From jpshack at gmail.com Sat Oct 1 18:03:45 2005 From: jpshack at gmail.com (John-Mason P. Shackelford) Date: Sat, 1 Oct 2005 17:03:45 -0500 Subject: [Rubygems-developers] ri support w/ gem rdoc install In-Reply-To: <1888.12.223.42.166.1128101805.squirrel@12.223.42.166> References: <83f770ff0509300923p272c2634w39ba2764191968b0@mail.gmail.com> <9e7db9110509300926j44cc1613j5969b939749b55a8@mail.gmail.com> <1888.12.223.42.166.1128101805.squirrel@12.223.42.166> Message-ID: <83f770ff0510011503r7a06124dxe7082e7364344bd9@mail.gmail.com> Since ri already loads data from multiple paths why wouldn't we implement this by: * Extending ri's path handling to search the gems directories for rdoc files. * Adding an option to rdoc to generate ri docs for a particular gem This would allow us to remove the docs when a gem is removed. Is the major concern the performance hit of searching though all the gem directories for the ri doc files? If so, couldn't this be addressed with a persisted index? The index could be updated when a gem is installed or removed so that it would have little performance impact on ri itself. If these are ignorant questions, forgive me. I have only just started looking at the rdoc/ri source. John-Mason Shackelford Software Developer Pearson Educational Measurement 2510 North Dodge St. Iowa City, IA 52245 ph. 319-354-9200x6214 john-mason.shackelford at pearson.com http://pearsonedmeasurement.com From halostatue at gmail.com Sat Oct 1 19:19:26 2005 From: halostatue at gmail.com (Austin Ziegler) Date: Sat, 1 Oct 2005 19:19:26 -0400 Subject: [Rubygems-developers] Packaging BOF on Friday the 14th? Message-ID: <9e7db9110510011619o5798cdeam2dfade82fd5e2a6d@mail.gmail.com> (Crossposted to both ruby-core and rubygems-developers for the benefit of both sets of subscribers.) It seems to me that a lot of the nonsense surrounding RubyGems may be best solved in an in-person meeting, preferably with Matz involved. Shall we plan on having a BOF session about packaging and RubyGems at RubyConf on the Friday evening? If you have any concerns that you would like to see specifically discussed and you are not going to be *at* RubyConf, might I suggest adding something to the RubyGarden webpage on the RubyGems to do list? -austin -- Austin Ziegler * halostatue at gmail.com * Alternate: austin at halostatue.ca From dblack at wobblini.net Sat Oct 1 19:49:05 2005 From: dblack at wobblini.net (David A. Black) Date: Sat, 1 Oct 2005 16:49:05 -0700 (PDT) Subject: [Rubygems-developers] Packaging BOF on Friday the 14th? In-Reply-To: <9e7db9110510011619o5798cdeam2dfade82fd5e2a6d@mail.gmail.com> References: <9e7db9110510011619o5798cdeam2dfade82fd5e2a6d@mail.gmail.com> Message-ID: Hi -- On Sat, 1 Oct 2005, Austin Ziegler wrote: > (Crossposted to both ruby-core and rubygems-developers for the benefit > of both sets of subscribers.) > > It seems to me that a lot of the nonsense surrounding RubyGems may be > best solved in an in-person meeting, preferably with Matz involved. > Shall we plan on having a BOF session about packaging and RubyGems at > RubyConf on the Friday evening? > > If you have any concerns that you would like to see specifically > discussed and you are not going to be *at* RubyConf, might I suggest > adding something to the RubyGarden webpage on the RubyGems to do list? We've already got a roundtable with Matz scheduled for that evening. Packaging things can certainly be brought up at the roundtable but I'm not in favor of pre-empting other topics or meeting as a small group. Remember, there will be 200 people there, all equally eager and entitled to spend some time talking with Matz :-) David -- David A. Black dblack at wobblini.net From eivind at FreeBSD.ORG Sun Oct 2 07:26:30 2005 From: eivind at FreeBSD.ORG (Eivind Eklund) Date: Sun, 2 Oct 2005 11:26:30 +0000 Subject: [Rubygems-developers] Packaging BOF on Friday the 14th? In-Reply-To: <9e7db9110510011619o5798cdeam2dfade82fd5e2a6d@mail.gmail.com> References: <9e7db9110510011619o5798cdeam2dfade82fd5e2a6d@mail.gmail.com> Message-ID: <20051002112630.GF39400@FreeBSD.org> On Sat, Oct 01, 2005 at 07:19:26PM -0400, Austin Ziegler wrote: > (Crossposted to both ruby-core and rubygems-developers for the benefit > of both sets of subscribers.) > > It seems to me that a lot of the nonsense surrounding RubyGems may be > best solved in an in-person meeting, preferably with Matz involved. > Shall we plan on having a BOF session about packaging and RubyGems