Now I come to think about it, it's kind of a WTF that PC hardware manufacturers still haven't come up with something more user-friendly than either beep codes or the two-digit figure on a POST card readout after all these years;

     

    Sounds like an opportunity. Maybe it's a good thing that the provenance of these ideas is recorded here...

     

     @Jake Grey said:

    MIDI-synthesised voice reading out the error instead of a beep code? (Admittedly the latter might add to the price of the board a bit, as the BIOS would have to go on a microSD card or something.) 

    Yeah, but that wouldn't have anything to do with MIDI, would it? My experience with MIDI is that it basically contains data like "turn on the Saxophone sound at 440hz." What you're talking about is a waveform sound.

    I like the idea of putting the failure information on the net (e.g. emailing some network admin type). The target market for the device would be people administering distant or isolated computers.

     



  • @beau29 said:

    There is something in Linguistics called the Saapir-Whorf Hypothesis (sp.?) which asserts that people's entire way of thinking is shaped by their language.

    Not having seen the theorem I agree whole heartedly with the spirit. You don't to go too far down the road of learning a foreign language to realise that there are concepts that can't be expressed in your own native tongue (or at least not neatly). Thus if you can't express it then your thinking really has been shaped by your language.



  • @Jake Grey said:

    I mean, how hard could it be to offer a POST card with an LCD screen that prints a scrolling line of text describing the error state, or a MIDI-synthesised voice reading out the error instead of a beep code? (Admittedly the latter might add to the price of the board a bit, as the BIOS would have to go on a microSD card or something.) 
    I've seen some MSI motherboards opt for LEDs instead of beep codes. Four LEDs switching from red to green and back identifying all the stages the machine goes through from the moment you press the power button to loading the OS. Of course, like beep codes, you need to consult the mobo manual to identify each color combination.



  • @OzPeter said:

    @beau29 said:
    There is something in Linguistics called the Saapir-Whorf Hypothesis (sp.?) which asserts that people's entire way of thinking is shaped by their language.
    Not having seen the theorem I agree whole heartedly with the spirit. You don't to go too far down the road of learning a foreign language to realise that there are concepts that can't be expressed in your own native tongue (or at least not neatly). Thus if you can't express it then your thinking really has been shaped by your language.

    But OzPeter, thinking and speaking are two different processes. While I can understand that a language would influence how you describe concepts to others, I don't see how it would prevent you from thinking about concepts which are not so easily described in your native tongue.



  • @beau29 said:

    "plz e-mail teh codez" phenomenon

     This phenomenon exists just about everywhere.  I work in the data acquisition arena and if you look at National Instrument's forums you will see similar requests.

     A friend of mine years back did tech support for Alpha 4 (yes - that long ago).  Any this "consultant" calls up asking how to code this project he is working on.  Due the next day.  He tells my friend his requirements and asks "just tell me what to type".  Obviously my friend refuses and asks if there is a specific problem.  The "consultant" goes on and on about how he won't use Alpha 4 ever again.

     "plz e-mail teh codez", is just the most recent version of please phone me the codes.



  • @Jake Grey said:

    Now I come to think about it, it's kind of a WTF that PC hardware manufacturers still haven't come up with something more user-friendly than either beep codes or the two-digit figure on a POST card readout after all these years; I mean, how hard could it be to offer a POST card with an LCD screen that prints a scrolling line of text describing the error state, or a MIDI-synthesised voice reading out the error instead of a beep code? (Admittedly the latter might add to the price of the board a bit, as the BIOS would have to go on a microSD card or something.)
    Many motherboards are equiped with LEDs that show status (usually 4 LEDs), some motherboards can show the error code on a 2-digit LED display, and certain motherboards can speak the error through the on-board sound card. Failing that, you can always buy a POST code display card, plug it in and get the 2-digit code.



  • @Jake Grey said:

    I mean, how hard could it be to offer a POST card with an LCD screen that prints a scrolling line of text describing the error state, or a MIDI-synthesised voice reading out the error instead of a beep code? (Admittedly the latter might add to the price of the board a bit, as the BIOS would have to go on a microSD card or something.) 

