From toad at amphibian.dyndns.org Mon Feb 4 11:15:03 2008 From: toad at amphibian.dyndns.org (Matthew Toseland) Date: Mon, 4 Feb 2008 11:15:03 +0000 Subject: [Tech] [Fwd: Re: [p2p-hackers] [tahoe-dev] Surely M$ can patent this process?] In-Reply-To: <479D3A48.5030709@cs.ucl.ac.uk> References: <479D3A48.5030709@cs.ucl.ac.uk> Message-ID: <200802041115.11057.toad@amphibian.dyndns.org> On Monday 28 January 2008 02:13, Michael Rogers wrote: > Does MS have a patent on CHKs? > ... > It's an interesting puzzle of intellectual history. The idea > certainly seems to have been "in the air", as both Mojo Nation and > Freenet were working on it before the May 2000 patent submission by > Doceur et al., but Mojo Nation and Freenet each published the idea > shortly after May 2000. According to my limited understanding of > patent law, this means that they don't count as prior art on that > patent. This ought to be evidence of obviousness... but probably isn't. :| -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: not available Url : http://emu.freenetproject.org/pipermail/tech/attachments/20080204/d1a03a79/attachment.pgp From jargonautti at hotmail.com Wed Feb 6 10:40:20 2008 From: jargonautti at hotmail.com (Jusa Saari) Date: Wed, 06 Feb 2008 12:40:20 +0200 Subject: [Tech] F2F news/blogging References: Message-ID: This is completely pointless. Simply post the article (with links to older editions) under an USK, and give that USK key to whoever you want to be able to read it. >From an application developers perspective, Freenet is not a network, it is a distributed file storage system. The concept of sending a file to a receiver is meaningless in this context; you simply write the file to Freenet, after which everyone who knows the key can read it, at least in theory. Trying to build a network on top of a filesystem (Freenet API) running over a network (Freenet FNP connections) running over a network (Internet) is an extremely misguided idea. On Wed, 16 Jan 2008 15:53:02 +0000, Michael Rogers wrote: > This is an old idea of mine that could be implemented with N2NMs - similar > to Syndie but with an F2F distribution mechanism. > > A journal is a series of articles signed with the author's private key. > The articles in the journal may be written by the user or reposted from > other journals, either manually or automatically (syndication). When a > user writes or reposts an article it's sent to her online friends and > queued for her offline friends. Articles can be marked 'friends only', in > which case they should never be reposted, but obviously this can't be > enforced - you just have to trust your friends. > > A user who reads a reposted article and wants to subscribe to the author's > journal (to read and/or syndicate it) sends a subscription request to the > friend in whose journal she read the article. If the friend is already > subscribed to the requested journal, she forwards subsequent articles to > the requester. If the friend reposted an individual article but doesn't > subscribe to the journal herself, she can forward the request to the > friend in whose journal she read the article, and so on. (Requests to > offline users are queued.) Eventually the publisher or a subscriber will > be found and the new subscription path can be established. > > Any thoughts? > > Cheers, > Michael From m.rogers at cs.ucl.ac.uk Wed Feb 6 10:51:00 2008 From: m.rogers at cs.ucl.ac.uk (Michael Rogers) Date: 06 Feb 2008 10:51:00 +0000 Subject: [Tech] F2F news/blogging In-Reply-To: References: Message-ID: On Feb 6 2008, Jusa Saari wrote: >This is completely pointless. Simply post the article (with links to older >editions) under an USK, and give that USK key to whoever you want to be >able to read it. And then send a message to Frost every time I update the USK, as people are currently doing? The point of my suggestion is to propagate new articles to subscribers without polling. >The concept of sending a file to a >receiver is meaningless in this context; you simply write the file to >Freenet, after which everyone who knows the key can read it, at least in >theory. And how do they get the key? This proposal isn't about sending files, it's about propagating keys. Sorry if I didn't make that clear. Cheers, Michael From jUrner at arcor.de Wed Feb 13 08:53:06 2008 From: jUrner at arcor.de (juergen urner) Date: Wed, 13 Feb 2008 09:53:06 +0100 Subject: [Tech] Fcp sizes and time Message-ID: <47B2AFF2.6040003@arcor.de> Hi, I have a question regarding byte sizes in Fcp. Along with some requests sizes are passed as 1G or 1000K. Quite obvious what it's about but nevertheless it is not documented. Can I assume Fcp always uses kibs and values are always passed following this patttern? -1 (not set) n n*1000 or (n)K n*1.000.000 or (n)M (...) The only time value (except from secs since epoch) I faced so far is one I got in response to GetConfig. It read '1HOUR'. Is there a similar pattern for time values to rely on? Juergen From jflesch at gmail.com Wed Feb 13 13:59:11 2008 From: jflesch at gmail.com (Jerome Flesch) Date: Wed, 13 Feb 2008 14:59:11 +0100 Subject: [Tech] Fcp sizes and time In-Reply-To: <47B2AFF2.6040003@arcor.de> References: <47B2AFF2.6040003@arcor.de> Message-ID: <200802131459.14479.jflesch@gmail.com> > Hi, > > I have a question regarding byte sizes in Fcp. Along with some requests > sizes are