Re: [freenet-support] Freenet Project health
Quoting Toad <[EMAIL PROTECTED]>: > Okay, you'll never get acceptable performance. Tough. Yeah! Get a proper operating system use schedule. -todd ___ Support mailing list [EMAIL PROTECTED] http://news.gmane.org/gmane.network.freenet.support Unsubscribe at http://dodo.freenetproject.org/cgi-bin/mailman/listinfo/support Or mailto:[EMAIL PROTECTED]
Re: [freenet-support] Freenet Project health
On Mon, Jul 12, 2004 at 09:31:23PM -0500, David Masover wrote: > -BEGIN PGP SIGNED MESSAGE- > Hash: SHA1 > > > > > Toad wrote: > | On Sun, Jul 11, 2004 at 10:12:19PM -0500, David Masover wrote: > | > | Strange. It didn't produce actual error messages? Usually the node > | responds in a reasonable time nowadays... of course if your browser is > | set to only use 2 connections it might take a while... > > My browser is set to a bit more, but it still takes at least a few > minutes to get an error. I think the error is usually either data not > found or (more likely) key not found. These are from the default > bookmarks -- things like FIND. Expect fewer 'I thinks' in my complaints > when I actually have a node running again. > > |>| How much RAM does it have? We have had reports of reasonable performance > |> > |>256 megs, but it's also running a lot of stuff -- > |>apache,squid,qmail,djbdns,dhcpd,samba,sshd,bincimap,stunnel > |> > |>Admittedly, there's not a lot of load on it, and a lot of that could be > |>swapped. But I'm running Linux 2.6, and Top shows 99% CPU in userland, > |>by Freenet -- not in IO-Wait, where it would be if RAM was an issue. > | > | How long after startup? I'd expect a CPU spike for the first hour or > | two... it should go away after that... but is this on a 200MHz system? > > It is a 200 mhz system, but I usually don't leave Freenet running for > very long. Like, I'll run it for an hour in the evening while I try to > get content (and browse other sites while I'm bored of waiting), and > then I'll shut it down for the night so my brother can game (he wakes up > earlier than me). Okay, you'll never get acceptable performance. Tough. Not much we can do about it. Well, maybe it'll improve with some future changes... > > | That's not been everyone else's experience :). Seriously, it's not a > | matter of local node optimisation. It's a matter of optimising THE > | NETWORK. That means getting routing working. > > Ah, I see. Still, they are sort of co-dependent. If a node has a > massive spike in CPU usage for the first hour, it hurts the network. If > the network is slow, it doesn't matter much how fast the local node is. -- Matthew J Toseland - [EMAIL PROTECTED] Freenet Project Official Codemonkey - http://freenetproject.org/ ICTHUS - Nothing is impossible. Our Boss says so. signature.asc Description: Digital signature ___ Support mailing list [EMAIL PROTECTED] http://news.gmane.org/gmane.network.freenet.support Unsubscribe at http://dodo.freenetproject.org/cgi-bin/mailman/listinfo/support Or mailto:[EMAIL PROTECTED]
Re: [freenet-support] Freenet Project health
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Toad wrote: | On Sun, Jul 11, 2004 at 10:12:19PM -0500, David Masover wrote: | | Strange. It didn't produce actual error messages? Usually the node | responds in a reasonable time nowadays... of course if your browser is | set to only use 2 connections it might take a while... My browser is set to a bit more, but it still takes at least a few minutes to get an error. I think the error is usually either data not found or (more likely) key not found. These are from the default bookmarks -- things like FIND. Expect fewer 'I thinks' in my complaints when I actually have a node running again. |>| How much RAM does it have? We have had reports of reasonable performance |> |>256 megs, but it's also running a lot of stuff -- |>apache,squid,qmail,djbdns,dhcpd,samba,sshd,bincimap,stunnel |> |>Admittedly, there's not a lot of load on it, and a lot of that could be |>swapped. But I'm running Linux 2.6, and Top shows 99% CPU in userland, |>by