| --- Log opened Wed Aug 16 00:00:02 2006 |
| 00:59 | | ReivZzz is now known as Reiver |
| 01:27 | | himi-sleeping is now known as himi |
| 01:35 | | * Vornicus starts putting together a Continuous Integration machine, gets it to show a yellow screen that says "I need more RAM! ;_;" |
| 01:35 | <@Chalcedon> | hehe |
| 01:36 | < Vornicus> | It does, too. |
| 01:36 | < Vornicus> | On one of the tests it runs, it's supposed to finish in a few seconds... |
| 01:36 | < Vornicus> | but after ten minutes of pegging the processor so hard that Task Manager hangs, I decided that it wasn't going to go anywhere. |
| 01:44 | <@ToxicFrog> | o.O |
| 01:46 | | MahalWork [~Mahal@Nightstar-5049.worldnet.co.nz] has quit [Connection reset by peer] |
| 01:47 | < Vornicus> | ...I think it's a statement on the quality of my friends list that I can make a post about my sheets desperately needing a wash because of all the dust, and get comments that educate me on the nature of haiku. |
| 01:49 | < himi> | Well, you said /dust/ |
| 01:49 | < himi> | Dust is relatively inoccuous |
| 01:49 | < himi> | Had you said, say, mould, you might have gotten more interesting comments |
| 01:50 | <@Chalcedon> | I would have thought someone would offer to wash them for you! |
| 01:50 | < Vornicus> | Yes, perhaps I would have had a treatise on Shakespeare's effects on the sonnet. |
| 01:51 | < Vornicus> | Or perhaps a dissertation on the orbital characteristics of Kuiper Belt objects. |
| 01:51 | < Vornicus> | :P |
| 01:52 | <@Chalcedon> | o.O |
| 01:52 | < Vornicus> | :P |
| 01:52 | < Vornicus> | I said that already. |
| 02:05 | | Syloq is now known as Syloqs-AFH |
| 03:01 | | Vornicus [~vorn@Nightstar-18307.slkc.qwest.net] has quit [Quit: ] |
| 03:42 | | MahalWork [~Mahal@Nightstar-5049.worldnet.co.nz] has joined #Code |
| 03:43 | | mode/#code [+o MahalWork] by ChanServ |
| 03:44 | | Vornicus-Latens is now known as Vornicus |
| 03:45 | | Vornicus is now known as NSGuest-722 |
| 03:45 | | NSGuest-722 is now known as Vornicus |
| 04:13 | | Syloqs-AFH [Syloq@NetAdmin.Nightstar.Net] has quit [Connection reset by peer] |
| 04:19 | | * himi idly ponders Vornicus |
| 04:19 | <@Vornicus> | yees? |
| 04:22 | < himi> | I could post comments about exterior ballistics, ocean waves, armour penetration of large caliber naval shells, modelling of power requirements for ship propulsion . . . . |
| 04:22 | < himi> | Probably a few other things . . . |
| 04:22 | <@Vornicus> | heh |
| 04:23 | < himi> | I'd need to do research to post about the orbital characteristics of Kuiper Belt objects, I'm afraid |
| 05:26 | | Chalcy [~Chalceon@Nightstar-869.bitstream.orcon.net.nz] has joined #code |
| 05:26 | | mode/#code [+o Chalcy] by ChanServ |
| 05:28 | | Chalcedon [~Chalceon@Nightstar-869.bitstream.orcon.net.nz] has quit [Ping Timeout] |
| 05:31 | | himi is now known as himi-coffee |
| 05:58 | | Mahal [~Mahal@Nightstar-5049.worldnet.co.nz] has quit [Quit: reboot] |
| 06:03 | | Mahal [~Mahal@Nightstar-5049.worldnet.co.nz] has joined #code |
| 06:03 | | mode/#code [+o Mahal] by ChanServ |
| 06:26 | | Mahal [~Mahal@Nightstar-5049.worldnet.co.nz] has quit [Quit: It's hard to be mad at someone who misses you while you're asleep. ] |
| 06:27 | | Chalc [~Chalceon@Nightstar-869.bitstream.orcon.net.nz] has joined #code |
| 06:29 | | Chalcy [~Chalceon@Nightstar-869.bitstream.orcon.net.nz] has quit [Ping Timeout] |
| 06:42 | | Serah [~Shemhazai@Nightstar-8502.ds1-ba.adsl.cybercity.dk] has quit [Quit: Connection reset by evil CHILD EATING MONSTERS!] |
| 06:52 | | Chalc is now known as Chalcedon |
| 06:53 | | Reiver is now known as ReivF0d |