passed as 1G or 1000K. > I would be interrested to know where you saw that. From what I know, FCP always provide/need sizes/times in bytes/seconds (except for things related to the node config, see below). > Quite obvious what it's about but > nevertheless it is not documented. > > Can I assume Fcp always uses kibs and values are always passed following > this patttern? > > -1 (not set) > n > n*1000 or (n)K > n*1.000.000 or (n)M > (...) > > The only time value (except from secs since epoch) I faced so far is one > I got in > response to GetConfig. It read '1HOUR'. Is there a similar pattern for > time values to rely on? > This value comes directly from the node configuration, and node configuration is made to be human readable. Imo, that's the only reason why this time is not given in seconds. -- Jerome Flesch -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: This is a digitally signed message part. Url : http://emu.freenetproject.org/pipermail/tech/attachments/20080213/b67d4040/attachment.pgp From jUrner at arcor.de Wed Feb 13 19:45:03 2008 From: jUrner at arcor.de (juergen urner) Date: Wed, 13 Feb 2008 20:45:03 +0100 Subject: [Tech] Fcp sizes and time In-Reply-To: <200802131459.14479.jflesch@gmail.com> References: <47B2AFF2.6040003@arcor.de> <200802131459.14479.jflesch@gmail.com> Message-ID: <47B348BF.8010807@arcor.de> Jerome Flesch schrieb: >> Hi, >> >> I have a question regarding byte sizes in Fcp. Along with some requests >> sizes are passed as 1G or 1000K. >> >> > I would be interrested to know where you saw that. From what I know, FCP > always provide/need sizes/times in bytes/seconds (except for things related > to the node config, see below). > ConfigData current.node.outputBandwidthLimit=12K current.node.storeSize=1G (...) EndMessage >> Quite obvious what it's about but >> nevertheless it is not documented. >> >> Can I assume Fcp always uses kibs and values are always passed following >> this patttern? >> >> -1 (not set) >> n >> n*1000 or (n)K >> n*1.000.000 or (n)M >> (...) >> >> The only time value (except from secs since epoch) I faced so far is one >> I got in >> response to GetConfig. It read '1HOUR'. Is there a similar pattern for >> time values to rely on? >> >> > This value comes directly from the node configuration, and node configuration > is made to be human readable. Imo, that's the only reason why this time is > not given in seconds. > > ConfigData default.logger.interval=1HOUR EndMessage To make these values adjustable by users I have to parse them. So there should be a spec somewhere. It would be enough to give me a pointer where to find it in freenets sources (still a bit lost in there) and I'd happily add a page regarding Fcp types to the wiki. Juergen From jflesch at gmail.com Wed Feb 13 20:49:05 2008 From: jflesch at gmail.com (Jerome Flesch) Date: Wed, 13 Feb 2008 21:49:05 +0100 Subject: [Tech] Fcp sizes and time In-Reply-To: <47B348BF.8010807@arcor.de> References: <47B2AFF2.6040003@arcor.de> <200802131459.14479.jflesch@gmail.com> <47B348BF.8010807@arcor.de> Message-ID: <200802132149.08341.jflesch@gmail.com> > Jerome Flesch schrieb: > >> Hi, > >> > >> I have a question regarding byte sizes in Fcp. Along with some requests > >> sizes are passed as 1G or 1000K. > > > > I would be interrested to know where you saw that. From what I know, FCP > > always provide/need sizes/times in bytes/seconds (except for things > > related to the node config, see below). > > ConfigData > current.node.outputBandwidthLimit=12K > current.node.storeSize=1G > (...) > EndMessage > "Except for things related to the node config". > >> Quite obvious what it's about but > >> nevertheless it is not documented. > >> > >> Can I assume Fcp always uses kibs and values are always passed following > >> this patttern? > >> > >> -1 (not set) > >> n > >> n*1000 or (n)K > >> n*1.000.000 or (n)M > >> (...) > >> > >> The only time value (except from secs since epoch) I faced so far is one > >> I got in > >> response to GetConfig. It read '1HOUR'. Is there a similar pattern for > >> time values to rely on? > > > > This value comes directly from the node configuration, and node > > configuration is made to be human readable. Imo, that's the only reason > > why this time is not given in seconds. > > ConfigData > default.logger.interval=1HOUR > EndMessage > > > To make these values adjustable by users I have to parse them. > So there should be a spec somewhere. It would be enough to give > me a pointer where to find it in freenets sources (still a bit lost > in there) and I'd happily add a page regarding Fcp types to the wiki. > http://freenet.googlecode.com/svn/trunk/freenet/src/freenet/config/ http://freenet.googlecode.com/svn/trunk/freenet/src/freenet/config/LongOption.java For the sizes: http://freenet.googlecode.com/svn/trunk/freenet/src/freenet/support/Fields.java (Fields.parseLong() ; I think it's the i18n standard (so 1000 and not 1024)) For the periods: http://freenet.googlecode.com/svn/trunk/freenet/src/freenet/support/FileLoggerHook.java (FileLoggerHook.setInterval() ; ok, this one is not really standard ...) But I agree, it would probably be better if FCP would be able to specify the type of input expected (boolean, long, string, period, etc). Anyway, if it's an user input issue, you can do like I did in Thaw (the dev version only atm) or like it's done in FProxy : let