    The main reason is cost. How often do you need a POST code decoder? Except during initial build/configuration of a PC, the answer is probably "very infrequently", so why spend more than a few cents for a couple of LEDs (or even less just to reuse the speaker).

    At the office, we have got one PC with a motherboard with a voice synthesiser (probably just a recording) which will read out a description of the error through a built-in speaker. All I can say is that we turned that option off... Beeps were much nicer!

    Remotely there's very little point in knowing the POST code, except you'd know what you'd probably need to take along (eg if it was a memory error, you'd know to take your fingers to re-seat the memory) - but you'd probably just take everything you could possibly need, in case the reported error was a red herring, so the fact the PC isn't working is probably sufficient knowledge...



  • I have a gigabyte motherboard (model escapes me at the moment) that speaks in clearly-synthesized-but-completely-understandable English when there are boot errors. It has yelled at me for a missing video card, keyboard, and cpu fan failure. So the voice thing has been done. It just isn't widespread.



  • @X-Cubed said:

    @OzPeter said:

    Thus if you can't express it then your thinking really has been shaped by your language.

    But OzPeter, thinking and speaking are two different processes. While I can understand that a language would influence how you describe concepts to others, I don't see how it would prevent you from thinking about concepts which are not so easily described in your native tongue.

    Indeed, one does not think exclusively in one's own language, as if one were incapable of though without language. That's just a silly notion. You have thoughts, and then maybe -- but usually -- translate them into language. It's true that some languages are more suited for expression of certain thoughts. The most basic example is a word that exists in one language, but not "really" in another, such as the Dutch 'gezellig', or the Martian 'grok'.

    But the relationship between thought and language is a strong one,and we all know that putting things into words can clear things up and, in a manner of speaking (H^H^H^), change your thoughts.



  • @Jake Grey said:

    At the office, we have got one PC with a motherboard with a voice synthesiser (probably just a recording) which will read out a description of the error through a built-in speaker. All I can say is that we turned that option off... Beeps were much nicer!

    So here's what the Navya guy should implement: get one of those speaking motherboards, put a VOICEMONROBOT next to it with a microphone that does speech recognition. The text is then printed out on continuous paper, rolling past on a wooden table in front of a webcam that periodically takes a picture of it, which is sent for photo development, and the photo gets mailed to the waiting sysadmin.



  •  @fbjon said:

    @Jake Grey said:
    At the office, we have got one PC with a motherboard with a voice synthesiser (probably just a recording) which will read out a description of the error through a built-in speaker. All I can say is that we turned that option off... Beeps were much nicer!
    So here's what the Navya guy should implement: get one of those speaking motherboards, put a VOICEMONROBOT next to it with a microphone that does speech recognition. The text is then printed out on continuous paper, rolling past on a wooden table in front of a webcam that periodically takes a picture of it, which is sent for photo development, and the photo gets mailed to the waiting sysadmin.

    I just had a sad mental image of a lonely, forgotten VOICEMONROBOT muttering away in an empty corner of the data center.

     

    In addition to the "talking" motherboard others have mentioned, I have seen dozens of different kinds of visual displays on various servers to indicate POST faults.  Proliants had optional LCD screens on them for years (they use simpler printed diagrams with LEDs now) and so did Alphas.  Integritys have Remote Management Consoles tied into their EFI BIOSes, so you already have your network interface from which to troubleshoot startup problems.



  • @X-Cubed said:

    But OzPeter, thinking and speaking are two different processes. While I can understand that a language would influence how you describe concepts to others, I don't see how it would prevent you from thinking about concepts which are not so easily described in your native tongue.

    I didn't mean it would prevent you, but it certainly can make things extremely difficult - to the point that it is not worth the effort to try - which could be the equivalent of stopping you. For (a fictional) example, take the concepts in "1984". Removal of words from the "official" language in order to control the thoughts of the population by denying them the ability to express ideas. In the real world you have propaganda where words/concepts are bound together in ways that are meant to direct the publics thoughts. If the propaganda campaign is successful it is very hard to not think in terms of the manufactured word/concept links.