Freenet -- not in IO-Wait, where it would be if RAM was an issue. | | How long after startup? I'd expect a CPU spike for the first hour or | two... it should go away after that... but is this on a 200MHz system? It is a 200 mhz system, but I usually don't leave Freenet running for very long. Like, I'll run it for an hour in the evening while I try to get content (and browse other sites while I'm bored of waiting), and then I'll shut it down for the night so my brother can game (he wakes up earlier than me). | That's not been everyone else's experience :). Seriously, it's not a | matter of local node optimisation. It's a matter of optimising THE | NETWORK. That means getting routing working. Ah, I see. Still, they are sort of co-dependent. If a node has a massive spike in CPU usage for the first hour, it hurts the network. If the network is slow, it doesn't matter much how fast the local node is. -BEGIN PGP SIGNATURE- Version: GnuPG v1.2.4 (GNU/Linux) Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org iQIVAwUBQPNJe3gHNmZLgCUhAQJyNA/+OgEMQuYpxVS71BbUzF/a+nMJ361ox70P VEr9R35zAMt/bY5z2VzbHafNQNCTCX64fC9osc9nxBZPZbUROU6s1duYZhbQx7H4 7kYpfFOMx6gPdUnhooVtomq0X4yx+fnO+sLW2pvTnS7IjFiHRPBx9EOoqd1G4Q1Q YOphGjSbM9zkKflavZzIEDeX06vUhpiq8ycZbBYR9Tw2lpVX/nG17T1J9KLTjod1 5JoHSf+mShkyVgcFPLsgLsQMqfyB50MtHOZ78SshPnKScAPL0oaNUWFU4Z6ltwYG DzefCDLCP3G5ns3EiycdBtiLrl6WW2LWze8s/zKj4pA0WqkgLbhVfHOtdXqgqBdU sJxQCGFlZ9dZ3QfFFldJNa15gX7YwKtM/3fWHRZKhhxQLDII7cFbYUzD+XYQQ+cD uMVhavSY1jdfBtsIVF+9MeECVWTxB2hCGY3tRnykl6I5FE1so6u7q5/fFXwugmSE d9xz0ZO3Mh/wOO4he53bs8WQ2BbMrUWmnaW2VrkE1h5qT+RGIZqORGbnrpq56DAo 5Zy2ce7MnC0RBkqtMaAS1VkNj+T/fPpV8TTtFS/ku/hjAYCsDaEUSjY6Gln1DCZm d6byD9ZkZdXGzVW7xrMM0RC8o5zfr8OTw+k5VjtWBcgB90WPxwKCpbQWSMoT4w/X 9KhTIMqVk/0= =6Tvp -END PGP SIGNATURE- ___ Support mailing list [EMAIL PROTECTED] http://news.gmane.org/gmane.network.freenet.support Unsubscribe at http://dodo.freenetproject.org/cgi-bin/mailman/listinfo/support Or mailto:[EMAIL PROTECTED]
Re: [freenet-support] Freenet Project health
On Sun, Jul 11, 2004 at 03:11:46AM -0700, Tracy R Reed wrote: > On Sat, Jul 10, 2004 at 01:40:33PM -0400, Paul spake thusly: > > Not nessessarly. Freenet requires a lot of horsepower because of all > > the crypto required for even simple connections. > > Which is why we need to use native BigInt and FEC encoders to get > something approaching reasonable performance. Fast as C++ my patootie. It depends what you're doing. Even the Java zealots admit this. Personally I see no reason why GCJ-compiled code (using GMP for BigInteger) shouldn't be reasonably fast. Of course Sun's BigInteger support is slow, they don't open source the code so they can't use libgmp! -- Matthew J Toseland - [EMAIL PROTECTED] Freenet Project Official Codemonkey - http://freenetproject.org/ ICTHUS - Nothing is impossible. Our Boss says so. signature.asc Description: Digital signature ___ Support mailing list [EMAIL PROTECTED] http://news.gmane.org/gmane.network.freenet.support Unsubscribe at http://dodo.freenetproject.org/cgi-bin/mailman/listinfo/support Or mailto:[EMAIL PROTECTED]
Re: [freenet-support] Freenet Project health
On Sun, Jul 11, 2004 at 09:58:52PM -0500, David Masover wrote: > -BEGIN PGP SIGNED MESSAGE- > Hash: SHA1 > Good luck (and I don't mean that sarcastically). I've never gotten > Kaffe to run a Hello World program. Kaffe runs loads of stuff, big stuff like JBoss. GCJ runs ECLIPSE! Seriously, it's only because we use the obscure non-blocking I/O APIs that Freenet doesn't work on Kaffe right now. -- Matthew J Toseland - [EMAIL PROTECTED] Freenet Project Official Codemonkey - http://freenetproject.org/ ICTHUS - Nothing is impossible. Our Boss says so. signature.asc Description: Digital signature ___ Support mailing list [EMAIL PROTECTED] http://news.gmane.org/gmane.network.freenet.support Unsubscribe at http://dodo.freenetproject.org/cgi-bin/mailman/listinfo/support Or mailto:[EMAIL PROTECTED]
Re: [freenet-support] Freenet Project health