the user enters whatever [s]he wants, and if the node complains / doesn't accept the value, then tell him (atm, in thaw, it doesn't tell him, it just put displays again the previous value). -- Jerome Flesch -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: This is a digitally signed message part. Url : http://emu.freenetproject.org/pipermail/tech/attachments/20080213/0e175cf7/attachment.pgp From jUrner at arcor.de Wed Feb 13 21:39:17 2008 From: jUrner at arcor.de (juergen urner) Date: Wed, 13 Feb 2008 22:39:17 +0100 Subject: [Tech] Fcp sizes and time In-Reply-To: <200802132149.08341.jflesch@gmail.com> References: <47B2AFF2.6040003@arcor.de> <200802131459.14479.jflesch@gmail.com> <47B348BF.8010807@arcor.de> <200802132149.08341.jflesch@gmail.com> Message-ID: <47B36385.6040603@arcor.de> Jerome Flesch schrieb: >> >> To make these values adjustable by users I have to parse them. >> So there should be a spec somewhere. It would be enough to give >> me a pointer where to find it in freenets sources (still a bit lost >> in there) and I'd happily add a page regarding Fcp types to the wiki. >> >> > http://freenet.googlecode.com/svn/trunk/freenet/src/freenet/config/ > http://freenet.googlecode.com/svn/trunk/freenet/src/freenet/config/LongOption.java > > For the sizes: > http://freenet.googlecode.com/svn/trunk/freenet/src/freenet/support/Fields.java > (Fields.parseLong() ; I think it's the i18n standard (so 1000 and not 1024)) > > For the periods: > http://freenet.googlecode.com/svn/trunk/freenet/src/freenet/support/FileLoggerHook.java > (FileLoggerHook.setInterval() ; ok, this one is not really standard ...) > Thanks for the pointers Jerome. I'll have a look. > But I agree, it would probably be better if FCP would be able to specify the > type of input expected (boolean, long, string, period, etc). > In the wiki or in the sources? For the sources a (parsable) header file containing all message / field / type decls would put a smile on my face. > Anyway, if it's an user input issue, you can do like I did in Thaw (the dev > version only atm) or like it's done in FProxy : let the user enters whatever > [s]he wants, and if the node complains / doesn't accept the value, then tell > him (atm, in thaw, it doesn't tell him, it just put displays again the > previous value). I don't trust users. You never know what they might be throwing at your input boxes ;-) Juergen From jflesch at gmail.com Wed Feb 13 21:44:15 2008 From: jflesch at gmail.com (Jerome Flesch) Date: Wed, 13 Feb 2008 22:44:15 +0100 Subject: [Tech] Fcp sizes and time In-Reply-To: <47B348BF.8010807@arcor.de> References: <47B2AFF2.6040003@arcor.de> <200802131459.14479.jflesch@gmail.com> <47B348BF.8010807@arcor.de> Message-ID: <200802132244.17937.jflesch@gmail.com> I commited a patch into the node to make it possible to ask for expected data types with FCP. See http://wiki.freenetproject.org/FCP2p0GetConfig , http://wiki.freenetproject.org/FCP2p0ConfigData . I hope it will help you (anyway it will help me :) For periods, I can't make the data type more specific than "string" at the moment, sorrry. Currently, I'm wondering if it would also be a good idea to convert all the values into basic units (-> in bytes) and let the client apps displays it in a more user-friendly way ? (I'm not really sure how the client would handle values like '1025' ...) > Jerome Flesch schrieb: > >> Hi, > >> > >> I have a question regarding byte sizes in Fcp. Along with some requests > >> sizes are passed as 1G or 1000K. > > > > I would be interrested to know where you saw that. From what I know, FCP > > always provide/need sizes/times in bytes/seconds (except for things > > related to the node config, see below). > > ConfigData > current.node.outputBandwidthLimit=12K > current.node.storeSize=1G > (...) > EndMessage > > >> Quite obvious what it's about but > >> nevertheless it is not documented. > >> > >> Can I assume Fcp always uses kibs and values are always passed following > >> this patttern? > >> > >> -1 (not set) > >> n > >> n*1000 or (n)K > >> n*1.000.000 or (n)M > >> (...) > >> > >> The only time value (except from secs since epoch) I faced so far is one > >> I got in > >> response to GetConfig. It read '1HOUR'. Is there a similar pattern for > >> time values to rely on? > > > > This value comes directly from the node configuration, and node > > configuration is made to be human readable. Imo, that's the only reason > > why this time is not given in seconds. > > ConfigData > default.logger.interval=1HOUR > EndMessage > > > To make these values adjustable by users I have to parse them. > So there should be a spec somewhere. It would be enough to give > me a pointer where to find it in freenets sources (still a bit lost > in there) and I'd happily add a page regarding Fcp types to the wiki. > > > Juergen > > > > > > _______________________________________________ > Tech mailing list > Tech at freenetproject.org > http://emu.freenetproject.org/cgi-bin/mailman/listinfo/tech -- Jerome Flesch -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: This is a digitally signed message part. Url : http://emu.freenetproject.org/pipermail/tech/attachments/20080213/4bba7f9d/attachment.pgp From jflesch at gmail.com Wed Feb 13 21:50:15 2008 From: jflesch at gmail.com (Jerome Flesch) Date: Wed, 13 Feb 2008 22:50:15 +0100 Subject: [Tech] Fcp sizes and time