at > RubyConf on the Friday evening? To make the work, I think it's very very important to make sure that there is somebody that can relay knowledge of repackager policies etc present. Is there anybody that can take that role reading? I won't be able to attend, otherwise I'd be glad to do it myself, of course. Eivind. From dblack at wobblini.net Sun Oct 2 08:28:48 2005 From: dblack at wobblini.net (David A. Black) Date: Sun, 2 Oct 2005 05:28:48 -0700 (PDT) Subject: [Rubygems-developers] Packaging BOF on Friday the 14th? In-Reply-To: <20051002112630.GF39400@FreeBSD.org> References: <9e7db9110510011619o5798cdeam2dfade82fd5e2a6d@mail.gmail.com> <20051002112630.GF39400@FreeBSD.org> Message-ID: Hi -- On Sun, 2 Oct 2005, Eivind Eklund wrote: > On Sat, Oct 01, 2005 at 07:19:26PM -0400, Austin Ziegler wrote: >> (Crossposted to both ruby-core and rubygems-developers for the benefit >> of both sets of subscribers.) >> >> It seems to me that a lot of the nonsense surrounding RubyGems may be >> best solved in an in-person meeting, preferably with Matz involved. >> Shall we plan on having a BOF session about packaging and RubyGems at >> RubyConf on the Friday evening? > > To make the work, I think it's very very important to make sure > that there is somebody that can relay knowledge of repackager > policies etc present. Is there anybody that can take that role > reading? Please remember that Friday evening (up until about 9:00) is already scheduled, with an open-ended roundtable discussion with Matz. This is an opportunity for everyone at the conference to ask Matz questions about a variety of topics, including (but not limited to) RubyGems. David -- David A. Black dblack at wobblini.net From dblack at wobblini.net Sun Oct 2 08:48:29 2005 From: dblack at wobblini.net (David A. Black) Date: Sun, 2 Oct 2005 05:48:29 -0700 (PDT) Subject: [Rubygems-developers] Packaging BOF on Friday the 14th? In-Reply-To: <20051002112630.GF39400@FreeBSD.org> References: <9e7db9110510011619o5798cdeam2dfade82fd5e2a6d@mail.gmail.com> <20051002112630.GF39400@FreeBSD.org> Message-ID: [I seem to be screwing up the cross-posting. Sorry if duplicated.] Hi -- On Sun, 2 Oct 2005, Eivind Eklund wrote: > On Sat, Oct 01, 2005 at 07:19:26PM -0400, Austin Ziegler wrote: >> (Crossposted to both ruby-core and rubygems-developers for the benefit >> of both sets of subscribers.) >> >> It seems to me that a lot of the nonsense surrounding RubyGems may be >> best solved in an in-person meeting, preferably with Matz involved. >> Shall we plan on having a BOF session about packaging and RubyGems at >> RubyConf on the Friday evening? > > To make the work, I think it's very very important to make sure > that there is somebody that can relay knowledge of repackager > policies etc present. Is there anybody that can take that role > reading? Please remember that Friday evening (up until about 9:00) is already scheduled, with an open-ended roundtable discussion with Matz. This is an opportunity for everyone at the conference to ask Matz questions about a variety of topics, including (but not limited to) RubyGems. David -- David A. Black dblack at wobblini.net From chad at chadfowler.com Sun Oct 2 09:26:01 2005 From: chad at chadfowler.com (Chad Fowler) Date: Sun, 2 Oct 2005 09:26:01 -0400 Subject: [Rubygems-developers] ri support w/ gem rdoc install In-Reply-To: <83f770ff0510011503r7a06124dxe7082e7364344bd9@mail.gmail.com> References: <83f770ff0509300923p272c2634w39ba2764191968b0@mail.gmail.com> <9e7db9110509300926j44cc1613j5969b939749b55a8@mail.gmail.com> <1888.12.223.42.166.1128101805.squirrel@12.223.42.166> <83f770ff0510011503r7a06124dxe7082e7364344bd9@mail.gmail.com> Message-ID: On 01-Oct-05, at 6:03 PM, John-Mason P. Shackelford wrote: > Since ri already loads data from multiple paths why wouldn't we > implement this by: > > * Extending ri's path handling to search the gems directories for > rdoc files. > * Adding an option to rdoc to generate ri docs for a particular gem > > This would allow us to remove the docs when a gem is removed. > > Is the major concern the performance hit of searching though all the > gem directories for the ri doc files? If so, couldn't this be > addressed with a persisted index? The index could be updated when a > gem is installed or removed so that it would have little performance > impact on ri itself. > > If these are ignorant questions, forgive me. I have only just started > looking at the rdoc/ri