On Sun, Jul 11, 2004 at 10:12:19PM -0500, David Masover wrote: > -BEGIN PGP SIGNED MESSAGE- > Hash: SHA1 > > > > > Toad wrote: > > | On stable? How many live connections? Do you get RNFs? DNFs? > > I think stable was actually better last I checked than unstable. Hmm. In what sense? Certainly there is more content on stable, because there are more nodes on stable.. > Don't > know RNF or DNF from Dionsaur. 5-10 connections, 2 or 3 loaded before > the browser timed out (guess). Strange. It didn't produce actual error messages? Usually the node responds in a reasonable time nowadays... of course if your browser is set to only use 2 connections it might take a while... > > | How much RAM does it have? We have had reports of reasonable performance > > 256 megs, but it's also running a lot of stuff -- > apache,squid,qmail,djbdns,dhcpd,samba,sshd,bincimap,stunnel > > Admittedly, there's not a lot of load on it, and a lot of that could be > swapped. But I'm running Linux 2.6, and Top shows 99% CPU in userland, > by Freenet -- not in IO-Wait, where it would be if RAM was an issue. How long after startup? I'd expect a CPU spike for the first hour or two... it should go away after that... but is this on a 200MHz system? > > | on that class of hardware. OTOH, it's not a big priority at the moment. > > What's the bigger priority right now? Because everything always seemed > to "just work" for me with Freenet -- it would just do it excruciatingly > slow. That's not been everyone else's experience :). Seriously, it's not a matter of local node optimisation. It's a matter of optimising THE NETWORK. That means getting routing working. > > | I don't think Freenet's bandwidth limiting is unobtrusive enough for > | gaming. Having said that, a gamer will tell you that ANYTHING else > | running on the connection will increase his ping. Even if it doesn't! :) > > I'm a gamer, and I run web and email on this. My bro notices when I run > freenet. > > | those will be rectified when GCJ works. If we had used C++, we'd have > | spent a year arguing over whether to include a garbage collector. If > | we had used Ocaml, we'd have had even fewer coders than we have now. > > If you'd used Perl? (Ok, that's not fair, and it's moot anyway.) ROFL. > > | I've seen all of the above. Get hold of a copy of Eclipse compiled under > | GCJ sometime... > > Will do! Sometime... -- Matthew J Toseland - [EMAIL PROTECTED] Freenet Project Official Codemonkey - http://freenetproject.org/ ICTHUS - Nothing is impossible. Our Boss says so. signature.asc Description: Digital signature ___ Support mailing list [EMAIL PROTECTED] http://news.gmane.org/gmane.network.freenet.support Unsubscribe at http://dodo.freenetproject.org/cgi-bin/mailman/listinfo/support Or mailto:[EMAIL PROTECTED]
Re: [freenet-support] Freenet Project health
On Sunday 11 July 2004 10:58 pm, David Masover wrote: > | not sure my archives even go back that far, but the basis for > | choosing Java should be obvious; platform independence and a > | rich API that comes standard with the language. > > As a purely academic argument, Parrot and .NET both do those things now, > and an API doesn't seem like it'd help that much with Freenet. It's not > the interface that's broken. So in 1999 what should have been chosen? Neither Parrot nor .NET were even around. Parrot itself is another virtual machine, being used for Perl6. And .NET is currently Microsoft specific, the GNU implementation Mono still far enough behind to be considered a stable platform on Unix, or even Windows itself. -- Jay Oliveri GnuPG ID: 0x5AA5DD54 FCPTools Maintainer www.sf.net/users/joliveri ___ Support mailing list [EMAIL PROTECTED] http://news.gmane.org/gmane.network.freenet.support Unsubscribe at http://dodo.freenetproject.org/cgi-bin/mailman/listinfo/support Or mailto:[EMAIL PROTECTED]
Re: [freenet-support] Freenet Project health
