From rhooker at cisco.com Tue Apr 1 22:17:04 2008 From: rhooker at cisco.com (Raymond Hooker (rhooker)) Date: Tue, 1 Apr 2008 22:17:04 -0400 Subject: [raleigh.rb] WSDL2Ruby.rb In-Reply-To: <47F158EA.6000005@lojic.com> References: <47F15480.5090501@lojic.com> <47F158EA.6000005@lojic.com> Message-ID: <848DB51F3E58E146B08D188F803DCB6F064395CA@xmb-rtp-20c.amer.cisco.com> Has anyone used this tool that comes with SOAP4R? I have tried several WSDL files, all of which work perfectly with the AXIS and AXIS2 generation tools for Java. They bomb with this tool. There is very little documentation, so any pointers would be great. I attached the WSDL file (I have tried others but this is a good example. The errors are listed below: C:\Project\CDWS\PremierMidOfficeWeb\src\ruby>wsdl2ruby.rb --wsdl RiskReportPort ervice.wsdk --type client --force F, [2008-04-01T18:41:26.937000 #8568] FATAL -- app: Detected an exception. Stop ing ... Bad file descriptor - connect(2) (://:0) (Errno::EBADF) c:/ruby/lib/ruby/gems/1.8/gems/httpclient-2.1.2/lib/httpclient.rb:1497:i n `init alize' c:/ruby/lib/ruby/gems/1.8/gems/httpclient-2.1.2/lib/httpclient.rb:1497:i n `new' c:/ruby/lib/ruby/gems/1.8/gems/httpclient-2.1.2/lib/httpclient.rb:1497:i n `crea e_socket' c:/ruby/lib/ruby/gems/1.8/gems/httpclient-2.1.2/lib/httpclient.rb:1455:i n `conn ct' c:/ruby/lib/ruby/1.8/timeout.rb:56:in `timeout' c:/ruby/lib/ruby/1.8/timeout.rb:76:in `timeout' c:/ruby/lib/ruby/gems/1.8/gems/httpclient-2.1.2/lib/httpclient.rb:1454:i n `conn ct' c:/ruby/lib/ruby/gems/1.8/gems/httpclient-2.1.2/lib/httpclient.rb:1311:i n `quer ' c:/ruby/lib/ruby/gems/1.8/gems/httpclient-2.1.2/lib/httpclient.rb:932:in `query c:/ruby/lib/ruby/gems/1.8/gems/httpclient-2.1.2/lib/httpclient.rb:2131:i n `do_g t_block' c:/ruby/lib/ruby/gems/1.8/gems/httpclient-2.1.2/lib/httpclient.rb:1956:i n `requ st' c:/ruby/lib/ruby/gems/1.8/gems/httpclient-2.1.2/lib/httpclient.rb:2068:i n `prep re_request' c:/ruby/lib/ruby/gems/1.8/gems/httpclient-2.1.2/lib/httpclient.rb:1955:i n `requ st' c:/ruby/lib/ruby/gems/1.8/gems/httpclient-2.1.2/lib/httpclient.rb:1925:i n `get' c:/ruby/lib/ruby/gems/1.8/gems/httpclient-2.1.2/lib/httpclient.rb:1891:i n `get_ ontent' c:/ruby/lib/ruby/gems/1.8/gems/httpclient-2.1.2/lib/httpclient.rb:2050:i n `foll w_redirect' c:/ruby/lib/ruby/gems/1.8/gems/httpclient-2.1.2/lib/httpclient.rb:1890:i n `get_ ontent' c:/ruby/lib/ruby/gems/1.8/gems/soap4r-1.5.8/lib/wsdl/xmlSchema/importer. rb:73:i `fetch' c:/ruby/lib/ruby/gems/1.8/gems/soap4r-1.5.8/lib/wsdl/xmlSchema/importer. rb:36:i `import' c:/ruby/lib/ruby/gems/1.8/gems/soap4r-1.5.8/lib/wsdl/importer.rb:18:in `import' c:/ruby/lib/ruby/gems/1.8/gems/soap4r-1.5.8/lib/wsdl/soap/wsdl2ruby.rb:2 06:in ` mport' c:/ruby/lib/ruby/gems/1.8/gems/soap4r-1.5.8/lib/wsdl/soap/wsdl2ruby.rb:3 6:in `r n' c:/ruby/lib/ruby/gems/1.8/gems/soap4r-1.5.8/bin/wsdl2ruby.rb:46:in `run' c:/ruby/lib/ruby/1.8/logger.rb:659:in `start' c:/ruby/lib/ruby/gems/1.8/gems/soap4r-1.5.8/bin/wsdl2ruby.rb:137 c:/ruby/bin/wsdl2ruby.rb:19:in `load' c:/ruby/bin/wsdl2ruby.rb:19 I, [2008-04-01T18:41:27.000000 #8568] INFO -- app: End of app. (status: -1) C:\Project\CDWS\PremierMidOfficeWeb\src\ruby> -------------- next part -------------- A non-text attachment was scrubbed... Name: RiskReportPortService.wsdl Type: application/octet-stream Size: 4182 bytes Desc: RiskReportPortService.wsdl Url : http://rubyforge.org/pipermail/raleigh-rb-members/attachments/20080401/6ac7cd51/attachment.obj From rhooker at cisco.com Tue Apr 1 23:02:28 2008 From: rhooker at cisco.com (Raymond Hooker (rhooker)) Date: Tue, 1 Apr 2008 23:02:28 -0400 Subject: [raleigh.rb] WSDL2Ruby.rb In-Reply-To: <848DB51F3E58E146B08D188F803DCB6F064395CA@xmb-rtp-20c.amer.cisco.com> References: <47F15480.5090501@lojic.com><47F158EA.6000005@lojic.com> <848DB51F3E58E146B08D188F803DCB6F064395CA@xmb-rtp-20c.amer.cisco.com> Message-ID: <848DB51F3E58E146B08D188F803DCB6F064395DE@xmb-rtp-20c.amer.cisco.com> Hmm.. It looks like there was a spelling error ("...wskd").. Amazing how you can miss something like that until you pop it in a note or show it to someone.. Ray -----Original Message----- From: raleigh-rb-members-bounces at rubyforge.org [mailto:raleigh-rb-members-bounces at rubyforge.org] On Behalf Of Raymond Hooker (rhooker) Sent: Tuesday, April 01, 2008 10:17 PM To: The mailing list of raleigh.rb Subject: [raleigh.rb] WSDL2Ruby.rb Has anyone used this tool that comes with SOAP4R? I have tried several WSDL files, all of which work perfectly with the AXIS and AXIS2 generation tools for Java. They bomb with this tool. There is very little documentation, so any pointers would be great. I attached the WSDL file (I have tried others but this is a good example. The errors are listed below: C:\Project\CDWS\PremierMidOfficeWeb\src\ruby>wsdl2ruby.rb --wsdl RiskReportPort ervice.wsdk --type client --force F, [2008-04-01T18:41:26.937000 #8568] FATAL -- app: Detected an exception. Stop ing ... Bad file descriptor - connect(2) (://:0) (Errno::EBADF) c:/ruby/lib/ruby/gems/1.8/gems/httpclient-2.1.2/lib/httpclient.rb:1497:i n `init alize' c:/ruby/lib/ruby/gems/1.8/gems/httpclient-2.1.2/lib/httpclient.rb:1497:i n `new' c:/ruby/lib/ruby/gems/1.8/gems/httpclient-2.1.2/lib/httpclient.rb:1497:i n `crea e_socket' c:/ruby/lib/ruby/gems/1.8/gems/httpclient-2.1.2/lib/httpclient.rb:1455:i n `conn ct' c:/ruby/lib/ruby/1.8/timeout.rb:56:in `timeout' c:/ruby/lib/ruby/1.8/timeout.rb:76:in `timeout' c:/ruby/lib/ruby/gems/1.8/gems/httpclient-2.1.2/lib/httpclient.rb:1454:i n `conn ct' c:/ruby/lib/ruby/gems/1.8/gems/httpclient-2.1.2/lib/httpclient.rb:1311:i n `quer ' c:/ruby/lib/ruby/gems/1.8/gems/httpclient-2.1.2/lib/httpclient.rb:932:in `query c:/ruby/lib/ruby/gems/1.8/gems/httpclient-2.1.2/lib/httpclient.rb:2131:i n `do_g t_block' c:/ruby/lib/ruby/gems/1.8/gems/httpclient-2.1.2/lib/httpclient.rb:1956:i n `requ st' c:/ruby/lib/ruby/gems/1.8/gems/httpclient-2.1.2/lib/httpclient.rb:2068:i n `prep re_request' c:/ruby/lib/ruby/gems/1.8/gems/httpclient-2.1.2/lib/httpclient.rb:1955:i n `requ st' c:/ruby/lib/ruby/gems/1.8/gems/httpclient-2.1.2/lib/httpclient.rb:1925:i n `get' c:/ruby/lib/ruby/gems/1.8/gems/httpclient-2.1.2/lib/httpclient.rb:1891:i n `get_ ontent' c:/ruby/lib/ruby/gems/1.8/gems/httpclient-2.1.2/lib/httpclient.rb:2050:i n `foll w_redirect' c:/ruby/lib/ruby/gems/1.8/gems/httpclient-2.1.2/lib/httpclient.rb:1890:i n `get_ ontent' c:/ruby/lib/ruby/gems/1.8/gems/soap4r-1.5.8/lib/wsdl/xmlSchema/importer. rb:73:i `fetch' c:/ruby/lib/ruby/gems/1.8/gems/soap4r-1.5.8/lib/wsdl/xmlSchema/importer. rb:36:i `import' c:/ruby/lib/ruby/gems/1.8/gems/soap4r-1.5.8/lib/wsdl/importer.rb:18:in `import' c:/ruby/lib/ruby/gems/1.8/gems/soap4r-1.5.8/lib/wsdl/soap/wsdl2ruby.rb:2 06:in ` mport' c:/ruby/lib/ruby/gems/1.8/gems/soap4r-1.5.8/lib/wsdl/soap/wsdl2ruby.rb:3 6:in `r n' c:/ruby/lib/ruby/gems/1.8/gems/soap4r-1.5.8/bin/wsdl2ruby.rb:46:in `run' c:/ruby/lib/ruby/1.8/logger.rb:659:in `start' c:/ruby/lib/ruby/gems/1.8/gems/soap4r-1.5.8/bin/wsdl2ruby.rb:137 c:/ruby/bin/wsdl2ruby.rb:19:in `load' c:/ruby/bin/wsdl2ruby.rb:19 I, [2008-04-01T18:41:27.000000 #8568] INFO -- app: End of app. (status: -1) C:\Project\CDWS\PremierMidOfficeWeb\src\ruby> From nathaniel at talbott.ws Wed Apr 2 11:24:52 2008 From: nathaniel at talbott.ws (Nathaniel Talbott) Date: Wed, 2 Apr 2008 11:24:52 -0400 Subject: [raleigh.rb] Advanced Ruby Studio - Reston, VA Message-ID: <4ce336a20804020824i31ed69f4m9642c58f0a73b62d@mail.gmail.com> >From Mike Clark: Don't miss an opportunity to learn advanced Ruby techniques from Dave Thomas and Chad Fowler. The Pragmatic Studio is offering an Advanced Ruby Studio on May 8-10 in Reston, VA. Register by April 11th (next Friday) and save $300! Details and registration info are at http://pragmaticstudio.com/ruby/ -- Nathaniel Talbott <:((>< From lindsay at TSMworks.com Wed Apr 2 15:40:58 2008 From: lindsay at TSMworks.com (Lindsay Morris) Date: Wed, 2 Apr 2008 15:40:58 -0400 Subject: [raleigh.rb] Looking for help at TSMworks In-Reply-To: <7ce1d32f0803300240ye627e6aja743c1e3946687cd@mail.gmail.com> References: <7ce1d32f0803300240ye627e6aja743c1e3946687cd@mail.gmail.com> Message-ID: Hello, all --- We need to put the finishing touches on our prototype product. So we need some CSS work, some web design, some Rails work. With great respect to all the wizards out there, we're not looking for rock stars right now. We're looking for people who understand the Rails conventions, have some Rails development experience, and have some other web development experience. More info at beta.tsmworks.com/jobs. Please email - we're pushing too hard to spend a lot of time on the phone. Thanks! -- Mr. Lindsay Morris Principal TSMworks, Inc. From pelargir at gmail.com Wed Apr 2 19:15:12 2008 From: pelargir at gmail.com (Matthew Bass) Date: Wed, 2 Apr 2008 19:15:12 -0400 Subject: [raleigh.rb] Seed 3 conference Message-ID: Has anyone ever been to 37 signals' Seed 3 conference before? If so, what did you think of it? Matthew -------------- next part -------------- An HTML attachment was scrubbed... URL: http://rubyforge.org/pipermail/raleigh-rb-members/attachments/20080402/7612aa59/attachment.html From curtis.duhn at gmail.com Thu Apr 3 01:27:31 2008 From: curtis.duhn at gmail.com (Curtis Duhn) Date: Thu, 3 Apr 2008 01:27:31 -0400 Subject: [raleigh.rb] Group insurance options In-Reply-To: References: <3e320180803240933i40b8f637y44b824c6e38f4a7c@mail.gmail.com> <47E7E9AC.3070804@lojic.com> <47E8271D.1000602@lojic.com> Message-ID: <3e320180804022227g6a95f2e9ia2891c08f0b26b93@mail.gmail.com> This is the first chance I've had to thank everyone for all the advice regarding group medical insurance for independent consultants. I did a ton of research into various options, so I thought I'd take a moment to give back some of what I've learned for anyone else who finds themselves in a similar situation. It was James who suggested that I might want to look into a company like iProfessional. A friend of mine had a good experience consulting through a similar company, so I ended up pursuing this solution. These companies were tough to find because there isn't one standard name for this type of company. Some call them Portable Employers of Record (PER). Others call them Passthrough Agencies. They're sometimes called Umbrella Companies, especially in the UK. I'll call them PERs here. A PER is a company that you can sign on with as an independent consultant. Legally you're their employee, but they expect you to find your own clients and negotiate your own contracts. They'll review your contracts and if everything looks ok they'll handle billing and collections from your clients. You then get a paycheck every week or two as a W-2 employee with taxes withheld, etc. Most have group medical, dental, and vision coverage, life insurance, 401K plans, etc. PERs are similar to Professional Employment Organizations (PEOs), which also hire employees on behalf of other companies, providing payroll services, group benefits, etc. The key difference is that PEOs are focused on selling their services to companies that want to have some or all of their employees work through the PEO to avoid all the administrative costs associated with having actual employees. While the services are similar, I found few PEOs that seemed interested in working with independent contractors. PERs are also similar to staffing agencies, many of which employ the talent they recruit, offering a regular salary and group benefits. They'll then lease these people out to clients they acquire. The obvious difference here is that unlike PERs, staffing agencies take an active lead in acquiring clients. These agencies are often very secretive about how much they mark up the services of their employees, whereas most PERs are very transparent about their fees. Here are some of the PERs I looked at: - *MBO Partners / MyBizOffice* (www.mybizoffice.com) - Probably the biggest player in the space. Seems to get credit (not sure if it's entirely deserved) for having invented the model. I didn't go with them because their medical insurance network didn't suit my needs. Medical: Aetna; Fee: 5% for first $125K billed, 1.5% after that. - *PACE* (www.pacepros.com) - Recognized as having adopted MBO's model and executed it well. My friend used this agency, but he didn't need their medical coverage, which is unfortuantely a regional plan. Medical: Aetna of California; Fee: 6% - *HelpMeWork* (www.helpmework.com) - Both on the web and on the phone, this company was pretty opaque about their policies, which made me nervous, so I didn't bother digging too deep. Medical: BCBS of CA; Fee: 7%-11% - *Independent Professional Services* (www.iprofessional.com) - Site's opaque. Ran out of time and didn't learn much about them. Medical: United Healthcare; Fee: ??? - *Solution Specialists* (www.solutionspecialists.net) - Tiny mom & pop PER run since the mid-nineties by a friendly guy named Steve and his wife up in Pennsylvania. I found them through a comment made by one of their consultants on a blog post. Be warned that the web site hasn't been maintained, so the insurance rates quoted on the site aren't accurate. The rate has gone up quite a bit. They have the lowest adminsitrative fee I've seen, and they seem like nice people, but unfortunately their medical plan didn't suit my needs. Medical: United Healthcare; Fee: 3% - *Yurcor* (www.yurcor.com) - This is the company I ended up going with, primarily because their medical plan worked best for my particular situation. They weren't quite as transparent as some of their competitors, which made me nervous, but so far things seem legit. Medical: Cigna; Fee: 4% (or 18% -- see below) While these companies provide a legitimate service and are a good solution for someone in my situation, it appears that these companies may unintentionally enable some more questionable employment hacks. For example, you can apparently set up a consulting contract through one of these agencies, then end the contract and go apply for unemployment. This also qualifies you for 18 months of group insurance coverage under COBRA. These abusive practices unfortunately translate into hidden costs that are born by those of us who would use these services legitimately. As a W-2 employer, the PER must pay for federal and state unemployment insurance in addition to workers compensation, medicare, and social security taxes. These