| 06:57 | | EvilZZZLord is now known as EvilDarkLord |
| 07:09 | | ReivF0d is now known as Reiver |
| 07:40 | | Vornicus is now known as Vornicus-Latens |
| 08:00 | | Serah [~Shemhazai@Nightstar-27025.bynxx4.adsl-dhcp.tele.dk] has joined #Code |
| 09:01 | | Mahal [~Mahal@Nightstar-3665.worldnet.co.nz] has joined #code |
| 09:01 | | mode/#code [+o Mahal] by ChanServ |
| 09:19 | | Mahal [~Mahal@Nightstar-3665.worldnet.co.nz] has quit [Quit: dessert] |
| 10:07 | | Chalcy [~Chalceon@Nightstar-869.bitstream.orcon.net.nz] has joined #code |
| 10:07 | | mode/#code [+o Chalcy] by ChanServ |
| 10:09 | | Chalcedon [~Chalceon@Nightstar-869.bitstream.orcon.net.nz] has quit [Ping Timeout] |
| 10:13 | | Chalcy is now known as Chalcedon |
| 10:29 | | Mahal [~Mahal@Nightstar-5049.worldnet.co.nz] has joined #code |
| 10:29 | | mode/#code [+o Mahal] by ChanServ |
| 10:33 | | EvilDarkLord is now known as EvilF0dLord |
| 11:06 | | EvilF0dLord is now known as EvilDarkLord |
| 11:21 | | Mahal [~Mahal@Nightstar-5049.worldnet.co.nz] has quit [Quit: This computer just fell asleep.] |
| 14:46 | | ToxicFrog|AFK is now known as ToxicFrog|W`rkn |
| 14:49 | | Chalcy [~Chalceon@Nightstar-869.bitstream.orcon.net.nz] has joined #code |
| 14:49 | | mode/#code [+o Chalcy] by ChanServ |
| 14:49 | | Chalcedon [~Chalceon@Nightstar-869.bitstream.orcon.net.nz] has quit [Ping Timeout] |
| 15:15 | | Reiver is now known as ReivZzz |
| 15:46 | | ReivZzz is now known as ReivSLEP |
| 16:22 | | Syloq [Syloq@NetAdmin.Nightstar.Net] has joined #code |
| 16:52 | | Vornicus [~vorn@Nightstar-18307.slkc.qwest.net] has joined #code |
| 16:52 | | mode/#code [+o Vornicus] by ChanServ |
| 17:44 | | Chalcy [~Chalceon@Nightstar-869.bitstream.orcon.net.nz] has quit [Ping Timeout] |
| 17:52 | | Serah [~Shemhazai@Nightstar-27025.bynxx4.adsl-dhcp.tele.dk] has quit [Quit: Connection reset by evil CHILD EATING MONSTERS!] |
| 18:18 | <@ToxicFrog|W`rkn> | Hmm. |
| 18:18 | <@ToxicFrog|W`rkn> | I need error handling advice. |
| 18:21 | <@ToxicFrog|W`rkn> | In specific, CK_Receive() has two success modes (message and signal) and no failure modes, and CK_Send() has one success mode (message) and two failure modes (invalid initial state, transition to invalid state while blocked). |
| 18:22 | <@ToxicFrog|W`rkn> | CK_Reply() has one of each. |
| 18:22 | <@ToxicFrog|W`rkn> | I'm wondering if I should go with exception throws, or simply return an error code (since I'm already returning some form of status code to distinguish between message and signal) |
| 18:26 | <@ToxicFrog|W`rkn> | An alternate approach is not to return exception throws, but instead have some kind of signal handler mechanism, although it'll still only get tripped when CK_Receive() is called. |
| 18:26 | | himi-coffee [~himi@Nightstar-29007.ACT.netspace.net.au] has quit [Operation timed out] |
| 18:28 | | himi-coffee [~himi@Nightstar-29007.ACT.netspace.net.au] has joined #code |
| 18:30 | <@Vornicus> | Do you have strict exception handling like Java? |
| 18:32 | <@ToxicFrog|W`rkn> | How do you mean 'strict'? |
| 18:32 | <@Vornicus> | In the sense that you know or can determine easily what functions would see the exception. |
| 18:33 | <@ToxicFrog|W`rkn> | Aah. No. You can throw from anywhere, and it goes up the call stack until it either gets caught or leaves the VM (which in this case terminates the calling thread) |
| 18:35 | <@Vornicus> | I think the world needs an IDE that figures out exception mapping on its own. |
| 18:35 | <@Vornicus> | I'd prefer error codes then. |
| 18:37 | | * ToxicFrog|W`rkn nods. |