On Sat, Jul 10, 2004 at 01:40:33PM -0400, Paul spake thusly: > Not nessessarly. Freenet requires a lot of horsepower because of all > the crypto required for even simple connections. Which is why we need to use native BigInt and FEC encoders to get something approaching reasonable performance. Fast as C++ my patootie. -- Tracy Reed The attachment is a digital signature. http://copilotconsulting.com More info: http://copilotconsulting.com/sig pgpeulinsYo8T.pgp Description: PGP signature ___ Support mailing list [EMAIL PROTECTED] http://news.gmane.org/gmane.network.freenet.support Unsubscribe at http://dodo.freenetproject.org/cgi-bin/mailman/listinfo/support Or mailto:[EMAIL PROTECTED]
Re: [freenet-support] Freenet Project health
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Toad wrote: | On stable? How many live connections? Do you get RNFs? DNFs? I think stable was actually better last I checked than unstable. Don't know RNF or DNF from Dionsaur. 5-10 connections, 2 or 3 loaded before the browser timed out (guess). | How much RAM does it have? We have had reports of reasonable performance 256 megs, but it's also running a lot of stuff -- apache,squid,qmail,djbdns,dhcpd,samba,sshd,bincimap,stunnel Admittedly, there's not a lot of load on it, and a lot of that could be swapped. But I'm running Linux 2.6, and Top shows 99% CPU in userland, by Freenet -- not in IO-Wait, where it would be if RAM was an issue. | on that class of hardware. OTOH, it's not a big priority at the moment. What's the bigger priority right now? Because everything always seemed to "just work" for me with Freenet -- it would just do it excruciatingly slow. | I don't think Freenet's bandwidth limiting is unobtrusive enough for | gaming. Having said that, a gamer will tell you that ANYTHING else | running on the connection will increase his ping. Even if it doesn't! :) I'm a gamer, and I run web and email on this. My bro notices when I run freenet. | those will be rectified when GCJ works. If we had used C++, we'd have | spent a year arguing over whether to include a garbage collector. If | we had used Ocaml, we'd have had even fewer coders than we have now. If you'd used Perl? (Ok, that's not fair, and it's moot anyway.) | I've seen all of the above. Get hold of a copy of Eclipse compiled under | GCJ sometime... Will do! Sometime... -BEGIN PGP SIGNATURE- Version: GnuPG v1.2.4 (GNU/Linux) Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org iQIVAwUBQPIBk3gHNmZLgCUhAQJ6iw//eN1HBBGJnpTUCpZIlyQn4BgP2p6xmh7F Y47zlqfIj0wbL4MGL2OnHpmdXfWHHehESjwnd1lYMSA/IfY9TkOfQGU4DC8DSJYC YhYjaw8PmVZVSo1LwdzUFwYt9fXtde2FcNCflh+qkUCT+gHQxkSKE7tLVzepZBo8 Eq3Hnh+OMV4lYkTmsUvOYkiyBUvXQTK21yO2PN4YGv/3wh1cVsrmE+dLgEChh/Hl u7xMlbWpjKYCMmoUWjkZNxh+5eJ2JVhnQrcI6VRMJFkGFvfnw5dcKTmAuezoTFlv t6vpCZxM8v9bNCb1dWHha9MN/a231kI0QNvFL2ORTrIwBowbhxyyQeSfntvUUsVM 3pk03O7mWDLzGSbiB3sv9PEW4NiI8fL3VBwUtq8IUwRuCtSJTJGO8Jku2lL9x+TO EAwNono8IsE2ztM+1ZpCckkLzlKpnKJTRk1LrH3XuMyXMJlnw44BDjrO5UPIhKgF HcCOfedMVT9fBsDlk325BTk1/8I6QCdvNng3EtYkZ34zXfPwcN9YBvs6jtqiG18g ruoyIP4IAouOSuyBkux5T00OOX0UayukOv1eQWs7OG3V62Syw1mwWqhzW+YJCerQ LpcEueMPRT4LVdy+73LS165yDbQ7qP3O6ycOqoIWGhLCaFcYXngB1nD1ExaOREGV U5HD4qPX9mk= =3AxR -END PGP SIGNATURE- ___ Support mailing list [EMAIL PROTECTED] http://news.gmane.org/gmane.network.freenet.support Unsubscribe at http://dodo.freenetproject.org/cgi-bin/mailman/listinfo/support Or mailto:[EMAIL PROTECTED]
Re: [freenet-support] Freenet Project health
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Jay Oliveri wrote: [...] | days. Did you leave it running while properly opening up the port it uses | for incoming connections in your firewall (if needed)? Yes. It's on the "firewall" machine, which really only does NAT anyway. | I have an 2.5G Athlon XP w/ 1G of RAM and I don't notice that it's running. | It tends to play nice on my system regardless of my success in getting any | data out of it. I'm talking about bandwidth, and it runs on said "firewall" machine -- a 200 mhz emachine. | not sure my archives even go back that far, but the basis for choosing Java | should be obvious; platform independence and a rich API that comes standard | with the language. As a purely academic argument, Parrot and .NET both do those things now, and an API doesn't seem like it'd help that much with Freenet. It's not the