Re: Mirror list
On 30 Oct 2005, lordSauron wrote: Cool. The initial rsync is done. I'm pulling from debian.csail.mit.edu currently twice a day. I'll set up to be a push server as soon as our networking group opens 22/tcp through the campus firewall. The URL is: --- http://mirror.tamu.edu/debian-amd64/ down/768Kbps up cable connection at home. I think we currently pay around $120,000USD for the OC-12 at work. A month or a year? That's an insane connection rate.. you've probably got a fibre LAN going in the building, right? Otherwise you'd be somewhat disabled by anything less than gigabit ethernet. They'll let you host that at your workplace though... That is per year. We have several pieces of fiber from our campus to Verizon (local phone provider). OC circuits use a different frame type separate from ethernet (see SONET/SDH information at Wikipedia if you want more information). We have a lot more than just data running across the fiber to the telco. Yes, Texas would be in my neck of the woods (I'm in California, so not *that* close, but closer than all the many German mirrors). I'm not sure if you were aware of debian.csail.mit.edu previously, but it should have been faster than the European mirrors for you. Anyway, try it or mine and see if things are better. -- Mark Nippere-contacts: 832 Tanglewood Drive[EMAIL PROTECTED] Bryan, Texas 77802-4013 http://nipsy.bitgnome.net/ (979)575-3193 AIM/Yahoo: texasnipsy ICQ: 66971617 -BEGIN GEEK CODE BLOCK- Version: 3.1 GG/IT d- s++:+ a- C++$ UBL$ P---+++ L+++$ !E--- W++(--) N+ o K++ w(---) O++ M V(--) PS+++(+) PE(--) Y+ PGP t+ 5 X R tv b+++@ DI+(++) D+ G e h r++ y+(**) --END GEEK CODE BLOCK-- ---begin random quote of the moment--- A good traveller has no fixed plans and is not intent on arriving. -- Lao Tzu (570-490 BC) end random quote of the moment -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Eclipse
Marc Petit-Huguenin wrote: Dalibor Topic wrote: Marcin Dbicki wrote: Can someone explain me why Eclipse depends on gij not java-virtual-machine or java-runtime. In order to move into main, eclipse needs to build and run with free software in main. gcj/gij are in main, the proprietary software you mentioned is not. How a virtual package can be considered, as you said, as a prorietary software? http://www.debian.org/doc/packaging-manuals/virtual-package-names-list.txt You mentioned Sun's and Blackdown's VMs, which are proprietary, so unfortunately they can't be used for the packaging task to move eclipse into main. The packagers used a specific free runtime to make the eclipse package build and work, so they made that runtime specifically part of the dependencies, as that's a configuration the packagers can focus on to support. You are most welcome to contribute, and help improve the eclipse packages. see you on debian-java, dalibor topic -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
problems with installation
Hello, I've just bought a new PC with amd64bit architecture. and wanted to install a debian linux. First I tried sarge for i386. I had installed it, but the installer couldn't find my network card. So with no possibility to get to the web, I tried an unstable for amd64. but on way of the installation of the base system, the installer threw an error (something like cannot chroot, no such mount point) even when I let automatically build the partitions. So my questions are: Is there a way, to run network and sound under sarge? Or can I do anything to install the base-system of the amd64 port? (I would prefer the second way) my system: proc: amd sempron64 3000+ mainboard: gigabyte k8NS nForce3 250 Chipset, Lan, Sound onboard graphic: ati radeon 9550 hd: samsung, spinpoint 160Gb SATA thanks for your help Jean -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Best approach for a system that needs desktop apps and 64bit address space
Hi all, we are making some tests with Linux operating systems for an environment where we need desktop apps (openoffice, mozilla w/ flash) and large memory (my machine has 8gigs) Also, we need openGL accelleration, nVidia driver. what would be the best approach? Install an ia32 system with 64SMP kernel or the whole 64 system with ia32 chroot? FYI, I am not an admin here, but the only guy with debian background. I am doing this tests without proper knowledge of the boss, under a sudo chroot set up by one of the admins here. -- Paulo Marcondes http://rj.debianbrasil.org
Re: Bug#314988: Still a problem?
On Fri, 2005-10-28 at 13:58 -0700, Steve Langasek wrote: On Fri, Oct 28, 2005 at 09:59:34PM +0200, Artur R. Czechowski wrote: On Fri, Oct 28, 2005 at 09:55:25PM +0200, Leopold Palomo Avellaneda wrote: I did a build-deps and and dpkg-buildpackage to make sure that I had all the build depens, but it seems as Ted said and I confirm that I needs the nvidia-glx-dev. The Question is: why the package needs nvidia-glx-dev on amd64 and on other platforms it builds succesfully without nvidia-glx-dev? Because of a longstanding bug in the nvidia lib package which creates a broken .so symlink when it's installed without nvidia-glx-dev as well. So should this be reassigned to nvidia-glx-dev? Is this the same issue as in 272438? -Adam -- GPG fingerprint: D54D 1AEE B11C CE9B A02B C5DD 526F 01E8 564E E4B6 Welcome to the best software in the world today cafe! http://www.take6.com/albums/greatesthits.html -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Best approach for a system that needs desktop apps and 64bit address space