employer costs are taken out of your consulting revenue, along with the company's administrative fees and your personal tax withholdings before they cut your paycheck. It is in the unemployment insurance component that the most insidious costs are hidden. Employers start with a fairly reasonable unemployment insurance rate (1.2% in NC), but after two years it gets adjusted based on the number of former employees that have applied for unemployment compensation. It seems that these PERs, by their nature, end up with much higher unemployment insurance rates than a standard employer would. The rate for Yurcor, for example, is 4.5%. Yurcor offers an optional flat fee of 18% that covers all of these employer costs as an alternative to their normal 4% administrative fee. I went with the 18% option, which I think saves me something like 0.35% versus the 4% plan, but this is clearly going to take a big bite out of my take home pay. Once I get some individual insurance quotes, I plan to re-run the numbers and decide whether or not this PER thing makes sense as a long term solution. There has been very little written about this little sub-industry, so I had to piece this information together through a couple dozen hours of research. If any of you are considering this option, I hope I've saved you some time and/or money. Curtis On undefined, Jeff Keating wrote: > Curtis, if your son's condition is enough to be "pre-existing" but isn't > likely to be expensive every year, you could pursue a high-deductible health > plan with HSA, and carry the risk yourself while saving in a tax-exempt > fashion to cover possible expenses down the road. On the surface, it might > seem sub-optimal compared to being employed and able to get group rates, but > some might argue that it's actually better for you and your family in the > long run... > > I'm sorry to hear of the difficulties. I spent a fair amount of time > researching this very issue this year and found no viable alternative to > paying full-price, non-group... if anybody has example or advice I would > also welcome hearing it. > > - jeff > > > On Mon, Mar 24, 2008 at 6:11 PM, Brian Adkins wrote: > > > Brian Adkins wrote, On 03/24/2008 01:49 PM: > > > I'm pretty sure that a small group health insurance plan can't be more > > > than 66% higher than a normal/average plan, so I expect a small group > > > plan will be *much* cheaper than an individual plan. You may need to > > > prove to the insurance company that you have a 'viable' business. > > Having > > > an LLC or other form of corporation would be quite helpful. > > > > I just talked to someone from Cigna, and she mentioned that for a single > > employee group policy in NC, they can only charge 25% more than the > > preliminary rate; whereas, with a 2-24 employee group policy, they can > > charge up to 67% more. Either way, it's much better than individual > > plans which can charge up to 7 times the preliminary rate! > > > > > > _______________________________________________ > > raleigh-rb-members mailing list > > raleigh-rb-members at rubyforge.org > > http://rubyforge.org/mailman/listinfo/raleigh-rb-members > > > > > _______________________________________________ > raleigh-rb-members mailing list > raleigh-rb-members at rubyforge.org > http://rubyforge.org/mailman/listinfo/raleigh-rb-members > -- Curtis Duhn -------------- next part -------------- An HTML attachment was scrubbed... URL: http://rubyforge.org/pipermail/raleigh-rb-members/attachments/20080403/f0969275/attachment.html From otto.hammersmith at gmail.com Thu Apr 3 11:17:56 2008 From: otto.hammersmith at gmail.com (Otto Hammersmith) Date: Thu, 3 Apr 2008 11:17:56 -0400 Subject: [raleigh.rb] Mongrel unsable on RHEL 5? Message-ID: Hey all, I'm having this issue where I need to bounce Mongrel after about 12 hours of in activity. This is only on RHEL5. I have the same code running on Ubuntu and it's perfectly stable for weeks at a time. I doubt anyone will have the magic bullet to solve my problem, so I'm looking for ways to track down the problem. Any suggestions? Thanks. -------------- next part -------------- An HTML attachment was scrubbed... URL: http://rubyforge.org/pipermail/raleigh-rb-members/attachments/20080403/4bc9d5b6/attachment-0001.html From aaron at aaronbedra.com Thu Apr 3 11:57:41 2008 From: aaron at aaronbedra.com (Aaron Bedra) Date: Thu, 3 Apr 2008 11:57:41 -0400 Subject: [raleigh.rb] Mongrel unsable on RHEL 5? In-Reply-To: References: Message-ID: Are you using the up2date package of ruby or your own compiled version? Aaron On Thu, Apr 3, 2008 at 11:17 AM, Otto Hammersmith < otto.hammersmith at gmail.com> wrote: > Hey all, > > I'm having this issue where I need to bounce Mongrel after about 12 hours > of in activity. This is only on RHEL5. I have the same code running on > Ubuntu and it's perfectly stable for weeks at a time. > > I doubt anyone will have the magic bullet to solve my problem, so I'm > looking for ways to track down the problem. > > Any suggestions? > > Thanks. > > > _______________________________________________ > raleigh-rb-members mailing list > raleigh-rb-members at rubyforge.org > http://rubyforge.org/mailman/listinfo/raleigh-rb-members > -------------- next part -------------- An HTML attachment was scrubbed... URL: http://rubyforge.org/pipermail/raleigh-rb-members/attachments/20080403/dc6d5686/attachment.html From otto.hammersmith at gmail.com Thu Apr 3 12:14:48 2008 From: otto.hammersmith at gmail.com (Otto Hammersmith) Date: Thu, 3 Apr 2008 12:14:48 -0400 Subject: [raleigh.rb] Mongrel unsable on RHEL 5? In-Reply-To: References: Message-ID: I knew I left out a *minor* detail... :) The up2date version of ruby was 1.8.1 so a bit too out of date. I thought that was our problem in the first place because a) mongrel would complain about an out-of-date ruby and use some patch b) I'd see a mongrel thread dump as the last thing in the log... something about thread not restarting. I wish I had saved that stack trace because I haven't seen it since I upgraded ruby. So the ruby I upgraded to was the ruby from RubyWorks. It's 1.8.6... I removed all the packages that RHEL had related to ruby and installed these. I'm still using a hand-installed gem, since I couldn't get the RubyWorks gem package to play nice with any of the gems not installed via RPM. On Thu, Apr 3, 2008 at 11:57 AM, Aaron Bedra wrote: > Are you using the up2date package of ruby or your own compiled version? > > Aaron > > On Thu, Apr 3, 2008 at 11:17 AM, Otto Hammersmith < > otto.hammersmith at gmail.com> wrote: > > > Hey all, > > > > I'm having this issue where I need to bounce Mongrel after about 12 > > hours of in activity. This is only on RHEL5. I have the same code running > > on Ubuntu and it's perfectly stable for weeks at a time. > > > > I doubt anyone will have the magic bullet to solve my problem, so I'm > > looking for ways to track down the problem. > > > > Any suggestions? > > > > Thanks. > > > > > > _______________________________________________ > > raleigh-rb-members mailing list > > raleigh-rb-members at rubyforge.org > > http://rubyforge.org/mailman/listinfo/raleigh-rb-members > > > > > _______________________________________________ > raleigh-rb-members mailing list > raleigh-rb-members at rubyforge.org > http://rubyforge.org/mailman/listinfo/raleigh-rb-members > -------------- next part -------------- An HTML attachment was scrubbed... URL: http://rubyforge.org/pipermail/raleigh-rb-members/attachments/20080403/5ea2382d/attachment.html From brenton.leanhardt at gmail.com Thu Apr 3 12:49:46 2008 From: brenton.leanhardt at gmail.com (brenton leanhardt) Date: Thu, 3 Apr 2008 12:49:46 -0400 Subject: [raleigh.rb] Mongrel unsable on RHEL 5? In-Reply-To: References: Message-ID: <7c0835400804030949s7db30579k368b5f18776f74a3@mail.gmail.com> On Thu, Apr 3, 2008 at 12:14 PM, Otto Hammersmith wrote: > I knew I left out a minor detail... :) > > The up2date version of ruby was 1.8.1 so a bit too out of date. I thought > that was our problem in the first place because a) mongrel would complain > about an out-of-date ruby and use some patch b) I'd see a mongrel thread > dump as the last thing in the log... something about thread not restarting. > I wish I had saved that stack trace because I haven't seen it since I > upgraded ruby. Are you certain you are on RHEL5? The last version of RHEL that used up2date was 4 (it now uses yum). I just checked one of my systems and RHEL5 ships with Ruby 1.8.5. --Brenton > > So the ruby I upgraded to was the ruby from RubyWorks. It's 1.8.6... I > removed all the packages that RHEL had related to ruby and installed these. > > I'm still using a hand-installed gem, since I couldn't get the RubyWorks gem > package to play nice with any of the gems not installed via RPM. > > > > On Thu, Apr 3, 2008 at 11:57 AM, Aaron Bedra wrote: > > Are you using the up2date package of ruby or your own compiled version? > > > > Aaron > > > > > > > > > > > > On Thu, Apr 3, 2008 at 11:17 AM, Otto Hammersmith > wrote: > > > > > > > > > > > > > > Hey all, > > > > > > I'm having this issue where I need to bounce Mongrel after about 12 > hours of in activity. This is only on RHEL5. I have the same code running > on Ubuntu and it's perfectly stable for weeks at a time. > > > > > > I doubt anyone will have the magic bullet to solve my problem, so I'm > looking for ways to track down the problem. > > > > > > Any suggestions? > > > > > > Thanks. > > > > > > > > > _______________________________________________ > > > raleigh-rb-members mailing list > > > raleigh-rb-members at rubyforge.org > > > http://rubyforge.org/mailman/listinfo/raleigh-rb-members > > > > > > > > > _______________________________________________ > > raleigh-rb-members mailing list > > raleigh-rb-members at rubyforge.org > > http://rubyforge.org/mailman/listinfo/raleigh-rb-members > > > > > _______________________________________________ > raleigh-rb-members mailing list > raleigh-rb-members at rubyforge.org > http://rubyforge.org/mailman/listinfo/raleigh-rb-members > From otto.hammersmith at gmail.com Thu Apr 3 14:33:44 2008 From: otto.hammersmith at gmail.com (Otto Hammersmith) Date: Thu, 3 Apr 2008 14:33:44 -0400 Subject: [raleigh.rb] Mongrel unsable on RHEL 5? In-Reply-To: <7c0835400804030949s7db30579k368b5f18776f74a3@mail.gmail.com> References: <7c0835400804030949s7db30579k368b5f18776f74a3@mail.gmail.com> Message-ID: Yes I'm sure I'm using RHEL 5. I probably mis-stated which version of ruby was there before... I do know that it was older than the one on my ubuntu system and got the mogrel warnings, so I was hoping that was the problem. I installed ruby with yum... I thought it was set up to out-of-the-box coexist with up2date & yum... I'd only used RHEL4's configured that way. # cat /etc/redhat-release Red Hat Enterprise Linux Server release 5.1 (Tikanga) Just to sanity check myself, I don't want to ignore any possibility... I have a feeling it's something simple I'm missing. :) Thanks. On Thu, Apr 3, 2008 at 12:49 PM, brenton leanhardt < brenton.leanhardt at gmail.com> wrote: > On Thu, Apr 3, 2008 at 12:14 PM, Otto Hammersmith > wrote: > > I knew I left out a minor detail... :) > > > > The up2date version of ruby was 1.8.1 so a bit too out of date. I > thought > > that was our problem in the first place because a) mongrel would > complain > > about an out-of-date ruby and use some patch b) I'd see a mongrel thread > > dump as the last thing in the log... something about thread not > restarting. > > I wish I had saved that stack trace because I haven't seen it since I > > upgraded ruby. > > Are you certain you are on RHEL5? The last version of RHEL that used > up2date was 4 (it now uses yum). I just checked one of my systems and > RHEL5 ships with Ruby 1.8.5. > > --Brenton > > > > So the ruby I upgraded to was the ruby from RubyWorks. It's 1.8.6... I > > removed all the packages that RHEL had related to ruby and installed > these. > > > > I'm still using a hand-installed gem, since I couldn't get the RubyWorks > gem > > package to play nice with any of the gems not installed via RPM. > > > > > > > > On Thu, Apr 3, 2008 at 11:57 AM, Aaron Bedra > wrote: > > > Are you using the up2date package of ruby or your own compiled > version? > > > > > > Aaron > > > > > > > > > > > > > > > > > > On Thu, Apr 3, 2008 at 11:17 AM, Otto Hammersmith > > wrote: > > > > > > > > > > > > > > > > > > > Hey all, > > > > > > > > I'm having this issue where I need to bounce Mongrel after about 12 > > hours of in activity. This is only on RHEL5. I have the same code > running > > on Ubuntu and it's perfectly stable for weeks at a time. > > > > > > > > I doubt anyone will have the magic bullet to solve my problem, so > I'm > > looking for ways to track down the problem. > > > > > > > > Any suggestions? > > > > > > > > Thanks. > > > > > > > > > > > > _______________________________________________ > > > > raleigh-rb-members mailing list > > > > raleigh-rb-members at rubyforge.org > > > > http://rubyforge.org/mailman/listinfo/raleigh-rb-members > > > > > > > > > > > > > _______________________________________________ > > > raleigh-rb-members mailing list > > > raleigh-rb-members at rubyforge.org > > > http://rubyforge.org/mailman/listinfo/raleigh-rb-members > > > > > > > > > _______________________________________________ > > raleigh-rb-members mailing list > > raleigh-rb-members at rubyforge.org > > http://rubyforge.org/mailman/listinfo/raleigh-rb-members > > > _______________________________________________ > raleigh-rb-members mailing list > raleigh-rb-members at rubyforge.org > http://rubyforge.org/mailman/listinfo/raleigh-rb-members > -------------- next part -------------- An HTML attachment was scrubbed... URL: http://rubyforge.org/pipermail/raleigh-rb-members/attachments/20080403/522e24ae/attachment.html From aaron at aaronbedra.com Thu Apr 3 15:46:18 2008 From: aaron at aaronbedra.com (Aaron Bedra) Date: Thu, 3 Apr 2008 15:46:18 -0400 Subject: [raleigh.rb] Mongrel unsable on RHEL 5? In-Reply-To: References: <7c0835400804030949s7db30579k368b5f18776f74a3@mail.gmail.com> Message-ID: I can assure you that the ruby that ships with RHEL is well broken. If you are going to use ruby with RHEL you need to either compile it from source or rpm package it up yourself with the proper items. The yum version doesn't contain the proper readline, zlib, or openssl