zigbee???
in [Design]
|
From: Michael A. Terrell on 17 May 2010 17:47 Joerg wrote: > > Michael A. Terrell wrote: > > Joerg wrote: > >> Michael A. Terrell wrote: > >>> D Yuniskis wrote: > >>>> Hi Michael, > >>>> > >>>> Michael A. Terrell wrote: > >> [...] > >> > >>>>> If I can scrape up the $100 before they others are gone, I'd prefer a > >>>>> newer and faster server that draws less power. :) > >>>> Why? :> Is there a reason you *need* a "server" instead of > >>>> a regular desktop machine? I.e., I only run my servers > >>>> because I want long term reliability (power supplies and > >>>> spindles). Most of my regular work I do on consumer > >>>> grade machines. Servers tend to get reliability at the > >>>> expense of noise and size. :< > >>> > >>> The server can go into a closet with the other network hardware, > >>> cable modem and netwoerk switch hardware. It free up desk space where I > >>> work on computers, as well. ... > >> And makes your electric meter really spin up :-) > > > > > > And? > > > > Oh yeah, I forgot, it just comes out of the wall outlet so it's free :-) No. 'And' means I can work faster from a local server. The quicker I finish what I'm working , the sooner it gets shut down. Without a dedicated server, files are scattered over hard drives, CD-ROM, DVD & USB sticks all over the place. I may have five computers up to find what I need. I also want to run a real NNTP server to allow me to filter out some idiots and countries before they ever hit my computer. Would you like to compare electric usage? -- Anyone wanting to run for any political office in the US should have to have a DD214, and a honorable discharge.
From: D Yuniskis on 17 May 2010 18:00 Hi Joerg, Joerg wrote: >>>>>> You can get around this by using "hosts" files (which is a >>>>>> maintenance headache) *or* force everything to deal with >>>>>> SMB/windows style names/protocols <frown> >>>>> >>>>> It's all automatic over here. Plug it in, wait ... works. >>>> >>>> Then how do you reference a particular printer/computer? >>> >>> I don't know all the details because I just let the mfg's software >>> figure that out. I believe it's DHCP. Essentially you plug them in >>> and after a few seconds the printers are being recognized. Once after >>> some >> >> But how do you *reference* them? I.e., how do you "pick" >> which printer you want to talk to? (since their names/addresses >> aren't "constant") > > They are mapped to always the same names. One is always "HP Laser 5L", > the other is always "Brother MFC7820N" and so on. That's how they show > up in the print menu on any program. Ah, OK. So, if you had *two* of a particular printer, there is no way to tell them apart (?) (i.e., in a large business) >>> Now they all show up in the printer selection and you just pick the >>> one you want. But I didn't give them names like you did :-) >> >> I'm still unsure how you know which printer you are talking to. :< >> My UN*X and Sun boxes need (want) name/IP addresses. So, if I let >> the devices take their IP addresses from a DHCP server, then >> *I* don't know what they are from one boot to the next... > > Maybe Windows is better in that respect :-) Windows broadcasts "device names" and expects "folks" to notice those broadcasts. I.e., it can't find your printer until your printer says "Hi, I'm here! My name is..." By contrast, I name everything explicitly (i.e., I can always find my web server or file server using its IP) and talk *to* them (instead of waiting for them to tell me they are available). Imagine if you had to wait for google.com to tell you "I'm here (in case you want to access me)", etc. > It never fails, since years, I see the correct printer names and not IP > addresses. Always. > >>>> Yup. I use Windows (2KS and XP) for anything that needs >>>> "expensive tools" (CAD, PCB, DTP, etc.) and NetBSD & Solaris >>>> for everything else. Since the "everything else" tends to >>>> take far more time than any of the other activities (i.e., >>>> I can design and layout a board in 10% of the time that it >>>> will take to write the firmware *for* that board!), I spend >>>> most of my time away from Windows. >>>> >>>> OTOH, I spend more *maintenance* time on the Windows machines >>>> than any other! :< >>> >>> Hmm, not here. It just works. The only time I needed to really deal >>> with Windows was a re-install. That was after the HD of a laptop >>> croaked with a loud grounding noise. Surprisingly the re-install took >>> less than 30 minutes and most of that unattended. >> >> For reinstalls, image your drive with Clonezilla (or it's ilk). >> This has been a lifesaver for me when trying to restore a >> munged system (usually the result of upgrading something and >> belatedly discovering that the update screwed up something >> else, etc.) >> >> A lot of my time is spent managing disk space. I don't like >> big drives -- preferring, instead, multiple spindles (70 - 140G >> per spindle). So, if I install something and "C:" fills up, >> I have to move something off to another spindle, etc. >> >> And, if that "something" was referenced anyplace (shortcuts, >> MRU lists, etc.) then there is some cost "fixing" those >> references. >> >> By contrast, the UN*X boxen let me move things freely and >> don't rely on as many "references" as Windows does. > > True. Although when this incident happened it wasn't really a big deal > to re-install stuff. Most of that ran in the background while I was > doing some design on another computer. In hindsight I was glad I didn't > mirror because Dell had furnished some new and truly better drivers for > this machine. For example, now it could display movie files in fast > realtime. I need this for DICOM. IOW afterwards I had a laptop that was > better and more valuable than when I bought it new. All for a new $60 > (and much bigger) hard drive and the modest effort of a re-install. When I build a new machine, I go through each piece of software that was on the *old* machine and ask myself: - do I want to keep this? - do I want to *upgrade* this? - do I want to abandon this? I record detailed notes about each installation (what got installed, where in the filesystem it resides, any tweeks I had to do to menus, authorization codes, etc.). So, if I want to build a new machine, I know what steps to follow. (often the order of installs makes a difference -- <frown> -- so I keep track of what works... and *why* if I can sort that out!)