pmarc wrote: Hi all, we are making some tests with Linux operating systems for an environment where we need desktop apps (openoffice, mozilla w/ flash) and large memory (my machine has 8gigs) Also, we need openGL accelleration, nVidia driver. what would be the best approach? Install an ia32 system with 64SMP kernel or the whole 64 system with ia32 chroot? FYI, I am not an admin here, but the only guy with debian background. I am doing this tests without proper knowledge of the boss, under a sudo chroot set up by one of the admins here. -- Paulo Marcondes http://rj.debianbrasil.org nVidia and IPTables don't work properly with the Install an ia32 system with 64SMP kernel option. What exactly do you plan on doing with all that memory? 32-bit apps can't (well, not really, earlier arguments on the mailing list withstanding) use all the memory, so the apps planning on using all that RAM would need to be 64bit --Jo Shields -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Eclipse
On Mon, Oct 31, 2005 at 08:22:31AM -0400, Ed Tomlinson wrote: On Monday 31 October 2005 05:28, Dalibor Topic wrote: The packagers used a specific free runtime to make the eclipse package build and work, so they made that runtime specifically part of the dependencies, as that's a configuration the packagers can focus on to support. You are most welcome to contribute, and help improve the eclipse packages. This does _not_ make a lot of sense. It would make much more sense to suggest gcj/gij and depend on java-virtual-machine. This leaves it up the the user to decide if he can use a non-free jvm. I my case many of the apps I use (non-debian) fail with the free jvms. In short this type of depends is, IMO a bug. It will force me, and many others, to bypass the packaging system, which is usually a bad idea. Your argument is only reasonable if your non-free Java environment is a complete drop-in replacement for building and running Eclipse. If not, then you're asking for extra work to be done to support multiple JVMs. If that's what you need, patches are probably welcome. Hamish -- Hamish Moffatt VK3SB [EMAIL PROTECTED] [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
unsubscribe
unsubscribe
Re: installing Oracle on Debian AMD64
On Tue, Oct 25, 2005 at 09:51:31PM +0200, Jean-Christophe Montigny wrote: Well, I am afraid I'm not quite postgresql-literate, and I live by the (perhaps false) assumption that PostgreSQL and MySQL are more or less the same : open source database projects, except PostgreSQL are supposed to be faster in reading and slower in writing than MySQL, and that they roughly have the same capabilities.. Postgres has much more complete SQL syntax support, and finer grain locking than mysql. They are quite different in features and their target markets. It is also easier to be fast writing if you lock the whole table and prevent others from accessing it while you update things. Slows down reading to stopped while you do a write though. Well, of course it doesn't mean anything to use Oracle in a small environment, as I said it only becomes good when you have a single DB that is being used by several clients and you need data consistency without having to modify all the clients when there's a structural change (for instance, say you add a table that needs updated when you do whatever action on the other tables -- Oracle allows you to code an event associated to that action - ie a procedure...). That's merely a scenario, of course. If you just need the standard functionalities of DB and don't mind having your client software ensuring data consistency, mysql and i guess postgresql perform fine and will even be faster than Oracle for very simple tasks. Merely a question of raw processing power. Postgres does have quite a lot of support for triggers and events to perform, although I have no idea how it compares to Oracle. It is a lot better than mysql though. Len Sorensen -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Best approach for a system that needs desktop apps and 64bit address space
nVidia and IPTables don't work properly with the Install an ia32 system with 64SMP kernel option. IPTables is not a problem here, I believe. However, doing without nVidia proprietary drivers is a no-go What exactly do you plan on doing with all that memory? 32-bit apps can't (well, not really, earlier arguments on the mailing list withstanding) use all the memory, so the apps planning on using all that RAM would need to be 64bit Simulation stuff. Some of the apps are 64, some 32, the desktop stuff should run with all the bells and whistles. -- Paulo Marcondes http://rj.debianbrasil.org
Thunderbird segmentation fault
Dear List!
My Thunderbird under Debian sarge amd64 keeps crashing giving the
following message:
[EMAIL PROTECTED]:~$ mozilla-thunderbird --g-fatal-warnings
selected locale: en-US
DOUBLE-CLICK: 250 -- -1 THRESHOLD: 8 -- -1
/usr/lib/mozilla-thunderbird/run-mozilla.sh: line 451: 4657
Segmentation fault $prog ${1+$@}
Can anybody tell me what to do?
regards, Ralph
PS: I searched the list archive without finding a solution. So hopefully
this issue hasn't been discussed in recent time :-)
PPS: My package information:
[EMAIL PROTECTED]:~$ dpkg-query -p mozilla-thunderbird
Package: mozilla-thunderbird
Priority: optional
Section: mail
Installed-Size: 37596
Maintainer: Alexander Sack [EMAIL PROTECTED]
Architecture: amd64
Version: 1.0.2-2
Provides: mail-reader, imap-client
Depends: libatk1.0-0 (= 1.7.2), libc6 (= 2.3.2.ds1-4), libfontconfig1
(= 2.3.0), libfreetype6 (= 2.1.5-1), libgcc1 (= 1:3.4.1-3),
libglib2.0-0 (= 2.6.0), libgtk2.0-0 (= 2.6.0), libpango1.0-0 (=
1.8.1), libstdc++6 (= 3.4.3-1), libx11-6 | xlibs ( 4.1.0), libxext6 |
xlibs ( 4.1.0), libxft2 ( 2.1.1), libxp6 | xlibs ( 4.1.0),
libxrender1, libxt6 | xlibs ( 4.1.0), zlib1g (= 1:1.2.1)
Recommends: myspell-en-us | myspell-dictionary, xprint
Suggests: mozilla-thunderbird-offline,
mozilla-thunderbird-typeaheadfind, mozilla-thunderbird-inspector,
mozilla-firefox, ttf-bitstream-vera, ttf-freefont,
mozilla-thunderbird-enigmail
Size: 12277684
Description: Mozilla Thunderbird standalone mail client
Mozilla Thunderbird is a redesign of the Mozilla mail component. The
goal is to produce a cross platform stand alone mail application using
the XUL user interface language. Mozilla Thunderbird leaves a somewhat
smaller memory footprint than the Mozilla suite.
