IRC log for #tomcat on 20080330

00:04.19*** join/#tomcat thiamteck (i=3c30667c@gateway/web/ajax/mibbit.com/x-3e72b2e36168961c)
00:05.29*** join/#tomcat thiamteck (i=3c30667c@gateway/web/ajax/mibbit.com/x-6c7c0f913d323db1)
00:08.09*** join/#tomcat thiamteck (i=3c30667c@gateway/web/ajax/mibbit.com/x-0a104ad40bcf8e1e)
00:42.03internat85can tomcat have a specific error log for each host context?
00:56.13*** join/#tomcat vikumar (n=Vimal@nat/redhat-in/x-6ee2eae38d20f4ee)
01:03.26*** join/#tomcat vikumar (n=Vimal@nat/redhat-in/x-6f6eb637710e36c5)
01:03.53*** join/#tomcat rossand (n=rossand@CPE000c413a19a3-CM00159a025ad4.cpe.net.cable.rogers.com)
01:04.03*** part/#tomcat rossand (n=rossand@CPE000c413a19a3-CM00159a025ad4.cpe.net.cable.rogers.com)
01:10.06jasonbinternat85: The answer is yes.  However, host != context.  They are two separate entities.
01:55.49*** join/#tomcat Hellaenergy (n=Hellaene@c-75-72-155-242.hsd1.mn.comcast.net)
02:00.29jasonbHi Hellaenergy.
02:05.33mistik1jasonb: I never did get that build to work
02:06.16Hellaenergyhi jasonb
02:06.47HellaenergyAnyone here using JSF?
02:08.56jasonbmistik1: Our book, Tomcat: The Definitive Guide, 2nd Edition (O'Reilly) goes over how to build Tomcat step by step.  That would probably help you.
02:09.22jasonbmistik1: It doesn't go over building the extras stuff, but does show how to build Tomcat itself.
02:13.19lyken_yeah ive read it, its pretty sweet
02:13.42jasonblyken_: Like it?  Did you read a large chunk of it?
02:13.56mistik1I just came back and looks like my brain started firing a bit better with a couple hours away
02:14.18mistik1Here I was trying compilerarg when all I needed was ANT_OPTS
02:14.24mistik1DUH!!
02:14.32mistik1build successfull
02:14.36jasonbahh, there.
02:14.37lyken_jasonb: i think im the only one that has read the entire thing, including you
02:15.27mistik1jasonb: Is it on safari yet?
02:15.33jasonblyken_: Excellent.  I have one main question then: while reading it, did it give you ideas about what features were missing from Tomcat, or what you could write for Tomcat that would be really popular?
02:15.39jasonbmistik1: Of course.  :)
02:15.44mistik1awesome
02:15.53mistik1Then I'll be checking it out
02:16.04jasonbmistik1: It's very detailed.
02:16.27mistik1Cool, there a quite few things I want to know how to do with old tommy
02:16.44lyken_jasonb: yeah i found some OS X sections would be alright ;)
02:17.02jasonblyken_: :)
02:17.51jasonblyken_: While writing I found like three main things that I thought were really missing, and that would get used if they weren't missing anymore.
02:18.31lyken_oh
02:18.56lyken_gotta go out in a minute
02:18.58lyken_im in miami atm
02:19.09jasonblyken_: 1) an elegant, performant way to proxy requests to another web server, 2) a separate program that automates performance testing, and 3) a new API library that allows webapps to have more control over Tomcat.
02:19.11jasonblyken_: Okay.
02:19.22jasonblyken_: I won't keep you.  Just thought I'd chat for a minute.
02:19.27lyken_jasonb: yup
02:19.32lyken_that sounds exactly whats missing
02:19.33jasonblyken_: Stay safe over there.. ugh.
02:19.40lyken_and it'll take till version 8 to get them too
02:19.46lyken_eh its been great
02:19.54lyken_i was pretty drunk last night and managed to only spent $30
02:19.56lyken_lol
02:19.59jasonbhah!
02:20.04Hellaenergyjasonb, a built in trouble shooting util (configtest) would be nice too.
02:20.08lyken_in a place that has $9 carona's
02:20.21lyken_booze soo expensive here
02:20.36jasonbHellaenergy: Well, there's already a way to validate the server.xml file.. though it's probably not real exhaustive.
02:20.48mistik1Will there be an embedded tomcat-6.x?
02:21.01jasonblyken_: Inflation has really hit the US hard.
02:21.06lyken_yes
02:21.09jasonblyken_: Prices have gone up sharply.
02:21.13lyken_indeed
02:21.29jasonbmistik1: You can already use Tomcat 6 as an embedded servlet container.  See the Embedded class.
02:21.34lyken_anywho, time to stop geeking out and get ready for this partay, adios
02:21.41jasonbmistik1: The Embedded class's Javadocs say how it works.
02:21.46lyken_later jason
02:21.51jasonblyken_: L8r!  Have fun.
02:22.09Hellaenergyjasonb, how is that?
02:22.13mistik1jasonb: I have done it with 5.5, unless its changed a lot
02:22.43jasonbmistik1: It hasn't changed all that much.
02:22.51mistik1nice
02:23.04jasonbIt might not have changed at all.
02:24.44jasonbHellaenergy: Hmm, my notes only show how to perform validation of web.xml.
02:26.19jasonbHellaenergy: And, doing that is very obscure.  My notes say how to do it (probably scraped from the tomcat-dev list), but it only works with an old xerces jar.
02:26.51jasonbHellaenergy: It might be the case that there's no way to validate a server.xml file, nor to troubleshoot what's wrong with one other than trying to run Tomcat with it.
02:27.22jasonbHellaenergy: Tomcat is supposed to tell you what's going wrong, but I'll admit it isn't real good at saying what's wrong in plain english.
02:35.53HellaenergyLooks like we found another thing that tomcat needs
02:36.01jasonbYes.