| 18:37 | <@Vornicus> | mnyeh. I don't know. Frankly error codes grate on me, because I really have to check against all of them right there, and that's a pain... but exceptions tend to go wild unless you're really careful with them. |
| 18:37 | <@Vornicus> | ...though, uh |
| 18:38 | <@ToxicFrog|W`rkn> | The convention I'm using right now is that the programmer misusing the API (eg, calling CK_Send() with no arguments) raises an error, and other conditions (such as a thread dying while you're reply blocked on it) are what I'm working on now... |
| 18:38 | <@Vornicus> | I don't really know which is worse: an ignored error code that causes problems later, or an ignored exception that kills your application. |
| 18:38 | | himi-coffee [~himi@Nightstar-29007.ACT.netspace.net.au] has quit [Operation timed out] |
| 18:38 | <@ToxicFrog|W`rkn> | I think the former, because the latter is immediately visible. |
| 18:39 | <@ToxicFrog|W`rkn> | And tells you exactly where the mistake is. |
| 18:39 | <@Vornicus> | True, this. |
| 18:39 | <@Vornicus> | BUt then, the latter also seems to encourage /wilfully/ ignoring exceptions using catch/pass |
| 18:39 | | himi-coffee [~himi@Nightstar-29007.ACT.netspace.net.au] has joined #code |
| 18:40 | <@ToxicFrog|W`rkn> | Which is why I'm using exceptions for reporting API misuse: because that's definitely a /bug/, not a valid failure mode. |
| 18:41 | <@ToxicFrog|W`rkn> | But it's less clear when, say, you call CK_Send() on a thread that isn't there. Or a thread that is there, but goes away before calling CK_Reply(). |
| 18:41 | <@ToxicFrog|W`rkn> | When a thread vanishes, should the threads blocked on it throw errors? Or should they return { <<invalid thread handle>>; "error"; "target thread has exited"; } |
| 18:43 | <@ToxicFrog|W`rkn> | Honestly, I would definitely be doing with exceptions if exception handling in Lua was easier. |
| 18:45 | <@ToxicFrog|W`rkn> | And then there's ITC return conventions. Argh. |
| 18:48 | <@Vornicus> | So many ways of returning values and handling errors. |
| 18:50 | <@ToxicFrog|W`rkn> | Yeah. |
| 18:51 | <@Vornicus> | And not one of them always works. |
| 18:53 | <@ToxicFrog|W`rkn> | And I'm getting into areas where my implementation (and the language capabilities) are sufficiently different from how SGOS handles things that I can't just look at the SGOS specs and model it on that. |
| 19:10 | | * ToxicFrog|W`rkn idly throttles the man pages for sem_* |
| 19:11 | <@ToxicFrog|W`rkn> | If I didn't already know the semantics of semaphores, these would be almost entirely useless |
| 19:15 | <@ToxicFrog|W`rkn> | ...dammit. I have this really interesting way of doing ITC, but it doesn't work if I need reference resolution. |
| 19:17 | <@ToxicFrog|W`rkn> | Hmm. Or maybe it does. |
| 19:18 | <@ToxicFrog|W`rkn> | Ok. One thread manipulating both thread's VMs, coordinating with barriers? Or both threads, one pushing onto a queue and one popping, coordinating with semaphores? |
| 19:19 | <@ToxicFrog|W`rkn> | ...oh damn. |
| 19:19 | <@ToxicFrog|W`rkn> | Reference resolution doesn't work for signals this way. |
| 19:20 | <@ToxicFrog|W`rkn> | Implementing signals a la SGOS is sounding better and better ;.; |
| 19:29 | | Chalcedon [~Chalceon@Nightstar-869.bitstream.orcon.net.nz] has joined #code |
| 19:29 | | mode/#code [+o Chalcedon] by ChanServ |
| 19:47 | <@ToxicFrog|W`rkn> | Yaye, more design questions. |
| 19:47 | <@ToxicFrog|W`rkn> | Thread handle to name mappings! |
| 19:48 | <@ToxicFrog|W`rkn> | The basic premise is thus: |
| 19:49 | <@ToxicFrog|W`rkn> | - a thread can be registered; it can have a name that you can use as a synonym for the thread handle in any ITC functions. |