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: installing Oracle on Debian AMD64
Lennart Sorensen wrote on 31/10/2005 15:41: On Tue, Oct 25, 2005 at 09:51:31PM +0200, Jean-Christophe Montigny wrote: Well, I am afraid I'm not quite postgresql-literate, and I live by the (perhaps false) assumption that PostgreSQL and MySQL are more or less the same : open source database projects, except PostgreSQL are supposed to be faster in reading and slower in writing than MySQL, and that they roughly have the same capabilities.. Postgres has much more complete SQL syntax support, That might have been the case before MySQL 5.0, but as far as I can tell, now MySQL has the more complete and more standard-compliant support for SQL 2003. and finer grain locking than mysql. It locks finer than a single column? They are quite different in features and their target markets. That's for certain. I wouldn't try using Postgres with database replication (i.e. automatically keeping two or more databases in sync with automatic updates when the master is updated etc.) currently. MySQL has done this quite fine for some years though. Full-text indexes and real-time replication support (or lack of mature solutions for this) really are downsides of Postgres. It is also easier to be fast writing if you lock the whole table and prevent others from accessing it while you update things. Slows down reading to stopped while you do a write though. True. But MySQL gives you the choice: MyISAM tables are fast in writing and might get slow if you concurrently try to read. InnoDB on the other hand is slower, but uses column-locking instead of table-locking. It's not really faster on reads than MyISAM if you do reads only, but it is faster on reads if you concurrently write to your tables. Postgres does have quite a lot of support for triggers and events to perform, although I have no idea how it compares to Oracle. It is a lot better than mysql though. Is it? What kind of trigger does Postgres support that MySQL (5) doesn't? I didn't find any when I browsed through the trigger documentation for each. I might have overlooked some trigger which isn't currently supported by MySQL, but is by Postgres. I really don't like this repeated Postgres is better than MySQL bashing that is mostly based on long-outdated versions of MySQL. Both Postgres and MySQL are good database systems. However, each seems to have its own pro´s and con´s, and it all depends on what _you_ want to do with your database. For me, real-time replication is far more important than some obscure ACID things (Adam Skutt said MySQL had tons of things which can cause transaction invalidation, but honestly: transactions are there specifically to catch these invalidations instead of running headlong into an inconsistant database, and I actually never experienced a single transaction failure with MySQL). And even though Adam said MySQL would lack working row-locking, this has never failed for me. Subqueries have been supported since MySQL 4.1 and 5.0 also gives stored procedures and proper trigger support. So all in all: Check _your_ needs regarding database functionality (both on SQL level and on the database management level). After that, you can decide which DB-software is the best for _you_ (i.e. fullfills your requirements and is relatively cheap). There simply is no best database software (though there might be the worst somewhere). cu, sven
Re: Problem with installing Debian AMD64 on hp workstation 6200
On Thu, Oct 27, 2005 at 11:21:17AM +0200, . wrote: Emmanuel Gamby schrieb: We have several hp workstation xw6200 (Xeon 3.20Ghz HT, EM64T) and we would like to install debian on them. I downloaded the debian-amd64-netinst.iso Shouldn?t you better use the ia64 distribution? That would be for an itanium. EM64T = AMD64 but made by intel. IA64 = Itanium. Len Sorensen -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Mirror list
On Sat, Oct 29, 2005 at 06:10:21PM -0700, mike wrote: it might be off topic for this list, but i think it may be advantageous to think about using some sort of p2p type mechanism in the future. imagine an infinite supply of mirrors to grab the files from, instead of centrally-hosted sites with what seems to be regular turnover. obviously would still need to have some sort of central servers with the ports/packages list to verify the md5sum's of the files to ensure integrity... could leverage existing protocols like bittorrent too. Someone suggested that in the past. As far as I recall here are some of the problems that were pointed out: You would need a .torrent per deb file Many files are small while bittorrent is best for larger files that can be divided into blocks for parallel transfers. Someone has to generate all these torrent files and then you have to mirror those around so apt-get can get the torrent to then connect to a tracker to then start downloading the deb. Often for smaller deb files the actual download from http would have been faster. bittorrent makes sense for cd images (although even there the jigdo system makes more sense since you can update a 3.1r0 cd image to a 3.1r1 cd image by only downloading the actual changes, while bittorrent would find most of the image had changed and do nothing useful with the old data). Besides the current set of mirrors seem pretty good to me. I often get 560KB/s on my 5Mbit link from the mirror I use. http works great for my needs. Len Sorensen -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: problems with installation
Jean Christoph Jung wrote: Hello, I've just bought a new PC with amd64bit architecture. and wanted to install a debian linux. First I tried sarge for i386. I had installed it, but the installer couldn't find my network card. So with no possibility to get to the web, I tried an unstable for amd64. but on way of the installation of the base system, the installer threw an error (something like cannot chroot, no such mount point) even when I let automatically build the partitions. So my questions are: Is there a way, to run network and sound under sarge? Or can I do anything to install the base-system of the amd64 port? (I would prefer the second way) my system: proc: amd sempron64 3000+ mainboard: gigabyte k8NS nForce3 250 Chipset, Lan, Sound onboard graphic: ati radeon 9550 hd: samsung, spinpoint 160Gb SATA thanks for your help Jean Forgot to reply to the list... Give one of these a try: http://tinyplanet.ca/~lsorense/amd64/ I have the same motherboard, and these installers worked for me. Though, I have had a lot of problems with the onboard NIC, so I eventually just disabled it in the BIOS and added a PCI NIC. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: installing Oracle on Debian AMD64
Sven Mueller wrote: It locks finer than a single column? MySQL only locks that fine if you're blessed enough not to be using MyISAM. Postgresql locks at less than table-granularity all the time. Full-text indexes and real-time replication support (or lack of mature solutions for this) really are downsides of Postgres. Too bad FTI on MySQL requires using nasty MyISAM It is also easier to be fast writing if you lock the whole table and prevent others from accessing it while you update things. Slows down reading to stopped while you do a write though. More like stops every other read and write. True. But MySQL gives you the choice: MyISAM tables are fast in writing and might get slow if you concurrently try to read. No, you have that backwards. They're incredibly quick to read, and absolutely kill /all/ concurrency when writes come in, because it's a table-write. Sure, that one write may be fast, but it doesn't take many concurrent writes or a high read load to drop your performance to almost nothing. That's ignoring all the other horrible things about MyISAM. I really don't like this repeated Postgres is better than MySQL bashing that is mostly based on long-outdated versions of MySQL. MySQL 5 isn't even a month old, and MySQL 4.1 