02:36.13jasonbApache httpd does offer at least one way to check the main config file.
02:36.39HellaenergyExactly thus my configtest reference :)
02:36.54HellaenergyAnd I use that all the time
02:39.06mistik1jasonb: The book is on my bookshelf ;)
02:39.17jasonbmistik1: Cool.  Let me know what you think of it.
02:39.28mistik1I will
02:40.24Hellaenergymine too :)
02:40.53HellaenergyI found the info on the eclipse compiler interesting
02:41.02jasonbIt's always been on my bookshelf.  :)
02:41.20HellaenergyWhy the hell did they do that? Licensing?
02:42.04jasonbdo what?
02:42.24jasonbReimplement a whole new Java compiler?
02:43.44HellaenergyYes
02:46.38jasonbWhat I heard was: Sun's compiler wasn't friendly at all when you needed to invoke it from within your Java application.  It had memory leaks, was slow, and didn't have what many developers wanted.  Also, IBM wants to control more of Java, even though Sun "owned" it, mainly because IBM can't just start a new language and try to overtake Java.
02:47.20jasonbJava already became the popular language.  So, IBM set out to own the IDE, own the compiler, and other things.
02:47.33HellaenergyVery interesting
02:47.49jasonbIt is in large part a struggle between IBM and Sun.
02:48.05HellaenergyTrying to beat Sun at their own game.
02:48.43HellaenergyThat seems unbelieveable to me that Sun's compiler sucks that bad.
02:48.45jasonbYes.  And, now that the JDK is open source, that means Sun tries to own less of Java.  But, Sun isn't just allowing IBM to own the Java IDE.
02:49.22jasonbNo, Sun's compiler has been around for a very long time, and has grown iteratively, so I believe that it has lots of design flaws that wouldn't be there if they wrote a brand new one today (like IBM did).
02:49.27jasonbIt does make sense.
02:50.03HellaenergyInteresting
02:50.13jasonbAlso, having their own compiler is very important for owning the IDE space.
02:50.17npmquestion: if fronting a bunch of java apps through apache (mod_proxy, proxy_ajp, ProxyPass) if i want to add HTML to do google AdSense to all my apps, what's the best way to do it through apache... for example, would mod_include work?  mod_layout ??
02:50.41jasonbThe eclipse IDE recompiles dynamically so well because IBM tightly integrated their own compiler.
02:50.52Hellaenergynpm try #apache
02:51.14npmtrue, but they won't know about mod_jk
02:51.33Hellaenergysure they will
02:51.36jasonbThey should know about mod_proxy.
02:52.09jasonbnpm: Why not do the google AdSense stuff inside of your webapp instead of inside Apache httpd?
02:52.12npmbut they're javaless heathens!!
02:52.53npmjasonb... because i don't want to have to maintain and customize difft code across multiple difft webapps
02:53.32npmi just want to slap on some ads on the top and bottom
02:54.18jasonbWell, okay.  You may have more options to do that in httpd.
02:54.55npmand it also seems like a good place to "gate" whether ads are displayed at all.
02:54.57jasonbBut, then, it won't have to do with mod_proxy nor mod_jk.  It would be the same as asking "how do I include a header and/or footer in every web page that httpd serves?"
02:55.29npmyes. any web page, and also mod_proxy/mod_jk
02:55.51npmbut alas for example mod_include expects .shtml files with directives...
02:55.56jasonbnpm: it is possible to do this in Tomcat using custom ServletFilter that gets automagically included in all webapps by configuring it in conf/web.xml.
02:56.04npmso chances are it won't work in a straightforward fashiohn
02:56.35npmjasonb -- that'd probably be my second option if i can't get it working in apache
02:56.59npmi think the java side will incur enough load as is :-)
02:57.08Hellaenergynpm how are  your pages constructed?
02:57.38HellaenergyDo you plan to use frames or something?
02:57.43npmsome in roller. some in xwiki, some in other tools
02:58.09npmno frames
02:59.35HellaenergySo how did you plan on structuring your proxy definitions?
03:00.35Hellaenergy| /webapp/
03:00.40npme.g. ProxyPass /roller/ ajp://127.0.0.1:8009/roller/
03:01.04Hellaenergyok so all references to / will be handled by apache
03:01.18npmyes, all references period will be handled by apache
03:01.23Hellaenergyso put your good stuff there
03:01.27npmyep
03:01.36npmmod_layout might do it
03:01.46npmhttp://tangent.org/362/mod_layout.html
03:02.02HellaenergyWhat is your delima again?
03:02.07npmbut i was hoping for something more supported/standard that tracks apache
03:02.32Hellaenergytracks apache?
03:02.35npmi don't want to add and support google ad additions in each webapp i subsume
03:02.43npmthus i want it in apache
03:03.13npmby tracks apache, i just mean if it's part of apache.org it has a better chance of being supported in the future and not suffering bitrot
03:03.48HellaenergyThus your questions should be directed to #apache
03:04.06HellaenergyThose folks are very helpful when it comes to this type of question.
03:04.07npmeither that or i just answered my own question :-)
03:04.17npmbut thanks i will ask over there
03:04.56HellaenergyI think you should go with jasonb's ServletFilter suggestion just because it's cool :)
03:05.18*** join/#tomcat ong_ (n=alex_g@p549FDA73.dip.t-dialin.net)
03:06.05jasonbIt would totally work, and if done well it would add very little overhead.
03:06.19npmthat''s option #2, but it won't do my legacy HTML pages (and i don't want to serve files from java)
03:06.47jasonbAnother option: write a custom Tomcat Valve, doing the same thing as a Servlet Filter would do.
03:07.14jasonbnpm: serving static files from Tomcat stand-alone is faster than serving them from Apache httpd.
03:07.18HellaenergyWhy don't you want Tomcat to serve yoru static content?