| 19:49 | <@ToxicFrog|W`rkn> | - a thread can be named: it can have some descriptive name that is used in debugging output and the like. |
| 19:50 | <@ToxicFrog|W`rkn> | So, are these the same thing? |
| 19:51 | <@ToxicFrog|W`rkn> | Every time I start leaning towards one answer, I realize that's wrong and I should use the other~ |
| 19:51 | | * Vornicus isn't sure. |
| 19:51 | <@ToxicFrog|W`rkn> | In SGOS, they're different. |
| 19:52 | <@ToxicFrog|W`rkn> | So you have, say, a hundred threads named "HTTP Worker", and the one thread named "HTTP Admin" registered as CT_HTTP_ADMINISTRATOR |
| 19:52 | <@ToxicFrog|W`rkn> | Which is what I'm presently leaning towards. |
| 19:52 | <@Vornicus> | I would say that having them be different is okay - but I'd also suggest that the name name defaults to the registered name. |
| 19:53 | <@ToxicFrog|W`rkn> | A thread's name is a property of the thread, which lets you map a thread handle to a human-readable name; a thread registration is a property of the process, which lets you map a human-readable name to a unique thread handle. |
| 19:53 | <@ToxicFrog|W`rkn> | Yes. |
| 19:53 | <@ToxicFrog|W`rkn> | CK_Register() implicitly sets the thread's name if it's not already set. |
| 19:53 | <@Vornicus> | ok |
| 19:54 | <@ToxicFrog|W`rkn> | So, following the above example, the admin would call CK_Register("HTTP Administrator"), and the workers would each call CK_Set_my_thread_name("HTTP Worker") |
| 19:54 | <@ToxicFrog|W`rkn> | Hmm. |
| 19:55 | <@ToxicFrog|W`rkn> | I might always want to implment CK_Wait_for_registration, which implies some kind of signaling mechanism for registration. |
| 19:56 | | * ToxicFrog|W`rkn finds himself adding a lot more ITC calls for this version. |
| 19:56 | <@Vornicus> | ITC = Inter-Thread Communication? |
| 19:56 | <@ToxicFrog|W`rkn> | Yes. |
| 19:57 | <@ToxicFrog|W`rkn> | Last one had Send, Receive, Reply, Create, My_thread, My_creator. This one adds Signal, Register, Register_wait, My_name, Set_my_name, and probably functions for mapping between names, registrations and thread handles. |
| 19:59 | <@ToxicFrog|W`rkn> | (oh, and CK_Wait(), which is actually based on pthread_join; there's no equivalent in SGOS) |
| 20:02 | | * ToxicFrog|W`rkn fiddles with the register_wait logic. |
| 20:03 | <@ToxicFrog|W`rkn> | pthread_barrier_wait, pthread_cond_wait or sem_wait? |
| 20:03 | <@ToxicFrog|W`rkn> | sem_wait is easiest, cond_wait is fastest. |
| 20:05 | <@Vornicus> | Start with the easiest. If it turns out to be too slow, use a faster one. |
| 20:06 | <@ToxicFrog|W`rkn> | It would be easier if I could find documentation for pthread_cond_wait that didn't suck. |
| 20:08 | <@ToxicFrog|W`rkn> | Such as, say, explaining what the hell the mutex is for, and the behaviour if a thread returns from pthread_cond_wait and doesn't unlock it. |
| 20:09 | | * Vornicus goes shopping for computers! |
| 20:11 | <@ToxicFrog|W`rkn> | Ok, on further reading, pthread_cond_wait is actually pretty damn scary. I'm sticking to semaphores. |
| 20:13 | <@Vornicus> | Hokay. need cusstons of RAM, a decent processor, and a big hard drive. Need it to run XP Pro. |
| 20:13 | <@ToxicFrog|W`rkn> | Why? |
| 20:13 | <@ToxicFrog|W`rkn> | (and are we talking desktop-cusstons, gaming-cusstons, or image-editing-cusstons?) |
| 20:14 | <@Vornicus> | DB server cusstons. |
| 20:14 | <@Vornicus> | I need it to run XP pro because our networking widget is apparently in C#. |
| 20:15 | <@ToxicFrog|W`rkn> | - me. |
| 20:15 | <@Vornicus> | We're not talking datacenter size, though. |