hasn't been GA for even a full year. It's not long-outdated. And tons of shops still have 3.3.x in production. Both Postgres and MySQL are good database systems. However, each seems to have its own pro´s and con´s, and it all depends on what _you_ want to do with your database. For me, real-time replication is far more important than some obscure ACID things (Adam Skutt said MySQL had tons of things which can cause transaction invalidation, but honestly: transactions are there specifically to catch these invalidations instead of running headlong into an inconsistant database, and I actually never experienced a single transaction failure with MySQL). This kind of statement shows you don't understand what a trasnscation is about. The whole point of a transaction is to prevent an inconsistent database in the first place. And MySQL forbids transactions in places that are totally nonsensical. For example, any user-created lock (i.e., CREATE LOCK) invalidates the transaction context. So you say, what's the big deal? No other transaction can mess with yours, because of the lock. But what if your transaction fails? Now, the onus of rolling back any changes is on the user, not the DBMS. For complicated transactions (hell, maybe even simple ones, depending on schema), that's a major PITA. So no, it is a big deal, if you want to do more than just shove data in/out of a DB. Because not being able to keep a transaction inside a lock will cause an inconsistent database. The point of a transaction isn't to catch these invalidations. Frankly, I don't even know what you mean there, it makes no sense--the purpose of a transaction isn't to just to note scope. It's to provide an atomic context. And MySQL forbids you from having an atomic context everywhere it's useful, which is a major shortcoming. To be fair, PostgreSQL has one major shortcoming with it's transactions I forgot to mention: a UDF always has an implicit BEGIN/END transaction associated with it, which makes them impractical for bulk data loading. This is a big deal, as committing transactions in PostgreSQL is slow. But at least my data is still intact. And even though Adam said MySQL would lack working row-locking, this has never failed for me. It doesn't, because the user based locking is useless, mostly due to the above. Adam -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: problems with installation
On Mon, Oct 31, 2005 at 01:20:42PM +0100, Jean Christoph Jung wrote: Hello, I've just bought a new PC with amd64bit architecture. and wanted to install a debian linux. First I tried sarge for i386. I had installed it, but the installer couldn't find my network card. So with no possibility to get to the web, I tried an unstable for amd64. but on way of the installation of the base system, the installer threw an error (something like cannot chroot, no such mount point) even when I let automatically build the partitions. So my questions are: Is there a way, to run network and sound under sarge? Or can I do anything to install the base-system of the amd64 port? (I would prefer the second way) my system: proc: amd sempron64 3000+ mainboard: gigabyte k8NS nForce3 250 Chipset, Lan, Sound onboard The onboard lan uses an ICS chip, which last I checked had really really poor linux support. graphic: ati radeon 9550 Well good luck. I generally highly anti-recommend anything by ati for linux use. hd: samsung, spinpoint 160Gb SATA Len Sorensen -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: installing Oracle on Debian AMD64
Adam Skutt wrote on 31/10/2005 18:33: Sven Mueller wrote: It locks finer than a single column? MySQL only locks that fine if you're blessed enough not to be using MyISAM. True. Postgresql locks at less than table-granularity all the time. Fine. I never said it wouldn't. Full-text indexes and real-time replication support (or lack of mature solutions for this) really are downsides of Postgres. Too bad FTI on MySQL requires using nasty MyISAM I agree with you there. And in my opinion, this is definately a major drawback. It is also easier to be fast writing if you lock the whole table and prevent others from accessing it while you update things. Slows down reading to stopped while you do a write though. More like stops every other read and write. for as long as the write takes. True. But MySQL gives you the choice: MyISAM tables are fast in writing and might get slow if you concurrently try to read. No, you have that backwards. They're incredibly quick to read, and absolutely kill /all/ concurrency when writes come in, because it's a table-write. Sure, that one write may be fast, but it doesn't take many concurrent writes or a high read load to drop your performance to almost nothing. Actually, I didn't exactly get it backwards. MyISAM _is_ fast to write (and fast to read) as long as you only write or only read. If you try to do both concurrently, reads and writes get in each others way, which slows things down a lot. Reads are not possible while writing and writes can't get their locks set while some process is still reading. That's ignoring all the other horrible things about MyISAM. There are certainly other drawbacks in MyISAM, but I wouldn't call them horrible things. I really don't like this repeated Postgres is better than MySQL bashing that is mostly based on long-outdated versions of MySQL. MySQL 5 isn't even a month old, and MySQL 4.1 hasn't been GA for even a full year. It's not long-outdated. And tons of shops still have 3.3.x in production. 4.0 is far older than a year and fixed many things which where problematic in 3.x. 5.0 has reached GA a few weeks ago and its release candidates (and even its betas) effectively were really stable. That a lot of shops still run 3.x is their problem, the upgrade can easily be done with only a few (like 2) minutes of downtime if you are careful. If you don't want to be careful, it still doesn't take more than 15 minutes of downtime (and including compiling the complete mysql suite, not more than 30 minutes). Both Postgres and MySQL are good database systems. However, each seems to have its own pro´s and con´s, and it all depends on what _you_ want to do with your database. For me, real-time replication is far more important than some obscure ACID things (Adam Skutt said MySQL had tons of things which can cause transaction invalidation, but honestly: transactions are there specifically to catch these invalidations instead of running headlong into an inconsistant database, and I actually never experienced a single transaction failure with MySQL). This kind of statement shows you don't understand what a trasnscation is about. I simplified and possibly chose the wrong words. A transaction is there to do multiple updates in an atomic way (i.e. they are executed without interuption by other updates or reads). The side effect (on which I perhaps put to much emphasis) is that if a single update in the transaction fails, the whole transaction fails (and should be rolled back). The whole point of a transaction is to prevent an inconsistent database in the first place. Which doesn't contradict me. And MySQL forbids transactions in places that are totally nonsensical. For example, any user-created lock (i.e., CREATE LOCK) invalidates the transaction context. I don't understand what you would want a lock for inside a transaction. I mean either the transaction is atomic (and therefor doesn't need a lock) or it isn't (and might need one). Maybe I overlook something, but as far as I can tell, a transaction shouldn't need to set locks. There might be other things which invalidate transactions in MySQL, but I didn't come across anything I would have needed inside a transaction until now. So you say, what's the big deal? No other transaction can mess with yours, because of the lock. But what if your transaction fails? Now, the onus of rolling back any changes is on the user, not the DBMS. For complicated transactions (hell, maybe even simple ones, depending on schema), that's a major PITA. Just to make sure I understand you correctly: You say that if you use a user set lock inside a transaction, the transaction context is invalidated. And if I understand you correctly, this means that the actions taken inside the transaction are executed as if outside a transaction and therefor not automatically rolled back anymore. Right? Of course, manually rolling back multiple changes is a major