In-Reply-To: <47B36385.6040603@arcor.de> References: <47B2AFF2.6040003@arcor.de> <200802132149.08341.jflesch@gmail.com> <47B36385.6040603@arcor.de> Message-ID: <200802132250.17745.jflesch@gmail.com> > > But I agree, it would probably be better if FCP would be able to specify > > the type of input expected (boolean, long, string, period, etc). > > In the wiki or in the sources? For the sources a (parsable) header file > containing > all message / field / type decls would put a smile on my face. > I was thinking in FCP (and so in the wiki :) I'm not sure to see what you mean an "parsable" header file (or at least, I'm not sure where you want to put this header in the source ?) > > Anyway, if it's an user input issue, you can do like I did in Thaw (the > > dev version only atm) or like it's done in FProxy : let the user enters > > whatever [s]he wants, and if the node complains / doesn't accept the > > value, then tell him (atm, in thaw, it doesn't tell him, it just put > > displays again the previous value). > > I don't trust users. You never know what they might be throwing at your > input boxes ;-) > Correct, but I think you can trust the node to reject any invalid inputs :) -- Jerome Flesch -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: This is a digitally signed message part. Url : http://emu.freenetproject.org/pipermail/tech/attachments/20080213/29d67531/attachment.pgp From jUrner at arcor.de Thu Feb 14 10:16:05 2008 From: jUrner at arcor.de (juergen urner) Date: Thu, 14 Feb 2008 11:16:05 +0100 Subject: [Tech] Fcp sizes and time In-Reply-To: <200802132244.17937.jflesch@gmail.com> References: <47B2AFF2.6040003@arcor.de> <200802131459.14479.jflesch@gmail.com> <47B348BF.8010807@arcor.de> <200802132244.17937.jflesch@gmail.com> Message-ID: <47B414E5.50304@arcor.de> Jerome Flesch schrieb: > I commited a patch into the node to make it possible to ask for expected data > types with FCP. > See http://wiki.freenetproject.org/FCP2p0GetConfig , > http://wiki.freenetproject.org/FCP2p0ConfigData . > > I hope it will help you (anyway it will help me :) > > For periods, I can't make the data type more specific than "string" at the > moment, sorrry. > > Currently, I'm wondering if it would also be a good idea to convert all the > values into basic units (-> in bytes) and let the client apps displays it in > a more user-friendly way ? (I'm not really sure how the client would handle > values like '1025' ...) > Probably a good idea. This would separate processing logic from representation logic. Different clients may have different needs how to present 1K to the user. This is nothing the node should care about. Same goes for time intervals. But it may break existing clients. > In the wiki or in the sources? For the sources a (parsable) header file > containing > all message / field / type decls would put a smile on my face. > > > I was thinking in FCP (and so in the wiki :) > I'm not sure to see what you mean an "parsable" header file (or at least, I'm > not sure where you want to put this header in the source ?) > On second thought, extensive documentation is better. Would be nice to have one single file in freenets sources you can parse for consts and the like but this is really of minor importance. Juergen ...almost forgot. in node/PeerManager.java I can see all the peer node status consts defined. >> public static final int PEER_NODE_STATUS_CONNECTED = 1; In the Peer message this translates to 'CONNECTED'. Do you have a pointer where this is done? From toad at amphibian.dyndns.org Wed Feb 13 22:40:06 2008 From: toad at amphibian.dyndns.org (Matthew Toseland) Date: Wed, 13 Feb 2008 22:40:06 +0000 Subject: [Tech] Fcp sizes and time In-Reply-To: <200802132244.17937.jflesch@gmail.com> References: <47B2AFF2.6040003@arcor.de> <47B348BF.8010807@arcor.de> <200802132244.17937.jflesch@gmail.com> Message-ID: <200802132240.06943.toad@amphibian.dyndns.org> On Wednesday 13 February 2008 21:44, Jerome Flesch wrote: > I commited a patch into the node to make it possible to ask for expected data > types with FCP. > See http://wiki.freenetproject.org/FCP2p0GetConfig , > http://wiki.freenetproject.org/FCP2p0ConfigData . > > I hope it will help you (anyway it will help me :) > > For periods, I can't make the data type more specific than "string" at the > moment, sorrry. > > Currently, I'm wondering if it would also be a good idea to convert all the > values into basic units (-> in bytes) and let the client apps displays it in > a more user-friendly way ? (I'm not really sure how the client would handle > values like '1025' ...) Exactly, there's a reason why we keep the original string. If the user wants 12 GiB then we should remember that it's 12GiB and not 12884901888 bytes. > > > > Jerome Flesch schrieb: > > >> Hi, > > >> > > >> I have a question regarding byte sizes in Fcp. Along with some requests > > >> sizes are passed as 1G or 1000K. > > > > > > I would be interrested to know where you saw that. From what I know, FCP > > > always provide/need sizes/times in bytes/seconds (except for things > > > related to the node config, see below). > > > > ConfigData > > current.node.outputBandwidthLimit=12K > > current.node.storeSize=1G > > (...) > > EndMessage > > > > >> Quite obvious what it's about but > > >> nevertheless it is not documented. > > >> > > >> Can I assume Fcp always uses kibs and