03:07.45npmthe thing is, no matter what, apache is going to be fronting it for auth and security and port80 reasons
03:08.30npmbut once it slaps on header/footer, it's no longer serving files, it's garbage collecting
03:08.33jasonbnpm: You can.  It won't be more secure that way though.
03:09.40HellaenergyI second that
03:09.40jasonbnpm: I don't know all the auth stuff you need, so it is possible you need auth implementations that only httpd has.
03:09.47npmmod_ssl
03:09.58Hellaenergyjasonb is there a ldap auth in tomcat?
03:10.05jasonbHellaenergy: Of course.
03:11.03Hellaenergyjasonb, i've never done ldap auth with tomcat
03:11.04*** join/#tomcat acidjnk (n=have@pD950B432.dip0.t-ipconnect.de)
03:11.25Hellaenergynpm tomcat can do ssl transactions
03:11.26npmjason : an implementation similar to the one i did http://ipssources.com for example :-)
03:11.29jasonbHellaenergy: I haven't either.. yet.  But, I hear it works fine.
03:12.30npmexcept I'd use http://myopenid.com instead of verisign's highway robbery
03:12.32jasonbTomcat does HTTPS quite well.  It *may* not be faster than httpd stand-alone, but Tomcat's HTTPS is inded faster than httpd(HTTPS) --> Tomcat and back.
03:13.13jasonbnpm: Ahh, client cert?
03:14.12jasonbnpm: You should read my text in Tomcat: The Definitive Guide, 2nd Edition, Chapter 6 where I give step by step instructions on getting CLIENT-CERT auth set up and running on Tomcat stand-alone.
03:15.06jasonbIt works great.
03:15.24jasonbknew ipssources.com sounded familiar for some reason. :)
03:15.35npmexcellent. thanks for the reference... i'll lookup the examples
03:15.44jasonbnpm: Good work on that, btw.
03:16.12npmthanks
03:53.48Hellaenergyjasonb, do you use awstats at all?
04:15.30jasonbHellaenergy: I think I have downloaded and tinkered with it.. long ago.
04:16.14Hellaenergywhat, if anything, do you use for log file crunching?
04:17.18jasonbWell, my work uses both yahoo and google analytics, plus LeadLander.  But so far we don't process the access logs we save.
04:18.16jasonbApparently, awstats has some vulnerabilities that are commonly scanned for when software scans a web site for vunlerabilities.
04:18.38HellaenergyI wouldn't doubt it
04:18.57jasonb(whether new versions of awstats are still vulnerable in the same way(s), I don't know)
04:19.10Hellaenergyit is a perl cgi
04:19.16jasonbahh.
04:19.41jasonbOne could run it from Tomcat.. but you'd have to be pretty careful with that
04:20.25Hellaenergyhow so?
04:21.24HellaenergyI mean can you explain why
04:24.05jasonbCGI means you're spawning another process and running a program in it.  If you have it set up so it can do that, it's potentially possible for a malicious user to use that in a clever way to gain access to something they're not supposed to have access to.
04:24.39HellaenergyOh yes that is the case with any server though
04:24.55HellaenergyI thought tomcat was particularly sensitive though
04:25.20Hellaenergyor I thought that was what you were getting at
04:25.26jasonbAlso, once you're running code outside the JVM, security rules change a bit, so you'd need to filter all sorts of user input to prevent all sorts of attacks that work in other languages.. like perl for instance.
04:25.51Hellaenergyoh yes good point
04:26.28jasonbI wrote some code to filter all user input in the form of request parameters, but there are other things to filter as well.  And, I've left it up to the administrator to write some regular expressions of their own to filter for what they need to prevent.
04:27.07HellaenergyWhats filtering?
04:27.19jasonbhuh?
04:28.40HellaenergyWhen you say you wrote some code that filters...
04:28.47HellaenergyWhere is this happening?
04:30.07jasonbI wrote a custom Tomcat Valve, and a custom Servlet Filter.  You can use either one.  They perform the same work.
04:30.27jasonbThis is all detailed in Chapter 6.
04:30.43HellaenergyAh I'll check it out
04:30.59HellaenergyI need to verse myself in Valves
04:31.19jasonbIf you don't want to modify anything in webapps you're running, or if you can't, then you can use the Valve.  If you have 1 webapp (or a small number of them), then you can just add a single filter to your web.xml and run the Filter.
04:31.19HellaenergyThere is so much I do not know :)
04:31.44jasonbValves were invented (in Tomcat) before Filters were added to the Servlet API.
04:31.54HellaenergyI have the advantage of starting from the begining here :)
04:31.54jasonbValves were a sort of a prototype for what could be done.
04:32.12jasonbValves are Tomcat specific, and Filters are servlet container implementation independent.
04:32.29HellaenergySo going forward use Filters
04:32.40jasonbFilters can do most things that Valves can do.  But, since Valves are Tomcat specific, Valves can more easily control Tomcat.
04:32.40Hellaenergysince that is portable in theory
04:32.49jasonbFilters are nearly always preferable.
04:32.51Hellaenergyhm...
04:32.59Hellaenergygood to know
04:33.02jasonbBut, the way I look at it, they're both different things, each with their pros and cons.
04:33.15jasonbValves run in Tomcat's own space.  Filters run in the webapp space.
04:33.31Hellaenergykinda like kernel vs user :)
04:33.34jasonbYes.
04:33.44jasonbSo, it really depends on what you need to do.
04:33.45Hellaenergyinteresting
04:33.59jasonbAnd, it also depends on how Tomcat-specific you want the code to be.
04:34.35jasonbMost people choose to write strictly to the Servlet API, but it's almost always the case that they only run their code on one servlet container.. ever.
04:34.57jasonbSo, in that case, it doesn't make much sense to fight adding any dependency on the servlet container you use.
04:35.00HellaenergyIs that based on you personal exp?
04:35.03jasonbYes.
04:35.11jasonbYour mileage may vary.  :)
04:35.16jasonb(but probably won't)
04:35.30HellaenergyWell if you sell a war
04:35.42Hellaenergybuyers always want what you don't provide
04:35.51jasonbIf you sell a webapp, but you never test it on other servlet containers, it may not work there anyway.