Re: Mirror list
On 10/31/05, Lennart Sorensen [EMAIL PROTECTED] wrote: On Sat, Oct 29, 2005 at 06:10:21PM -0700, mike wrote: it might be off topic for this list, but i think it may be advantageous to think about using some sort of p2p type mechanism in the future. imagine an infinite supply of mirrors to grab the files from, instead of centrally-hosted sites with what seems to be regular turnover. Well, if it's infinite, then EVERYONE is gonna need a bigger hard drive (I for one wouldn't mind if they sent me a nice happy 500gig SATA150 - I'd be more than happy to dedicate 300 gigs of that to a mirror, but as we all know, I'm in no place to host so much as a paperclip, much less a debian mirror!) obviously would still need to have some sort of central servers with the ports/packages list to verify the md5sum's of the files to ensure integrity... could leverage existing protocols like bittorrent too. Someone suggested that in the past. As far as I recall here are some of the problems that were pointed out: You would need a .torrent per deb file Many files are small while bittorrent is best for larger files that can be divided into blocks for parallel transfers. Someone has to generate all these torrent files and then you have to mirror those around so apt-get can get the torrent to then connect to a tracker to then start downloading the deb. Often for smaller deb files the actual download from http would have been faster. bittorrent makes sense for cd images (although even there the jigdo system makes more sense since you can update a 3.1r0 cd image to a 3.1r1 cd image by only downloading the actual changes, while bittorrent would find most of the image had changed and do nothing useful with the old data). Besides the current set of mirrors seem pretty good to me. I often get 560KB/s on my 5Mbit link from the mirror I use. http works great for my needs. I hate you not really, I just hate that you get such fast downloads... y'know, that evious sorta hate? -- === GCB v3.1 === GCS d-(+) s+:- a? C+() UL+++() P L++(+++) E- W+(+++) N++ w--- M++ PS-- PE Y+ PGP- t++(+++) 5? X? R !tv-- b++ DI+++ D-- G !e h(*) !r x--- === EGCB v3.1 ===
Re: Mirror list
well the idea was reusing the idea of a distributed/p2p mirroring concept. not necessarily tied to a specific protocol, but it is a standard... (and the .torrent files could be compiled just like the md5 signature files, still on a central mirror/mirrors, but those would only run trackers and not do a lot of seeding, it would be up to the rest of the world to seed) i know that i would run one of these. i would put an upload cap on it, since i don't have unmetered connectivity yet on my colocated cluster, but i would contribute. :) i'm just thinking of ways to offload reliance on the mirrors for -all- types of transfering. - mike On 10/31/05, Lennart Sorensen [EMAIL PROTECTED] wrote: Someone suggested that in the past. As far as I recall here are some of the problems that were pointed out: You would need a .torrent per deb file Many files are small while bittorrent is best for larger files that can be divided into blocks for parallel transfers. Someone has to generate all these torrent files and then you have to mirror those around so apt-get can get the torrent to then connect to a tracker to then start downloading the deb. Often for smaller deb files the actual download from http would have been faster. ...
Re: Mirror list
On 10/31/05, Mark Nipper [EMAIL PROTECTED] wrote: On 30 Oct 2005, lordSauron wrote: Cool. The initial rsync is done. I'm pulling from debian.csail.mit.edu currently twice a day. I'll set up to be a push server as soon as our networking group opens 22/tcp through the campus firewall. The URL is: --- http://mirror.tamu.edu/debian-amd64/ Awesome. I'll try plugging that into my sources.list file later after a bit more boring, required work is done. down/768Kbps up cable connection at home. I think we currently pay around $120,000USD for the OC-12 at work. A month or a year? That's an insane connection rate.. you've probably got a fibre LAN going in the building, right? Otherwise you'd be somewhat disabled by anything less than gigabit ethernet. They'll let you host that at your workplace though... That is per year. We have several pieces of fiber from our campus to Verizon (local phone provider). OC circuits use a different frame type separate from ethernet (see SONET/SDH information at Wikipedia if you want more information). We have a lot more than just data running across the fiber to the telco. SEVERAL Now I can believe $120,000/year. Give me your building addresses - I'm going to find you a cheaper provider. I want to see if a really good company called Hurricane Electric is right for you - they offer GIG-E connections, which is basically a gigabit ethernet connection to the internet. But no, you shouldn't be spending a eighth the cost of my house annually for internet. If you can get me this info, I can find you something cheaper: 1) building addresses 2) number of connections required per address (or the total bandwidth required) No, I'm not some weird Rasputin-esque rapist, I'm just one of the worlds greatest ranking tightwads who can't bear to see money spent, and I chafe greatly at the sight of money wasted. You're paying far too much and I need to find you some better service, even if I end up having to drive out to Texas and install it myself (I've offered to break through sheetrock to install a network connection in a public school before, so I'm normally this way, okay?) Yes, Texas would be in my neck of the woods (I'm in California, so not *that* close, but closer than all the many German mirrors). I'm not sure if you were aware of debian.csail.mit.edu previously, but it should have been faster than the European mirrors for you. Anyway, try it or mine and see if things are better. Yeah, goto Google Maps (www.maps.google.com) and goto Pleasanton California. That's where I live. Then find Amador Valley High School. That's where I go to school. I suggest veiwing in composite map mode, since that's the best blend of the local veiw and street directions. I won't tell you driectly where I live, since that's not smart on the internet, but I think you can safely tell me where your college is (you're just saving me the time of hunting down the addresses myself, and telling me how many connections you need). I tell you my city location b/c the espri.arizona mirror is closer than Michigan, but it's still not that fast... I'll try your new mirror to see how good it is. Since I'm rather new to Debian (only had it for about 2 months now) could you tell me the deb and deb-src entries I'd want to tell apt-get to use? I'm not terribly good at these things yet... I'm still rather proud of myself for getting beyond text mode and getting KDE installed. -- Mark Nippere-contacts: 832 Tanglewood Drive[EMAIL PROTECTED] Bryan, Texas 77802-4013 http://nipsy.bitgnome.net/ (979)575-3193 AIM/Yahoo: texasnipsy ICQ: 66971617 -BEGIN GEEK CODE BLOCK- Version: 3.1 GG/IT d- s++:+ a- C++$ UBL$ P---+++ L+++$ !E--- W++(--) N+ o K++ w(---) O++ M V(--) PS+++(+) PE(--) Y+ PGP t+ 5 X R tv b+++@ DI+(++) D+ G e h r++ y+(**) --END GEEK CODE BLOCK-- ---begin random quote of the moment--- A good traveller has no fixed plans and is not intent on arriving. -- Lao Tzu (570-490 BC) end random quote of the moment -- === GCB v3.1 === GCS d-(+) s+:- a? C+() UL+++() P L++(+++) E- W+(+++) N++ w--- M++ PS-- PE Y+ PGP- t++(+++) 5? X? R !tv-- b++ DI+++ D-- G !e h(*) !r x--- === EGCB v3.1 ===
Re: I love you!