| 20:20 | <@Vornicus> | ...I wish HP's site were more easily navigable, I want to check over multiple models with similar capabilities in certain areas. |
| 20:21 | <@Vornicus> | I'm actually halfway tempted to build my own, but that has never ended well. |
| 20:28 | <@ToxicFrog|W`rkn> | ...it hasn't? |
| 20:30 | <@Vornicus> | Nope. |
| 20:31 | <@Vornicus> | I have yet to build a machine that has lasted more than a year. |
| 20:32 | <@ToxicFrog|W`rkn> | Freaky. |
| 20:33 | <@Vornicus> | Which is a pain, because apparently getting the kind of machine I want costs a thoroughly batshit amount. |
| 20:34 | <@ToxicFrog|W`rkn> | Hrm. Well, if it's a database server you don't need lots of video or anything...I'd say you could probably put it together for under $1000. |
| 20:34 | <@Vornicus> | Again, this would technically be easier if I could set some requirements and have the site make me offers. |
| 20:35 | <@Vornicus> | As opposed to having to pick a model and then add stuff to it. |
| 20:35 | <@ToxicFrog|W`rkn> | Depending on how much load this thing has to stand, anyways, but that'll get you a fast dualcore processor, 2-3GB of memory, 300-600GB of disk, and motherboard/case/PSU, depending on how you divide it up. |
| 20:36 | <@Vornicus> | Yeah I'm not getting near that on any prebuilts. |
| 20:36 | <@ToxicFrog|W`rkn> | Yes, it doesn't go as far with the 50% markup~ |
| 20:37 | <@ToxicFrog|W`rkn> | Also check that the prebuilts aren't trying to sell you monitors as well? |
| 20:48 | | Mahal [~Mahal@Nightstar-5049.worldnet.co.nz] has joined #code |
| 20:48 | | mode/#code [+o Mahal] by ChanServ |
| 20:56 | <@ToxicFrog|W`rkn> | 'morning, Mahal. |
| 20:56 | <@ToxicFrog|W`rkn> | Have a sem_t. |
| 20:56 | <@Mahal> | Morning Frog. |
| 20:56 | <@Mahal> | Have a ? |
| 20:57 | <@ToxicFrog|W`rkn> | sem_t == Semaphore. It's a construct used for thread synchronization. |
| 20:59 | | Chalcedon is now known as Chalcicle |
| 20:59 | <@ToxicFrog|W`rkn> | o.O |
| 21:07 | <@Vornicus> | THat would be a good point. the db server does not need a monitor. |
| 21:08 | <@Mahal> | Servers rarely do |
| 21:08 | | Chalcicle is now known as Chalcedon |
| 21:08 | <@Mahal> | Although we do have one that /will not boot/ without a keyboard. |
| 21:08 | <@Vornicus> | ...but the number I'm getting ($1300 from HP) has no monitor. |
| 21:13 | <@ToxicFrog|W`rkn> | Keh. |
| 21:13 | <@ToxicFrog|W`rkn> | Mahal: you can probably turn that off in BIOS. |
| 21:13 | <@ToxicFrog|W`rkn> | OTOH, we have a prototype here that won't boot with a keyboard, not because it won't /boot/ per se, but because it takes three and a half hours to do so... |
| 21:14 | <@Mahal> | Frog: we probably can turn it off. |
| 21:27 | | * ToxicFrog|W`rkn arghs. |
| 21:27 | <@ToxicFrog|W`rkn> | There is no god, only lots and lots of race conditions. |
| 21:31 | <@Chalcedon> | ? |
| 21:33 | <@ToxicFrog|W`rkn> | I'm trying to implement CK_Register (registers a thread's name, which can then be used by other threads in place of the thread's handle) and CK_Lookup (translates a registered thread name into a thread handle; if the name isn't registered, blocks until it is) |
| 21:35 | <@ToxicFrog|W`rkn> | I am running into issues where it is remotely possible to, for example, call CK_Lookup(), and in between it releasing the lock on the thread registry and grabbing the semaphore for the name it's waiting on, the corresponding thread registers the name, the semaphore is destroyed and we end up calling sem_wait on either an unitialized semaphore or a block of freed memory. |
| 21:40 | <@ToxicFrog|W`rkn> | This is not a good thing. |
| 21:45 | <@Mahal> | Oh dear. |