From: Joerg on 17 May 2010 18:15 D Yuniskis wrote: > Hi Joerg, > > Joerg wrote: >>>>>>> You can get around this by using "hosts" files (which is a >>>>>>> maintenance headache) *or* force everything to deal with >>>>>>> SMB/windows style names/protocols <frown> >>>>>> >>>>>> It's all automatic over here. Plug it in, wait ... works. >>>>> >>>>> Then how do you reference a particular printer/computer? >>>> >>>> I don't know all the details because I just let the mfg's software >>>> figure that out. I believe it's DHCP. Essentially you plug them in >>>> and after a few seconds the printers are being recognized. Once >>>> after some >>> >>> But how do you *reference* them? I.e., how do you "pick" >>> which printer you want to talk to? (since their names/addresses >>> aren't "constant") >> >> They are mapped to always the same names. One is always "HP Laser 5L", >> the other is always "Brother MFC7820N" and so on. That's how they show >> up in the print menu on any program. > > Ah, OK. So, if you had *two* of a particular printer, there > is no way to tell them apart (?) (i.e., in a large business) > Sure you can. I accepted the default names because there's only one each. But you could also name one Alpha, the next one Bravo, or whatever. Or HP5L_Office, HP5L_Lab. >>>> Now they all show up in the printer selection and you just pick the >>>> one you want. But I didn't give them names like you did :-) >>> >>> I'm still unsure how you know which printer you are talking to. :< >>> My UN*X and Sun boxes need (want) name/IP addresses. So, if I let >>> the devices take their IP addresses from a DHCP server, then >>> *I* don't know what they are from one boot to the next... >> >> Maybe Windows is better in that respect :-) > > Windows broadcasts "device names" and expects "folks" to notice > those broadcasts. I.e., it can't find your printer until your > printer says "Hi, I'm here! My name is..." > But what's wrong with that? This way if it doesn't answer a light may come on in your head such as "Oh, yes, I serviced it yesterday and probably forgot to turn it back on". > By contrast, I name everything explicitly (i.e., I can > always find my web server or file server using its IP) > and talk *to* them (instead of waiting for them to tell > me they are available). > I don't want to talk to them before they are available. Else you'll have print jobs hanging in the queue forever and when in a rush that's not good. I'd rather see that a printer has reported for duty and _then_ send a job to it. After all, that's how the whole military works, once a soldier reaches his station he must report to the sergeant that he did :-) > Imagine if you had to wait for google.com to tell you > "I'm here (in case you want to access me)", etc. > I don't see anything wrong on a LAN. In fact, it's good. If the printer doesn't show that means something electrical must be broken. >> It never fails, since years, I see the correct printer names and not >> IP addresses. Always. >> >>>>> Yup. I use Windows (2KS and XP) for anything that needs >>>>> "expensive tools" (CAD, PCB, DTP, etc.) and NetBSD & Solaris >>>>> for everything else. Since the "everything else" tends to >>>>> take far more time than any of the other activities (i.e., >>>>> I can design and layout a board in 10% of the time that it >>>>> will take to write the firmware *for* that board!), I spend >>>>> most of my time away from Windows. >>>>> >>>>> OTOH, I spend more *maintenance* time on the Windows machines >>>>> than any other! :< >>>> >>>> Hmm, not here. It just works. The only time I needed to really deal >>>> with Windows was a re-install. That was after the HD of a laptop >>>> croaked with a loud grounding noise. Surprisingly the re-install >>>> took less than 30 minutes and most of that unattended. >>> >>> For reinstalls, image your drive with Clonezilla (or it's ilk). >>> This has been a lifesaver for me when trying to restore a >>> munged system (usually the result of upgrading something and >>> belatedly discovering that the update screwed up something >>> else, etc.) >>> >>> A lot of my time is spent managing disk space. I don't like >>> big drives -- preferring, instead, multiple spindles (70 - 140G >>> per spindle). So, if I install something and "C:" fills up, >>> I have to move something off to another spindle, etc. >>> >>> And, if that "something" was referenced anyplace (shortcuts, >>> MRU lists, etc.) then there is some cost "fixing" those >>> references. >>> >>> By contrast, the UN*X boxen let me move things freely and >>> don't rely on as many "references" as Windows does. >> >> True. Although when this incident happened it wasn't really a big deal >> to re-install stuff. Most of that ran in the background while I was >> doing some design on another computer. In hindsight I was glad I >> didn't mirror because Dell had furnished some new and truly better >> drivers for this machine. For example, now it could display movie >> files in fast realtime. I need this for DICOM. IOW afterwards I had a >> laptop that was better and more valuable than when I bought it new. >> All for a new $60 (and much bigger) hard drive and the modest effort >> of a re-install. > > When I build a new machine, I go through each piece of software > that was on the *old* machine and ask myself: > - do I want to keep this? > - do I want to *upgrade* this? > - do I want to abandon this? > > I record detailed notes about each installation (what got installed, > where in the filesystem it resides, any tweeks I had to do to > menus, authorization codes, etc.). So, if I want to build a new > machine, I know what steps to follow. > > (often the order of installs makes a difference -- <frown> -- so > I keep track of what works... and *why* if I can sort that out!) Looks like we are very similar in that respect. The directory pattern is the same on all computers here, except that many softwares are only installed on one because of license restrictions. Each computer also has its own manila folder with the manuals, notes, disks and so on in there. If there's a problem I turn around on my chair, open a file hanger drawer and pull the respective file. -- Regards, Joerg http://www.analogconsultants.com/ "gmail" domain blocked because of excessive spam. Use another domain or send PM.
From: Joerg on 17 May 2010 19:23 Michael A. Terrell wrote: > Joerg wrote: >> Michael A. Terrell wrote: >>> Joerg wrote: >>>> Michael A. Terrell wrote: >>>>> D Yuniskis wrote: >>>>>> Hi Michael, >>>>>> >>>>>> Michael A. Terrell wrote: >>>> [...] >>>> >>>>>>> If I can scrape up the $100 before they others are gone, I'd prefer a >>>>>>> newer and faster server that draws less power. :) >>>>>> Why? :> Is there a reason you *need* a "server" instead of >>>>>> a regular desktop machine? I.e., I only run my servers >>>>>> because I want long term reliability (power supplies and >>>>>> spindles). Most of my regular work I do on consumer >>>>>> grade machines. Servers tend to get reliability at the >>>>>> expense of noise and size. :< >>>>> The server can go into a closet with the other network hardware, >>>>> cable modem and netwoerk switch hardware. It free up desk space where I >>>>> work on computers, as well. ... >>>> And makes your electric meter really spin up :-) >>> >>> And? >>> >> Oh yeah, I forgot, it just comes out of the wall outlet so it's free :-) > > No. 'And' means I can work faster from a local server. The quicker > I finish what I'm working , the sooner it gets shut down. Without a > dedicated server, files are scattered over hard drives, CD-ROM, DVD & > USB sticks all over the place. I may have five computers up to find > what I need. I also want to run a real NNTP server to allow me to > filter out some idiots and countries before they ever hit my computer. > That can also be done by a non server-grade PC, they usually use less electricity. > > Would you like to compare electric usage? > WRT my file server, yes, anytime. Hint: It runs on a wee 12V line. -- Regards, Joerg http://www.analogconsultants.com/ "gmail" domain blocked because of excessive spam. Use another domain or send PM.