interface that's broken. | You don't have much experience with Java then. Freenet is atypical of a I'm glad I don't have much experience with Java. I'll say no more. | All this aside, when routing doesn't work in Freenet it can't be blamed on I like to blame Java, but you're right. | helping matters much. It's hoped this will change when one of the Free | Software implementations of Java (gcj, Kaffe) becomes more stable wrt | Freenet. Good luck (and I don't mean that sarcastically). I've never gotten Kaffe to run a Hello World program. -BEGIN PGP SIGNATURE- Version: GnuPG v1.2.4 (GNU/Linux) Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org iQIVAwUBQPH+bHgHNmZLgCUhAQI6NQ/8DLNghbAFGeltbxDutvB3D3N6f7hSPtkR sDTMyXUnTjgNRtiLMVlifcJbYLKhX7r90HXGCJR1s0iDD36WzDrCtgEnKTydhD4T j+N8IdvcP046ncASiNzjtHMIg4ZD923gsV747U6eHmpfFHzXkgC73do0iHHxpQMS L7gjzOtJb6P/p1A1DF1PBGPRP8FOJDGV240oWS3q+V7pQzbe4ozHiIumt5pEMm2x wCoqy8+CkxSaAWoQcHPV6SZ3nihXDNv0kh/jCeUl/FH7RdTgFbGd+Psp6dK1bqrZ LTkYxrp3gw47HzgzQq5Qy1pHv/MMd6I/m0z0uHdSk2vk75FnQ/dsWCNY9tIgQ3yD 2FmaKw0Vdi5xbnoZgfZS8O4i7rqvxnOV8Rfs4+wM11yczXI9lZbi+rTCoCpG3DtA 48lc9pHTIBCjrVCXBksi4lSFLHxnQ4Jt1dcUJ9QQVWokJdfD1s2yY7LraH6DLnaD zppSuMti4FNPuEpKS4C4V6l1E1/fhF5UuoyFOT6wHXROqQ5Ar9jQPqxo4Ty6Q2y6 BzS7JNTbdi7e2Gth8agl4LgX1Fz1USRToevR/IZX6UAxJ485bTX9rqHBPKLSkY8M d6vNLc8dLvl+s1GEL9FU9PSUU+f4HxV474U+x3F0Ib2dzxY4f3E1qVpyONwEKBVk ugBkhpWRgfo= =Kt4w -END PGP SIGNATURE- ___ Support mailing list [EMAIL PROTECTED] http://news.gmane.org/gmane.network.freenet.support Unsubscribe at http://dodo.freenetproject.org/cgi-bin/mailman/listinfo/support Or mailto:[EMAIL PROTECTED]
Re: [freenet-support] Freenet Project health
On Sat, 10 Jul 2004, Nicholas Sturm wrote: > >helping matters much. It's hoped this will change when one of the Free > >Software implementations of Java (gcj, Kaffe) becomes more stable wrt > >Freenet. > > > I lost the meaning in the last sentence. What was intended by "wrt"? with-respect-to ___ Support mailing list [EMAIL PROTECTED] http://news.gmane.org/gmane.network.freenet.support Unsubscribe at http://dodo.freenetproject.org/cgi-bin/mailman/listinfo/support Or mailto:[EMAIL PROTECTED]
Re: [freenet-support] Freenet Project health
On Thu, Jul 08, 2004 at 11:15:50PM -0500, David Masover wrote: > -BEGIN PGP SIGNED MESSAGE- > Hash: SHA1 > > > > What's the health of Freenet as a whole right now? I'm not sure. It depends on various factors. For example, which branch you are running. > I'm getting lots of > pages taking forever to load (or never loading), On stable? How many live connections? Do you get RNFs? DNFs? > and I think it's still > using 100% CPU on my 200 mhz router on 768k (up and down) DSL, even > though the browser is on another machine... :( How much RAM does it have? We have had reports of reasonable performance on that class of hardware. OTOH, it's not a big priority at the moment. "First make it work. Then make it work fast.". Most machines running Freenet are probably 5 times faster than the above hardware... > > I was planning to make a permanent node, but I don't run it much > anymore, because my brother games (so he needs high bandwidth and low > latency), and Freenet is still the most costly service I run on that > thing (in terms of CPU, bandwidth, etc.) I don't think Freenet's bandwidth limiting is unobtrusive enough for gaming. Having said that, a gamer will tell you that ANYTHING else running on the connection will increase his ping. Even if it doesn't! :) > > I know it's been mentioned before, but I'll state for the record that I > think Java was a bad choice. Rather than start a new flame war, I'd > like to go read up on why it was chosen (any archives I should look at?). Ian liked it. Ian was the original coder. Java is a reasonable OO-procedural language which has a number of eccentricities. Most of those will be rectified when GCJ works. If we had used C++, we'd have spent a year arguing over whether to include a garbage collector. If we had used Ocaml, we'd have had even fewer coders than we have now. Etc etc. > > For the record, I have