| 21:45 | <@ToxicFrog|W`rkn> | pthread_cond_wait() might solve this problem - the documentation is unclear - but introduces some very scary issues of its own. |
| 21:47 | <@ToxicFrog|W`rkn> | Mainly stemming from the fact that when it returns, there's no guarantee that it was actually signalled or broadcast. |
| 21:47 | <@ToxicFrog|W`rkn> | It might just randomly return without warning. |
| 21:51 | <@Chalcedon> | oops |
| 21:51 | <@Chalcedon> | o.O |
| 21:57 | <@ToxicFrog|W`rkn> | The really irritating bit is that if pthreads already used SRR, this is a trivially easy problem to solve~ |
| 22:02 | | Syloq is now known as Syloqs-AFH |
| 22:21 | | Serah [~Shemhazai@Nightstar-8502.ds1-ba.adsl.cybercity.dk] has joined #Code |
| 22:34 | <@Vornicus> | Arg. I suck at shopping! |
| 22:34 | | ToxicFrog|W`rkn is now known as ToxicFrog|AFK |
| 22:34 | | * Vornicus blings |
| 22:37 | | * Serah PatPats Vornicus. |
| 22:37 | < Serah> | Gave in to temptation? |
| 22:37 | <@Vornicus> | No, I just need to buy a server. |
| 22:38 | <@Vornicus> | And I'm shopping around. |
| 22:38 | <@Vornicus> | And I suck at it. |
| 22:40 | < Serah> | Oh, one which is shipped or leased? |
| 22:40 | <@Vornicus> | SHipped. |
| 22:40 | < Serah> | Want me to make people help? |
| 22:41 | <@Vornicus> | I need to get it, install software on it, stuff it back in the box, and send it halfway across the country. |
| 22:41 | < Serah> | -.^ Colocation? |
| 22:41 | <@Vornicus> | No |
| 22:41 | < Serah> | Hmm? |
| 22:42 | <@Vornicus> | Setting up a database server for a customer with no experience at all with computers. |
| 22:42 | < Serah> | Oh. |
| 22:43 | | * Mahal hugs Vornicus. |
| 22:43 | <@Chalcedon> | oh dear |
| 22:43 | | * Chalcedon proffers cookies |
| 22:43 | <@Vornicus> | But yeah, I just suck at shopping. |
| 22:43 | | * Serah dances with Vornicus. |
| 22:44 | <@Chalcedon> | first you need to know what you want |
| 22:44 | <@Vornicus> | I know what I want. Sorta. |
| 22:44 | <@Chalcedon> | write it down somewhere |
| 22:44 | <@Vornicus> | It's all the extras that these folks bundle in that I can't get rid of, and the secondaries (customer service) that I can't easily evaluate, that's the problem. |
| 22:44 | <@Chalcedon> | a spreadsheet would be good (for comparisons later) |
| 22:45 | <@Chalcedon> | what are you looking at? |
| 22:45 | < Serah> | I made people on my MSN list find something for me. |
| 22:45 | <@Chalcedon> | (in terms of extras) |
| 22:45 | <@Vornicus> | Technically I want /no/ extras. |
| 22:45 | <@Vornicus> | ALl it needs is a fat pipe, lots of RAM, a decent processor, and a big hard drive. |
| 22:46 | | * Mahal nod |
| 22:47 | <@Chalcedon> | so write a list: pipe > ideal size / RAM > ideal quantity (or minumum) / processor > minimum / HD : ideal size or range |
| 22:47 | <@Chalcedon> | you'll need an extra category for extras the ideal being none |
| 22:48 | <@Chalcedon> | then go round each of the manufacturers that you're looking at, pick the best offering from each and list it's specs along side the original |
| 22:48 | <@Chalcedon> | then you can score each for meeting or exceeding your requirements |
| 22:48 | < Serah> | Well, Holler if you need me, you might need to awake me since I'm sleepybeddybye. I offer to make people find stuff. Other than that I would only suggest what Chalcy is guiding you through. |
| 22:48 | <@Chalcedon> | night Serah |
| 22:48 | <@Vornicus> | And I'd build the thing myself, but that has never ended well. |
| 22:48 | < Serah> | Nighters. |
| 22:49 | <@Chalcedon> | so you're stuck with buying one. |