bits. As far as crashing is concerned I am not sure but you should just start there. If you are compiling it from source I suggest you do so in an unobtrusive way by using --prefix=/opt/local or something of that nature so that you can quickly remove it if you get a solid package. Aaron On Thu, Apr 3, 2008 at 2:33 PM, Otto Hammersmith wrote: > Yes I'm sure I'm using RHEL 5. I probably mis-stated which version of ruby > was there before... I do know that it was older than the one on my ubuntu > system and got the mogrel warnings, so I was hoping that was the problem. > > I installed ruby with yum... I thought it was set up to out-of-the-box > coexist with up2date & yum... I'd only used RHEL4's configured that way. > > # cat /etc/redhat-release > Red Hat Enterprise Linux Server release 5.1 (Tikanga) > > Just to sanity check myself, I don't want to ignore any possibility... I > have a feeling it's something simple I'm missing. :) > > Thanks. > > > On Thu, Apr 3, 2008 at 12:49 PM, brenton leanhardt < > brenton.leanhardt at gmail.com> wrote: > > > On Thu, Apr 3, 2008 at 12:14 PM, Otto Hammersmith > > wrote: > > > I knew I left out a minor detail... :) > > > > > > The up2date version of ruby was 1.8.1 so a bit too out of date. I > > thought > > > that was our problem in the first place because a) mongrel would > > complain > > > about an out-of-date ruby and use some patch b) I'd see a mongrel > > thread > > > dump as the last thing in the log... something about thread not > > restarting. > > > I wish I had saved that stack trace because I haven't seen it since I > > > upgraded ruby. > > > > Are you certain you are on RHEL5? The last version of RHEL that used > > up2date was 4 (it now uses yum). I just checked one of my systems and > > RHEL5 ships with Ruby 1.8.5. > > > > --Brenton > > > > > > So the ruby I upgraded to was the ruby from RubyWorks. It's 1.8.6... > > I > > > removed all the packages that RHEL had related to ruby and installed > > these. > > > > > > I'm still using a hand-installed gem, since I couldn't get the > > RubyWorks gem > > > package to play nice with any of the gems not installed via RPM. > > > > > > > > > > > > On Thu, Apr 3, 2008 at 11:57 AM, Aaron Bedra > > wrote: > > > > Are you using the up2date package of ruby or your own compiled > > version? > > > > > > > > Aaron > > > > > > > > > > > > > > > > > > > > > > > > On Thu, Apr 3, 2008 at 11:17 AM, Otto Hammersmith > > > wrote: > > > > > > > > > > > > > > > > > > > > > > > > Hey all, > > > > > > > > > > I'm having this issue where I need to bounce Mongrel after about > > 12 > > > hours of in activity. This is only on RHEL5. I have the same code > > running > > > on Ubuntu and it's perfectly stable for weeks at a time. > > > > > > > > > > I doubt anyone will have the magic bullet to solve my problem, so > > I'm > > > looking for ways to track down the problem. > > > > > > > > > > Any suggestions? > > > > > > > > > > Thanks. > > > > > > > > > > > > > > > _______________________________________________ > > > > > raleigh-rb-members mailing list > > > > > raleigh-rb-members at rubyforge.org > > > > > http://rubyforge.org/mailman/listinfo/raleigh-rb-members > > > > > > > > > > > > > > > > > _______________________________________________ > > > > raleigh-rb-members mailing list > > > > raleigh-rb-members at rubyforge.org > > > > http://rubyforge.org/mailman/listinfo/raleigh-rb-members > > > > > > > > > > > > > _______________________________________________ > > > raleigh-rb-members mailing list > > > raleigh-rb-members at rubyforge.org > > > http://rubyforge.org/mailman/listinfo/raleigh-rb-members > > > > > _______________________________________________ > > raleigh-rb-members mailing list > > raleigh-rb-members at rubyforge.org > > http://rubyforge.org/mailman/listinfo/raleigh-rb-members > > > > > _______________________________________________ > raleigh-rb-members mailing list > raleigh-rb-members at rubyforge.org > http://rubyforge.org/mailman/listinfo/raleigh-rb-members > -------------- next part -------------- An HTML attachment was scrubbed... URL: http://rubyforge.org/pipermail/raleigh-rb-members/attachments/20080403/54e43e8a/attachment-0001.html From otto.hammersmith at gmail.com Thu Apr 3 15:50:02 2008 From: otto.hammersmith at gmail.com (Otto Hammersmith) Date: Thu, 3 Apr 2008 15:50:02 -0400 Subject: [raleigh.rb] Mongrel unsable on RHEL 5? In-Reply-To: References: <7c0835400804030949s7db30579k368b5f18776f74a3@mail.gmail.com> Message-ID: I'd rather package it up myself... I was going down that route by updating the RHEL src.rpm when I found the RubyWorks package. Maybe that's my next step, build from scratch and install ruby in /usr/local... just to see if it runs stably on that version of ruby. On Thu, Apr 3, 2008 at 3:46 PM, Aaron Bedra wrote: > I can assure you that the ruby that ships with RHEL is well broken. If > you are going to use ruby with RHEL you need to either compile it from > source or rpm package it up yourself with the proper items. The yum version > doesn't contain the proper readline, zlib, or openssl bits. As far as > crashing is concerned I am not sure but you should just start there. If you > are compiling it from source I suggest you do so in an unobtrusive way by > using --prefix=/opt/local or something of that nature so that you can > quickly remove it if you get a solid package. > > Aaron > > > On Thu, Apr 3, 2008 at 2:33 PM, Otto Hammersmith < > otto.hammersmith at gmail.com> wrote: > > > Yes I'm sure I'm using RHEL 5. I probably mis-stated which version of > > ruby was there before... I do know that it was older than the one on my > > ubuntu system and got the mogrel warnings, so I was hoping that was the > > problem. > > > > I installed ruby with yum... I thought it was set up to out-of-the-box > > coexist with up2date & yum... I'd only used RHEL4's configured that way. > > > > # cat /etc/redhat-release > > Red Hat Enterprise Linux Server release 5.1 (Tikanga) > > > > Just to sanity check myself, I don't want to ignore any possibility... I > > have a feeling it's something simple I'm missing. :) > > > > Thanks. > > > > > > On Thu, Apr 3, 2008 at 12:49 PM, brenton leanhardt < > > brenton.leanhardt at gmail.com> wrote: > > > > > On Thu, Apr 3, 2008 at 12:14 PM, Otto Hammersmith > > > wrote: > > > > I knew I left out a minor detail... :) > > > > > > > > The up2date version of ruby was 1.8.1 so a bit too out of date. I > > > thought > > > > that was our problem in the first place because a) mongrel would > > > complain > > > > about an out-of-date ruby and use some patch b) I'd see a mongrel > > > thread > > > > dump as the last thing in the log... something about thread not > > > restarting. > > > > I wish I had saved that stack trace because I haven't seen it since > > > I > > > > upgraded ruby. > > > > > > Are you certain you are on RHEL5? The last version of RHEL that used > > > up2date was 4 (it now uses yum). I just checked one of my systems and > > > RHEL5 ships with Ruby 1.8.5. > > > > > > --Brenton > > > > > > > > So the ruby I upgraded to was the ruby from RubyWorks. It's 1.8.6... > > > I > > > > removed all the packages that RHEL had related to ruby and installed > > > these. > > > > > > > > I'm still using a hand-installed gem, since I couldn't get the > > > RubyWorks gem > > > > package to play nice with any of the gems not installed via RPM. > > > > > > > > > > > > > > > > On Thu, Apr 3, 2008 at 11:57 AM, Aaron Bedra > > > wrote: > > > > > Are you using the up2date package of ruby or your own compiled > > > version? > > > > > > > > > > Aaron > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > On Thu, Apr 3, 2008 at 11:17 AM, Otto Hammersmith > > > > wrote: > > > > > > > > > > > > > > > > > > > > > > > > > > > > > Hey all, > > > > > > > > > > > > I'm having this issue where I need to bounce Mongrel after about > > > 12 > > > > hours of in activity. This is only on RHEL5. I have the same code > > > running > > > > on Ubuntu and it's perfectly stable for weeks at a time. > > > > > > > > > > > > I doubt anyone will have the magic bullet to solve my problem, > > > so I'm > > > > looking for ways to track down the problem. > > > > > > > > > > > > Any suggestions? > > > > > > > > > > > > Thanks. > > > > > > > > > > > > > > > > > > _______________________________________________ > > > > > > raleigh-rb-members mailing list > > > > > > raleigh-rb-members at rubyforge.org > > > > > > http://rubyforge.org/mailman/listinfo/raleigh-rb-members > > > > > > > > > > > > > > > > > > > > > _______________________________________________ > > > > > raleigh-rb-members mailing list > > > > > raleigh-rb-members at rubyforge.org > > > > > http://rubyforge.org/mailman/listinfo/raleigh-rb-members > > > > > > > > > > > > > > > > > _______________________________________________ > > > > raleigh-rb-members mailing list > > > > raleigh-rb-members at rubyforge.org > > > > http://rubyforge.org/mailman/listinfo/raleigh-rb-members > > > > > > > _______________________________________________ > > > raleigh-rb-members mailing list > > > raleigh-rb-members at rubyforge.org > > > http://rubyforge.org/mailman/listinfo/raleigh-rb-members > > > > > > > > > _______________________________________________ > > raleigh-rb-members mailing list > > raleigh-rb-members at rubyforge.org > > http://rubyforge.org/mailman/listinfo/raleigh-rb-members > > > > > _______________________________________________ > raleigh-rb-members mailing list > raleigh-rb-members at rubyforge.org > http://rubyforge.org/mailman/listinfo/raleigh-rb-members > -------------- next part -------------- An HTML attachment was scrubbed... URL: http://rubyforge.org/pipermail/raleigh-rb-members/attachments/20080403/56b5a49f/attachment.html From larry.karnowski at gmail.com Thu Apr 3 19:53:23 2008 From: larry.karnowski at gmail.com (Larry Karnowski) Date: Thu, 3 Apr 2008 19:53:23 -0400 Subject: [raleigh.rb] Mongrel unsable on RHEL 5? In-Reply-To: References: <7c0835400804030949s7db30579k368b5f18776f74a3@mail.gmail.com> Message-ID: <2b27183e0804031653x2ef60ce1ha72b7a9d42564c33@mail.gmail.com> Otto, We ran Ruby 1.8.6 on Centos 4.4 at my last job. We rebuilt the FC7 Ruby source RPMs (ruby-1.8.6.36-3.src.rpm) and had no problems. Stay away from the FC6 RPMs, though. HTH, Larry On Thu, Apr 3, 2008 at 3:50 PM, Otto Hammersmith wrote: > I'd rather package it up myself... I was going down that route by updating > the RHEL src.rpm when I found the RubyWorks package. > > Maybe that's my next step, build from scratch and install ruby in > /usr/local... just to see if it runs stably on that version of ruby. > > > > On Thu, Apr 3, 2008 at 3:46 PM, Aaron Bedra wrote: > > I can assure you that the ruby that ships with RHEL is well broken. If > you are going to use ruby with RHEL you need to either compile it from > source or rpm package it up yourself with the proper items. The yum version > doesn't contain the proper readline, zlib, or openssl bits. As far as > crashing is concerned I am not sure but you should just start there. If you > are compiling it from source I suggest you do so in an unobtrusive way by > using --prefix=/opt/local or something of that nature so that you can > quickly remove it if you get a solid package. > > > > Aaron > > > > > > > > > > > > On Thu, Apr 3, 2008 at 2:33 PM, Otto Hammersmith > wrote: > > > > > Yes I'm sure I'm using RHEL 5. I probably mis-stated which version of > ruby was there before... I do know that it was older than the one on my > ubuntu system and got the mogrel warnings, so I was hoping that was the > problem. > > > > > > I installed ruby with yum... I thought it was set up to out-of-the-box > coexist with up2date & yum... I'd only used RHEL4's configured that way. > > > > > > # cat /etc/redhat-release > > > Red Hat Enterprise Linux Server release 5.1 (Tikanga) > > > > > > Just to sanity check myself, I don't want to ignore any possibility... I > have a feeling it's something simple I'm missing. :) > > > > > > Thanks. > > > > > > > > > > > > > > > > > > On Thu, Apr 3, 2008 at 12:49 PM, brenton leanhardt > wrote: > > > > > > > On Thu, Apr 3, 2008 at 12:14 PM, Otto Hammersmith > > > > > > > > wrote: > > > > > > > > > I knew I left out a minor detail... :) > > > > > > > > > > The up2date version of ruby was 1.8.1 so a bit too out of date. I > thought > > > > > that was our problem in the first place because a) mongrel would > complain > > > > > about an out-of-date ruby and use some patch b) I'd see a mongrel > thread > > > > > dump as the last thing in the log... something about thread not > restarting. > > > > > I wish I had saved that stack trace because I haven't seen it since > I > > > > > upgraded ruby. > > > > > > > > Are you certain you are on RHEL5? The last version of RHEL that used > > > > up2date was 4 (it now uses yum). I just checked one of my systems and > > > > RHEL5 ships with Ruby 1.8.5. > > > > > > > > --Brenton > > > > > > > > > > > > > > > > > > > > > > So the ruby I upgraded to was the ruby from RubyWorks. It's > 1.8.6... I > > > > > removed all the packages that RHEL had related to ruby and installed > these. > > > > > > > > > > I'm still using a hand-installed gem, since I couldn't get the > RubyWorks gem > > > > > package to play nice with any of the gems not installed via RPM. > > > > > > > > > > > > > > > > > > > > On Thu, Apr 3, 2008 at 11:57 AM, Aaron Bedra > wrote: > > > > > > Are you using the up2date package of ruby or your own compiled > version? > > > > > > > > > > > > Aaron > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > On Thu, Apr 3, 2008 at 11:17 AM, Otto Hammersmith > > > > > wrote: > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > Hey all, > > > > > > > > > > > > > > I'm having this issue where I need to bounce Mongrel after about > 12 > > > > > hours of in activity. This is only on RHEL5. I have the same code > running > > > > > on Ubuntu and it's perfectly stable for weeks at a time. > > > > > > > > > > > > > > I doubt anyone will have the magic bullet to solve my problem, > so I'm > > > > > looking for ways to track down the problem. > > > > > > > > > > > > > > Any suggestions? > > > > > > > > > > > > > > Thanks. > > > > > > > > > > > > > > > > > > > > > _______________________________________________ > > > > > > > raleigh-rb-members mailing list > > > > > > > raleigh-rb-members at rubyforge.org > > > > > > > http://rubyforge.org/mailman/listinfo/raleigh-rb-members > > > > > > > > > > > > > > > > > > > > > > > > > _______________________________________________ > > > > > > raleigh-rb-members mailing list > > > > > > raleigh-rb-members at rubyforge.org > > > > > > http://rubyforge.org/mailman/listinfo/raleigh-rb-members > > > > > > > > > > > > > > > > > > > > > _______________________________________________ > > > > > raleigh-rb-members mailing list > > > > > raleigh-rb-members at rubyforge.org > > > > > http://rubyforge.org/mailman/listinfo/raleigh-rb-members > > > > > > > > > _______________________________________________ > > > > raleigh-rb-members mailing list > > > > raleigh-rb-members at rubyforge.org > > > > http://rubyforge.org/mailman/listinfo/raleigh-rb-members > > > > > > > > > > > > > _______________________________________________ > > > raleigh-rb-members mailing list > > > raleigh-rb-members at rubyforge.org > > > http://rubyforge.org/mailman/listinfo/raleigh-rb-members > > > > > > > > > _______________________________________________ > > raleigh-rb-members mailing list > > raleigh-rb-members at rubyforge.org > > http://rubyforge.org/mailman/listinfo/raleigh-rb-members > > > > > _______________________________________________ > raleigh-rb-members mailing list > raleigh-rb-members at rubyforge.org > http://rubyforge.org/mailman/listinfo/raleigh-rb-members > From ryan.daigle at gmail.com Fri Apr 4 07:54:50 2008 From: ryan.daigle at gmail.com (Ryan Daigle) Date: Fri, 4 Apr 2008 07:54:50 -0400 Subject: [raleigh.rb] Full time Rails job Message-ID: <748ff3ff0804040454s5417ae22vff2dd9e02413eeb2@mail.gmail.com> If anybody on the list is interested in breaking free from their full time Java gig and wants to work with Ruby 100% while still pulling down a consistent salary and benefits - let me know. A client I've worked with in the past is looking to add a full time developer to their small team. Pretty sure you could still work from home, set your own hours and all that fun stuff. If you're interested in hearing more, go ahead and send me your resume (off the list, pls) and I'll get you in touch with them. Thanks, guys. -Ryan -------------- next part -------------- An HTML attachment was scrubbed... URL: http://rubyforge.org/pipermail/raleigh-rb-members/attachments/20080404/66141695/attachment.html From ryan.daigle at gmail.com Fri Apr 4 08:25:38 2008 From: ryan.daigle at gmail.com (Ryan Daigle) Date: Fri, 4 Apr 2008 08:25:38 -0400 Subject: [raleigh.rb] Full time Rails job In-Reply-To: <748ff3ff0804040454s5417ae22vff2dd9e02413eeb2@mail.gmail.com> References: <748ff3ff0804040454s5417ae22vff2dd9e02413eeb2@mail.gmail.com> Message-ID: <748ff3ff0804040525s738b66femf9308e74edc98429@mail.gmail.com> To be clear - Java experience isn't required. I just figured those would be the people most likely to bolt from their current positions :) -Ryan On 4/4/08, Ryan Daigle wrote: > > If anybody on the list is interested in breaking free from their full time > Java gig and wants to work with Ruby 100% while still pulling down a > consistent salary and benefits - let me know. A client I've worked with in > the past is looking to add a full time developer to their small team. > Pretty sure you could still work from home, set your own hours and all that > fun stuff. > > If you're interested in hearing more, go ahead and send me your resume > (off the list, pls) and I'll get you in touch with them. > > Thanks, guys. > > -Ryan > -------------- next part -------------- An HTML attachment was scrubbed... URL: http://rubyforge.org/pipermail/raleigh-rb-members/attachments/20080404/21656524/attachment-0001.html From randomutterings at gmail.com Fri Apr 4 08:30:46 2008 From: randomutterings at gmail.com (Chris Barnes) Date: Fri, 4 Apr 2008 08:30:46 -0400 Subject: [raleigh.rb] Full time Rails job In-Reply-To: <748ff3ff0804040525s738b66femf9308e74edc98429@mail.gmail.com> References: <748ff3ff0804040454s5417ae22vff2dd9e02413eeb2@mail.gmail.com> <748ff3ff0804040525s738b66femf9308e74edc98429@mail.gmail.com> Message-ID: <52c57d810804040530x6f930db3o731d10e8cae84562@mail.gmail.com> Yeah I figured that, I actually like what I'm doing at my present job but I'm interested in telecommuting. On Fri, Apr 4, 2008 at 8:25 AM, Ryan Daigle wrote: > To be clear - Java experience isn't required. I just figured those would > be the people most likely to bolt from their current positions :) > > -Ryan > > > On 4/4/08, Ryan Daigle wrote: > > > > If anybody on the list is interested in breaking free from their full > > time Java gig and wants to work with Ruby 100% while still pulling down a > > consistent salary and benefits - let me know. A client I've worked with in > > the past is looking to add a full time developer to their small team. > > Pretty sure you could still work from home, set your own hours and all that > > fun stuff. > > > > If you're interested in hearing more, go ahead and send me your resume > > (off the list, pls) and I'll get you in touch with them. > > > > Thanks, guys. > > > > -Ryan > > > > > _______________________________________________ > raleigh-rb-members mailing list > raleigh-rb-members at rubyforge.org > http://rubyforge.org/mailman/listinfo/raleigh-rb-members > -------------- next part -------------- An HTML attachment was scrubbed... URL: http://rubyforge.org/pipermail/raleigh-rb-members/attachments/20080404/d752e18e/attachment.html From info at lojic.com Fri Apr 4 14:17:15 2008 From: info at lojic.com (Brian Adkins) Date: Fri, 04 Apr 2008 14:17:15 -0400 Subject: [raleigh.rb] Group insurance options In-Reply-To: <3e320180804022227g6a95f2e9ia2891c08f0b26b93@mail.gmail.com> References: <3e320180803240933i40b8f637y44b824c6e38f4a7c@mail.gmail.com> <47E7E9AC.3070804@lojic.com> <47E8271D.1000602@lojic.com> <3e320180804022227g6a95f2e9ia2891c08f0b26b93@mail.gmail.com> Message-ID: <47F670AB.6070001@lojic.com> Curtis Duhn wrote, On 04/03/2008 01:27 AM: > This is the first chance I've had to thank everyone for all the advice > regarding group medical insurance for independent consultants. I did a > ton of research into various options, so I thought I'd take a moment to > give back some of what I've learned for anyone else who finds themselves > in a similar situation. Thanks for posting the results of your research. > ... > PERs are similar to Professional Employment Organizations (PEOs), which > also hire employees on behalf of other companies, providing payroll > services, group benefits, etc. The key difference is that PEOs are > focused on selling their services to companies that want to have some or > all of their employees work through the PEO to avoid all the > administrative costs associated with having actual employees. While the > services are similar, I found few PEOs that seemed interested in working > with independent contractors. That was my experience also. It seems that you need at least 5 people or so for a PEO relationship to make sense financially, and many smaller PEO companies don't offer a group plan but underwrite each client separately which kind of defeats the purpose. Couple that with the fact that you have to introduce worker's comp, etc. which don't apply to an owner of a single-member LLC and it didn't make sense. I'm still trying to get all my ducks in a row to apply for small group health insurance for my LLC, but a PER might be the way to go when all the numbers are in. I think I was actually looking for a PER originally, but didn't know it and went down the PEO road which was a dead end. Brian From redinger at gmail.com Fri Apr 4 18:19:59 2008 From: redinger at gmail.com (Christopher Redinger) Date: Fri, 4 Apr 2008 18:19:59 -0400 Subject: [raleigh.rb] Another Full Time Rails Job Message-ID: <681d4e8f0804041519q7f364d0fpbaa81dedf2759766@mail.gmail.com> Not to distract from Ryan's posting, but if that job doesn't work out for you, my company, PrepChamps, is also hiring another full time Rails developer. http://agiledisciple.com/2008/4/4/prepchamps-is-hiring From mindcrime at cpphacker.co.uk Fri Apr 4 19:05:10 2008 From: mindcrime at cpphacker.co.uk (Phillip Rhodes) Date: Fri, 04 Apr 2008 19:05:10 -0400 Subject: [raleigh.rb] Full time Rails job In-Reply-To: <748ff3ff0804040525s738b66femf9308e74edc98429@mail.gmail.com> References: <748ff3ff0804040454s5417ae22vff2dd9e02413eeb2@mail.gmail.com> <748ff3ff0804040525s738b66femf9308e74edc98429@mail.gmail.com> Message-ID: <47F6B426.1020500@cpphacker.co.uk> Ryan Daigle wrote: > To be clear - Java experience isn't required. I just figured those > would be the people most likely to bolt from their current positions :) Isn't it wonderful how jokes about language bigotry never get old? :-) I would think that average COBOL, RPG or APL programmers would all be more interested in bolting their jobs than average Java programmers, but hey, to each his own... TTYL, -- Phillip Rhodes Chief Architect - OpenQabal https://openqabal.dev.java.net LinkedIn: http://www.linkedin.com/in/philliprhodes -------------- next part -------------- A non-text attachment was scrubbed... Name: mindcrime.vcf Type: text/x-vcard Size: 224 bytes Desc: not available Url : http://rubyforge.org/pipermail/raleigh-rb-members/attachments/20080404/ca161d4c/attachment.vcf From steve at iannopollo.com Fri Apr 4 22:02:56 2008 From: steve at iannopollo.com (Steve Iannopollo) Date: Fri, 4 Apr 2008 22:02:56 -0400 Subject: [raleigh.rb] Full time Rails job In-Reply-To: <47F6B426.1020500@cpphacker.co.uk> References: <748ff3ff0804040454s5417ae22vff2dd9e02413eeb2@mail.gmail.com> <748ff3ff0804040525s738b66femf9308e74edc98429@mail.gmail.com> <47F6B426.1020500@cpphacker.co.uk> Message-ID: <19183CAA-A2F9-41D1-8E7E-437B4B00460B@iannopollo.com> > I would think that average COBOL, RPG or APL programmers would all > be more interested in bolting their jobs than average Java > programmers, but hey, to each his own... Are you kidding? Not with the advent of Cobol on Cogs HTTP://WWW.COBOLONCOGS.ORG/INDEX.HTM -Steve From mindcrime at cpphacker.co.uk Sun Apr 6 22:09:02 2008 From: mindcrime at cpphacker.co.uk (Phillip Rhodes) Date: Sun, 06 Apr 2008 22:09:02 -0400 Subject: [raleigh.rb] [OT] Vote to bring Startup Weekend back to NC Message-ID: <47F9823E.9040700@cpphacker.co.uk> Hi all, if you missed the first Startup Weekend in NC (see for info) or would like to repeat the experience, please take a minute and go vote for a NC location (preferably RTP!) for a future SW event. TTYL, -- Phillip Rhodes Chief Architect - OpenQabal https://openqabal.dev.java.net LinkedIn: http://www.linkedin.com/in/philliprhodes -------------- next part -------------- A non-text attachment was scrubbed... Name: mindcrime.vcf Type: text/x-vcard Size: 224 bytes Desc: not available Url : http://rubyforge.org/pipermail/raleigh-rb-members/attachments/20080406/a67fde25/attachment.vcf From otto.hammersmith at gmail.com Mon Apr 7 13:52:09 2008 From: otto.hammersmith at gmail.com (Otto Hammersmith) Date: Mon, 7 Apr 2008 13:52:09 -0400 Subject: [raleigh.rb] Mongrel unsable on RHEL 5? In-Reply-To: <2b27183e0804031653x2ef60ce1ha72b7a9d42564c33@mail.gmail.com> References: <7c0835400804030949s7db30579k368b5f18776f74a3@mail.gmail.com> <2b27183e0804031653x2ef60ce1ha72b7a9d42564c33@mail.gmail.com> Message-ID: I may have solved my problem. I need to let the environment sit overnight again... but I found out why I wasn't seeing any exceptions. They show up in the mongrel log after I issue he shutdown so when I was looking for exceptions there was nothing. Googling for the message ultimately leads me to Mongrel FAQ . The long & short of it is that there's a timing problem between MySQL and the Ruby MySQL connector. Specifically this question... http://mongrel.rubyforge.org/wiki/FAQ#Q:Mongrelstopsworkingifitsleftaloneforalongtime . I won't know until tomorrow morning if this goes away, but I suspect it will. Thanks, all. -------------- next part -------------- An HTML attachment was scrubbed... URL: http://rubyforge.org/pipermail/raleigh-rb-members/attachments/20080407/000b0885/attachment.html From randomutterings at gmail.com Mon Apr 7 21:59:07 2008 From: randomutterings at gmail.com (Chris Barnes) Date: Mon, 7 Apr 2008 21:59:07 -0400 Subject: [raleigh.rb] Radiant Message-ID: <52c57d810804071859m17cc3932m80d93734a7682046@mail.gmail.com> Hi, I'm currently moving my blog over to Radiant. I would like to preserve the old links if possible but at the same time move the blog part of the site to a subdomain blog.randomutterings.com. For example I wrote an article on redundant firewalls and the old url is http://randomutterings.com/articles/2007/06/15/redundant-failover-firewall-with-pf-pfsync-and-carp-on-freebsdI want the new url to be http://blog.randomutterings.com/2007/06/15/redundant-failover-firewall-with-pf-pfsync-and-carp-on-freebsdbut at the same time I would like to have the old url forward the user to the new one. Does anyone with Radiant experience have a suggestion for this? -- Chris Barnes http://www.randomutterings.com -------------- next part -------------- An HTML attachment was scrubbed... URL: http://rubyforge.org/pipermail/raleigh-rb-members/attachments/20080407/79b48247/attachment.html From nathaniel at talbott.ws Tue Apr 15 16:33:19 2008 From: nathaniel at talbott.ws (Nathaniel Talbott) Date: Tue, 15 Apr 2008 16:33:19 -0400 Subject: [raleigh.rb] Pre-meeting Chow Message-ID: <4ce336a20804151333p5650e4deq744b647e3bfc934@mail.gmail.com> Sorry for the late notice, but as usual, anyone who's available is invited to join me at 5:30 tonight at Baja Burrito (http://tinyurl.com/2o2luk) to grab dinner and some Ruby chatter before heading over to Red Hat for the meeting. Looking forward to it, -- Nathaniel Talbott <:((>< From info at lojic.com Tue Apr 15 22:25:50 2008 From: info at lojic.com (Brian Adkins) Date: Tue, 15 Apr 2008 22:25:50 -0400 Subject: [raleigh.rb] RESTfullness Message-ID: <480563AE.2070206@lojic.com> I have two questions regarding modeling in a RESTful way. I asked the first during the Raleigh.rb meeting, and thought of the second as soon as I left: 1) What is the best way to model modifying the state of a resource in a way other than simply providing a list of new attribute values in the PUT for the resource? I referred to allowing a user to deny an application, but I think an example that Nathaniel and I discussed is better. Consider a geometric resource (e.g. a rectangle). What is the best REST way to model scaling the shape? I don't