SPAM IN THE FORUMS!!! Sorry, I just love yelling that.
HP nx6125: ACPI thermal events (or lack thereof)
Hi, I know that this is more of a laptop issue than a 64bit one per se. However, I posted the message below to debian-laptop and got no replies. I am hoping that there is someone on this list with a similar 64bit laptop (HP nx6125) who may be able to verify the strange behaviour described below, or at least point me in the right direction as to why my thermal events are only getting processed when I do a cat /proc/acpi/thermal_zone/TZ?/temperature or an acpi -t (I thought that this might be a latency issue and have since re-compiled with option 3 --preemptable kernel-- with no change in the behaviour mentioned below.) Many thanks if someone can shed some light on this and sorry for the long post, but I thought it best to give all the facts. Richard -- Forwarded Message -- I've been working steadily over the past few weeks to get my new HP nx6125 working under Debian (amd64 port) and have made significant progress. However, there is one considerable problem: thermal events don't seem to be recognised or processed by the kernel (until I do a cat /proc/acpi/thermal_zone/TZ?/temperature). As soon as I do anything CPU intensive I really run the risk of frying my laptop :-( To be more specific, I am running kernel 2.6.13.4 (www.kernel.org vanilla) with the double timer patch applied (see http://bugzilla.kernel.org/attachment.cgi?id=6061action=view). When I boot up, my thermal trip points get set nicely. The first is at 58 *C, then 65*C, then 75 *C, and 80*C (S5 = 95 *C). When I was testing things I was doing frequent executions of cat /proc/acpi/thermal_zone/TZ?/temperature and I would observe the temp rise to 58 C, then the fan would kick in, the first trip point would then (automatically) re-set to 50 C and the CPU would cool through 8 C before the fan turned off (nice, I thought, and very clever this re-setting of trip points--sorry I'm very new to ACPI). When the fan turned off, the trip point would again re-set to 58 C. So, I thought all was working well. However, subsequent tests done by running glxgears and not executing the above cat command allowed the CPU temp to rise above several trip points without the fans kicking in! Only when I ran the above cat command did the fans start!? So, I stopped acpid and did a cat /proc/acpi/event while running glxgears. I waited a while and then did a cat /proc/acpi/thermal_zone/TZ?/temperature to see that indeed the temp of TZ1 had exceeded 58C---and immediately /proc/acpi/event received a thermal event (note: the temp had already exceeded 58 C, my first thermal trip point; the thermal event only occurred when I did the 'cat'). So, in order for thermal events to get through/processed I need to keep doing cat /proc/acpi/thermal_zone/TZ?/temperature Can anybody shed some light on this behaviour. I don't know much about ACPI, but it seems (?) like the linux kernel is not processing the thermal events properly. Incidentally, I am also seeing spurious syslog errors that read APIC error on CPU0: 40(40) meaning that some interrupts presumably are not being correctly identified by the interrupt controller (thermal ones? could there be some correlation here?). Other info: the HP nx6125 is a Turion 64 based laptop with ATI chipset (yes, I know). I am running acpid and have just installed powernowd (doesn't fix it). I have also observed the above behaviour running the standard Debian 2.6.12-1-amd64 kernel (booting with no_timer_check to avoid double timer interrupts). Any help, suggestions would be greatly appreciated. If anyone has any ideas why catting /proc/acpi/thermal_zone/TZ?/temperature gets things to work, I'd be very happy to hear an explanation, too. Richard -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Mirror list
On 10/31/05, mike [EMAIL PROTECTED] wrote: it's pretty off-topic but i don't think that HE's bandwidth will suit their needs. they need proper frame relay, not bandwidth from a colo provider on steroids. Perhaps. But nonetheless, HE does offer some of the best bandwidth avaliable. I've looked into frame relay, and I haven't seen it go that fast compared to HE's GIG-E. Can you imagine *gigabit* ethernet - to the Internet!!! HE's a very professional company and I'm absolutely positive they'd allow multiple lines into the same building - for a price, of course. I know censored Comcast wouldn't do that unless you paid them well more than you're able to! Furthermore, $120,000.00/year is far too much for Internet, no matter how fast. I think that it's time to do some research as to better options, since that's far too expensive sounding. I just wish Verizon were in my area - I'd LOVE to get their new FIOS service! With them I could actually host something (yay!). But noo, Comcast maintains an unfair stranglehold monopoly on my area Whatever, I'm raving. And no, HE is more than a colo company on steriods (though probably not far from it!)