| 22:49 | <@Chalcedon> | you may find that the best one has lots of extras, which you'll just have to put up with or you may find that it has the least. |
| 22:49 | <@Chalcedon> | is price an issue? |
| 22:50 | <@Chalcedon> | can you give me a link to one of the ones you're looking at? |
| 22:52 | <@Vornicus> | Price is a sort of issue. |
| 22:53 | <@Vornicus> | Links are hard - these are those active pages, you choose stuff and the price updates automatically. |
| 22:54 | <@Chalcedon> | :s |
| 22:54 | <@Chalcedon> | maker and model number then? |
| 22:54 | <@Chalcedon> | (or base model number) |
| 22:54 | <@Vornicus> | RIght now I'm looking at HP's small business desktops and workstations, and the equivalent at Dell. |
| 22:55 | | Mahal [~Mahal@Nightstar-5049.worldnet.co.nz] has quit [Quit: off to work. ] |
| 22:57 | <@Chalcedon> | btw, what is a fat pipe? |
| 22:57 | <@Vornicus> | a nice ethernet card. |
| 22:57 | <@Vornicus> | like, gigabit. |
| 22:57 | <@Chalcedon> | k |
| 22:57 | <@Chalcedon> | ... actually that makes sense now.... |
| 22:58 | <@Vornicus> | heh |
| 22:58 | <@Chalcedon> | so what do you need in terms of RAM/processor/HD? |
| 22:59 | <@Chalcedon> | probably as minimums. |
| 23:00 | <@Vornicus> | 2GB 2GHz Notsurebutbig. |
| 23:01 | <@Vornicus> | I think I might have one... |
| 23:01 | <@Vornicus> | OKay, yes. |
| 23:01 | <@Chalcedon> | ? |
| 23:04 | <@Vornicus> | HP dx2200, 2GB RAM, 250GB hard drive, XP Pro, Celeron D 346 processor, Realtek 10/100 Ethernet (apparently gigabit would be overkill), 48x CD-ROM. $997. |
| 23:04 | <@Vornicus> | Hokay. |
| 23:05 | <@Chalcedon> | thats great :) |
| 23:05 | <@Vornicus> | It is. |
| 23:05 | | * Vornicus fiddles. |
| 23:06 | <@Chalcedon> | are you worried about checking for a better deal elsewhere? or is that merely going to complicate the situation? |
| 23:06 | <@Vornicus> | I think that will do, actually. Now I have to get a "better" and a "best" system figured out. |
| 23:08 | <@Chalcedon> | so, you're giving the customer several options? |
| 23:08 | <@Vornicus> | Yes. |
| 23:10 | | * Chalcedon whaps the Dell website |
| 23:10 | <@Chalcedon> | I wondered why it was all so expensive... it put me in NZ instead of US |
| 23:10 | <@Vornicus> | heh |
| 23:10 | <@Chalcedon> | do you need any more help or, have you figured out how you want to do things? |
| 23:11 | <@Vornicus> | I think I figured it out. |
| 23:12 | <@Chalcedon> | :) |
| 23:12 | | * Chalcedon gives Vorn a cookie |
| 23:12 | <@Vornicus> | Now I just have to take what I've listed and figure out how to improve on it. |
| 23:16 | <@Chalcedon> | better processor? |
| 23:16 | <@Chalcedon> | is your customer likely to want to upgrade to Vista? |
| 23:18 | <@Vornicus> | I don't think that will be an issue - this is the server. OS upgrades are very careful things. |
| 23:18 | < Serah> | Eee-ew. He shouldn't, Vista is a 64 thread gaming oriented OS. |
| 23:19 | <@Vornicus> | I would wait until Vista is at least three years old before doing that sort of madness. |
| 23:19 | < Serah> | Even when Vista is 3 years old, I might not want to use it anyway :p |
| 23:20 | <@Vornicus> | In any case I do not think my customer will even think about that. |
| 23:20 | <@Vornicus> | This machine will not even have a monitor. |
| 23:20 | <@Chalcedon> | right |
| 23:20 | <@Chalcedon> | I maintain that the question was worth asking. |
| 23:21 | <@Chalcedon> | (just to be sure that the answer really was no) |
| 23:26 | <@Vornicus> | ...now if I could tell what "HT" stands for I can choose an appropriate processor. |
| 23:27 | <@Chalcedon> | hyper threading technology? |
| 23:27 | | * Chalcedon contemplates tea |