source. > > Hi! I think it would be worth doing some performance testing, but I seem to remember Dave Thomas being concerned about performance when this came up previously. It would be wonderful to be able install/ uninstall ri data for gems. Chad From chad at chadfowler.com Sun Oct 2 14:17:14 2005 From: chad at chadfowler.com (Chad Fowler) Date: Sun, 2 Oct 2005 14:17:14 -0400 Subject: [Rubygems-developers] Packaging BOF on Friday the 14th? In-Reply-To: <9e7db9110510011619o5798cdeam2dfade82fd5e2a6d@mail.gmail.com> References: <9e7db9110510011619o5798cdeam2dfade82fd5e2a6d@mail.gmail.com> Message-ID: <51BC36E1-03C1-48CA-9D66-F452436C76C6@chadfowler.com> On 01-Oct-05, at 7:19 PM, Austin Ziegler wrote: > (Crossposted to both ruby-core and rubygems-developers for the benefit > of both sets of subscribers.) > > It seems to me that a lot of the nonsense surrounding RubyGems may be > best solved in an in-person meeting, preferably with Matz involved. > Shall we plan on having a BOF session about packaging and RubyGems at > RubyConf on the Friday evening? > > If you have any concerns that you would like to see specifically > discussed and you are not going to be *at* RubyConf, might I suggest > adding something to the RubyGarden webpage on the RubyGems to do list? > I agree that it would be nice to do, but as David pointed out, this would conflict with the already-arranged stuff for Friday night. A bigger problem for me is that I imagine I'll have somewhere close to zero free moments at the conference, so it might be hard for me to get away for this. Maybe we could do it Thursday night depending on when people are coming in? Chad From halostatue at gmail.com Sun Oct 2 14:20:28 2005 From: halostatue at gmail.com (Austin Ziegler) Date: Sun, 2 Oct 2005 14:20:28 -0400 Subject: [Rubygems-developers] Packaging BOF on Friday the 14th? In-Reply-To: <51BC36E1-03C1-48CA-9D66-F452436C76C6@chadfowler.com> References: <9e7db9110510011619o5798cdeam2dfade82fd5e2a6d@mail.gmail.com> <51BC36E1-03C1-48CA-9D66-F452436C76C6@chadfowler.com> Message-ID: <9e7db9110510021120rc1c067fs8f3bb944d6156a66@mail.gmail.com> On 10/2/05, Chad Fowler wrote: > On 01-Oct-05, at 7:19 PM, Austin Ziegler wrote: > > (Crossposted to both ruby-core and rubygems-developers for the benefit > > of both sets of subscribers.) > > > > It seems to me that a lot of the nonsense surrounding RubyGems may be > > best solved in an in-person meeting, preferably with Matz involved. > > Shall we plan on having a BOF session about packaging and RubyGems at > > RubyConf on the Friday evening? > > > > If you have any concerns that you would like to see specifically > > discussed and you are not going to be *at* RubyConf, might I suggest > > adding something to the RubyGarden webpage on the RubyGems to do list? > I agree that it would be nice to do, but as David pointed out, this > would conflict with the already-arranged > stuff for Friday night. A bigger problem for me is that I imagine > I'll have somewhere close to zero free moments at the conference, > so it might be hard for me to get away for this. > > Maybe we could do it Thursday night depending on when people are > coming in? I'm in around 9:30pm, but I should be good until midnight ;) -austin -- Austin Ziegler * halostatue at gmail.com * Alternate: austin at halostatue.ca From wekempf at gmail.com Fri Oct 7 11:35:37 2005 From: wekempf at gmail.com (Bill Kempf) Date: Fri, 7 Oct 2005 10:35:37 -0500 Subject: [Rubygems-developers] buffer errors Message-ID: <2090b5ac0510070835l4ef04742r79076a20ab5620f0@mail.gmail.com> This is driving me nuts. I've created some utility methods and I'm packaging them up for ease of distribution to multiple computers. Everything worked great for quite some time. Then I updated to rubygems 0.8.11. Ever since I've been having issues with "buffer errors". gem build project.gemspec works fine. gem inst project would fail with: Attempting local installation of 'project' ERROR: Error installing gem project[.gem]: buffer error Interestingly, random changes to the source code of any scripts within the project can cause everything to work flawlessly. I've Googled myself to death on this one, and the closest thing I could find was a thread about a similar issue with installing RMagic. The suggested fix was to modify package.rb changing "if Zlib::ZLIB_VERSION < '1.2.1'" to "<=". Running with --debug reveals that this change causes the location of the exception to change from line 615 to 