want to require the user to pass a new length & height. Basically, I'd like to see the equivalent to my_object.scale(2.0) 2) Smalltalk/Ruby have a single dispatch object model: foo.fun(bar) Lisp/CLOS has a multiple dispatch model: fun(foo, bar) It seems that REST fits better with the former than the latter since it's resource-centric. How would you model with REST a situation involving a logical operation that spans multiple objects/resources? In other words, an operation that makes sense from a user's perspective that results in state modifications to multiple resources? Or would you simply not use REST in this scenario? Thanks, Brian Adkins From rick.denatale at gmail.com Tue Apr 15 23:26:45 2008 From: rick.denatale at gmail.com (Rick DeNatale) Date: Tue, 15 Apr 2008 23:26:45 -0400 Subject: [raleigh.rb] RESTfullness In-Reply-To: <480563AE.2070206@lojic.com> References: <480563AE.2070206@lojic.com> Message-ID: On Tue, Apr 15, 2008 at 10:25 PM, Brian Adkins wrote: > I have two questions regarding modeling in a RESTful way. I asked the > first during the Raleigh.rb meeting, and thought of the second as soon > as I left: > > 1) What is the best way to model modifying the state of a resource in a > way other than simply providing a list of new attribute values in the > PUT for the resource? > > I referred to allowing a user to deny an application, but I think an > example that Nathaniel and I discussed is better. Consider a geometric > resource (e.g. a rectangle). What is the best REST way to model scaling > the shape? I don't want to require the user to pass a new length & > height. Basically, I'd like to see the equivalent to my_object.scale(2.0) To some extent, as usual, it depends. Here's a strawman: map.resources :shapes do |shape | shape.resource :scale end You could model scaling as an operation or an attribute. The first would change the length(width?) and height. The second would set a scale attribute used to transform the object. If you model it as an operation, then the ScaleController would only need a put action, and perhaps a new action to provide a browser ui, although it's more likely that the UI would be provided as part of another view. It you model it as an attribute then put, update and show actions would probably be appropriate. Another option, more suited for the operation case, would be to put the action for putting the scale operation resource into the shapes controller with map.resources :shapes, :member => {:scale => :put} This points out that there isn't a strict mapping between controllers and resources, or even between A controller and A collection of resources and the members of THAT collection, but this is something best done in moderation. It really is better to use a separate controller, and you can make the controller really thin using a plugin like resource_controller, which would also allow you to reuse the scale controller nested under different parent resources, if that made sense. > 2) Smalltalk/Ruby have a single dispatch object model: foo.fun(bar) > Lisp/CLOS has a multiple dispatch model: fun(foo, bar) It seems that > REST fits better with the former than the latter since it's > resource-centric. I guess, although the REST 'dispatch' model is really: resource.get resource.put(parameters) resource.update(parameters) resource.delete Just those four methods. Where the resource is identified by the uri (and possibly some other parts of the http request) and the parameters are carried by the HTTP request. As an example of the resource being partially identified by other parts of the request, consider Ben's dashboard example where, presumably, the user isn't identified by the URI, but by the session information transported via a cookie. In Rails there's a second dispatch which maps the http request to a controller, action, and parameters. > How would you model with REST a situation involving a logical operation > that spans multiple objects/resources? In other words, an operation that > makes sense from a user's perspective that results in state > modifications to multiple resources? Or would you simply not use REST in > this scenario? I think that, by definition, a rest request deals with one resource, any RESTful http request is a transfer of a representation of the logical state of a single resource, either from the server (get), or to the server(post, update, delete). But that doesn't mean that the state of a resource can't be complex and span multiple objects on the server. Don't fall back into the trap of equating resources with models, while some resources might map to single model objects, that's not the sole, or perhaps, even the most common case. -- Rick DeNatale My blog on Ruby http://talklikeaduck.denhaven2.com/ From rick.denatale at gmail.com Wed Apr 16 00:03:59 2008 From: rick.denatale at gmail.com (Rick DeNatale) Date: Wed, 16 Apr 2008 00:03:59 -0400 Subject: [raleigh.rb] RESTfullness In-Reply-To: References: <480563AE.2070206@lojic.com> Message-ID: I was going to mention this, but I forgot. Chapter 4 of Obie's "The Rails Way" has some good discussion germane to this thread. -- Rick DeNatale My blog on Ruby http://talklikeaduck.denhaven2.com/ From raleighrb at turrean.com Wed Apr 16 04:49:52 2008 From: raleighrb at turrean.com (Ben Scofield) Date: Wed, 16 Apr 2008 04:49:52 -0400 Subject: [raleigh.rb] RESTfullness In-Reply-To: <480563AE.2070206@lojic.com> References: <480563AE.2070206@lojic.com> Message-ID: Hi Brian, Thanks for your question during the talk - and for this clarification; I think I understand what you were asking much better now. 1) At first glance, I'm not terribly happy with either of the options Rick provided - though of the two, identifying scale as its own resource (scoped to shape) is better, I think. Instead, I'd fall back on the standard PUT update to a specific shape, like this: # in the controller def update # where params[:shape] is {:scale => '2.0'} in this case Shape.update_attributes(params[:shape]) # ... end # in the model def scale=(value) self.length = self.length * value.to_f self.width = self.wdth * value.to_f end To me, this makes more sense; the downside is that you now have 'attributes' of the resource that are only used to accomplish these specific actions, but I don't see that as a major problem. If you're concerned about how the client will know they can manipulate scale in this fashion, I think we can fall back on the discoverability constraint - essentially, the edit form should show scale as a valid manipulable attribute. Of course, YMMV. Remember that for any modeling exercise, you have to be aware of the level of description that's most appropriate for your particular needs. It could very well be that you need a Scale resource scoped to each of your shapes, as Rick suggested - nevertheless, this is still a solvable problem for a RESTful system. 2a) I think this is right, though I'm not sure what it means. You could certainly build a RESTful application in any given language - in Lisp/CLOS it would just look like `put(resource, representation)` instead of `resource.put(representation)`. 2b) This is slightly reminiscent of the question Clinton asked during the talk, where a single nested resource 'belongs' to two parents of the same resource type (his example was a Friendship shared by two Users). In that case, it's perfectly acceptable to access and manipulate the same underlying Friendship information via two different URIs/resources, and I'd say the same holds for your example. There's nothing preventing the creation or update of a specific resource on the server from having side-effects on other resources, so one approach would be to identify a 'primary' resource to trigger the action. Alternatively, you could provide triggers through each of the affected resources. A third option would be to identify a higher-level resource that spans the affected ones, and use it as an independent trigger. Without a specific example, I can't say which of these would be the most appropriate, but I think they're all viable choices. Thanks again! Ben From adam at thewilliams.ws Wed Apr 16 06:51:53 2008 From: adam at thewilliams.ws (Adam Williams) Date: Wed, 16 Apr 2008 06:51:53 -0400 Subject: [raleigh.rb] RESTfullness In-Reply-To: <480563AE.2070206@lojic.com> References: <480563AE.2070206@lojic.com> Message-ID: Ugh, I'm so sad I missed this discussion! On Apr 15, 2008, at 10:25 PM, Brian Adkins wrote: > I have two questions regarding modeling in a RESTful way. I asked the > first during the Raleigh.rb meeting, and thought of the second as soon > as I left: > > 1) What is the best way to model modifying the state of a resource > in a > way other than simply providing a list of new attribute values in the > PUT for the resource? > > I referred to allowing a user to deny an application, but I think an > example that Nathaniel and I discussed is better. Consider a geometric > resource (e.g. a rectangle). What is the best REST way to model > scaling > the shape? I don't want to require the user to pass a new length & > height. Basically, I'd like to see the equivalent to > my_object.scale(2.0) > > 2) Smalltalk/Ruby have a single dispatch object model: foo.fun(bar) > Lisp/CLOS has a multiple dispatch model: fun(foo, bar) It seems that > REST fits better with the former than the latter since it's > resource-centric. > > How would you model with REST a situation involving a logical > operation > that spans multiple objects/resources? In other words, an operation > that > makes sense from a user's perspective that results in state > modifications to multiple resources? Or would you simply not use > REST in > this scenario? > > Thanks, > Brian Adkins > > _______________________________________________ > raleigh-rb-members mailing list > raleigh-rb-members at rubyforge.org > http://rubyforge.org/mailman/listinfo/raleigh-rb-members From jrubyforge-raleigh-rb at indythinker.com Wed Apr 16 02:08:34 2008 From: jrubyforge-raleigh-rb at indythinker.com (jrubyforge-raleigh-rb at indythinker.com) Date: Wed, 16 Apr 2008 02:08:34 -0400 Subject: [raleigh.rb] RESTfullness In-Reply-To: <480563AE.2070206@lojic.com> References: <480563AE.2070206@lojic.com> Message-ID: <480597E2.4020707@indythinker.com> Brian, Let me take a shot at these questions. Please keep in mind that I'm mostly new to REST and am a total n00b with regard to Ruby. I might be totally wrong. I hope you find my answers useful, anyhow. Brian Adkins wrote: > I have two questions regarding modeling in a RESTful way. I asked the > first during the Raleigh.rb meeting, and thought of the second as soon > as I left: > > 1) What is the best way to model modifying the state of a resource in a > way other than simply providing a list of new attribute values in the > PUT for the resource? > > I referred to allowing a user to deny an application, but I think an > example that Nathaniel and I discussed is better. Consider a geometric > resource (e.g. a rectangle). What is the best REST way to model scaling > the shape? I don't want to require the user to pass a new length & > height. Basically, I'd like to see the equivalent to my_object.scale(2.0) What you seem to be asking for is a "complex transformation", where the resource will be transformed in a way that is more complex than simply substituting one value for another. This raises the design question of where your computational load should reside and how much you trust the client software to behave correctly. There are a few different ways to approach the issue and I'm not sure which one is "right". Approach #1: Perform the calculation on the client side --------------------------------------------------------------------- The popularity of REST has risen along with RIAs (rich internet applications), partly because it's very well suited to the task. In a typical AJAX application (* see below), you would GET a resource that is represented as XML, use Javascript and HTML to provide an interface to the user, and use PUT to store the changes if the user chooses to transform it in any way. It should not matter to the RESTful application as to whether the user made the changes or the client application computed a complex transformation on the user's behalf. The only thing that should matter is that the XML that is returned with the PUT request be a valid representation of that resource. So, here is one way to do this approach: 1) User triggers event in RIA that retrieves the resource. HTTP request: GET /rectangle/209 (request body empty, response contains XML) 2) Javascript transforms the XML representation of the resource into HTML and displays it in a div 3) User triggers another event which signals that it should scale the rectangle by a factor of 2. 4) Javascript calculates the new rectangle, updates the user's display to show it, and asks if the user wants to save changes. 