Re: Mirror list
On Mon, 31 Oct 2005, lordSauron wrote: On 10/31/05, mike [EMAIL PROTECTED] wrote: it's pretty off-topic but i don't think that HE's bandwidth will suit their needs. they need proper frame relay, not bandwidth from a colo provider on steroids. Perhaps. But nonetheless, HE does offer some of the best bandwidth avaliable. I've looked into frame relay, and I haven't seen it go that fast compared to HE's GIG-E. Can you imagine *gigabit* ethernet - to the Internet!!! I have more than that. A snapshot of our university uplink at the ubuntu breezy release: http://stats.sunet.se/stat-q/plot-all/umea2.umea-srp,2005-10-13,raw,traffic-kbit There is nothing magic about multi-gigabit pipes. It just costs a fair chunk of money. HE's a very professional company and I'm absolutely positive they'd allow multiple lines into the same building - for a price, of course. I know censored Comcast wouldn't do that unless you paid them well more than you're able to! Furthermore, $120,000.00/year is far too much for Internet, no matter how fast. I think that it's time to do some research as to better options, since that's far too expensive sounding. I just wish Verizon were in my area - I'd LOVE to get their new FIOS service! With them I could actually host something (yay!). But noo, Comcast maintains an unfair stranglehold monopoly on my area You seem to have a very limited ISP experience. I would suggest that you take a wider look at reality before going off stating such things as certain. And no, you don't have a stronghold monopoly from comcast if you are willing to spend the cash to put down a fiber of your own to a real isp. This is what you do if you are a university. Whatever, I'm raving. And no, HE is more than a colo company on steriods (though probably not far from it!) SLAs cost more than bits too, btw. As does redundancy. And routers. I have 100Mbit/s at home, at a very resonable price. But the same capacity for a company would be much more expensive, due to less oversubscribing and higher availability demands. /Mattias Wadenstein PS. Could you please stop responding to every piece of spam. It just doubles the annoyance and noise mailflow. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: installing Oracle on Debian AMD64
Sven Mueller wrote: Postgresql locks at less than table-granularity all the time. Fine. I never said it wouldn't. The point is (and I wasn't clear) is that the locking in MySQL can become a major gotcha if you're not paying attention in MySQL, while it's a constant in PostgreSQL. It is also easier to be fast writing if you lock the whole table and prevent others from accessing it while you update things. Slows down reading to stopped while you do a write though. More like stops every other read and write. for as long as the write takes. Which means you won't be doing any bulk-data loading on a busy database. In fact, it's enough of a problem MySQL AB had to had the special 'INSERT DELAYED' syntax to work around it for batch loads. Actually, I didn't exactly get it backwards. MyISAM _is_ fast to write (and fast to read) as long as you only write or only read. It's only fast to write compared to anything else because it doesn't bother with any sort of transactional state. Write speed of a single transaction isn't generally interesting, save for the single case of bulk loads/updates. Even then, it's not that interesting, as in most databases you'll end up structuring your bulk loads into some sort of batch transaction to increase speed. So yes, it's fast for a single write. But no one really cares about that. We care how fast it is with concurrent writes, and it's jsut plain terrible. There are certainly other drawbacks in MyISAM, but I wouldn't call them horrible things. Funny, most DBAs on any other platform would call lack of transactions (ignoring replications' binary log psuedo-transactions) and referential integrity, and the ability to mount and use totally corrupted tables all horrible things. 4.0 is far older than a year and fixed many things which where problematic in 3.x. Except the subqueries and the stored procedures. Which are big ticket items. 5.0 has reached GA a few weeks ago and its release candidates (and even its betas) effectively were really stable. Not interested. If it's not gold, it's not worth using here. Sorry ;p That a lot of shops still run 3.x is their problem, the upgrade can easily be done with only a few (like 2) minutes of downtime if you are careful. The code upgrade, sure. But 5.x isn't totally compatible with 3.x, and to use the new, nifty features, you have to move to InnoDB which isn't just a drop and go thing. So no, it takes way more than 2 minutes of downtime. It takes an almost complete revaluation of what you're currently storing and doing. The whole point of a transaction is to prevent an inconsistent database in the first place. Which doesn't contradict me. You're right, it doesn't, but your use of the word 'instead' confused me. I don't understand what you would want a lock for inside a transaction. Consider the simple case of a forum, and a column that indicates posts per thread. That must be updated on every post, and the code might look something like this (psuedo code): BEGIN VAR pc; INSERT INTO posts VALUES (22, 'my post', 'my name', 'some post text'); SELECT INTO pc post_count FROM threads WHERE tid = 22; UPDATE threads SET post_count = pc+1 WHERE tid=22; COMMIT; Now, if all transactions are serial (i.e., always executed one after another) this will work correctly. But, no database runs in serializable mode, because it's worse than slow. They run in READ COMMITTED or REPEATED READ modes. The former ensures a single read will be consistent, the latter, all reads will be consistent within a single transaction. But they say nothing about other transactions. As such, it's possible for two transactions to run at the same time, see the same value of pc, and update it to the same value. That's a bug. As such, you'd traditionally either use serialized mode, which kills concurrency, or use a row-level lock. Only you can't do that in MySQL for complicated transactions, as you lose your atomic context. Well, you can, if you want to write the cleanup code yourself. But that's a major PITA, and why we use techniques like RAII in programming languages. Now, to be fair to MySQL, it does provide a solution in the SELECT ... FOR UPDATE syntax[1], but that's just another fine example of it deviating from the way everyone else does things. I mean either the transaction is atomic (and therefor doesn't need a lock) or it isn't (and might need one). Maybe I overlook something, but as far as I can tell, a transaction shouldn't need to set locks. Yes, you should read any database guide (even the MySQL one for InnoDB) on transaction isolation levels, and what they mean. It's a gotcha for many people. Just to make sure I understand you correctly: You say that if you use a user set lock inside a transaction, the transaction context is invalidated. Yes, LOCK TABLES really looks like: COMMIT REALLY LOCK TABLES And if I understand you correctly,
Re: Eclipse
Hamish Moffatt kiedys napisal: On Mon, Oct 31, 2005 at 08:22:31AM -0400, Ed Tomlinson wrote: On Monday 31 October 2005 05:28, Dalibor Topic wrote: The packagers used a specific free runtime to make the eclipse package build and work, so they made that runtime specifically part of the dependencies, as that's a configuration the packagers can focus on to support. You are most welcome to contribute, and help improve the eclipse packages. This does _not_ make a lot of sense. It would make much more sense to suggest gcj/gij and depend on java-virtual-machine. This leaves it up the the user to decide if he can use a non-free jvm. I my case many of the apps I use (non-debian) fail with the free jvms. In short this type of depends is, IMO a bug. It will force me, and many others, to bypass the packaging system, which is usually a bad idea. Your argument is only reasonable if your non-free Java environment is a complete drop-in replacement for building and running Eclipse. If not, then you're asking for extra work to be done to support multiple JVMs. If that's what you need, patches are probably welcome. Hamish Maybe not. Original (downloaded) Eclipse version works with both gcj/gij and Sun JDK. And I think that with Sable and kaffe it could also work wothout patching as far as I know. Maybe when all Eclipse packages will be available, I will repack it and try with each virtual machine. -- Registered Linux User 369908 [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Mirror list
I agree with Mattias - it sounds like your experience with ISPs and such is limited. If you'd like I can walk you through some of HE's history, off the list. They are a colo company on steroids, for all intents and purposes (look at their About Us page even) - the reason I know this? My friend and I ran a server out of HE back when you were still in middle or elementary school :) If I didn't know any better, I'd say you're trying to refer them business; nobody here is looking to be sold and you're trying to sell - hardcore. There must be some sort of motivation for this kind of plugging. - mike On 10/31/05, lordSauron [EMAIL PROTECTED] wrote: Perhaps. But nonetheless, HE does offer some of the best bandwidth avaliable. I've looked into frame relay, and I haven't seen it go that fast compared to HE's GIG-E. Can you imagine *gigabit* ethernet - to the Internet!!! HE's a very professional company and I'm absolutely positive they'd allow multiple lines into the same building - for a price, of course. I know censored Comcast wouldn't do that unless you paid them well more than you're able to!