04:35.52Hellaenergysuch as WebSphere or WebLogic support
04:36.08Hellaenergyvery ture
04:36.12Hellaenergytrue
04:36.39Hellaenergyl lot of times you need to have specific configuration logic for each
04:36.51jasonbYup.
04:37.02HellaenergyThat is what i've seen in fact
04:37.41jasonbSo, you *can* write or use a Filter that just depends on the Servlet API, but only if you don't need to have more integrated control over the servlet container (web server) than Filters offer you.
04:37.53HellaenergySo coding to the servlet container really isn't out of hand as long as you can manage it.
04:38.01jasonbMost of the time Filters work fine.
04:38.10jasonbYes, right.
04:38.28HellaenergySo are Valves an aspect of Tomcat that are focused on still?
04:38.38jasonbIt's all a program inside the JVM.. as far as the JVM is concerned, there are no real borders to things.. it's one huge program.
04:38.41HellaenergyIs there development being done with them in the base still?
04:39.28HellaenergyOr is it a dieing feature?
04:39.28jasonbSince 5.5 or so, Valves haven't changed.  Some of the Tomcat committers consider Valves deprecated in favor of Filters, but they don't get it.
04:39.45HellaenergyInteresting
04:39.59jasonbOne of them actually began marking Valves as deprecated until I spoke up about it.  :)
04:40.08HellaenergyIf anyone should get it you would think they should
04:40.21jasonbWell, Craig was the one who invented Valves.
04:40.30jasonbIt was a replacement for Tomcat 3 Interceptors.
04:40.44HellaenergyWhy the name Valve
04:40.52jasonbRemy supported it all, but doesn't appear to care about them.
04:41.13jasonbThe name came as a play on the whole request processing "pipeline" idea.
04:41.23jasonbYou have a pipe, and you have valves, that let some through.  :)
04:41.38Hellaenergymakes sense
04:41.47jasonbIt's a short word, and nobody else was using it for anything.
04:42.22jasonbSince Craig left the project, the others have maintained the Valve code, but they don't develop it further.
04:42.39jasonbMainly, it's an interface, and if the interface changes, all implementations break (usually).
04:43.08HellaenergyWhat happend to Craig?
04:43.09jasonbSo, they've left that alone, mostly.  It actually did change from 4.1 to 5.0 (or was it 5.0 to 5.5?) but it wasn't a real bad change.
04:43.24jasonbCraig works for Sun.
04:43.44jasonbSun and the ASF had a falling out around the Tomcat 5.0 timeframe.
04:44.02HellaenergyWhat to make way for Glassfish?
04:44.02jasonbI forget if I told you about that or not.
04:44.11jasonbYes, he's been working on Glassfish.
04:44.13HellaenergyI don't recall
04:44.32jasonbGlassfish's servlet container is a fork of Tomcat's codebase, with heavy modifications.
04:44.36HellaenergySo Tomcat is no longer the implementation server for the specs?
04:44.53jasonbRight.  It's no longer the reference implementation.. as of Tomcat 5.0.
04:45.09jasonbActually, I don't remember if Tomcat 5.0 was or not, but 5.5 was not.
04:45.25HellaenergyThey should probably take that off the front page then :D
04:45.36jasonbYes, they should.  They don't want to.  :)
04:45.38HellaenergyOr alter the wording
04:46.25HellaenergyThat's not even right
04:46.54HellaenergyVery misleading
04:47.09HellaenergySo is Glassfish any good?
04:47.15jasonbThe good news is: Tomcat is still, by far, the most popular servlet container implementation.
04:47.33jasonbGlassfish is probably good.  They do have a good team of developers.
04:47.54jasonbBut, the code is GPL'd.
04:48.06Hellaenergylol
04:48.26Hellaenergyeven sun gpl'd their java code
04:48.27jasonbAlso, there's no Glassfish: The Definitive Guide.
04:48.37jasonbYeah, it's sort of strange.
04:49.11HellaenergyYou should really put up an article on that
04:49.24Hellaenergyand make it public
04:50.18jasonbI wouldn't have much to lose at this point in doing that, but as time goes on, this is sort of all water under the bridge.  :)
04:50.47Hellaenergywater under the bridge? how so?
04:50.48jasonbAlso, lots of what I know is not exactly directly from the source.  :)
04:51.28HellaenergyOr just an article on your take on the GPL and Java
04:51.32jasonbWell, I want Sun and the ASF to get along.  If they didn't before, for certain personal reasons, then that's bad, but after lots of time goes by it matters far less.
04:52.16jasonbI could indeed write an article about GPL and Java.  I've strongly considered doing that on a number of occasions.
04:52.43HellaenergyYou would be doing us all a favor by doing that ;)
04:52.45jasonbOne thing that would come out of that is: I'd have to do a bunch more research and know solid answers about things I've gathered, and that in itself would be enlightening.
04:53.13HellaenergyYou're already quite pasionate about it :)
04:53.51jasonbThe main problem there is: if you're not actively listening in on the fire hose of discussion forums between the ASF and the FSF, you lose track of where everything stands, and then you're out of date and people will say you don't know what you're talking about (and it would be true).
04:55.23jasonbSomething very interesting to note is: after all the discussion around Apache Harmony and the FSF, they have still not been able to "meet in the middle" somehow as they promised to do, and Harmony still includes *none* of the GPL'd Classpath source -- the Harmony guys had to reimplement all their own (duplicate effort, big time!).
04:55.50jasonbSo, even after more than a year of discussion on how they could just do that, nobody really budged.
04:56.02jasonb(and that's exactly my point, that they won't)
04:56.22jasonbEach side has their reasons.
04:56.36jasonbAnd, both side's reasons make sense.
04:56.40HellaenergySo what is the argumenting all about from what you've seen?