611, but the exception is the same. Any idea what would be causing this frustrating issue and how to fix it? From chad at chadfowler.com Fri Oct 7 11:51:25 2005 From: chad at chadfowler.com (Chad Fowler) Date: Fri, 7 Oct 2005 11:51:25 -0400 Subject: [Rubygems-developers] buffer errors In-Reply-To: <2090b5ac0510070835l4ef04742r79076a20ab5620f0@mail.gmail.com> References: <2090b5ac0510070835l4ef04742r79076a20ab5620f0@mail.gmail.com> Message-ID: <034E435C-5286-4C49-8689-635A0703F59F@chadfowler.com> On 07-Oct-05, at 11:35 AM, Bill Kempf wrote: > This is driving me nuts. I've created some utility methods and I'm > packaging them up for ease of distribution to multiple computers. > Everything worked great for quite some time. Then I updated to > rubygems 0.8.11. Ever since I've been having issues with "buffer > errors". > > gem build project.gemspec works fine. gem inst project would fail > with: > > Attempting local installation of 'project' > ERROR: Error installing gem project[.gem]: buffer error > > Interestingly, random changes to the source code of any scripts within > the project can cause everything to work flawlessly. > > I've Googled myself to death on this one, and the closest thing I > could find was a thread about a similar issue with installing RMagic. > The suggested fix was to modify package.rb changing "if > Zlib::ZLIB_VERSION < '1.2.1'" to "<=". Running with --debug reveals > that this change causes the location of the exception to change from > line 615 to 611, but the exception is the same. > > Any idea what would be causing this frustrating issue and how to > fix it? > Bill, sorry you're being driven crazy with this one. Let's see if we can figure it out. Can you show us your "ruby -v" output? Also, you can run: gem inst project --backtrace to see more information. Thanks, Chad -------------- next part -------------- An HTML attachment was scrubbed... URL: http://rubyforge.org/pipermail/rubygems-developers/attachments/20051007/8f4ad864/attachment.htm From wekempf at gmail.com Fri Oct 7 14:55:09 2005 From: wekempf at gmail.com (Bill Kempf) Date: Fri, 7 Oct 2005 13:55:09 -0500 Subject: [Rubygems-developers] buffer errors In-Reply-To: <034E435C-5286-4C49-8689-635A0703F59F@chadfowler.com> References: <2090b5ac0510070835l4ef04742r79076a20ab5620f0@mail.gmail.com> <034E435C-5286-4C49-8689-635A0703F59F@chadfowler.com> Message-ID: <2090b5ac0510071155s120d5368nf2f0d822ba6b758d@mail.gmail.com> On 10/7/05, Chad Fowler wrote: > Bill, sorry you're being driven crazy with this one. Let's see if we can > figure it out. > > Can you show us your "ruby -v" output? Also, you can run: > > gem inst project --backtrace > > to see more information. > > Thanks, > Chad I'd love to... but after struggling with this for an entire morning, it's stopped doing it. I have no idea why, but until it starts happening again, I obviously can't diagnose the issue any further :(. From chad at chadfowler.com Fri Oct 7 15:25:18 2005 From: chad at chadfowler.com (Chad Fowler) Date: Fri, 7 Oct 2005 15:25:18 -0400 Subject: [Rubygems-developers] buffer errors In-Reply-To: <2090b5ac0510071155s120d5368nf2f0d822ba6b758d@mail.gmail.com> References: <2090b5ac0510070835l4ef04742r79076a20ab5620f0@mail.gmail.com> <034E435C-5286-4C49-8689-635A0703F59F@chadfowler.com> <2090b5ac0510071155s120d5368nf2f0d822ba6b758d@mail.gmail.com> Message-ID: <0DD14AF7-9CF6-485D-AE91-C169B4932CEC@chadfowler.com> On 07-Oct-05, at 2:55 PM, Bill Kempf wrote: > On 10/7/05, Chad Fowler wrote: > >> Bill, sorry you're being driven crazy with this one. Let's see if >> we can >> figure it out. >> >> Can you show us your "ruby -v" output? Also, you can run: >> >> gem inst project --backtrace >> >> to see more information. >> >> Thanks, >> Chad >> > > I'd love to... but after struggling with this for an entire morning, > it's stopped doing it. I have no idea why, but until it starts > happening again, I obviously can't diagnose the issue any further :(. > For future reference, can you give us your "ruby -v"? Thanks, Chad From wekempf at gmail.com Fri Oct 7 15:26:10 2005 From: wekempf at gmail.com (Bill Kempf) Date: Fri, 7 Oct 2005 14:26:10 -0500 Subject: [Rubygems-developers] buffer errors In-Reply-To: <0DD14AF7-9CF6-485D-AE91-C169B4932CEC@chadfowler.com> References: <2090b5ac0510070835l4ef04742r79076a20ab5620f0@mail.gmail.com> <034E435C-5286-4C49-8689-635A0703F59F@chadfowler.com> <2090b5ac0510071155s120d5368nf2f0d822ba6b758d@mail.gmail.com> <0DD14AF7-9CF6-485D-AE91-C169B4932CEC@chadfowler.com> Message-ID: <2090b5ac0510071226s46fb158u76c7aa0f180b0e6@mail.gmail.com> ruby 1.8.2 (2004-12-25) [i386-mswin32] On 10/7/05, Chad Fowler wrote: > > > On 07-Oct-05, at 2:55 PM, Bill Kempf wrote: > > > On 10/7/05, Chad Fowler wrote: > > > >> Bill, sorry you're being driven crazy with this one. Let's see if > >> we can > >> figure it out. > >> > >> Can you show us your "ruby -v" output? Also, you can run: > >> > >> gem inst project --backtrace > >> > >> to see more information. > >> > >> Thanks, > >> Chad > >> > > > > I'd love to... but after struggling with this for an entire morning, > > it's stopped doing it. I have no idea why, but until it starts > > happening again, I obviously can't diagnose the issue any further :(. > > > > For future reference, can you give us your "ruby -v"? > > Thanks, > Chad > -------------- next part -------------- An HTML attachment was scrubbed... URL: http://rubyforge.org/pipermail/rubygems-developers/attachments/20051007/fd6686ab/attachment.htm From lumbergh at gmail.com Sat Oct 15 16:50:43 2005 From: lumbergh at gmail.com (Peter Marreck) Date: Sat, 15 Oct 2005 16:50:43 -0400 Subject: [Rubygems-developers] Question about issue w/unfound header file Message-ID: Hi all, I'm trying to set up an install of ruby/rails/gems on a new box (haven't done a fresh install in a while) and I'm running into an issue I don't know how to get around. I'm sure a gem developer would have a quick answer though. Rails installs fine. I installed sqlite via DarwinPorts which by default goes into /opt/local/. but when I do a "gem install sqlite" I get the following error: Attempting local installation of 'sqlite' Local gem file not found: sqlite*.gem Attempting remote installation of 'sqlite' Updating Gem source index for: http://gems.rubyforge.org Building native extensions. This could take a while... ERROR: While executing gem ... (RuntimeError) ERROR: Failed to build gem native extension. Gem files will remain installed in /usr/lib/ruby/gems/1.8/gems/sqlite-2.0.1 for inspection. ruby extconf.rb install sqlite\nchecking for main() in -lsqlite... no checking for sqlite.h... no I know where sqlite.h is (where DarwinPorts put it) but how do I tell the gem installer where to look to find it? Thanks for any assistance!! Regards, Peter From halostatue at gmail.com Mon Oct 17 17:30:25 2005 From: halostatue at gmail.com (Austin Ziegler) Date: Mon, 17 Oct 2005 17:30:25 -0400 Subject: [Rubygems-developers] gem server horked? In-Reply-To: <1129582071.16428.11.camel@hal> References: <8D6CF5C2-4264-432D-85FC-09DC7EBF7545@grayproductions.net> <1129582071.16428.11.camel@hal> Message-ID: <9e7db9110510171430x20f9702eufc08731607223ed0@mail.gmail.com> On 10/17/05, Tom Copeland wrote: > On Tue, 2005-10-18 at 05:36 +0900, Jeff Wood wrote: > > Has anybody that's part of that project caught this yet ??? I'm surprised > > more people aren't yelling ... I've got a bunch of Ruby-Nubies that I'm > > trying to get setup, and this is a pretty bad first experience. > > I know the discussion about RubyGems being part of the standard distro is > > still open ... but, it's just a black eye to have something "standard" be > > down like this. > > I may be interested in mirroring them ( at least for me and the team I work > > with ) ... is there some mechanism for me to do that? > > I need a working solution soon, if theres anything I can do to help, please > > let me know. > > The problem is that RubyForge is on a relatively small network > connection - 768K. Mirroring out large files helps tremendously, but > there's still a ton of content being served through that pipe. What would it take for RubyGems to recognise a 302-redirect and be mirrored the same way that everything else is on RubyForge, leaving the real bandwidth of RubyForge to the website? -austin -- Austin Ziegler * halostatue at gmail.com * Alternate: austin at halostatue.ca From tom at infoether.com Mon Oct 17 17:35:04 2005 From: tom at infoether.com (Tom Copeland) Date: Mon, 17 Oct 2005 17:35:04 -0400 Subject: [Rubygems-developers] gem server horked? In-Reply-To: <9e7db9110510171430x20f9702eufc08731607223ed0@mail.gmail.com> References: <8D6CF5C2-4264-432D-85FC-09DC7EBF7545@grayproductions.net> <1129582071.16428.11.camel@hal> <9e7db9110510171430x20f9702eufc08731607223ed0@mail.gmail.com> Message-ID: <1129584904.16428.36.camel@hal> On Mon, 2005-10-17 at 17:30 -0400, Austin Ziegler wrote: > On 10/17/05, Tom