values are always passed following > > >> this patttern? > > >> > > >> -1 (not set) > > >> n > > >> n*1000 or (n)K > > >> n*1.000.000 or (n)M > > >> (...) > > >> > > >> The only time value (except from secs since epoch) I faced so far is one > > >> I got in > > >> response to GetConfig. It read '1HOUR'. Is there a similar pattern for > > >> time values to rely on? > > > > > > This value comes directly from the node configuration, and node > > > configuration is made to be human readable. Imo, that's the only reason > > > why this time is not given in seconds. > > > > ConfigData > > default.logger.interval=1HOUR > > EndMessage > > > > > > To make these values adjustable by users I have to parse them. > > So there should be a spec somewhere. It would be enough to give > > me a pointer where to find it in freenets sources (still a bit lost > > in there) and I'd happily add a page regarding Fcp types to the wiki. > > > > > > Juergen > > > > > > > > > > > > _______________________________________________ > > Tech mailing list > > Tech at freenetproject.org > > http://emu.freenetproject.org/cgi-bin/mailman/listinfo/tech > > > > -- > Jerome Flesch > -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: not available Url : http://emu.freenetproject.org/pipermail/tech/attachments/20080213/941f7edf/attachment.pgp From toad at amphibian.dyndns.org Thu Feb 14 14:37:42 2008 From: toad at amphibian.dyndns.org (Matthew Toseland) Date: Thu, 14 Feb 2008 14:37:42 +0000 Subject: [Tech] Fcp sizes and time In-Reply-To: <47B414E5.50304@arcor.de> References: <47B2AFF2.6040003@arcor.de> <200802132244.17937.jflesch@gmail.com> <47B414E5.50304@arcor.de> Message-ID: <200802141437.43484.toad@amphibian.dyndns.org> On Thursday 14 February 2008 10:16, juergen urner wrote: > ...almost forgot. in node/PeerManager.java I can see all the peer node > status > consts defined. > > >> public static final int PEER_NODE_STATUS_CONNECTED = 1; > > In the Peer message this translates to 'CONNECTED'. > Do you have a pointer where this is done? PeerNode.getPeerNodeStatusString(int) -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: not available Url : http://emu.freenetproject.org/pipermail/tech/attachments/20080214/2ff63d1c/attachment.pgp From jUrner at arcor.de Thu Feb 14 14:45:24 2008 From: jUrner at arcor.de (juergen urner) Date: Thu, 14 Feb 2008 15:45:24 +0100 Subject: [Tech] Fcp sizes and time In-Reply-To: <200802132240.06943.toad@amphibian.dyndns.org> References: <47B2AFF2.6040003@arcor.de> <47B348BF.8010807@arcor.de> <200802132244.17937.jflesch@gmail.com> <200802132240.06943.toad@amphibian.dyndns.org> Message-ID: <47B45404.6040404@arcor.de> Matthew Toseland schrieb: > On Wednesday 13 February 2008 21:44, Jerome Flesch wrote: > >> I commited a patch into the node to make it possible to ask for expected >> > data > >> types with FCP. >> See http://wiki.freenetproject.org/FCP2p0GetConfig , >> http://wiki.freenetproject.org/FCP2p0ConfigData . >> >> I hope it will help you (anyway it will help me :) >> >> For periods, I can't make the data type more specific than "string" at the >> moment, sorrry. >> >> Currently, I'm wondering if it would also be a good idea to convert all the >> values into basic units (-> in bytes) and let the client apps displays it in >> a more user-friendly way ? (I'm not really sure how the client would handle >> values like '1025' ...) >> > > Exactly, there's a reason why we keep the original string. If the user wants > 12 GiB then we should remember that it's 12GiB and not 12884901888 bytes. > You are right. I change my question to: why are there kbytes and* kibiytes? I wouldn't mind forcing one or the other upon every other client around by the way my client handles it. Juergen From toad at amphibian.dyndns.org Thu Feb 14 15:07:09 2008 From: toad at amphibian.dyndns.org (Matthew Toseland) Date: Thu, 14 Feb 2008 15:07:09 +0000 Subject: [Tech] Fcp sizes and time In-Reply-To: <47B45404.6040404@arcor.de> References: <47B2AFF2.6040003@arcor.de> <200802132240.06943.toad@amphibian.dyndns.org> <47B45404.6040404@arcor.de> Message-ID: <200802141507.16028.toad@amphibian.dyndns.org> On Thursday 14 February 2008 14:45, juergen urner wrote: > Matthew Toseland schrieb: > > On Wednesday 13 February 2008 21:44, Jerome Flesch wrote: > > > >> I commited a patch into the node to make it possible to ask for expected > >> > > data > > > >> types with FCP. > >> See http://wiki.freenetproject.org/FCP2p0GetConfig , > >> http://wiki.freenetproject.org/FCP2p0ConfigData . > >> > >> I hope it will help you (anyway it will help me :) > >> > >> For periods, I can't make the data type more specific than "string" at the > >> moment, sorrry. > >> > >> Currently, I'm wondering if it would also be a good idea to convert all the > >> values into basic units (-> in bytes) and let the client apps displays it in > >> a more user-friendly way ? (I'm not really sure how the client would handle > >> values like '1025' ...) > >> > > > > Exactly, there's a reason why we keep the original string. If the user wants > > 12 GiB then we should remember that it's 12GiB and not 12884901888 bytes. > > > > You are right. I change my question to: why are there kbytes and* kibiytes? > I wouldn't mind forcing one or the other upon every other client around by > the way my client handles it. Because there was no real reason to force one or