5) If the user saves changes, Javascript creates an XML representation of the new resource and saves it. HTTP request: PUT /rectangle/209 (request body is XML, response is only relevant for error handling) Pros: > scales well to many users, because each user carries part of his/her own burden for computational transformations. > RESTful operations can be queued for later execution, even if internet connection is slow or absent. Cons: > client side scripting is still unpopular in some circles. > if someone else PUTs while you are transforming client-side, you have to resolve the conflict. > if you have complex domain logic that restricts the possible transformations, you may end up duplicating rules on both server and client. * Note: RIA with REST doesn't have to mean AJAX. This same strategy should work for any setup where you have a rich application on the client. The GUI could be provided in the browser by many things (HTML, Flash, Flex, Java, ActiveX) or could even be a desktop application, using hundreds of languages (including Ruby). The "representation" of the resource doesn't have to be XML, either. It could also be JSON, "URL encoded" data, proprietary binary data, or just about anything that can be described with a MIME type. Approach #2: Server does transformation at second resource ---------------------------------------------------------------------------- This is a rather "chatty" approach, but I think it preserves the RESTful approach pretty well. Someone will have to check me on this, because I've never read anything on this, I simply dreamed it up. The suggestion is to create a second "resource" on the server which performs transformations on rectangles. It would be a read-only (no PUT or POST allowed) resource. It would look something like this: GET /scaled/rectangle/209/scale/2.0 Or, perhaps this, but I'm still learning where the line is between URL and query parameters: GET /scaled/rectangle/209?scale=2.0 Or, perhaps you would include an XML representation of the rectangle in the body of the request. This would allow you to scale rectangles that haven't even had a POST yet. Note that I haven't specified "rectangle" in the URL, because it could theoretically handle representations of other resource types, as well: GET /scaler?scale=2.0 You can either display the newly scaled rectangle to the user or simply turn around and save it. Either way, you take the results of the GET from the scaler and then PUT it at the resource URI where you got the original. Pros: > the transformation is performed in a trusted environment where you can dictate software versions and settings. > you can run the transformation on a different server from the persistence layer, thus providing more flexibility in load balancing and scaling. Cons: > creates more network traffic than is probably necessary. > uses more server resources when the client could be doing the work. > the client must be online to perform a transformation. > if the network is slow, the UI might appear sluggish. Approach #3: Add an operator to the query string -------------------------------------------------------------- This one is kind of cheating, in my mind, but may be practical in many ways. You could add a bit onto the query string in order to signal to your controller that you want to perform a transformation. You're still sending a valid representation of the resource, but you're instructing the server to modify it before it saves it: PUT /rectangle/209?scale=2.0 Pros: > simple for most developers to understand. > transformation performed in trusted environment. Cons: > if you have lots of possible transformations, your controller will become cluttered. > uses server resources instead of client again, but this time it's harder to separate the persistence load from the transformation load. > can't show a preview to the user until you've saved the new representation. > 2) Smalltalk/Ruby have a single dispatch object model: foo.fun(bar) > Lisp/CLOS has a multiple dispatch model: fun(foo, bar) It seems that > REST fits better with the former than the latter since it's > resource-centric. > > How would you model with REST a situation involving a logical operation > that spans multiple objects/resources? In other words, an operation that > makes sense from a user's perspective that results in state > modifications to multiple resources? Or would you simply not use REST in > this scenario? Perhaps I'm oversimplifying the answer, but I think it's just this: You can GET a resource which is a collection of other resources, so why can't you PUT a collection of resources? If you're representing your resources with XML, you can just nest the data and send it all in one bundle. You have to deploy a URI that explicitly handles a PUT for a resource of this nature, but it should be very doable. It also seems that you're still caught up on "verbs", such as the "fun" in fun(foo, bar), rather than "nouns". Hopefully, my "approach #1" from above helped shift that mindset a bit. Try thinking of it this way, using an AJAX app as an example: - The HTML created by your Javascript is the "view". - Quite a bit of your Javascript functions as the "controller". - The RESTful web application itself is your "model". Then, within the web application, you have another model, view, and controller: - The XML created by your webapp is the "view". - The controller at this level checks for valid data, confirms user permissions, and triggers domain logic that can't be trusted to client apps. - The business objects or domain objects are your "model". So, there are my answers. If anybody else can fill in where I am lacking or confused, that would be great. I'm really excited about REST and am trying my best to "get it". Kind regards, Justis Peters http://justis.indythinker.com From rick.denatale at gmail.com Wed Apr 16 09:12:37 2008 From: rick.denatale at gmail.com (Rick DeNatale) Date: Wed, 16 Apr 2008 09:12:37 -0400 Subject: [raleigh.rb] RESTfullness In-Reply-To: References: <480563AE.2070206@lojic.com> Message-ID: On Wed, Apr 16, 2008 at 4:49 AM, Ben Scofield wrote: > Hi Brian, > > Thanks for your question during the talk - and for this clarification; > I think I understand what you were asking much better now. > > 1) At first glance, I'm not terribly happy with either of the options > Rick provided - Of course the purpose of a strawman, particularly one created as one long day turns to another, it to provide a target for sticks, and to stimulate discussion. > though of the two, identifying scale as its own > resource (scoped to shape) is better, I think. Instead, I'd fall back > on the standard PUT update to a specific shape, like this: > > # in the controller > def update > # where params[:shape] is {:scale => '2.0'} in this case > Shape.update_attributes(params[:shape]) > > # ... > end > > # in the model > def scale=(value) > self.length = self.length * value.to_f > self.width = self.wdth * value.to_f > end > > To me, this makes more sense; the downside is that you now have > 'attributes' of the resource that are only used to accomplish these > specific actions, but I don't see that as a major problem. If you're > concerned about how the client will know they can manipulate scale in > this fashion, I think we can fall back on the discoverability > constraint - essentially, the edit form should show scale as a valid > manipulable attribute. Yep, and I don't think that it's unreasonable to have multiple actions like edit which can provide specific forms to do an update of a specific attribute or set of attributes. On the other hand, if we're talking about a web service api rather and a ui, then we've got more documentation to do about what's expected in post/put requests. > Of course, YMMV. Remember that for any modeling exercise, you have to > be aware of the level of description that's most appropriate for your > particular needs. It could very well be that you need a Scale resource > scoped to each of your shapes, as Rick suggested - nevertheless, this > is still a solvable problem for a RESTful system. As usual there are multiple alternatives, each with it's own set of pros, cons, indications and contraindications. At the risk of raising the hackles of those who are allergic to the term, I suspect that there's the potential of a pattern language for RESTful design here, perhaps it already exists somewhere, even in nascent form. > 2a) I think this is right, though I'm not sure what it means. You > could certainly build a RESTful application in any given language - in > Lisp/CLOS it would just look like `put(resource, representation)` > instead of `resource.put(representation)`. > > 2b) This is slightly reminiscent of the question Clinton asked during > the talk, where a single nested resource 'belongs' to two parents of > the same resource type (his example was a Friendship shared by two > Users). In that case, it's perfectly acceptable to access and > manipulate the same underlying Friendship information via two > different URIs/resources, and I'd say the same holds for your example. > > There's nothing preventing the creation or update of a specific > resource on the server from having side-effects on other resources, so > one approach would be to identify a 'primary' resource to trigger the > action. Alternatively, you could provide triggers through each of the > affected resources. A third option would be to identify a higher-level > resource that spans the affected ones, and use it as an independent > trigger. Without a specific example, I can't say which of these would > be the most appropriate, but I think they're all viable choices. Another way of looking at this is, keeping in mind that resources aren't (necessarily) model objects, is that a resource is a view, in an analogy to the DB sense of a view rather than the MVC sense, of the state on the server, and that some aspects of that state might be present on multiple views/resources. Getting back to Rails and REST patterns, I was thinking this morning about the relationship of rails form and field helpers to REST. REST leaves the interpretation of the contents of an HTTP request up to the server. As we know, Rails has the mechanisms for building a form which generates a post/put request with contents which can be interpreted by the controller/model(s) as operating on a tree of models linked by attributes, using a request notation which gets mapped to nested hashes in params. This is perfectly RESTful, but the details aren't standardized (I think) outside of rails. Again, this works well for browser interactions since the notations carried via (X)HTML through the browser dom and back again can be taken on a round trip without the browser knowing about the notational conventions. In the case of a service API, where the client needs to post/put the right form of request without starting with a form, then more documentation of the API would be necessary. Perhaps there are standards for doing this documentation, if so I'd love to hear about them. -- Rick DeNatale My blog on Ruby http://talklikeaduck.denhaven2.com/ From russell.christopher at gmail.com Wed Apr 16 10:50:25 2008 From: russell.christopher at gmail.com (Russell Christopher) Date: Wed, 16 Apr 2008 10:50:25 -0400 Subject: [raleigh.rb] RESTfullness In-Reply-To: References: <480563AE.2070206@lojic.com> Message-ID: <575ed91c0804160750n798258cdq5f6b87271c903088@mail.gmail.com> Firstly, great talk Ben. Second, you straw man is good Rick, gets me thinking in new ways I was going to bring this up last nite but decided that this might be a better place. The APP (atom) people have recently discussed a PATCH verb to do On Wed, Apr 16, 2008 at 9:12 AM, Rick DeNatale wrote: > On Wed, Apr 16, 2008 at 4:49 AM, Ben Scofield > wrote: > > Hi Brian, > > > > Thanks for your question during the talk - and for this clarification; > > I think I understand what you were asking much better now. > > > > 1) At first glance, I'm not terribly happy with either of the options > > Rick provided - > > Of course the purpose of a strawman, particularly one created as one > long day turns to another, it to provide a target for sticks, and to > stimulate discussion. > > > though of the two, identifying scale as its own > > resource (scoped to shape) is better, I think. Instead, I'd fall back > > on the standard PUT update to a specific shape, like this: > > > > # in the controller > > def update > > # where params[:shape] is {:scale => '2.0'} in this case > > Shape.update_attributes(params[:shape]) > > > > # ... > > end > > > > # in the model > > def scale=(value) > > self.length = self.length * value.to_f > > self.width = self.wdth * value.to_f > > end > > > > To me, this makes more sense; the downside is that you now have > > 'attributes' of the resource that are only used to accomplish these > > specific actions, but I don't see that as a major problem. If you're > > concerned about how the client will know they can manipulate scale in > > this fashion, I think we can fall back on the discoverability > > constraint - essentially, the edit form should show scale as a valid > > manipulable attribute. > > Yep, and I don't think that it's unreasonable to have multiple actions > like edit which can provide specific forms to do an update of a > specific attribute or set of attributes. > > On the other hand, if we're talking about a web service api rather and > a ui, then we've got more documentation to do about what's expected in > post/put requests. > > > Of course, YMMV. Remember that for any modeling exercise, you have to > > be aware of the level of description that's most appropriate for your > > particular needs. It could very well be that you need a Scale resource > > scoped to each of your shapes, as Rick suggested - nevertheless, this > > is still a solvable problem for a RESTful system. > > As usual there are multiple alternatives, each with it's own set of > pros, cons, indications and contraindications. At the risk of raising > the hackles of those who are allergic to the term, I suspect that > there's the potential of a pattern language for RESTful design here, > perhaps it already exists somewhere, even in nascent form. > > > 2a) I think this is right, though I'm not sure what it means. You > > could certainly build a RESTful application in any given language - in > > Lisp/CLOS it would just look like `put(resource, representation)` > > instead of `resource.put(representation)`. > > > > 2b) This is slightly reminiscent of the question Clinton asked during > > the talk, where a single nested resource 'belongs' to two parents of > > the same resource type (his example was a Friendship shared by two > > Users). In that case, it's perfectly acceptable to access and > > manipulate the same underlying Friendship information via two > > different URIs/resources, and I'd say the same holds for your example. > > > > There's nothing preventing the creation or update of a specific > > resource on the server from having side-effects on other resources, so > > one approach would be to identify a 'primary' resource to trigger the > > action. Alternatively, you could provide triggers through each of the > > affected resources. A third option would be to identify a higher-level > > resource that spans the affected ones, and use it as an independent > > trigger. Without a specific example, I can't say which of these would > > be the most appropriate, but I think they're all viable choices. > > Another way of looking at this is, keeping in mind that resources > aren't (necessarily) model objects, is that a resource is a view, in > an analogy to the DB sense of a view rather than the MVC sense, of the > state on the server, and that some aspects of that state might be > present on multiple views/resources. > > Getting back to Rails and REST patterns, I was thinking this morning > about the relationship of rails form and field helpers to REST. REST > leaves the interpretation