Copeland wrote: > > The problem is that RubyForge is on a relatively small network > > connection - 768K. Mirroring out large files helps tremendously, but > > there's still a ton of content being served through that pipe. > > What would it take for RubyGems to recognise a 302-redirect and be > mirrored the same way that everything else is on RubyForge, leaving > the real bandwidth of RubyForge to the website? * tom defers to Gems fellas... From chad at chadfowler.com Mon Oct 17 17:50:03 2005 From: chad at chadfowler.com (Chad Fowler) Date: Mon, 17 Oct 2005 17:50:03 -0400 Subject: [Rubygems-developers] gem server horked? In-Reply-To: <1129584904.16428.36.camel@hal> References: <8D6CF5C2-4264-432D-85FC-09DC7EBF7545@grayproductions.net> <1129582071.16428.11.camel@hal> <9e7db9110510171430x20f9702eufc08731607223ed0@mail.gmail.com> <1129584904.16428.36.camel@hal> Message-ID: <6DDB784A-8771-4269-AE73-EC868AE95FE0@chadfowler.com> On 17-Oct-05, at 5:35 PM, Tom Copeland wrote: > On Mon, 2005-10-17 at 17:30 -0400, Austin Ziegler wrote: > >> On 10/17/05, Tom Copeland wrote: >> >>> The problem is that RubyForge is on a relatively small network >>> connection - 768K. Mirroring out large files helps tremendously, >>> but >>> there's still a ton of content being served through that pipe. >>> >> >> What would it take for RubyGems to recognise a 302-redirect and be >> mirrored the same way that everything else is on RubyForge, leaving >> the real bandwidth of RubyForge to the website? >> > > * tom defers to Gems fellas... > We could actually do it with the builtin "sources" feature. We just need to get round-robining thrown in and do a release. Chad -------------- next part -------------- An HTML attachment was scrubbed... URL: http://rubyforge.org/pipermail/rubygems-developers/attachments/20051017/462a348b/attachment.htm From halostatue at gmail.com Mon Oct 17 18:06:10 2005 From: halostatue at gmail.com (Austin Ziegler) Date: Mon, 17 Oct 2005 18:06:10 -0400 Subject: [Rubygems-developers] gem server horked? In-Reply-To: <6DDB784A-8771-4269-AE73-EC868AE95FE0@chadfowler.com> References: <8D6CF5C2-4264-432D-85FC-09DC7EBF7545@grayproductions.net> <1129582071.16428.11.camel@hal> <9e7db9110510171430x20f9702eufc08731607223ed0@mail.gmail.com> <1129584904.16428.36.camel@hal> <6DDB784A-8771-4269-AE73-EC868AE95FE0@chadfowler.com> Message-ID: <9e7db9110510171506g53b7af35x2c8a7e747788bde1@mail.gmail.com> On 10/17/05, Chad Fowler wrote: > We could actually do it with the builtin "sources" feature. We just need to > get round-robining thrown in and do a release. Can that be done quickly? I know Jim is working toward fixing the size of the downloads, but I think that -- especially with the impending release of Rails 1 -- we need to probably distribute gem delivery as well as large file delivery. -austin -- Austin Ziegler * halostatue at gmail.com * Alternate: austin at halostatue.ca From Alun.Eyre at aemedia.com Fri Oct 21 11:52:32 2005 From: Alun.Eyre at aemedia.com (Alun.Eyre@aemedia.com) Date: Fri, 21 Oct 2005 16:52:32 +0100 Subject: [Rubygems-developers] autorequire functionality Message-ID: Hi, This is a bit of a newbie question (new to ruby and new to gems!), but having read through the comp.lang.ruby posts, and some of the digests of this group, I could not see a similiar question, so if there is a better place to ask, or a more appropriate group, please let me know, and I will ask there :-) I wanted to get a better understanding of the usage of the 'autorequire' specification, as we saw an issue in-house with some small gems we were playing with, summarised as: (gem x) has a dependancy on (gem f) which in turn expects a constant to have been defined within (gem a) before it runs. Now, the gem spec for (gem f) had an autorequire for itself (gem f) within it's gem spec, which caused the issue where the constant was not defined. Removing this line from the gem spec solved the issue. Thanks, Alun. ********************************************************************** Aegis Media, Parker Tower, 43-49 Parker Street, London, WC2B 5PS. Telephone: 020 7430 6300 Facsimile: 020 7430 6319 www.aemedia.com This e-mail contains information which is confidential and which may be legally privileged and protected from disclosure. If you are not the intended recipient of this e-mail or any parts of it, please telephone us immediately on receipt. You should not disclose the contents to any other person, nor take any