04:57.16jasonbIt's multi-pronged.
04:57.20HellaenergyShit I forgot about Harmony
04:57.34jasonbDon't forget Harmony (nor the Alamo).
04:57.45Hellaenergy:D
04:58.28jasonbMainly the disagreement seems to be around the ability to fork and close the source, plus other issues.
04:58.48HellaenergyWhy close the dang source?
04:59.01jasonbCompanies often need to.
04:59.24jasonbAlso, there is huge disagreement on what the viralness of the GPL means, exactly.
04:59.36jasonb(inside the JVM, for example)
05:00.04jasonbAnd, what the definition of "derivative work" should be.. in each language, in each situation.
05:00.20jasonbNone of this stuff is simple.
05:00.37HellaenergyYes I see that
05:00.53jasonbAnd, it comes down to a sort of "party line" set of opinions.
05:01.13jasonbEach individual decides what they think about each issue, and then finds themselves in one camp or the other.
05:01.27HellaenergyYou could easily write a dissertation on this :)
05:01.32jasonbOf course!
05:01.42jasonbIt's fascinating stuff.
05:02.21Hellaenergyit would be good historical documentation
05:02.27Hellaenergyyou could get in the history books ;)
05:02.59jasonbWhoever writes about it, though, gets branded as being in one camp or the other, and then the writing is instantly assumed to be biased (it probably is).
05:03.35jasonbI'm actually not a member of either group, officially, but I do have opinions.  :)
05:03.56Hellaenergythats fair
05:04.19jasonbAnd, there are people who know far, far more about this than I do.
05:04.47jasonbThere are plenty of reasons to feel unqualified to write about it.
05:05.17HellaenergyDoes it have to be a definitive take?
05:05.28jasonbNo.
05:06.01jasonbIt could be a "here's how this looks to me" text.. but that's a lot like a blog, and I don't really blog (even though it sounds fun).
05:06.18HellaenergyI'd read it :)
05:06.31jasonbAs I said, I've been tempted.
05:07.30jasonbFor quite a while there I was hopeful that the FSF and ASF would somehow change their license text in whatever ways that would allow merging of GPL and ASF code, but I'm not hopeful of that anymore.
05:07.40HellaenergyLet it all out jasonb, let it all out.
05:07.45jasonbheh
05:08.31HellaenergyGPL is a buzz word now.
05:09.15Hellaenergy99% of people don't know whats in it.
05:09.31jasonbIt might be 99.99%.
05:09.50jasonbNobody wants to read it.  I'm not sure I blame them.
05:10.07Hellaenergyseriously you got to think about how many people actually have read any portion of it.
05:10.09jasonbIt's like reading all of the fine print when you buy a car or a house.
05:10.28Hellaenergynods
05:11.34Hellaenergyjasonb, do you use JSF?
05:11.51jasonbI don't.  I haven't yet found a good reason to.
05:11.53HellaenergyWhat is your take on that project?
05:12.02Hellaenergyah
05:12.14HellaenergyI just found out they have their own channel :)
05:12.16jasonbI think it's a mostly failed project.  But, not entirely, since it's still being used.. a little.
05:12.38Hellaenergythere's a v 2 being worked on apparently
05:12.52jasonbI'd look at that.
05:13.47Hellaenergyhttp://blogs.sun.com/rlubke/entry/jsf_2_0_new_feature2
05:14.21HellaenergyWhy do you say its a mostly failed proj?
05:14.24jasonbreads
05:14.30HellaenergyNot much buzz?
05:14.36jasonbWell, few webapps use JSF.
05:14.51jasonbYou don't hear much about it.. there aren't many web pages about it.
05:14.58jasonbetc.
05:16.49HellaenergyThe Glassfish folks documented some stuff: https://javaserverfaces.dev.java.net/users.html
05:16.59HellaenergyMajarra is their project
05:17.14HellaenergyBTW the channel is ##jsf
05:19.38jasonbThere's also http://myfaces.apache.org
05:20.19HellaenergyCool I haven't seen that yet
05:20.58HellaenergyThanks
05:23.32jasonbThat's an active project, from what I understand.
05:23.43jasonbor, set of projects.
05:27.49HellaenergyI just asked if JSF is being used by a lot of people and kenpaulsen answered "Yes" :)
05:27.59HellaenergyI think he works for sun
05:28.19Hellaenergy:)
05:29.15jasonbYeah, consider the source.  :)
05:29.31jasonbAlso, "a lot" is a relative term.
05:30.10HellaenergyHe pointed me to the forums
05:30.15Hellaenergyhttp://forum.java.sun.com/forum.jspa?forumID=427&start=0
05:31.05HellaenergyHe also (not suprisingly) said that MyFaces isn't as active as Mojarra
05:34.30Hellaenergydude we got some Tomcat bashers over here :) Jump in :)
05:35.28jasonbNo, that's alright.  Let them bash Tomcat, while wishing theirs was more popular.  :)
05:36.02jasonbWe all owe Sun a lot, IMHO, for giving us Java.  Oh, and for giving us Tomcat!
05:36.28jasonbYou might want to remind them that Tomcat came from Sun.
05:36.28HellaenergyThey aren't really bashing
05:36.35Hellaenergyjust promoting Glassfish
05:36.45Hellaenergysaying it has all the JSF stuff  you need
05:36.51Hellaenergyand Tomcat doen't
05:36.56jasonbTomcat isn't trying to.
05:37.19jasonbJSF is not mandatory, nor necessary.
05:37.42jasonbTomcat is actually on a diet.  :)
05:38.04jasonbIt's trying not to be such a fat cat!
05:40.24HellaenergyThat is good
05:40.31jasonbyup.
05:40.32HellaenergyI was going to ask you about that
05:40.43HellaenergyHow to trim it down to just what you need
05:40.57HellaenergyI'm a  minimalist when it comes to servers :)
05:41.00jasonbWell.. hmm.