of the contents of an HTTP request up to the > server. As we know, Rails has the mechanisms for building a form > which generates a post/put request with contents which can be > interpreted by the controller/model(s) as operating on a tree of > models linked by attributes, using a request notation which gets > mapped to nested hashes in params. This is perfectly RESTful, but the > details aren't standardized (I think) outside of rails. Again, this > works well for browser interactions since the notations carried via > (X)HTML through the browser dom and back again can be taken on a round > trip without the browser knowing about the notational conventions. > > In the case of a service API, where the client needs to post/put the > right form of request without starting with a form, then more > documentation of the API would be necessary. Perhaps there are > standards for doing this documentation, if so I'd love to hear about > them. > > -- > Rick DeNatale > > My blog on Ruby > http://talklikeaduck.denhaven2.com/ > _______________________________________________ > raleigh-rb-members mailing list > raleigh-rb-members at rubyforge.org > http://rubyforge.org/mailman/listinfo/raleigh-rb-members > -------------- next part -------------- An HTML attachment was scrubbed... URL: http://rubyforge.org/pipermail/raleigh-rb-members/attachments/20080416/cf1aa87e/attachment.html From russell.christopher at gmail.com Wed Apr 16 10:53:55 2008 From: russell.christopher at gmail.com (Russell Christopher) Date: Wed, 16 Apr 2008 10:53:55 -0400 Subject: [raleigh.rb] RESTfullness In-Reply-To: <575ed91c0804160750n798258cdq5f6b87271c903088@mail.gmail.com> References: <480563AE.2070206@lojic.com> <575ed91c0804160750n798258cdq5f6b87271c903088@mail.gmail.com> Message-ID: <575ed91c0804160753o79f1f889sfe53b6ce020ad694@mail.gmail.com> Sorry guys, I inadvertently hit send I was just wondering if anyone has been following the APP discussion about PATCH and if they have any opinions about it in a Rails context. The discussion has gone quiet AFAIC,S so maybe it's no longer actively being considered but it was an attempt to deal with updating selected items of an APP feed which seems relevant to this discussion On Wed, Apr 16, 2008 at 10:50 AM, Russell Christopher < russell.christopher at gmail.com> wrote: > Firstly, great talk Ben. Second, you straw man is good Rick, gets me > thinking in new ways > > I was going to bring this up last nite but decided that this might be a > better place. > > The APP (atom) people have recently discussed a PATCH verb to do > > On Wed, Apr 16, 2008 at 9:12 AM, Rick DeNatale > wrote: > > > On Wed, Apr 16, 2008 at 4:49 AM, Ben Scofield > > wrote: > > > Hi Brian, > > > > > > Thanks for your question during the talk - and for this > > clarification; > > > I think I understand what you were asking much better now. > > > > > > 1) At first glance, I'm not terribly happy with either of the options > > > Rick provided - > > > > Of course the purpose of a strawman, particularly one created as one > > long day turns to another, it to provide a target for sticks, and to > > stimulate discussion. > > > > > though of the two, identifying scale as its own > > > resource (scoped to shape) is better, I think. Instead, I'd fall back > > > on the standard PUT update to a specific shape, like this: > > > > > > # in the controller > > > def update > > > # where params[:shape] is {:scale => '2.0'} in this case > > > Shape.update_attributes(params[:shape]) > > > > > > # ... > > > end > > > > > > # in the model > > > def scale=(value) > > > self.length = self.length * value.to_f > > > self.width = self.wdth * value.to_f > > > end > > > > > > To me, this makes more sense; the downside is that you now have > > > 'attributes' of the resource that are only used to accomplish these > > > specific actions, but I don't see that as a major problem. If you're > > > concerned about how the client will know they can manipulate scale in > > > this fashion, I think we can fall back on the discoverability > > > constraint - essentially, the edit form should show scale as a valid > > > manipulable attribute. > > > > Yep, and I don't think that it's unreasonable to have multiple actions > > like edit which can provide specific forms to do an update of a > > specific attribute or set of attributes. > > > > On the other hand, if we're talking about a web service api rather and > > a ui, then we've got more documentation to do about what's expected in > > post/put requests. > > > > > Of course, YMMV. Remember that for any modeling exercise, you have to > > > be aware of the level of description that's most appropriate for your > > > particular needs. It could very well be that you need a Scale > > resource > > > scoped to each of your shapes, as Rick suggested - nevertheless, this > > > is still a solvable problem for a RESTful system. > > > > As usual there are multiple alternatives, each with it's own set of > > pros, cons, indications and contraindications. At the risk of raising > > the hackles of those who are allergic to the term, I suspect that > > there's the potential of a pattern language for RESTful design here, > > perhaps it already exists somewhere, even in nascent form. > > > > > 2a) I think this is right, though I'm not sure what it means. You > > > could certainly build a RESTful application in any given language - > > in > > > Lisp/CLOS it would just look like `put(resource, representation)` > > > instead of `resource.put(representation)`. > > > > > > 2b) This is slightly reminiscent of the question Clinton asked during > > > the talk, where a single nested resource 'belongs' to two parents of > > > the same resource type (his example was a Friendship shared by two > > > Users). In that case, it's perfectly acceptable to access and > > > manipulate the same underlying Friendship information via two > > > different URIs/resources, and I'd say the same holds for your > > example. > > > > > > There's nothing preventing the creation or update of a specific > > > resource on the server from having side-effects on other resources, > > so > > > one approach would be to identify a 'primary' resource to trigger the > > > action. Alternatively, you could provide triggers through each of the > > > affected resources. A third option would be to identify a > > higher-level > > > resource that spans the affected ones, and use it as an independent > > > trigger. Without a specific example, I can't say which of these would > > > be the most appropriate, but I think they're all viable choices. > > > > Another way of looking at this is, keeping in mind that resources > > aren't (necessarily) model objects, is that a resource is a view, in > > an analogy to the DB sense of a view rather than the MVC sense, of the > > state on the server, and that some aspects of that state might be > > present on multiple views/resources. > > > > Getting back to Rails and REST patterns, I was thinking this morning > > about the relationship of rails form and field helpers to REST. REST > > leaves the interpretation of the contents of an HTTP request up to the > > server. As we know, Rails has the mechanisms for building a form > > which generates a post/put request with contents which can be > > interpreted by the controller/model(s) as operating on a tree of > > models linked by attributes, using a request notation which gets > > mapped to nested hashes in params. This is perfectly RESTful, but the > > details aren't standardized (I think) outside of rails. Again, this > > works well for browser interactions since the notations carried via > > (X)HTML through the browser dom and back again can be taken on a round > > trip without the browser knowing about the notational conventions. > > > > In the case of a service API, where the client needs to post/put the > > right form of request without starting with a form, then more > > documentation of the API would be necessary. Perhaps there are > > standards for doing this documentation, if so I'd love to hear about > > them. > > > > -- > > Rick DeNatale > > > > My blog on Ruby > > http://talklikeaduck.denhaven2.com/ > > _______________________________________________ > > raleigh-rb-members mailing list > > raleigh-rb-members at rubyforge.org > > http://rubyforge.org/mailman/listinfo/raleigh-rb-members > > > > -------------- next part -------------- An HTML attachment was scrubbed... URL: http://rubyforge.org/pipermail/raleigh-rb-members/attachments/20080416/0da96636/attachment-0001.html From info at lojic.com Wed Apr 16 17:32:42 2008 From: info at lojic.com (Brian Adkins) Date: Wed, 16 Apr 2008 17:32:42 -0400 Subject: [raleigh.rb] RESTfulness In-Reply-To: <480563AE.2070206@lojic.com> References: <480563AE.2070206@lojic.com> Message-ID: <4806707A.9030407@lojic.com> Thanks to everyone for the replies; they've given me some things to chew on. I just wrote a fairly lengthy reply addressing some points and then deleted it because I think that it would better for me to simply read through Fielding's dissertation (I've only browsed it, and that was several years ago) before commenting further. Brian From jrubyforge-raleigh-rb at indythinker.com Wed Apr 16 18:48:54 2008 From: jrubyforge-raleigh-rb at indythinker.com (Justis Peters) Date: Wed, 16 Apr 2008 18:48:54 -0400 Subject: [raleigh.rb] A sincere thanks to Ben Scofield for Apr 15 raleigh.rb Message-ID: <48068256.9060908@indythinker.com> To Ben and the broader raleigh.rb members, So, yesterday was my first time attending raleigh.rb and it was very much worth it. Your presentation on REST was entertaining, informative, and very useful to me. I am currently designing the final API for a special server appliance I am building for a client and my new understanding of REST has already helped many times. This is really great stuff and I'm glad that I took the time to deepen my knowledge. Thanks so much for the presentation, Ben. I look forward to coming back to more raleigh.rb meetings. Ruby looks like a very interesting language and the size and enthusiasm of the crowd certainly makes it more attractive. You have a great meetup going here. I think I may have to prioritize learning a thing or two about Ruby. Now I just need to remember where I left my spare time. Kind regards, Justis Peters http://justis.indythinker.com From rsanheim at gmail.com Thu Apr 17 01:34:58 2008 From: rsanheim at gmail.com (Rob Sanheim) Date: Thu, 17 Apr 2008 01:34:58 -0400 Subject: [raleigh.rb] slides for REST talk Message-ID: Any chance of getting the presentation from the REST talk from week? I couldn't make it due to family things, but would love to listen/follow along in the slides. thanks, - Rob From raleighrb at turrean.com Thu Apr 17 06:10:16 2008 From: raleighrb at turrean.com (Ben Scofield) Date: Thu, 17 Apr 2008 06:10:16 -0400 Subject: [raleigh.rb] slides for REST talk In-Reply-To: References: Message-ID: I uploaded the slides to Slideshare: http://www.slideshare.net/bscofield/domain-modeling-with-rest Questions and comments are welcome! Ben On Thu, Apr 17, 2008 at 1:34 AM, Rob Sanheim wrote: > Any chance of getting the presentation from the REST talk from week? > I couldn't make it due to family things, but would love to > listen/follow along in the slides. > > thanks, > - Rob > _______________________________________________ > raleigh-rb-members mailing list > raleigh-rb-members at rubyforge.org > http://rubyforge.org/mailman/listinfo/raleigh-rb-members > From nospam at tonyspencer.com Thu Apr 17 13:17:44 2008 From: nospam at tonyspencer.com (Tony Spencer) Date: Thu, 17 Apr 2008 13:17:44 -0400 Subject: [raleigh.rb] Lighthouse users: Did you lose data in the latest upgrade? Message-ID: <058EF269-5A0A-452C-A481-CF861FC0680E@tonyspencer.com> Lighthouse bug tracking just released a fresh new design and I love the functionality but one of my users was deleted. :( I hope they can restore it and at least his comments are still available (under some generic lighthouse user account) but I am really concerned that it could be worse next time they lose data. Has anyone else experienced problems with the ugrade? Tony From nospam at tonyspencer.com Thu Apr 17 13:34:20 2008 From: nospam at tonyspencer.com (Tony Spencer) Date: Thu, 17 Apr 2008 13:34:20 -0400 Subject: [raleigh.rb] Lighthouse users: Did you lose data in the latest upgrade? In-Reply-To: <058EF269-5A0A-452C-A481-CF861FC0680E@tonyspencer.com> References: <058EF269-5A0A-452C-A481-CF861FC0680E@tonyspencer.com> Message-ID: Nevermind. It was user error. He just told me he was playing with the new interface and accidently deleted his account. :( On Apr 17, 2008, at 1:17 PM, Tony Spencer wrote: > Lighthouse bug tracking just released a fresh new design and I love > the functionality but one of my users was deleted. :( I hope they can > restore it and at least his comments are still available (under some > generic lighthouse user account) but I am really concerned that it > could be worse next time they lose data. Has anyone else experienced > problems with the ugrade? > > Tony > _______________________________________________ > raleigh-rb-members mailing list > raleigh-rb-members at rubyforge.org > http://rubyforge.org/mailman/listinfo/raleigh-rb-members > From javery at infozerk.com Wed Apr 23 10:22:01 2008 From: javery at infozerk.com (James Avery) Date: Wed, 23 Apr 2008 10:22:01 -0400 Subject: [raleigh.rb] Refresh the Triangle 006: 6:30 - 8:00 tomorrow In-Reply-To: <480f43d265f03@refreshthetriangle.org> References: <480f43d265f03@refreshthetriangle.org> Message-ID: <20af90580804230722n32ba3fb5ib49f1ce44f668035@mail.gmail.com> This meeting is an interesting blend of designers and developers, this months topic seems pretty relevant to Ruby devs: Refresh the Triangle meets tomorrow, Thursday, April 24th, from 6:30 to 8:00pm at Blackwell Interactive's office in Durham. Nathan Huening will present his talk, "Dr jQuery (Or, How I Learned to Stop Worrying and Love the DOM)," and Blackwell Interactive will provide refreshments. If you haven't already, please RSVP at Upcoming. http://upcoming.yahoo.com/event/491350/?ps=5 DIRECTIONS: Blackwell Interactive / Duke OIT is at the American Tobacco Historic District, which is across from the Durham Bulls Stadium. The office is located within Strickland Building. The entrance is opposite the courtyard from Tyler's Taproom. Signs will direct you once you find the