Re: Best approach for a system that needs desktop apps and 64bit address space
Which Nvidia drivers are you speaking of? Graphics or Nforce ethernet, etc? I have had no problem compiling and using Nvidia graphics on my 64 bit dual opteron system at work (Suse). nathan pmarc wrote: nVidia and IPTables don't work properly with the Install an ia32 system with 64SMP kernel option. IPTables is not a problem here, I believe. However, doing without nVidia proprietary drivers is a no-go What exactly do you plan on doing with all that memory? 32-bit apps can't (well, not really, earlier arguments on the mailing list withstanding) use all the memory, so the apps planning on using all that RAM would need to be 64bit Simulation stuff. Some of the apps are 64, some 32, the desktop stuff should run with all the bells and whistles. -- Paulo Marcondes http://rj.debianbrasil.org -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Mirror list
On 10/31/05, mike [EMAIL PROTECTED] wrote: I agree with Mattias - it sounds like your experience with ISPs and such is limited. Most likely is. Let's just revise my ignorance to a speciality in low-cost residential internet, okay? 'Cause I do know quite a bit about that (despite the fact that there's not that much to know) If you'd like I can walk you through some of HE's history, off the list. They are a colo company on steroids, for all intents and purposes (look at their About Us page even) - the reason I know this? My friend and I ran a server out of HE back when you were still in middle or elementary school :) so about 1995-1998 -ish? Nice. I still wish I had that kind of money as it is: I don't. If I didn't know any better, I'd say you're trying to refer them business; nobody here is looking to be sold and you're trying to sell First of all, I'm a hard-core tightwad, and the cheapest guy gets the contract. If no one is cheap enough, no one gets the contract. HE impressed me with prices that were low, but still just above my price range. sigh... Plus, they're the only company I know of that is almost entirely dedicated to enterprise-class connections. They don't offer any residential connections anymore (when I last contacte them, that is) - hardcore. There must be some sort of motivation for this kind of plugging. Their sales rep spoke English. You have any idea how rare it is to get someone who speaks English? With all the outsourcing going on, it's amazing they still sell things in the US!
Re: Bug#314988: Still a problem?
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 On Mon, Oct 31, 2005 at 08:33:40AM -0500, Adam C Powell IV wrote: On Fri, 2005-10-28 at 13:58 -0700, Steve Langasek wrote: On Fri, Oct 28, 2005 at 09:59:34PM +0200, Artur R. Czechowski wrote: On Fri, Oct 28, 2005 at 09:55:25PM +0200, Leopold Palomo Avellaneda wrote: I did a build-deps and and dpkg-buildpackage to make sure that I had all the build depens, but it seems as Ted said and I confirm that I needs the nvidia-glx-dev. The Question is: why the package needs nvidia-glx-dev on amd64 and on other platforms it builds succesfully without nvidia-glx-dev? Because of a longstanding bug in the nvidia lib package which creates a broken .so symlink when it's installed without nvidia-glx-dev as well. Is this the same issue as in 272438? Yes, that's the same bug. So should this be reassigned to nvidia-glx-dev? Uh, the build issue doesn't seem to have much to do with the reason for this bug report. - -- Steve Langasek Give me a lever long enough and a Free OS Debian Developer to set it on, and I can move the world. [EMAIL PROTECTED] http://www.debian.org/ -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.1 (GNU/Linux) iD8DBQFDZtvdKN6ufymYLloRAkL+AJ9XxSh2Pd5qRESx+/gRD98huTEyyQCgqzSq 0naitiNPEL10yjWAVOMUZhY= =9INz -END PGP SIGNATURE- -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Mirror list
Well, since there's a need, I thought I'd just ask any who're interested to help me and alleviate some of my ignorance by helping with Wikipedia, most particularly at http://en.wikipedia.org/wiki/Optical_Carrier there isn't much of a article there. Keep in mind that it's for those who want to, Lord Sauron won't force you : )
Re: Mirror list
On Mon, Oct 31, 2005 at 07:53:22PM -0800, lordSauron wrote: Well, since there's a need, I thought I'd just ask any who're interested to help me and alleviate some of my ignorance by helping with Wikipedia, most particularly at http://en.wikipedia.org/wiki/Optical_Carrier there isn't much of a article there. Keep in mind that it's for those who want to, Lord Sauron won't force you : ) There's plenty of inaccurate information on that page and the linked SONET page. AFAIK, only OC-3, 12, 48, 192, 768 etc are used. The other numbers (256, 384, 1536) are made up, though they could exist in theory. The SONET page lists other junk like OC-9, 12, 24, 96. OC-n is not n * 51.8 MBit/sec; OC-n is actually (n/3) * 155.52, because OC-3 is the base rate. OC-1 is the odd one out. The SONET page fails to explain this. This is quite off-topic for this list. Hamish -- Hamish Moffatt VK3SB [EMAIL PROTECTED] [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Mirror list
yeah, I suppose you're right. This discussion (of ISPs) is over as far as I'm concerned. The (patent pending) Lord Sauron's Veil of Silence is now in effect. Any further OT chatter on this thread directed at me might as well be read by blind eyes.
Spam on list posts was Re: Mirror list
On November 1, 2005 01:26 am, lordSauron wrote: The (patent pending) Lord Sauron's Veil of Silence is now in effect. Any chance we can get the same treatment for the spam on the list posts you seem to be fond of making as well. Thanks, Stephen -- Debian the choice of a GNU generation GPG Public Key: http://users.eastlink.ca/~stephencormier/publickey.asc pgptzmnzBw8x1.pgp Description: PGP signature