the other? Feel free to convert it into bytes and offer a slider in your app. > > Juergen -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: not available Url : http://emu.freenetproject.org/pipermail/tech/attachments/20080214/cb7a44e7/attachment.pgp From toad at amphibian.dyndns.org Tue Feb 19 15:32:26 2008 From: toad at amphibian.dyndns.org (Matthew Toseland) Date: Tue, 19 Feb 2008 15:32:26 +0000 Subject: [Tech] Pocket Switched Networking Message-ID: <200802191532.33044.toad@amphibian.dyndns.org> http://www.cl.cam.ac.uk/~ph315/ "A Socio-Aware Overlay for Publish/Subscribe Communication in Delay Tolerant Networks" and others. http://www.cl.cam.ac.uk/research/srg/netos/haggle/ (many broken links, the researcher promised he'd fix them soon). Might be interesting; long term IMHO Freenet is going to have to use this sort of stuff for a lot of traffic in more hostile environments: If the authorities implement traffic flow analysis on the internet, you have to use non-internet transports, which means wireless, PDA rendezvous, sneakernet, and the occasional parasitic internet transport (e.g. steal the video stream on skype). But the papers may be interesting even in the nearer term; community detection (like robert's network coloring code?), social networks ... Caveat: I haven't read the papers, I don't have time right now. -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: not available Url : http://emu.freenetproject.org/pipermail/tech/attachments/20080219/eda635aa/attachment.pgp From sandos at home.se Tue Feb 19 15:59:57 2008 From: sandos at home.se (=?windows-1252?Q?John_B=E4ckstrand?=) Date: Tue, 19 Feb 2008 16:59:57 +0100 Subject: [Tech] Pocket Switched Networking In-Reply-To: <200802191532.33044.toad@amphibian.dyndns.org> References: <200802191532.33044.toad@amphibian.dyndns.org> Message-ID: <47BAFCFD.1000605@home.se> Matthew Toseland wrote: > http://www.cl.cam.ac.uk/~ph315/ > > "A Socio-Aware Overlay for Publish/Subscribe Communication in Delay Tolerant > Networks" and others. > > http://www.cl.cam.ac.uk/research/srg/netos/haggle/ (many broken links, the > researcher promised he'd fix them soon). > > Might be interesting; long term IMHO Freenet is going to have to use this sort > of stuff for a lot of traffic in more hostile environments: If the > authorities implement traffic flow analysis on the internet, you have to use > non-internet transports, which means wireless, PDA rendezvous, sneakernet, > and the occasional parasitic internet transport (e.g. steal the video stream > on skype). But the papers may be interesting even in the nearer term; > community detection (like robert's network coloring code?), social > networks ... > > Caveat: I haven't read the papers, I don't have time right now. I was actually thinking about freenet/sneakernets the other day, realizing that sneakernet bandwidth is going up much, much faster than bandwidth over dsl, ethernet or fibre connections: I bought a 8GB memory card for my phone for about 60 ?. I carry this to and from work every day: 8GB full-duplex. Not a whole lot compared to many fibre connections in Sweden, but compared to say DSL upstream it starts looking pretty good. This is without me having to even think about carrying that data, as opposed to carrying a 3.5" disk. Now, if freenet would just run on J2ME... ;) I guess I will have to get a phone with a --- John B?ckstrand From m.rogers at cs.ucl.ac.uk Wed Feb 20 14:42:02 2008 From: m.rogers at cs.ucl.ac.uk (Michael Rogers) Date: 20 Feb 2008 14:42:02 +0000 Subject: [Tech] I'm an idiot Message-ID: Just realised the overridden method wasn't getting called, which is why the results looked rather similar! Re-running the sims now. Cheers, Michael From peteheist at yahoo.com Fri Feb 22 17:07:22 2008 From: peteheist at yahoo.com (Pete Heist) Date: Fri, 22 Feb 2008 18:07:22 +0100 Subject: [Tech] how does swapping work? Message-ID: <000001c87575$65e97040$6415a8c0@heistlt> I'm looking for a basic explanation of how the swapping of the locations of two nodes works (details aside, as it seems the implementation may still be in flux?) I'm confused as to how nodes can swap locations in the network without either changing the nodes they're connected to or changing their IDs (which would be a security problem in the first case or a routing complication in the second)...but I may just have a fundamental misunderstanding of how the 0.7 routing works... From m.rogers at cs.ucl.ac.uk Sat Feb 23 18:59:18 2008 From: m.rogers at cs.ucl.ac.uk (Michael Rogers) Date: Sat, 23 Feb 2008 18:59:18 +0000 Subject: [Tech] how does swapping work? In-Reply-To: <000001c87575$65e97040$6415a8c0@heistlt> References: <000001c87575$65e97040$6415a8c0@heistlt> Message-ID: <47C06D06.1060501@cs.ucl.ac.uk> Pete Heist wrote: > I'm looking for a basic explanation of how the swapping of the locations of > two nodes works (details aside, as it seems the implementation may still be > in flux?) Each node has a routing location, which is a number between 0 and 1 representing a point on the perimeter of a circle. For efficient routing, the distance between neighbouring nodes' routing locations should be minimised. The swapping algorithm tries to find a globally efficient solution to this problem using only local information, by swapping the locations of nodes without changing their connections. * Each node periodically