i2p.www/www.i2p2/pages/meeting61.html

{% extends "_layout.html" %}
{% block title %}I2P Development Meeting 61{% endblock %}
{% block content %}<div class="irclog">
<p>[23:04] &lt;jrand0m&gt; 0.0) welcome </p>
<p>[23:04] &lt;jrand0m&gt; 1.0) dev status </p>
<p>[23:04] &lt;jrand0m&gt; 2.0) new server </p>
<p>[23:04] &lt;jrand0m&gt; 3.0) questions? </p>
<p>[23:04] &lt;jrand0m&gt; 0.0) welcome to the 61st iip meeting </p>
<p>[23:04] &lt;jrand0m&gt; hi. </p>
<p>[23:04] &lt;mihi&gt; hi all</p>
<p>[23:04] &lt;dm&gt; hi you two ;)</p>
<p>[23:04] &lt;jrand0m&gt; ok, 1.0) dev status </p>
<p>[23:05] &lt;jrand0m&gt; we're pretty much ready for 0.2 - which is where people would be able to run i2p.   </p>
<p>[23:06] &lt;dm&gt; great!</p>
<p>[23:06] &lt;dm&gt; ETA?</p>
<p>[23:06] &lt;jrand0m&gt; it won't scale, and it is currently implemented with the insanely secure style of the ElG+AES (rather than the optimized w/ session tags) </p>
<p>[23:06] &lt;jrand0m&gt; eta: a few days (related point: agenda item 2.0) </p>
<p>[23:07] &lt;jrand0m&gt; but it will be functional.  I'm able to send messages of 10 bytes as well as 100K in ~10s </p>
<p>[23:07] &lt;jrand0m&gt; (both on the same CPU, so its cpu contention) </p>
<p>[23:07] &lt;jrand0m&gt; still some tunnel failback I'm working through, then its time to build the ant scripts, etc </p>
<p>[23:07] &lt;jrand0m&gt; actually </p>
<p>[23:08] &lt;jrand0m&gt; a few seconds ago I built everything using Kaffe 1.1.1 on freebsd with not even a warning. </p>
<p>[23:08] * jrand0m &lt;3 not having to make changes to code</p>
<p>[23:08] &lt;mihi&gt; but it wasn't working either, was it? ;)</p>
<p>[23:08] &lt;jrand0m&gt; heh </p>
<p>[23:09] &lt;jrand0m&gt; no, it works </p>
<p>[23:09] &lt;jrand0m&gt; the prob I'm working through is a failover (aka if one is shut down and starts back up, comm recovers cleanly) </p>
<p>[23:10] &lt;mihi&gt; have you tried to send a complete ASCII table (bytes 0x00 to 0xFF)? And does it appear as it should on the other end?</p>
<p>[23:10] &lt;jrand0m&gt; not yet, havent tried i2ptunnel yet either.  but shall tonight </p>
<p>[23:10] &lt;mihi&gt; try i2ptunnel with a binary connection (like ssh), no text based one.</p>
<p>[23:11] &lt;jrand0m&gt; 'k </p>
<p>[23:11] &lt;mihi&gt; what if a node shuts down and does not come up again?</p>
<p>[23:11] &lt;jrand0m&gt; then communication with it will fail ;) </p>
<p>[23:12] &lt;jrand0m&gt; (guaranteed delivery now fails correctly and the SDK has been patched to act on that) </p>
<p>[23:13] &lt;jrand0m&gt; ok, thats it for status atm.  I've been pulled in other directions for the last two days, but that should be quieting down shortly. </p>
<p>[23:13] &lt;jrand0m&gt; 2.0) new server </p>
<p>[23:13] &lt;jrand0m&gt; we have a new server on a triple homed T3.   </p>
<p>[23:14] &lt;jrand0m&gt; its going to host webcvs, a default i2p router, a webserver with seed router references, and a few default services (e.g. an i2p tunnel pointing at a squid instance, an i2ptunnel pointing at cvs, an i2ptunnel pointing at a webserver) </p>
<p>[23:15] *** Signoff: mihi (EOF From client)</p>
<p>[23:15] *** mihi_ (~mihi@anon.iip) has joined channel #iip-dev</p>
<p>[23:15] &lt;jrand0m&gt; i'll also be tossing on a public isproxy later </p>
<p>[23:15] &lt;dm&gt; damn, how much is that costing you?</p>
<p>[23:15] &lt;jrand0m&gt; 65/mo </p>
<p>[23:15] &lt;jrand0m&gt; 40Gb transfer, 2Gb data </p>
<p>[23:15] &lt;dm&gt; bandwidth limit?</p>
<p>[23:15] &lt;dm&gt; cool.</p>
<p>[23:16] *** mihi_ is now known as mihi</p>
<p>[23:16] &lt;jrand0m&gt; perhaps that server will become our cvs, but no need to switch right now.   </p>
<p>[23:16] &lt;jrand0m&gt; ok, 3.0) questions </p>
<p>[23:17] &lt;jrand0m&gt; [yes, this has been the fastest meeting I've ever seen ;)] </p>
<p>[23:17] &lt;mihi_backup&gt; do you know anything about naming service progress?</p>
<p>[23:17] &lt;jrand0m&gt; I hear wiht has committed a few files, but I haven't checked them out yet </p>
<p>[23:17] &lt;dm&gt; I'd like to offer my (mental) support to all the developers. Stay humble yet determined.</p>
<p>[23:18] &lt;mihi_backup&gt; they are only dummies afaik.</p>
<p>[23:18] &lt;jrand0m&gt; clueless wankers, for sure. </p>
<p>[23:18] &lt;mihi_backup&gt; interfaces to implement against.</p>
<p>[23:18] &lt;dm&gt; dummies?</p>
<p>[23:18] &lt;jrand0m&gt; lol </p>
<p>[23:19] &lt;dm&gt; ah</p>
<p>[23:19] * jrand0m extends MyParents</p>
<p>[23:19] &lt;mihi_backup&gt; java does not allow multiple inheritance ;)</p>
<p>[23:19] &lt;dm&gt; no bastards...</p>
<p>[23:20] * jrand0m implements MyDad, MyMom // just seems wrong</p>
<p>[23:20] &lt;mihi_backup&gt; although you do not extend either your mother nor your father (see the liskov (sp?) substitution principle)</p>
<p>[23:20] * dm implements jrandom'sMom</p>
<p>[23:20] &lt;dm&gt; ;)</p>
<p>[23:20] &lt;jrand0m&gt; oh, good point mihi </p>
<p>[23:20] * dm now implements mihi'smother</p>
<p>[23:20] &lt;jrand0m&gt; heh </p>
<p>[23:21] *** Signoff: mihi (Ping timeout)</p>
<p>[23:21] &lt;mihi_backup&gt; you just need a constructor public Human(Human father, Human mother) throws SameSexException</p>
<p>[23:22] &lt;dm&gt; nice one</p>
<p>[23:22] &lt;jrand0m&gt; I think modern science requires a few more constructors </p>
<p>[23:22] &lt;dm&gt; although it's not public</p>
<p>[23:22] &lt;dm&gt; it's pubic</p>
<p>[23:22] *** mihi (mihi@anon.iip) has joined channel #iip-dev</p>
<p>[23:22] &lt;dm&gt; I crack myself up.</p>
<p>[23:23] &lt;jrand0m&gt; ok, on that note </p>
<p>[23:23] &lt;jrand0m&gt; I think we've just gotten away with a 23 minute meeting :) </p>
<p>[23:23] * jrand0m slips the *baf*er out of my back pocket and *baf*s the meeting closed</p>
</div>
{% endblock %}