entrance. Parking will be most convenient in the South Parking Deck, accessible from Willard Street. This map might be helpful - http://www.americantobaccohistoricdistrict.com/directory.html -- James Avery Infozerk Inc. http://www.infozerk.com -------------- next part -------------- An HTML attachment was scrubbed... URL: http://rubyforge.org/pipermail/raleigh-rb-members/attachments/20080423/1f2e4e90/attachment.html From rhooker at cisco.com Wed Apr 23 10:35:48 2008 From: rhooker at cisco.com (Raymond Hooker (rhooker)) Date: Wed, 23 Apr 2008 10:35:48 -0400 Subject: [raleigh.rb] LDAP Authentication In-Reply-To: <480563AE.2070206@lojic.com> References: <480563AE.2070206@lojic.com> Message-ID: <848DB51F3E58E146B08D188F803DCB6F0671B0AB@xmb-rtp-20c.amer.cisco.com> I am trying to get LDAP authentication working. Since we have a fairly complex path, I have tried the code below (with of course my organization specific parameters). I tried a user that I know (my own of course) in IRB and see that it finds the user in the call to ldap.search. The ldap.bind returns true.. BUT when I change my password to something incorrect, it still returns true. Clearly that is not very adequate authentication.. Anyone have a clue about these libraries or a suggestion on a better approach? http://wiki.rubyonrails.org/rails/pages/HowtoAuthenticateWithRubyNetLdap require 'rubygems' require 'net/ldap' username = 'jdoe' password = 'mypass' ldap = Net::LDAP.new(:host => 'ldap.example.com', :base => 'dc=example,dc=com') filter = Net::LDAP::Filter.eq('uid', username) ldap.search(:filter => filter) {|entry| username = entry.dn} ldap.auth(username, password) if ldap.bind # authentication succeeded else # authentication failed end From mark.bennett.mail at gmail.com Wed Apr 23 11:07:05 2008 From: mark.bennett.mail at gmail.com (Mark Bennett) Date: Wed, 23 Apr 2008 11:07:05 -0400 Subject: [raleigh.rb] LDAP Authentication In-Reply-To: <848DB51F3E58E146B08D188F803DCB6F0671B0AB@xmb-rtp-20c.amer.cisco.com> References: <480563AE.2070206@lojic.com> <848DB51F3E58E146B08D188F803DCB6F0671B0AB@xmb-rtp-20c.amer.cisco.com> Message-ID: In this example I first connect as a ldap manager then authorize the user. I'm not sure if that's necessary but it is what I did. require 'net/ldap' def initialize_ldap_con(user_name, password) Net::LDAP.new( {:host => '10.1.6.23', :port => 389, :auth => { :method => :simple, :username => user_name, :password => password }} ) end def authenticate(username, password) ldap_con = initialize_ldap_con('MURDOCH\LDAPUser','#####') treebase = "DC=murdoch,DC=gov" user_filter = Net::LDAP::Filter.eq( "sAMAccountName", username ) op_filter = Net::LDAP::Filter.eq( "objectClass", "organizationalPerson" ) dn = String.new ldap_con.search( :base => treebase, :filter => op_filter & user_filter, :attributes=> 'dn') do |entry| pp entry dn = entry.dn end login_succeeded = false unless dn.empty? ldap_con = initialize_ldap_con(dn,password) login_succeeded = true if ldap_con.bind end login_succeeded end username = 'markbennett' password = ##### puts authenticate(username, password) Mark On Wed, Apr 23, 2008 at 10:35 AM, Raymond Hooker (rhooker) < rhooker at cisco.com> wrote: > I am trying to get LDAP authentication working. Since we have a fairly > complex path, I have tried the code below (with of course my > organization specific parameters). I tried a user that I know (my own > of course) in IRB and see that it finds the user in the call to > ldap.search. The ldap.bind returns true.. BUT when I change my password > to something incorrect, it still returns true. Clearly that is not very > adequate authentication.. Anyone have a clue about these libraries or > a suggestion on a better approach? > > http://wiki.rubyonrails.org/rails/pages/HowtoAuthenticateWithRubyNetLdap > > require 'rubygems' > require 'net/ldap' > > username = 'jdoe' > password = 'mypass' > ldap = Net::LDAP.new(:host => 'ldap.example.com', :base => > 'dc=example,dc=com') > filter = Net::LDAP::Filter.eq('uid', username) > ldap.search(:filter => filter) {|entry| username = entry.dn} > ldap.auth(username, password) > > if ldap.bind > # authentication succeeded > else > # authentication failed > end > _______________________________________________ > raleigh-rb-members mailing list > raleigh-rb-members at rubyforge.org > http://rubyforge.org/mailman/listinfo/raleigh-rb-members > -------------- next part -------------- An HTML attachment was scrubbed... URL: http://rubyforge.org/pipermail/raleigh-rb-members/attachments/20080423/bc79e0e6/attachment-0001.html From nospam at tonyspencer.com Wed Apr 23 13:46:31 2008 From: nospam at tonyspencer.com (Tony Spencer) Date: Wed, 23 Apr 2008 13:46:31 -0400 Subject: [raleigh.rb] gems are driving me crazy Message-ID: <06311996-680A-4680-9834-9FBB73EAF90F@tonyspencer.com> Does anybody know how I can revert my installation of gem to an earlier version? I'm on v 1.0.1 but the rest of my team is on 0.95 and as a result my app is complaining about the deprecated way to require a gem: /usr/local/lib/ruby/site_ruby/1.8/rubygems/custom_require.rb:27:in `gem_original_require': no such file to load -- RMagick.so (MissingSourceFile) So then I try to revert with: gem update --system -v '0.9.5' and my Mac complains: Attempting remote update of 0.9.5 ERROR: could not find 0.9.5 locally or in a repository Attempting remote update of system ERROR: could not find system locally or in a repository Gems updated: 0.9.5, system Arrgghh. Gem is no longer my friend. From scott at nextlot.com Thu Apr 24 18:42:20 2008 From: scott at nextlot.com (Scott Finkelstein) Date: Thu, 24 Apr 2008 18:42:20 -0400 Subject: [raleigh.rb] Ruby on Rails Developer - Full Time Position with Fast-Growing RTP Company In-Reply-To: References: Message-ID: Rails Developer Type: Full-Time Position Description: NextLot, an exciting and growing internet technology company in the auction software space, is seeking a Ruby on Rails developer to grow our development team in order to meet customer demand. NextLot is located in the heart of the Research Triangle Park in Durham, NC. The perfect applicant will enjoy a range of tasks and challenges, including but not limited to application enhancements, website design, quality assurance testing, user interface design, API design, performance tuning, and e-commerce. This position will require a confident mind and an adventurous spirit! Required: - Graduate of a 4-year college or university - Impressive grasp of Object Oriented design and principles - At least 6 months experience with all areas of Ruby on Rails development - Unquestionable mastery of HTML and CSS - At least 1 year experience with SQL - Strong understanding of software performance - Ability to meet challenges - Desire to learn Additionally Desired: - Quality Assurance experience - User Interface design experience - Flash experience - MySQL performance tuning experience - Web Services API experience - Streaming audio/video experience ** Will consider on-site full-time employees only. Contractors and recruiters are requested not to apply. Apply: Please apply by email to careers at nextlot.com. A cover letter or cover email is required. Location United States, RTP, NC This role requires onsite work Salary Pay: Competitive, commensurate with strength and range of talents and experience Further Information http://www.nextlot.com -- Scott Finkelstein NextLot, Inc. Toll-Free Number: 1-800-516-3830, ext. 3 Direct Dial: 919-361-1111 Fax: 484-770-4452 Cell: 516-315-2806 Email: scott at nextlot.com Website: www.nextlot.com -- Scott Finkelstein NextLot, Inc. Toll-Free Number: 1-800-516-3830, ext. 3 Direct Dial: 919-361-1111 Fax: 484-770-4452 Cell: 516-315-2806 Email: scott at nextlot.com Website: www.nextlot.com -------------- next part -------------- An HTML attachment was scrubbed... URL: http://rubyforge.org/pipermail/raleigh-rb-members/attachments/20080424/75c8372e/attachment.html From nospam at tonyspencer.com Mon Apr 28 17:07:12 2008 From: nospam at tonyspencer.com (Tony Spencer) Date: Mon, 28 Apr 2008 17:07:12 -0400 Subject: [raleigh.rb] Presentation on scaling Rails Message-ID: Too bad we can't see the video of this presentation but the slides are pretty informative if you find yourself investigating scaling Rails. http://www.scribd.com/doc/49575/Scaling-Rails-Presentation Also, if you find yourself needing to attempt to convince someone that Rails is perfectly scalable Scribd is a great real world example. I know until now I've had trouble pointing to high traffic sites and sometimes it can be tough with all the "Ruby is so slow" articles out there: 17 million pageviews a month Alexa rank: 538 http://siteanalytics.compete.com/scribd.com/?metric=uv -------------- next part -------------- An HTML attachment was scrubbed... URL: From nospam at tonyspencer.com Mon Apr 28 20:32:45 2008 From: nospam at tonyspencer.com (Tony Spencer) Date: Mon, 28 Apr 2008 20:32:45 -0400 Subject: [raleigh.rb] gems are driving me crazy In-Reply-To: <06311996-680A-4680-9834-9FBB73EAF90F@tonyspencer.com> References: <06311996-680A-4680-9834-9FBB73EAF90F@tonyspencer.com> Message-ID: <8D422564-15A3-4F87-A640-3E8A3F55BF61@tonyspencer.com> I'm replying to my own post here in case anyone else runs into this problem. Following this guys post solved the problem on my Mac (though I had to locate the .gz files from servers other than the ones he used): http://rubyforge.org/forum/message.php?msg_id=26160 On Apr 23, 2008, at 1:46 PM, Tony Spencer wrote: > Does anybody know how I can revert my installation of gem to an > earlier version? I'm on v 1.0.1 but the rest of my team is on 0.95 > and as a result my app is complaining about the deprecated way to > require a gem: > > /usr/local/lib/ruby/site_ruby/1.8/rubygems/custom_require.rb:27:in > `gem_original_require': no such file to load -- RMagick.so > (MissingSourceFile) > > So then I try to revert with: > > gem update --system -v '0.9.5' > > and my Mac complains: > > Attempting remote update of 0.9.5 > ERROR: could not find 0.9.5 locally or in a repository > Attempting remote update of system > ERROR: could not find system locally or in a repository > Gems updated: 0.9.5, system > > > Arrgghh. Gem is no longer my friend. > > _______________________________________________ > raleigh-rb-members mailing list > raleigh-rb-members at rubyforge.org > http://rubyforge.org/mailman/listinfo/raleigh-rb-members > From seancribbs at gmail.com Wed Apr 30 10:19:32 2008 From: seancribbs at gmail.com (Sean Cribbs) Date: Wed, 30 Apr 2008 09:19:32 -0500 Subject: [raleigh.rb] Moving to the area Message-ID: <48187FF4.8010003@gmail.com> Rubyists, It looks like I'll be moving the RDU-CH area come July, as my wife will be going to UNC in the fall. Thanks to Adam Williams, I've read the thread from January with your assessments of the different cities. Since money is going to be tight before the move, it's not likely that my wife and I will be able to see our options in person until maybe a day or two before the move. Does anyone know of a realtor or house-hunter who would be able to help us find a place to rent in Chapel Hill/Carrboro (or maybe even Durham)? We'd like to be fairly close to UNC or near a bus line so we can manage with just one car. Thanks! Sean Cribbs From ben at near-time.com Wed Apr 30 11:18:07 2008 From: ben at near-time.com (Ben Burdick) Date: Wed, 30 Apr 2008 11:18:07 -0400 Subject: [raleigh.rb] Moving to the area In-Reply-To: <48187FF4.8010003@gmail.com> References: <48187FF4.8010003@gmail.com> Message-ID: <15346C80-6C72-41B3-AC58-3C46BBDF562F@near-time.com> Might want to check out http://www.chapelhillrent.com/ - thats how I have generally found apartments in the area. Also - Chapel Hill/Carrboro > Durham IMO ;) On Apr 30, 2008, at 10:19 AM, Sean Cribbs wrote: > Rubyists, > > It looks like I'll be moving the RDU-CH area come July, as my wife > will be going to UNC in the fall. Thanks to Adam Williams, I've > read the thread from January with your assessments of the different > cities. Since money is going to be tight before the move, it's not > likely that my wife and I will be able to see our options in person > until maybe a day or two before the move. Does anyone know of a > realtor or house-hunter who would be able to help us find a place to > rent in Chapel Hill/Carrboro (or maybe even Durham)? We'd like to > be fairly close to UNC or near a bus line so we can manage with just > one car. > > Thanks! > > Sean Cribbs > _______________________________________________ > raleigh-rb-members mailing list > raleigh-rb-members at rubyforge.org > http://rubyforge.org/mailman/listinfo/raleigh-rb-members From rhooker at cisco.com Wed Apr 30 11:39:36 2008 From: rhooker at cisco.com (Raymond Hooker (rhooker)) Date: Wed, 30 Apr 2008 11:39:36 -0400 Subject: [raleigh.rb] Moving to the area In-Reply-To: <48187FF4.8010003@gmail.com> References: <48187FF4.8010003@gmail.com> Message-ID: <848DB51F3E58E146B08D188F803DCB6F067999BE@xmb-rtp-20c.amer.cisco.com> If you are renting an apartment, then it may not make as much of a difference but for a home, being slightly over the Durham County line can be very helpful in terms of cost. I can certainly recommend Jaime McFarlane at: Jamie McFarlane Cell: (919) 360-6369 Fax: (919)313-9832 McFarlaneJ at hpw.com Good luck in your search. Ray -----Original Message----- From: raleigh-rb-members-bounces at rubyforge.org [mailto:raleigh-rb-members-bounces at rubyforge.org] On Behalf Of Sean Cribbs Sent: Wednesday, April 30, 2008 10:20 AM To: raleigh-rb-members at rubyforge.org Subject: [raleigh.rb] Moving to the area Rubyists, It looks like I'll be moving the RDU-CH area come July, as my wife will be going to UNC in the fall. Thanks to Adam Williams, I've read the thread from January with your assessments of the different cities. Since money is going to be tight before the move, it's not likely that my wife and I will be able to see our options in person until maybe a day or two before the move. Does anyone know of a realtor or house-hunter who would be able to help us find a place to rent in Chapel Hill/Carrboro (or maybe even Durham)? We'd like to be fairly close to UNC or near a bus line so we can manage with just one car. Thanks! Sean Cribbs _______________________________________________ raleigh-rb-members mailing list raleigh-rb-members at rubyforge.org http://rubyforge.org/mailman/listinfo/raleigh-rb-members