never, ever seen a java program load quickly, run > even tolerably fast for anything beyond the most basic things, and I've > never seen an open source implementation of Java work firsthand. I > don't like the syntax, but that's a personal issue -- I'd love to be > proven wrong on this. I've seen all of the above. Get hold of a copy of Eclipse compiled under GCJ sometime... -- Matthew J Toseland - [EMAIL PROTECTED] Freenet Project Official Codemonkey - http://freenetproject.org/ ICTHUS - Nothing is impossible. Our Boss says so. signature.asc Description: Digital signature ___ Support mailing list [EMAIL PROTECTED] http://news.gmane.org/gmane.network.freenet.support Unsubscribe at http://dodo.freenetproject.org/cgi-bin/mailman/listinfo/support Or mailto:[EMAIL PROTECTED]
Re: [freenet-support] Freenet Project health
All this aside, when routing doesn't work in Freenet it can't be blamed on the language it was implemented in. Broken routing can easily be coded in C, Python, assembler or whatever language you desire. On the other hand being tied to a proprietary language like Java under Sun's control isn't helping matters much. It's hoped this will change when one of the Free Software implementations of Java (gcj, Kaffe) becomes more stable wrt Freenet. I lost the meaning in the last sentence. What was intended by "wrt"? ___ Support mailing list [EMAIL PROTECTED] http://news.gmane.org/gmane.network.freenet.support Unsubscribe at http://dodo.freenetproject.org/cgi-bin/mailman/listinfo/support Or mailto:[EMAIL PROTECTED]
Re: [freenet-support] Freenet Project health
On Sat, Jul 10, 2004 at 01:40:33PM -0400, Paul wrote: > Not nessessarly. Freenet requires a lot of horsepower because of all > the crypto required for even simple connections. No. Well... okay, asymmetric crypto is a good deal slower than it should be because Sun can't use gmp. We have code to fix that, by pulling in a native modPow() from GMP; this is still being worked on and was originally written for I2P. But I don't think connection negotiation is a big issue in Fred any more. I suspect routing is the bigger CPU issue. There are, naturally, performance issues with CPU usage in routing as well (Big*.doubleValue()). Anyway, I very much doubt that either asymmetric or symmetric crypto is the majority of the CPU cost of running Fred nowadays. I'd be delighted to be proven wrong on this though. -- Matthew J Toseland - [EMAIL PROTECTED] Freenet Project Official Codemonkey - http://freenetproject.org/ ICTHUS - Nothing is impossible. Our Boss says so. signature.asc Description: Digital signature ___ Support mailing list [EMAIL PROTECTED] http://news.gmane.org/gmane.network.freenet.support Unsubscribe at http://dodo.freenetproject.org/cgi-bin/mailman/listinfo/support Or mailto:[EMAIL PROTECTED]
Re: [freenet-support] Freenet Project health
On Thu, 08 Jul 2004 23:15:50 -0500, David Masover <[EMAIL PROTECTED]> wrote: For the record, I have never, ever seen a java program load quickly, run ... most probably because you've never seen _just_ a java program load. What you see is a complete virtual machine load and _then_ the application. Come newer versions of Java, that will be solved by loading one VM at boot and then java programs will start extremely fast. even tolerably fast for anything beyond the most basic things, and I've never seen an open source implementation of Java work firsthand. I don't like the syntax, but that's a personal issue -- I'd love to be proven wrong on this. If _Java_ was such a resource hungry and lousy language, it wouldn't be the largest environment in the world in smartcards and cellphones. Disclaimer: I do "virtual machine/native environment" Java work in cellphones for a living. regards, ___/ _/ -- http://troed.se - controversial views or common sense? ___ Support mailing list [EMAIL PROTECTED] http://news.gmane.org/gmane.network.freenet.support Unsubscribe at http://dodo.freenetproject.org/cgi-bin/mailman/listinfo/support Or mailto:[EMAIL PROTECTED]
Re: [freenet-support] Freenet Project health