05:41.21jasonbWhy would you want to subset Tomcat?
05:41.28jasonbWhat benefit do you think you'd get?
05:41.47jasonb(you probably wouldn't)
05:42.25*** join/#tomcat nmatrix9 (n=nmatrix9@bas17-toronto12-1177894353.dsl.bell.ca)
05:42.34jasonbKeep in mind that hard drive capacities get larger and larger each year, and also that CPUs get faster, and networks get faster.
05:42.46nmatrix9anyone know how to submit code corrections to the tomcat team?
05:42.57nmatrix9they have a example on their online docs which is borked
05:43.14Hellaenergyhttp://tomcat.apache.org/bugreport.html
05:43.17jasonbnmatrix9: Make a unified diff of your change, and submit an ASF Bugzilla bug and attach your diff to the bug.
05:44.01nmatrix9I've heard used dif before but a "unified diff" that Iam not familiar with
05:44.30nmatrix9jasonb, it's actually a example on setting up a datasource for postgresql
05:45.51jasonbnmatrix9: Unified diff can be generated with the command: diff -u  (on Linux)
05:46.10jasonbnmatrix9: Also, svn diff is always a unified diff.
05:47.12jasonblooks for some ice cream
05:47.27HellaenergyThis is interesting: http://www.ilikespam.com/blog/glassfish-vs-tomcat
05:47.50Hellaenergyv
05:50.03jasonbneat!
05:52.31jasonb2) is wrong.  Tomcat has better scalability than Glassfish.. from the benchmarks I've seen.
05:52.58HellaenergyGot those?
05:53.39jasonbNo.. they were mainly tomcat-dev messages that flew by.
05:53.55jasonbIt was APR vs. JIO vs. Grizzly, and Grizzly lost.
05:53.59jasonb(at the time)
05:54.16jasonbThat was a while ago, but the Tomcat connectors are actively developed.. still.. and I'd bet so is Grizzly.
05:54.54jasonbOn the "Deploy same webapp to multiple hosts" comparison, that actually would depend on what one means by that.
05:56.06jasonbHe's just wrong about the "Each host can listen on a different port" comparison.  Tomcat's should be a "Yes".
05:57.24jasonb"Each host can listen on an arbitrary combination of ports" is also wrong about what it says about Tomcat.
05:58.21jasonbHe seems mostly clueful though.
06:01.39HellaenergyGood to know the facts :)
06:05.58jasonbloves comparison matrices.
06:06.25jasonbHellaenergy: Have you seen wikimatrix or cmsmatrix?
06:06.32HellaenergySeems like they are always vague and bias
06:06.44Hellaenergyno
06:06.51Hellaenergychecks em out
06:07.03jasonbWell, but it's also very nice to know who is judging these implementations in what ways.
06:07.17*** join/#tomcat Fivestar (n=Administ@218.91.218.202)
06:07.21Hellaenergygood point
06:07.38Fivestarhello all, i reference the tomcat docs and try it. i mkdir 'docs','src','web' direcotries follow the docs instruction at /docs/appdev/processes.html, then put html,jsp into web and put java servlet source file into src, then i run 'ant', 'ant install', to watch it in browser, everything is ok , then i try to modify some source files under my workspace dir. i run 'ant', but there's nothing happen when i
06:07.38Fivestarrun 'ant reload', the browser shows the file content same as before. how to do ? [tomcat 6.0.x, Ant 1.7.0, my workspace dir: C:\ws\myapp, tomcat dir: C:\Program Files\Apache Software Foundation\Tomcat 6.0]
06:08.45jasonbFivestar: Which web browser?
06:11.19Fivestarany one ?
06:12.38jasonbHellaenergy: http://www.cmsmatrix.org and http://www.wikimatrix.org
06:13.00jasonbHellaenergy: Tell me that those aren't helpful web sites.  :)
06:13.27HellaenergyThat is a powerful idea
06:13.37Hellaenergywish I would have thougt of it first :)
06:16.41jasonbWhen I saw those sites I was glad I *didn't* think of it first since it's already done and usable, and works well (and seems like it must have been a lot of work).
06:18.22*** join/#tomcat rambo123 (n=rocky@59.94.104.176)
06:19.54Fivestarie
06:19.55Fivestarthe file in the webapps/test  is not reload ..not same as ws/myapp
06:20.03*** join/#tomcat antipattern (i=user@c-69-181-54-225.hsd1.ca.comcast.net)
06:20.27*** join/#tomcat antipattern (i=user@c-69-181-54-225.hsd1.ca.comcast.net)
06:21.06*** join/#tomcat antipattern (i=user@c-69-181-54-225.hsd1.ca.comcast.net)
06:23.25Hellaenergyjasonb, You do have a point there :)
06:23.52HellaenergyFivestar, try stopping tomcat, removing the work dir and starting it again.
06:38.35Hellaenergyjasonb, what was that small fast DB you were telling me about a while back? Was it derby or javadb?
06:39.17jasonbHellaenergy: There is hsqldb and now H2.
06:39.46jasonbHellaenergy: H2 is newer, but may not be as mature as hsqldb yet.  But, H2 is catching up.
06:40.25FivestarHellaenergy,  yes , this manner is useful, but is there better one ?
06:40.25Fivestarbut the tomcat docs says: just run 'ant reload' ?
06:48.51Hellaenergyjasonb, any comments on derby?
06:49.43jasonbHellaenergy: It's far slower than hsqldb or H2.  It comes with Java, but it doesn't seem like a better database impl to use.  And, hsqldb has been open source for longer, I believe.
06:49.57jasonbIt seems like there's more support for hsqldb.
06:51.23HellaenergyWhat are the advantage to using hsqldb over mysql?
07:04.20jasonbHellaenergy: hsqldb can significantly outperform mysql because hsqldb is extremely small code, and can run inside the same JVM as the application (no network IO would need to happen to use it, and communication happens all within the JVM in that case).