    And to take a computer language analogy, in the original versions of BASIC it would have been near impossible to describe OOP techniques, yet for Smalltalk or Simula of the same era it would have been trivial.



  • @operagost said:

    Integritys have Remote Management Consoles tied into their EFI BIOSes, so you already have your network interface from which to troubleshoot startup problems.

    From reading the various postings by Navya, think that this is actually the problem he/she is trying to solve - some sort of remote logging/diagnostics of POST codes. However I think that Navya was fixated on a specific solution for some reason or another (class project?) and couldn't generalize the problem domain to be able to include practical solutions.



  • I think this has been mentioned at some time on thedailyWTF, but it is still extremely relevant.

    http://www.apa.org/journals/features/psp7761121.pdf



  • @OzPeter said:

    @X-Cubed said:
    But OzPeter, thinking and speaking are two different processes. While I can understand that a language would influence how you describe concepts to others, I don't see how it would prevent you from thinking about concepts which are not so easily described in your native tongue.
    I didn't mean it would prevent you, but it certainly can make things extremely difficult - to the point that it is not worth the effort to try - which could be the equivalent of stopping you. For (a fictional) example, take the concepts in "1984". Removal of words from the "official" language in order to control the thoughts of the population by denying them the ability to express ideas. In the real world you have propaganda where words/concepts are bound together in ways that are meant to direct the publics thoughts. If the propaganda campaign is successful it is very hard to not think in terms of the manufactured word/concept links.


    True. Take Spanish, for example. There is no direct translation for "engineers" as used in context with "engineer a solution for ..." as well as "override". Some other concepts do exist in Spanish, but require 3, 4 or 5+ words to describe the same concept in Spanish, which makes it ugly to express some things (especially technical things.) In Mexico we tend to "switch" to English for some stuff, or even use loanwords; in Spain, however, they love to attempt translation of technical terms, more often than not the wrong way. Example: Archivo = File, but in Spain its Fichero (card) = File; ordenador (sorter/sorting machine) = computer instead of computadora... oh well.

    But the thing is that I do feel that language can actually help or hinder thought; and for technical stuff I feel English an easier language than Spanish for that.

    @OzPeter said:


    And to take a computer language analogy, in the original versions of BASIC it would have been near impossible to describe OOP techniques, yet for Smalltalk or Simula of the same era it would have been trivial.

    Isn't that what they call "Visual Basic" these days? Basically Microsoft took BASIC, created QuickBASIC which hacked some stuff into the original BASIC; then hacked it a bit more and made "Visual Basic". Finally, they added OOP somewhere along the line, but it wasn't really OOP until .NET, and that would be 2.0 in any case.

    I think COBOL also had some bastardization to some degree on making it do things it was not designed for, too.



  • @beau29 said:

    a CGA display adapter can only display 4 colors at 320x200 resolution... asking one's supposedly brilliant grandkid for a second opinion is not going to change that fact.

    Well, CGA is only designed to support 4 colors at 320x200, but that doesn't necessarily mean that it can't be cajoled into more through clever hacks -- for example, the mid-frame palette switching used by California Games, or the composite color artifacting supported by several Sierra On-Line adventure games.

    That being said, is it usually smarter, when support for more than 4 colors is part of the requirements, to spend the extra dough and upgrade to an EGA adapter?  Yes.

     



  • @danixdefcon5 said:

    In Mexico we tend to "switch" to English for some stuff, or even use loanwords; in Spain, however, they love to attempt translation of technical terms, more often than not the wrong way.

    You should take note of what happened to Portugal .. they sort of got forced down the path of adopting Brazilian spellings over their traditional European spellings. Something about 188 Million Brazilians vs 10 Million European Portuguese. Surely you have enough people to reign in those old fashioned Europeans. Maybe you could team up with some other Central or South American country and have a linguistic war???



  • @OzPeter said:

    Something about 188 Million Brazilians vs 10 Million European Portuguese.

    It's the warm weather: keeps people fucking like rabbits. 



  • Damnit, its simple:

    Create a USB device that detects failures and sends some sort of signal over a wire (say a microphone jack) which is detected by the soundcard of a receiver computer and emails you teh notification. In the event that I am completely missunderstanding the problem, just post it on java.com with the same question.