On Friday 09 July 2004 12:15 am, David Masover wrote: > What's the health of Freenet as a whole right now? I'm getting lots of > pages taking forever to load (or never loading), and I think it's still > using 100% CPU on my 200 mhz router on 768k (up and down) DSL, even > though the browser is on another machine... My node isn't running too well either, even after leaving it on for a few days. Did you leave it running while properly opening up the port it uses for incoming connections in your firewall (if needed)? > I was planning to make a permanent node, but I don't run it much > anymore, because my brother games (so he needs high bandwidth and low > latency), and Freenet is still the most costly service I run on that > thing (in terms of CPU, bandwidth, etc.) I have an 2.5G Athlon XP w/ 1G of RAM and I don't notice that it's running. It tends to play nice on my system regardless of my success in getting any data out of it. > I know it's been mentioned before, but I'll state for the record that I > think Java was a bad choice. Rather than start a new flame war, I'd > like to go read up on why it was chosen (any archives I should look at?). That's a good idea. A discussion based on backing out of Java at this point is totally pointless and would set the project back another X years. I'm not sure my archives even go back that far, but the basis for choosing Java should be obvious; platform independence and a rich API that comes standard with the language. > For the record, I have never, ever seen a java program load quickly, run > even tolerably fast for anything beyond the most basic things, and I've > never seen an open source implementation of Java work firsthand. I > don't like the syntax, but that's a personal issue -- I'd love to be > proven wrong on this. You don't have much experience with Java then. Freenet is atypical of a Java application IMO. I would argue that Java has found a nice home in the Web Services market (JSP, Servlets, EJB), but Freenet attempts to be all things that Java isn't necessarily good at (for starters, NIO is something not necessary for most web apps). All this aside, when routing doesn't work in Freenet it can't be blamed on the language it was implemented in. Broken routing can easily be coded in C, Python, assembler or whatever language you desire. On the other hand being tied to a proprietary language like Java under Sun's control isn't helping matters much. It's hoped this will change when one of the Free Software implementations of Java (gcj, Kaffe) becomes more stable wrt Freenet. -- Jay Oliveri GnuPG ID: 0x5AA5DD54 FCPTools Maintainer www.sf.net/users/joliveri ___ Support mailing list [EMAIL PROTECTED] http://news.gmane.org/gmane.network.freenet.support Unsubscribe at http://dodo.freenetproject.org/cgi-bin/mailman/listinfo/support Or mailto:[EMAIL PROTECTED]
Re: [freenet-support] Freenet Project health
Not nessessarly. Freenet requires a lot of horsepower because of all the crypto required for even simple connections. http://www.sys-con.com/story/?storyid=45250";>Java vs C++ Java vs C++ "Shootout" Revisited June 15, 2004 Summary "I was sick of hearing people say Java was slow," says Keith Lea, "so I took the benchmark code for C++ and Java from the now outdated Great Computer Language Shootout (Fall 2001) and ran the tests myself." Lea's results three years on? Java, he finds, is significantly faster than optimized C++ in many cases. http://www.kano.net/javabench/graph";> On Thu, 08 Jul 2004 23:15:50 -0500, David Masover <[EMAIL PROTECTED]> wrote: > -BEGIN PGP SIGNED MESSAGE- > Hash: SHA1 > > What's the health of Freenet as a whole right now? I'm getting lots of > pages taking forever to load (or never loading), and I think it's still > using 100% CPU on my 200 mhz router on 768k (up and down) DSL, even > though the browser is on another machine... > > I was planning to make a permanent node, but I don't run it much > anymore, because my brother games (so he needs high bandwidth and low > latency), and Freenet is still the most costly service I run on that > thing (in terms