07:05.16jasonbHellaenergy: hsqldb supports a few different runtime modes, one being "server" like mysql, another being embedded, and the only configuration difference in the application is the JDBC URL.
07:06.20jasonbHellaenergy: See this page: http://h2database.com/html/frame.html?performance.html&main
07:07.05jasonbHellaenergy: Also, hsqldb, H2, and even Derby are pure Java database implementations, and that's good for reasons you already know about.
07:07.45jasonbHellaenergy: Oh, and hsqldb & H2 support a column data type that mysql doesn't support: java.lang.Object.
07:07.46Hellaenergyjasonb, is it useful for data that needs to be kept for long periods of time?
07:08.11jasonbHellaenergy: It writes the data to disk just like any other relational database engine.  So, yes.
07:09.06jasonbAlso, stored procedures in hsqldb and H2 are written in Java.
07:09.28jasonb(in case anyone needs stored procedures)
07:10.13jasonbAlso, H2 supports some "compatibility modes" that try to be compatible with mysql and postgres.
07:10.24jasonbI haven't tried those out yet.
07:11.39HellaenergyWhich one would you suggest using? hsqldb or h2?
07:11.59Hellaenergyhis h2 the successor for hsqldb?
07:13.14jasonbWell, if you need maximum interoperability with lots of tools and frameworks and stuff, I'd say hsqldb.  But, if you're not worried so much about that, then H2 because it's really the next generation and is better in some ways.  Also, Hibernate, JPOX, and OpenJPA apparently support H2 already.
07:13.48HellaenergyHow is security?
07:14.21jasonbSo hsqldb used to be called HypersonicSQL.  The guy who originally wrote it named it that.  Then, he stopped working on it at some point (I was already using it then, in 1999).  But, some other developers kept maintaining it without him.. they renamed it "hsqldb" and put it on sourceforge.
07:14.48jasonbhsqldb continued without the original author.. who went on to work on other pure Java commercial databases, or at least I think that's what I read.
07:15.10jasonbhsqldb matured quite a bit, and lots of other projects supported it.
07:15.25jasonbThen, one day I found the H2 web site.. and it appeared it was pretty new.
07:15.54jasonbThe guy who had initially written Hypersonic started from scratch and wrote a new one, learning from the mistakes he made on the original HypersonicSQL design.
07:16.04jasonbSo now he's playing catch-up with hsqldb.
07:16.12jasonbThey're separate projects.
07:16.36jasonbI don't know if anyone is helping him work on H2, but I think he can make something significant by himself (he did before).
07:17.15jasonbAbout security of these, I don't really know.  It seems good from what I've seen, and I've never heard anyone complain that it is insecure somehow.
07:17.19jasonb(either of them)
07:18.14HellaenergyThanks again for the helpful insight :)
07:18.22HellaenergyI should get to sleep
07:18.26jasonbFor development, either hsqldb or H2 is ideal because they are SO fast and light.
07:18.33jasonbOkay.
07:18.36HellaenergyHave a good night
07:18.36jasonbYou're quite welcome.
07:18.41jasonbHave a nice night.
07:18.41HellaenergyYes seems that way
07:19.06HellaenergyYou too
07:33.54*** join/#tomcat kjkoster5489 (n=kjkoster@kjkoster.org)
07:47.20*** join/#tomcat cofeineSunshine (n=justinas@78-62-4-27.ip.zebra.lt)
08:19.25*** part/#tomcat rambo123 (n=rocky@59.94.104.176)
08:36.09*** join/#tomcat gregor_k (n=a@stgt-4dbdbfe3.pool.einsundeins.de)
10:01.46*** join/#tomcat Hali_303 (n=Hali_303@dsl5400CC3A.pool.t-online.hu)
10:22.50*** join/#tomcat golqm-noob (n=golqm-no@78.90.41.45)
10:35.36*** join/#tomcat Slihp (n=IceChat7@82-36-205-124.cable.ubr02.sutt.blueyonder.co.uk)
10:35.57Slihpheya people
10:36.09Slihpcan anyone offer any help with eclipse and SOAP services?
10:45.55*** join/#tomcat rambo123 (n=rocky@59.96.163.219)
10:55.33Slihpis there anyone out there?
11:02.03juanezdid you actually read the topic?
11:07.54Oliberthey just dribbled into their boots and went away :/
11:13.17*** join/#tomcat MHSL (i=Linux@116.88.6.47)
11:47.11*** join/#tomcat rambo123 (n=rocky@59.94.108.178)
11:52.51*** join/#tomcat ngong (n=ngong@thales.august.de)
12:01.35ngongjust "apt-get install tomcat5.5" - it works, but does anybody know what this MBeanExceptiong thing likes to tell me? http://paste.debian.net/52155
12:04.08*** join/#tomcat cofeineSunshine (n=justinas@78-62-4-27.ip.zebra.lt)
12:04.46kjkoster5489ngong: try using the distribution from Apache.org and not the apt-get package.
12:05.13kjkoster5489re-packaged tomcats tend to be a pain in the butt
12:05.53kjkoster5489In your case, there seems to be a read access problem on /usr/share/tomcat5.5-webapps/jsp-examples/WEB-INF/classes/logging.properties
12:06.12kjkoster5489It's in the stack trace, line 119 or so
12:20.44*** join/#tomcat cofeineSunshine (n=justinas@78-62-4-27.ip.zebra.lt)
12:30.05ngongkjkoster5489: ok, 6.0.16 works out of the box without error. It's a pitty. After a while you get used to aptitude and the like: wait and update for a bunch of packages, in theory consistent and working.
12:30.56kjkoster5489Any reason not to use 6.0?
12:31.28kjkoster5489I agree on the apt-thing. I'm on FreeBSD and I use the ports for everything, except tomcat and jboss and such.
12:31.31ngongkjkoster5489: the only one would be, because it is not Debian packaged
12:32.07kjkoster5489Installation of Tomcat is trivial (unzip a zip and chmod 755 bin/*.sh) and you can move it anywhere.