     

    expecting ur early reply,

    Me



  • @DOA said:

    @Jake Grey said:

    I mean, how hard could it be to offer a POST card with an LCD screen that prints a scrolling line of text describing the error state, or a MIDI-synthesised voice reading out the error instead of a beep code? (Admittedly the latter might add to the price of the board a bit, as the BIOS would have to go on a microSD card or something.) 
    I've seen some MSI motherboards opt for LEDs instead of beep codes. Four LEDs switching from red to green and back identifying all the stages the machine goes through from the moment you press the power button to loading the OS. Of course, like beep codes, you need to consult the mobo manual to identify each color combination.

     

    My 6-year-old motherboard with my Athlon XP1600 has the 4 LED solution. Those LEDs are really bright when the room is dark. :)

    My housemate just bought a Stiker II Formula motherboard which has a LCD screen called the "poster". When the computer is on standby it displays the time.

    A computer at work had the voice error report and said "CPU test fail" a few weeks ago. 



  • I once worked on a computer for someone that gave some kind of waveform response to POST messages through the speakers connected to the sound card. The particular problem I was fixing was software related, so I never got to hear anything else, other than "Loading operating system," immediately after which the Windows boot loader would kick in. I'm not sure how they obtained this capability, the computer was custom-built for them and had a Megabyte motherboard, if that helps anyone who is curious.



  • @The Vicar said:

    @beau29 said:

    @dhromed said:

    I really have a hard time putting myself in the heads of the Navya's of the world. How do you get through to them, if not in their native tongue?

     

    There is something in Linguistics called the Saapir-Whorf Hypothesis (sp.?) which asserts that people's entire way of thinking is shaped by their language. To some extent, we think in our language(s) even when we're not speaking or hearing them. Saapir-Whorf notices the big differences amongst languages and tries to draw parallels in attitude and behavior.

    I'm not sure I completely believe Saapir-Whorf. I feel capable of "thinking" things I can't neatly express in English, and in amateur hands Saapir-Whorf can seem like a thinly disguised form of ethnic stereotyping. But there's probably some truth to it. I mean, reading those posts, I can't help but think Navya's native language must be much more suited to the practice of law than the practice of engineering. 

    I'd say your skepticism is justified. If you go and peruse Not Always Right, you'll see plenty of instances of English-speaking customers of all types going through approximately:

    Customer: I want [request which is impossible for whatever reason]
    Business: Sorry, we can't do that because [explanation, possibly involving the fact that the customer is in entirely the wrong type of store]
    Customer: Okay, just give me [same request possibly with minor changes which do not alter the obvious impossibility of the base request]
    Business: We still can't do that because [variation on the explanation]
    [Repeat some number of times until either customer walks out in a huff, business kicks them out because there are other people waiting, or business manages to come up with a red herring of some kind, such as referring them to another business which also will not be able to comply with the request but which at least will get the customer out of the way for now.]

    I think it's just a particular type of stupidity, combined with arrogance of the "the world revolves around me, so naturally anything I want must be possible" type. Language has little to do with it, beyond making it harder for people to understand "no", not that they'd listen to it anyway.

     

     

    Welcome to my world - unfortunatly this is common in Embedded Electronic Engineering 



  • @Jake Grey said:

    Now I come to think about it, it's kind of a WTF that PC hardware manufacturers still haven't come up with something more user-friendly than either beep codes or the two-digit figure on a POST card readout after all these years; I mean, how hard could it be to offer a POST card with an LCD screen that prints a scrolling line of text describing the error state, or a MIDI-synthesised voice reading out the error instead of a beep code? (Admittedly the latter might add to the price of the board a bit, as the BIOS would have to go on a microSD card or something.) 

    because it wouldn't be good marketing. "buy our motherboards, we have highly detailed error messages". Now unless they offset this with storing the BIOS in an Oracle 11g db and of course xml... better yet, xml stored inside more xml that's stored inside an Oracle 11g db, all stored inside a very large CMOS chip.


Log in to reply