of CPU, bandwidth, etc.) > > I know it's been mentioned before, but I'll state for the record that I > think Java was a bad choice. Rather than start a new flame war, I'd > like to go read up on why it was chosen (any archives I should look at?). > > For the record, I have never, ever seen a java program load quickly, run > even tolerably fast for anything beyond the most basic things, and I've > never seen an open source implementation of Java work firsthand. I > don't like the syntax, but that's a personal issue -- I'd love to be > proven wrong on this. > > -BEGIN PGP SIGNATURE- > Version: GnuPG v1.2.4 (GNU/Linux) > Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org > > iQIVAwUBQO4b9ngHNmZLgCUhAQLCTA//frjR/duxAbReDA13WmWZQLA9LQdzcJ2v > 2GJ+kv80fJC+P1U6h6/ghv+oTrNUDCnZweLNYRm5G6WNDiNPdnyFiNRe3+arfQsh > TR3BCc1CwIn/TuO/7s0V4pjIpGnQgySx3VAy9i8ujYtCV5t6Kkdaps6tC4x32tzM > ZFo3gXWjuQ0zGF1psBRlwH84+ia5lDjSqBjOK7P47FKAXk1/vOocSYAztrCuIgw0 > qYK5oYStm/3miLzid08B4Mni0GL5m4+HlmitBgt2nVgqNhaKZN/HtUWtCVU8wNhR > 55XhmkR3+nHhO994R7U2/DRyYo7w7Yv6obaZsqEZE+ddUYErjfwb2XLr+T1OgM0F > hQaXPzGY1Onc9Yb9a3iZPkdVVTCPbcKHK0nrjZkei6VJ90oANurMNUgo4ZyHtuZT > BFMq44b6AVqa6quandxEcQzhqy6uqhvXVin6l0PCzxtcUZHXQacjrEvgqNngXZl0 > gvcOiPoKN/rH7vAXdHG9C4F+ZuSXAenWUWsgspyxvoPXItTMIMxbKQBG7wnYJu9w > ahEHIGGdhks6eRgnq/hrroL/yzV3uEa5qhrmSqdRqW68Nr90kgHkJitUCqSKG/LZ > fLTR75QC+L+n148Z3V5uHaTQSwjazDrnJ65Ln7apzANpdR9dmO+7cpm4EBFCmZiH > NbTMfmg3wHQ= > =pgvD > -END PGP SIGNATURE- > ___ > Support mailing list > [EMAIL PROTECTED] > http://news.gmane.org/gmane.network.freenet.support > Unsubscribe at http://dodo.freenetproject.org/cgi-bin/mailman/listinfo/support > Or mailto:[EMAIL PROTECTED] > ___ Support mailing list [EMAIL PROTECTED] http://news.gmane.org/gmane.network.freenet.support Unsubscribe at http://dodo.freenetproject.org/cgi-bin/mailman/listinfo/support Or mailto:[EMAIL PROTECTED]
[freenet-support] Freenet Project health
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 What's the health of Freenet as a whole right now? I'm getting lots of pages taking forever to load (or never loading), and I think it's still using 100% CPU on my 200 mhz router on 768k (up and down) DSL, even though the browser is on another machine... I was planning to make a permanent node, but I don't run it much anymore, because my brother games (so he needs high bandwidth and low latency), and Freenet is still the most costly service I run on that thing (in terms of CPU, bandwidth, etc.) I know it's been mentioned before, but I'll state for the record that I think Java was a bad choice. Rather than start a new flame war, I'd like to go read up on why it was chosen (any archives I should look at?). For the record, I have never, ever seen a java program load quickly, run even tolerably fast for anything beyond the most basic things, and I've never seen an open source implementation of Java work firsthand. I don't like the syntax, but that's a personal issue -- I'd love to be proven wrong on this. -BEGIN PGP SIGNATURE- Version: GnuPG v1.2.4 (GNU/Linux) Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org iQIVAwUBQO4b9ngHNmZLgCUhAQLCTA//frjR/duxAbReDA13WmWZQLA9LQdzcJ2v 2GJ+kv80fJC+P1U6h6/ghv+oTrNUDCnZweLNYRm5G6WNDiNPdnyFiNRe3+arfQsh TR3BCc1CwIn/TuO/7s0V4pjIpGnQgySx3VAy9i8ujYtCV5t6Kkdaps6tC4x32tzM ZFo3gXWjuQ0zGF1psBRlwH84+ia5lDjSqBjOK7P47FKAXk1/vOocSYAztrCuIgw0 qYK5oYStm/3miLzid08B4Mni0GL5m4+HlmitBgt2nVgqNhaKZN/HtUWtCVU8wNhR 55XhmkR3+nHhO994R7U2/DRyYo7w7Yv6obaZsqEZE+ddUYErjfwb2XLr+T1OgM0F hQaXPzGY1Onc9Yb9a3iZPkdVVTCPbcKHK0nrjZkei6VJ90oANurMNUgo4ZyHtuZT BFMq44b6AVqa6quandxEcQzhqy6uqhvXVin6l0PCzxtcUZHXQacjrEvgqNngXZl0 gvcOiPoKN/rH7vAXdHG9C4F+ZuSXAenWUWsgspyxvoPXItTMIMxbKQBG7wnYJu9w ahEHIGGdhks6eRgnq/hrroL/yzV3uEa5qhrmSqdRqW68Nr90kgHkJitUCqSKG/LZ fLTR75QC+L+n148Z3V5uHaTQSwjazDrnJ65Ln7apzANpdR9dmO+7cpm4EBFCmZiH NbTMfmg3wHQ= =pgvD -END PGP SIGNATURE- ___ Support mailing list [EMAIL PROTECTED] http://news.gmane.org/gmane.network.freenet.support Unsubscribe at http://dodo.freenetproject.org/cgi-bin/mailman/listinfo/support Or mailto:[EMAIL PROTECTED]