12:32.46kjkoster5489Plus, the re-packaged tomcats tend to be scattered all over the file system, making it harder for non-your-distro devs to help you root out bugs in the configuration.
12:34.38ngongkjkoster5489: "all over the file system" lacks in documentation, I agree. However, giving it a closer look the places are chosen wisely and it is too much work to do that manually.
12:38.35*** join/#tomcat cofeineSunshine (n=justinas@78-62-4-27.ip.zebra.lt)
12:44.34kjkoster5489ngong: I would just leave Tomcat as it is.
12:44.48kjkoster5489But then, we all have our preferences.
12:55.04ngongkjkoster5489: that's what I'm just doing, too, and as I did for years. Now I found tomcat and jspwiki packaged in Debian. Imagine to setup a new laptop just by hitting "+" beside those two packages. Do not underestimate the difference to downloading 2 tar.gzs and configuring all the locations.
12:57.30kjkoster5489Agreed.
12:57.54kjkoster5489It's different for me, because I am a developer, not a sysadmin.
13:16.09*** join/#tomcat rossand (n=rossand@CPE000c413a19a3-CM00159a025ad4.cpe.net.cable.rogers.com)
13:16.19*** part/#tomcat rossand (n=rossand@CPE000c413a19a3-CM00159a025ad4.cpe.net.cable.rogers.com)
14:56.03jasonbkjkoster5489: You would probably like my Tomcat RPM package then.  It's all in one directory, and it works.  But, it's not for freebsd.
14:56.51kjkoster5489:-)
14:57.03kjkoster5489Glad you preservd the dirs.
14:57.19kjkoster5489How do yoy deal with the 'all config must be in /etc'--crowd?
14:57.25kjkoster5489s/yuy/you/
15:00.53jasonbkjkoster5489: Strategic symlinks. :)
15:01.06*** join/#tomcat a4akba (n=a4akb@62.215.156.215)
15:01.49kjkoster5489So I have to be careful with blanket "no re-packaged tomcat" statements from now on.
15:11.19*** join/#tomcat streetpc (n=adrien@per92-3-82-224-242-177.fbx.proxad.net)
15:11.35streetpchello
15:12.54streetpcas a good & docile tomcat newbie, I want to install 5.5 on a customized RedHat Entreprise 4
15:13.11streetpcjust to launch some WebApps
15:13.51streetpcI think I must get the "deployer" bin release
15:14.15streetpcbut the README page isn't that clear for a non-native English speaker
15:14.43streetpcand I read "all non-embedded users download this" for the "base" version
15:14.54streetpcwhat do they ùean by "embeeded" ?
15:14.58streetpcmean*
15:18.27kjkoster5489You're probably best off using a stock Tomcat from http://tomcat.apache.org/download-55.cgi (download the core .zip)
15:19.45streetpckijkoster5489 > I'm already on this page
15:20.10streetpcI wanted to know the difference between the "deployer" and the "core" releases
15:20.41kjkoster5489Dunno. I only use Core.
15:22.01streetpcand it is enough to get some WebServices running (if possible with auto-deploy) without having to install Apache ?
15:22.51kjkoster5489Yes.
15:23.46streetpccool
15:23.50streetpcthanks :)
15:35.04*** join/#tomcat flashboss (n=lstancap@62.123.101.197)
15:35.12flashbosshi
15:35.16flashbosshow are you?
15:50.53flashbosswho works with jspx?
15:59.09*** join/#tomcat Hali_304 (n=Hali_303@dsl51B6CABF.pool.t-online.hu)
16:29.12*** join/#tomcat prgrmr (n=prgrmr@bzq-79-176-160-107.red.bezeqint.net)
16:32.43*** join/#tomcat Hellaenergy (n=Hellaene@c-75-72-155-242.hsd1.mn.comcast.net)
16:56.07*** join/#tomcat inhiesies (n=inh@207.226.33.20)
16:56.14inhiesiesew!
16:56.17*** part/#tomcat inhiesies (n=inh@207.226.33.20)
17:53.44*** join/#tomcat rambo123 (n=rocky@59.94.96.99)
18:43.38*** join/#tomcat keith80403 (n=keith804@71-208-207-165.hlrn.qwest.net) [NETSPLIT VICTIM]
18:59.22*** join/#tomcat streetpc (n=adrien@per92-3-82-224-242-177.fbx.proxad.net) [NETSPLIT VICTIM]
18:59.22*** join/#tomcat jorgenhg (n=jorgenhg@user.skolelinux.no) [NETSPLIT VICTIM]
18:59.22*** join/#tomcat lyken_ (n=dc@majesticgecko.com) [NETSPLIT VICTIM]
18:59.22*** join/#tomcat teite (i=teite@trough.maeh.org) [NETSPLIT VICTIM]
18:59.38*** join/#tomcat Infinito_ (n=argos@200-140-66-14.gnace701.dsl.brasiltelecom.net.br)
19:25.05*** join/#tomcat kjkoster5489 (n=kjkoster@kjkoster.org)
19:43.13*** join/#tomcat Infinito_ (n=argos@200-140-66-14.gnace701.dsl.brasiltelecom.net.br)
20:15.26*** join/#tomcat que (n=a4akb@62.215.156.215)
20:36.04*** part/#tomcat rambo123 (n=rocky@59.94.96.99)
21:54.54*** part/#tomcat flashboss (n=lstancap@62.123.101.197)
22:16.14*** join/#tomcat nmatrix9_ (n=nmatrix9@bas17-toronto12-1096795916.dsl.bell.ca)
22:39.53*** part/#tomcat Hali_304 (n=Hali_303@dsl51B6CABF.pool.t-online.hu)
22:53.22*** join/#tomcat jasonb (n=jasonb@75.144.23.117)

Generated by irclog2html.pl Modified by Tim Riker to work with infobot.