| 23:28 | <@ToxicFrog> | Hyperthreading. |
| 23:28 | <@Vornicus> | ...okay now I feel dumb. |
| 23:28 | <@ToxicFrog> | Intel's dualcore hardware emulation. |
| 23:29 | <@Vornicus> | okay. |
| 23:29 | | * ToxicFrog would, personally, suggest either an Ath64 X2 or a Core 2, either of which will give you both better absolute performance and better performance/$, and in the case of the X2 may also cost less -- if these are options. |
| 23:30 | <@Vornicus> | I haven't found anybody selling AMD systems. |
| 23:30 | <@Chalcedon> | generally speaking... they don |
| 23:30 | <@Chalcedon> | *don't |
| 23:30 | <@ToxicFrog> | (Err. That is to say, it's hardware that emulates a dualcore processor, not a dualcore processor that emulates some other bit of hardware. But you may know that already.) |
| 23:30 | <@Vornicus> | (I knew that) |
| 23:31 | <@ToxicFrog> | As far as individual parts go, everyone does. As far as prefabs go, I know Gateway does, but I don't know about others. |
| 23:31 | <@Vornicus> | Actually, scratch that. I /have/ seen it, but it was rather expensive. Seeing as I've found my "good" system, I figure splurging on better ones would work. |
| 23:32 | <@Chalcedon> | that was part of our problem with Forj's CAD machine. The program runs better on AMD than intel, but few package manufacturers sell AMD computers |
| 23:33 | <@ToxicFrog> | IIRC, the current "best" processor is the Core 2, followed by the A64 X2, A64, P4, ferret wheel, and Celeron. |
| 23:36 | <@Chalcedon> | ferret wheel? |
| 23:37 | <@ToxicFrog> | Yes. It's a little wheel attached to the motherboard with a ferret in it, driving the pumps and levers that do the calculation. |
| 23:38 | | * Chalcedon gigles |
| 23:38 | <@Chalcedon> | *giggles |
| 23:38 | <@Vornicus> | Core2 being Intel Core Duo? |
| 23:38 | <@Chalcedon> | kind of like a mouse wheel then? |
| 23:38 | <@Vornicus> | hamster, surely. |
| 23:39 | <@Chalcedon> | or a hamster |
| 23:39 | <@Vornicus> | Except that ferrets are much more hyper. |
| 23:39 | <@Chalcedon> | i don't think of those because we don't ahve them here |
| 23:39 | <@Chalcedon> | I don't think a ferret would be very happy in a computer case... nor would the computer be happy with it in there... they're a bit to big, warm and furry. |
| 23:39 | <@Chalcedon> | possibly a stoat? |
| 23:39 | <@Chalcedon> | (or a weasel?) |
| 23:40 | <@Vornicus> | Stoats and weasels are also big warm and furry. |
| 23:40 | <@Vornicus> | Anyway, the wheel is on the outside of the case. |
| 23:40 | <@Chalcedon> | but not as big as a ferret |
| 23:40 | <@Vornicus> | Like a waterwheel. |
| 23:40 | | * Vornicus imagines a hydroelectric computer. |
| 23:40 | <@Chalcedon> | I think the wheel would be bigger than the case |
| 23:43 | <@Chalcedon> | mmmm.... tea...... warm |
| 23:44 | | * Vornicus imagines, really, a big waterwheel with its axle driving directly into an awesome steampunk computer case. |
| 23:45 | | * Chalcedon tiptoes very carefully around Vorn |
| 23:46 | <@Vornicus> | :D |
| 23:46 | <@Vornicus> | Okay. Good computer was $1,000. Better computer should probably be, oh, $1250-$1750. |
| 23:49 | <@ToxicFrog> | Vornicus: Intel Core 2 Duo if you want to be pedantic. |
| 23:49 | <@ToxicFrog> | Core 2 is the pattern, Duo indicates dualcoreness. |
| 23:49 | <@Vornicus> | ok. |
| 23:49 | <@Vornicus> | Mainly I was wondering if it was that Intel thing. |
| 23:56 | < Syloqs-AFH> | heh funny...i spent 700 on my puter and its lasted almost 3 years...go fig..heh |
| 23:58 | <@Vornicus> | This is a database server. |
| 23:59 | <@Vornicus> | What's a Pentium D? |
| --- Log closed Thu Aug 17 00:00:02 2006 |