From: D Yuniskis on 17 May 2010 19:41
Hi Joerg, >>>>> I don't know all the details because I just let the mfg's software >>>>> figure that out. I believe it's DHCP. Essentially you plug them in >>>>> and after a few seconds the printers are being recognized. Once >>>>> after some >>>> >>>> But how do you *reference* them? I.e., how do you "pick" >>>> which printer you want to talk to? (since their names/addresses >>>> aren't "constant") >>> >>> They are mapped to always the same names. One is always "HP Laser >>> 5L", the other is always "Brother MFC7820N" and so on. That's how >>> they show up in the print menu on any program. >> >> Ah, OK. So, if you had *two* of a particular printer, there >> is no way to tell them apart (?) (i.e., in a large business) > > Sure you can. I accepted the default names because there's only one > each. But you could also name one Alpha, the next one Bravo, or > whatever. Or HP5L_Office, HP5L_Lab. So, it gives you some way of *naming* them? I.e., if I you went out tomorrow and bought *two* MORE LJ5L's and plugged both of them in at the same time, how would they appear in your "printers list"? How would you know which is A and which B? >>>>> Now they all show up in the printer selection and you just pick the >>>>> one you want. But I didn't give them names like you did :-) >>>> >>>> I'm still unsure how you know which printer you are talking to. :< >>>> My UN*X and Sun boxes need (want) name/IP addresses. So, if I let >>>> the devices take their IP addresses from a DHCP server, then >>>> *I* don't know what they are from one boot to the next... >>> >>> Maybe Windows is better in that respect :-) >> >> Windows broadcasts "device names" and expects "folks" to notice >> those broadcasts. I.e., it can't find your printer until your >> printer says "Hi, I'm here! My name is..." > > But what's wrong with that? This way if it doesn't answer a light may > come on in your head such as "Oh, yes, I serviced it yesterday and > probably forgot to turn it back on". It doesn't scale well. Imagine 500 nodes all announcing themselves to the world *periodically* -- even if no one cares about them! And, everyone has to remember who has announced themselves (or, force the user to wait for a cycle before you know who/what's available) >> By contrast, I name everything explicitly (i.e., I can >> always find my web server or file server using its IP) >> and talk *to* them (instead of waiting for them to tell >> me they are available). > > I don't want to talk to them before they are available. Else you'll have > print jobs hanging in the queue forever and when in a rush that's not > good. I'd rather see that a printer has reported for duty and _then_ > send a job to it. After all, that's how the whole military works, once a > soldier reaches his station he must report to the sergeant that he did :-) But the same naming mechanism is used for machines as well. And, just because something announces itself as "available", doesn't mean it will *still* be available when you use it. >> Imagine if you had to wait for google.com to tell you >> "I'm here (in case you want to access me)", etc. > > I don't see anything wrong on a LAN. In fact, it's good. If the printer > doesn't show that means something electrical must be broken. MS protocols aren't used on the 'net because they don't scale. There's be nothing but hundreds of thousands of announcements (which have to be BROADCAST since you have no idea who might want to reference a particular name, etc.) Even MS realized their folly when they belatedly added support for the TCP/IP protocols. Different for SOHO applications in which you can eliminate the need for a name server, etc. >> When I build a new machine, I go through each piece of software >> that was on the *old* machine and ask myself: >> - do I want to keep this? >> - do I want to *upgrade* this? >> - do I want to abandon this? >> >> I record detailed notes about each installation (what got installed, >> where in the filesystem it resides, any tweeks I had to do to >> menus, authorization codes, etc.). So, if I want to build a new >> machine, I know what steps to follow. >> >> (often the order of installs makes a difference -- <frown> -- so >> I keep track of what works... and *why* if I can sort that out!) > > Looks like we are very similar in that respect. The directory pattern is > the same on all computers here, except that many softwares are only > installed on one because of license restrictions. Each computer also has > its own manila folder with the manuals, notes, disks and so on in there. > If there's a problem I turn around on my chair, open a file hanger > drawer and pull the respective file. I have been trying to separate functionality between machines. Just too damn hard to keep a (Windows) machine running when you have a boatload of different applications on it. Things like DirectX always seem to get wonky based on who installs first, etc. (I think Windows remembers too much state whereas other OS's rediscover things each time they run) I've discarded (after scanning) almost all of my manuals to save space. Aside from service manuals, most of the stuff that I might need I can usually find online (e.g., "What's error 50?"). I used to keep my logs in bound books. But, that was hard to maintain. Build a new machine, have to write a whole new book?? So, now I keep each machine's log on the machine itself. Of course, if the machine dies, the log is inaccessible. So, I push a copy onto my FTP server each time I make a change to it (this is a weak link as there is nothing to force me to do this :< ). When I build a new machine, first thing I do is retrieve the log from FTP server for whichever machine is closest in personality to the one I am building. <shrug> I wouldn't recommend this approach to larger businesses. But, for me, it is a good balance between recordkeeping and efficiency (I can't afford the time to *be* an IT department!) |