uses a random walk to select a partner and decides whether to swap locations with it. * If swapping would reduce the product of the partners' distances to their neighbours, they swap. * If swapping would increase the product of the partners' distances to their neighbours, they swap with a probability that decreases with the increase in distance. The last part prevents the algorithm from getting stuck in suboptimal solutions. Oskar's paper has all the details: http://www.math.chalmers.se/~ossa/swroute.pdf Cheers, Michael From toad at amphibian.dyndns.org Sat Feb 23 19:00:40 2008 From: toad at amphibian.dyndns.org (Matthew Toseland) Date: Sat, 23 Feb 2008 19:00:40 +0000 Subject: [Tech] how does swapping work? In-Reply-To: <000001c87575$65e97040$6415a8c0@heistlt> References: <000001c87575$65e97040$6415a8c0@heistlt> Message-ID: <200802231900.49663.toad@amphibian.dyndns.org> On Friday 22 February 2008 17:07, Pete Heist wrote: > I'm looking for a basic explanation of how the swapping of the locations of > two nodes works (details aside, as it seems the implementation may still be > in flux?) > > I'm confused as to how nodes can swap locations in the network without > either changing the nodes they're connected to or changing their IDs (which > would be a security problem in the first case or a routing complication in > the second)...but I may just have a fundamental misunderstanding of how the > 0.7 routing works... A node's "location" is simply a number between 0 and 1. We route requests towards the location closest to the content we are trying to find. Does that clarify things a bit? -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: not available Url : http://emu.freenetproject.org/pipermail/tech/attachments/20080223/41870d00/attachment.pgp From peteheist at yahoo.com Tue Feb 26 15:36:58 2008 From: peteheist at yahoo.com (Pete Heist) Date: Tue, 26 Feb 2008 16:36:58 +0100 Subject: [Tech] how does swapping work? Message-ID: <000301c8788d$6de20530$6415a8c0@heistlt> On 2008-02-23 18:59, Michael Rogers wrote: > Pete Heist wrote: > > I'm looking for a basic explanation of how the swapping of the locations of > > two nodes works (details aside, as it seems the implementation may still be > > in flux?) > > Each node has a routing location, which is a number between 0 and 1 > representing a point on the perimeter of a circle. For efficient > routing, the distance between neighbouring nodes' routing locations > should be minimised. The swapping algorithm tries to find a globally > efficient solution to this problem using only local information, by > swapping the locations of nodes without changing their connections. > > ... That helps a lot, and I read Oskar's paper. The part I'm still not understanding is how swapping works. If two nodes swap locations, those two nodes are the only nodes that know about the swap (right, or are neighbors also informed, and if so how can they be informed reliably)? So if someone goes to find a piece of content, they may end up at a node that has swapped with another one. Do they then have to start the routing process over again to find the node with the content, and possibly do this multiple times? Thanks for the help... Pete From toad at amphibian.dyndns.org Tue Feb 26 16:00:18 2008 From: toad at amphibian.dyndns.org (Matthew Toseland) Date: Tue, 26 Feb 2008 16:00:18 +0000 Subject: [Tech] how does swapping work? In-Reply-To: <000301c8788d$6de20530$6415a8c0@heistlt> References: <000301c8788d$6de20530$6415a8c0@heistlt> Message-ID: <200802261600.24223.toad@amphibian.dyndns.org> On Tuesday 26 February 2008 15:36, Pete Heist wrote: > On 2008-02-23 18:59, Michael Rogers wrote: > > Pete Heist wrote: > > > I'm looking for a basic explanation of how the swapping of the locations > of > > > two nodes works (details aside, as it seems the implementation may still > be > > > in flux?) > > > > Each node has a routing location, which is a number between 0 and 1 > > representing a point on the perimeter of a circle. For efficient > > routing, the distance between neighbouring nodes' routing locations > > should be minimised. The swapping algorithm tries to find a globally > > efficient solution to this problem using only local information, by > > swapping the locations of nodes without changing their connections. > > > > ... > > That helps a lot, and I read Oskar's paper. > > The part I'm still not understanding is how swapping works. If two nodes > swap locations, those two nodes are the only nodes that know about the swap > (right, or are neighbors also informed, and if so how can they be informed > reliably)? So if someone goes to find a piece of content, they may end up at > a node that has swapped with another one. Do they then have to start the > routing process over again to find the node with the content, and possibly > do this multiple times? Only their direct neighbours are informed, and only their direct neighbours need to know. > > Thanks for the help... > > Pete -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: not available Url : http://emu.freenetproject.org/pipermail/tech/attachments/20080226/8d27cda7/attachment.pgp From toad at amphibian.dyndns.org Wed Feb 27 18:51:32 2008 From: toad at amphibian.dyndns.org (Matthew Toseland) Date: Wed, 27 Feb 2008 18:51:32 +0000 Subject: [Tech] Opera JPEG vulnerability Message-ID: <200802271851.38357.toad@amphibian.dyndns.org> This is interesting because it came at the end of a thread on Frost where the OP was arguing that Freenet shouldn't filter JPEGs. (Freenet strips out EXIF data and other unknown chunks from JPEGs on download to maximize security; in the future we will do something similar on inserts). http://www.opera.com/support/search/view/879/ -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: not available Url : http://emu.freenetproject.org/pipermail/tech/attachments/20080227/8ca9a129/attachment.pgp From stwa4647000 at yahoo.co.uk Wed Feb 27 22:19:52 2008 From: stwa4647000 at yahoo.co.uk (Stephen Walford) Date: Wed, 27 Feb 2008 22:19:52 +0000 (GMT) Subject: [Tech] Outside web access? Message-ID: <994755.82027.qm@web25412.mail.ukl.yahoo.com> An HTML attachment was scrubbed... URL: http://emu.freenetproject.org/pipermail/tech/attachments/20080227/51de8dca/attachment.htm From sandos at home.se Thu Feb 28 08:40:39 2008 From: sandos at home.se (=?UTF-8?B?Sm9obiBCw6Rja3N0cmFuZA==?=) Date: Thu, 28 Feb 2008 09:40:39 +0100 Subject: [Tech] Outside web access? In-Reply-To: <994755.82027.qm@web25412.mail.ukl.yahoo.com> References: <994755.82027.qm@web25412.mail.ukl.yahoo.com> Message-ID: <47C67387.6020405@home.se> Stephen Walford wrote: > Is there any way possible for someone to access Freenet sites with a web > browser WITHOUT first having installed the Freenet software, ie. general > http access? If not, will future developments in Freenet facilitate this? Not unless someone opens up a public fproxy I guess, and who would want to? Actually I considered this not more than 8 hours ago, when thinking about the wikileaks "takedown"/censoring. It went something like this: People _really_ should start using freenet for this sort of information. The problem is that people are too lazy to install freenet "just" too see some sensitive information. Hey, I should open up a http proxy for freenet, and then manually add filters when government agencies start complaining. The filtered pages would simply say: "please install freenet to see this page, since the legal system has asked me not to show this". The hope is that by moderating things manually, I would not be prosecuted for anything (hopefully) and people will at least get a chance to see information before it is banned. Removal of domainnames though, as for wikileaks, is really not a working measure as can be seen by the numerous alternative domain-names now up instead. So, the internet, for the moment at least, seems to do sort-of-fine without freenet. For this sort of information, at least. --- John B?ckstrand From marcoc1 at dada.it Fri Feb 29 18:49:07 2008 From: marcoc1 at dada.it (Marco A. Calamari) Date: Fri, 29 Feb 2008 19:49:07 +0100 Subject: [Tech] Opera JPEG vulnerability In-Reply-To: <200802271851.38357.toad@amphibian.dyndns.org> References: <200802271851.38357.toad@amphibian.dyndns.org> Message-ID: <1204310947.17396.34.camel@mustafar.winstonsmith.info> On Wed, 2008-02-27 at 18:51 +0000, Matthew Toseland wrote: > This is interesting because it came at the end of a thread on Frost where the > OP was arguing that Freenet shouldn't filter JPEGs. (Freenet strips out EXIF > data and other unknown chunks from JPEGs on download to maximize security; in > the future we will do something similar on inserts). IMHO changing in any way information inserted in Freenet *must* be documented, evident in user interface, up by default but easily user selectable. -- +--------------- http://www.winstonsmith.info ---------------+ | il Progetto Winston Smith: scolleghiamo il Grande Fratello | | the Winston Smith Project: unplug the Big Brother | | Marco A. Calamari marcoc at marcoc.it http://www.marcoc.it | | DSS/DH: 8F3E 5BAE 906F B416 9242 1C10 8661 24A9 BFCE 822B | + PGP RSA: ED84 3839 6C4D 3FFE 389F 209E 3128 5698 ----------+ -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 307 bytes Desc: This is a digitally signed message part Url : http://emu.freenetproject.org/pipermail/tech/attachments/20080229/0ea04961/attachment.pgp From toad at amphibian.dyndns.org Fri Feb 29 18:53:47 2008 From: toad at amphibian.dyndns.org (Matthew Toseland) Date: Fri, 29 Feb 2008 18:53:47 +0000 Subject: [Tech] Opera JPEG vulnerability In-Reply-To: <1204310947.17396.34.camel@mustafar.winstonsmith.info> References: <200802271851.38357.toad@amphibian.dyndns.org> <1204310947.17396.34.camel@mustafar.winstonsmith.info> Message-ID: <200802291853.48789.toad@amphibian.dyndns.org> On Friday 29 February 2008 18:49, Marco A. Calamari wrote: > On Wed, 2008-02-27 at 18:51 +0000, Matthew Toseland wrote: > > This is interesting because it came at the end of a thread on Frost where the > > OP was arguing that Freenet shouldn't filter JPEGs. (Freenet strips out EXIF > > data and other unknown chunks from JPEGs on download to maximize security; in > > the future we will do something similar on inserts). > > IMHO changing in any way information inserted in Freenet *must* > be documented, evident in user interface, up by default > but easily user selectable. The intention is that there should be a checkbox in jSite etc to turn insert-time filtering on or off. Maybe a global one and a per-file one too. -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: not available Url : http://emu.freenetproject.org/pipermail/tech/attachments/20080229/0e5fc666/attachment.pgp