copies. ********************************************************************** -------------- next part -------------- An HTML attachment was scrubbed... URL: http://rubyforge.org/pipermail/rubygems-developers/attachments/20051021/46ff6bd5/attachment-0001.htm From halostatue at gmail.com Fri Oct 21 12:08:06 2005 From: halostatue at gmail.com (Austin Ziegler) Date: Fri, 21 Oct 2005 12:08:06 -0400 Subject: [Rubygems-developers] autorequire functionality In-Reply-To: References: Message-ID: <9e7db9110510210908k26d76329k2cd51488064f071d@mail.gmail.com> On 10/21/05, Alun.Eyre at aemedia.com wrote: > I wanted to get a better understanding of the usage of the 'autorequire' > specification, The best usage, at this point, is not to use it. It will be disappearing from my gems on their next releases. -austin -- Austin Ziegler * halostatue at gmail.com * Alternate: austin at halostatue.ca From jim at weirichhouse.org Fri Oct 21 13:13:40 2005 From: jim at weirichhouse.org (Jim Weirich) Date: Fri, 21 Oct 2005 17:13:40 -0000 (UTC) Subject: [Rubygems-developers] autorequire functionality In-Reply-To: References: Message-ID: <47841.192.223.163.6.1129914820.squirrel@weirichhouse.org> Austin is correct that "autorequire" was a bad idea and is deprecated at the moment. However ... Alun.Eyre at aemedia.com said: > (gem x) has a dependancy on (gem f) which in turn expects a constant to > have been defined within (gem a) before it runs. If I understand this correctly, this sounds like a circular dependency. X depends upon F, which in turns depends upon X. I strongly recommend one of: (1) Move the constant in question to F, so it no longer has to depend upon X. This would break the dependency cycle. (2) Or, move the constant in question to gem C, and have both X and F depend upon C. Again, this breaks the cycle. (3) Or, merge X and F into a single gem XF, which breaks the cycle by the fact you now only have one gem. Dependency cycles at a package level is likely to cause other headaches down the line, and not only with RubyGems. -- -- Jim Weirich jim at weirichhouse.org http://onestepback.org ----------------------------------------------------------------- "Beware of bugs in the above code; I have only proved it correct, not tried it." -- Donald Knuth (in a memo to Peter van Emde Boas) From zstanaka at archer.livedoor.com Mon Oct 24 09:18:27 2005 From: zstanaka at archer.livedoor.com (TANAKA Shin-ya) Date: Mon, 24 Oct 2005 22:18:27 +0900 Subject: [Rubygems-developers] Problem while installing the Gems with tests from remote gem-server Message-ID: <20051024221827.97231822@archer.livedoor.com> Hi all, I got following error while installing the Gems with tests from remote gem-server. $ sudo gem install -t trie Attempting local installation of 'trie' Local gem file not found: trie*.gem Attempting remote installation of 'trie' Successfully installed trie-0.0.1 Installing RDoc documentation for trie-0.0.1... /usr/pkg/lib/ruby/site_ruby/1.8/rubygems.rb:204:in `report_activate_error': Could not find RubyGem trie (= 0.0.1) (Gem::LoadError) from /usr/pkg/lib/ruby/site_ruby/1.8/rubygems.rb:141:in `activate' from /usr/pkg/lib/ruby/site_ruby/1.8/rubygems.rb:37:in `require_gem_with_options' from /usr/pkg/lib/ruby/site_ruby/1.8/rubygems.rb:31:in `require_gem' from /usr/pkg/lib/ruby/site_ruby/1.8/rubygems/validator.rb:127:in `unit_test' from /usr/pkg/lib/ruby/site_ruby/1.8/rubygems/gem_commands.rb:222:in `execute' from /usr/pkg/lib/ruby/site_ruby/1.8/rubygems/gem_commands.rb:220:in `each' from /usr/pkg/lib/ruby/site_ruby/1.8/rubygems/gem_commands.rb:220:in `execute' from /usr/pkg/lib/ruby/site_ruby/1.8/rubygems/gem_commands.rb:153:in `each' from /usr/pkg/lib/ruby/site_ruby/1.8/rubygems/gem_commands.rb:153:in `execute' from /usr/pkg/lib/ruby/site_ruby/1.8/rubygems/command.rb:49:in `invoke' from /usr/pkg/lib/ruby/site_ruby/1.8/rubygems/cmd_manager.rb:94:in `process_args' from /usr/pkg/lib/ruby/site_ruby/1.8/rubygems/cmd_manager.rb:67:in `run' from /usr/pkg/lib/ruby/site_ruby/1.8/rubygems/gem_runner.rb:13:in `run' from /usr/pkg/bin/gem:17 In this case, installed gem isn't registered to Gem#source_index. Gem#source_index is cached at RemoteInstaller#install(rubygems/remote_installer.rb:399). To resolve this error, apply following patch. diff -ur rubygems/gem_commands.rb.orig rubygems/gem_commands.rb --- 1.8.orig/rubygems/gem_commands.rb 2005-10-19 05:50:56.000000000 +0900 +++ 1.8/rubygems/gem_commands.rb 2005-10-24 22:02:32.000000000 +0900 @@ -217,6 +217,7 @@ # TODO: catch exceptions and inform user that doc generation was not successful. end if options[:test] + Gem.clear_source_index installed_gems.each do |spec| gem_spec = Gem::SourceIndex.from_installed_gems.search(spec.name, spec.version.version).first result = Gem::Validator.new.unit_test(gem_spec) diff -ur rubygems.rb.orig rubygems.rb --- 1.8.orig/rubygems.rb 2005-10-19 05:50:56.000000000 +0900 +++ 1.8/rubygems.rb 2005-10-24 22:02:32.000000000 +0900 @@ -214,6 +214,12 @@ @gem_path = nil @@source_index = nil end + + # Clear source_index cache. + # + def clear_source_index + @@source_index = nil + end # Use the +home+ and (optional) +paths+ values for +dir+ and +path+. # Used mainly by the unit tests to provide environment isolation. I think it's ugly patch. Please suggest better solutions. Regards, TANAKA Shin-ya ----------------------------------------------- ?????????????????????????? livedoor ?? http://present.livedoor.com/ From hgs at dmu.ac.uk Fri Oct 28 14:09:04 2005 From: hgs at dmu.ac.uk (Hugh Sasse) Date: Fri, 28 Oct 2005 19:09:04 +0100 (WEST) Subject: [Rubygems-developers] Cache files etc. Message-ID: As far as I can see from the code (and there's a lot going on in this area with different cache locations etc) the cache just holds the yaml from yaml.Z and holds no metadata about it. It checks to see if it has the current version by comparing sizes with the one at gems.rubyforge.org. Am I right in thinking that the cache is one flat file? I'm wondering if it would spoil anything to have a metafile along with it. I see that the server which gives us the yaml.Z file is presently running Apache 1.x.y, and is thus limited in some respects as to how we can modify it (in comparison to what we could do to a pure ruby implemented server, I mean). But it is sending out Etags with the Yaml file, which means we should be able to avoid simple size comparisons, and actually get the server to send us the file iff it is modified. Use the GET with the "If-Modified-Since: #{etag}" header. But, also, should the server ever support it we would be able to use RFC3229 with this, provided we kept the etag somewhere, and provided the server had enough space to keep N back issues of the yaml.Z file and, either store the differences, or if it has enoguh power, generate the differences on demand. But none of this can happen without storing metadata in the cache. Anyone more familiar with the code (that's most of you :-)) care to comment? Thank you Hugh From tilman at code-monkey.de Mon Oct 31 13:59:38 2005 From: tilman at code-monkey.de (Tilman Sauerbeck) Date: Mon, 31 Oct 2005 19:59:38 +0100 Subject: [Rubygems-developers] [PATCH] Support extensions built with rake In-Reply-To: <20050925081121.GA5808@code-monkey.de> References: <20050812135812.GA5828@code-monkey.de> <3FF6741F-6FD5-4B26-B0E6-622CF62F772E@chadfowler.com> <20050830215620.22432409.tilman@code-monkey.de> <20050830230621.4fe3add9.tilman@code-monkey.de> <20050907211432.2b2007fd.tilman@code-monkey.de> <20050925081121.GA5808@code-monkey.de> Message-ID: <20051031185917.GA5631@code-monkey.de> Tilman Sauerbeck [2005-09-25 10:11]: > [patch] I know it's only ~5 seeks since I sent the last patch in but it would be really awesome if I could get some kind of feedback on this. Of course if that's a problem I'll happily wait a few more months. Regards, Tilman -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: not available Url : http://rubyforge.org/pipermail/rubygems-developers/attachments/20051031/98e68c55/attachment.bin From chad at chadfowler.com Mon Oct 31 14:09:52 2005 From: chad at chadfowler.com (Chad Fowler) Date: Mon, 31 Oct 2005 14:09:52 -0500 Subject: [Rubygems-developers] [PATCH] Support extensions built with rake In-Reply-To: <20051031185917.GA5631@code-monkey.de> References: <20050812135812.GA5828@code-monkey.de> <3FF6741F-6FD5-4B26-B0E6-622CF62F772E@chadfowler.com> <20050830215620.22432409.tilman@code-monkey.de> <20050830230621.4fe3add9.tilman@code-monkey.de> <20050907211432.2b2007fd.tilman@code-monkey.de> <20050925081121.GA5808@code-monkey.de> <20051031185917.GA5631@code-monkey.de> Message-ID: On 31-Oct-05, at 1:59 PM, Tilman Sauerbeck wrote: > Tilman Sauerbeck [2005-09-25 10:11]: > > >> [patch] >> > > I know it's only ~5 seeks since I sent the last patch in but it would > be really awesome if I could get some kind of feedback on this. > Of course if that's a problem I'll happily wait a few more months. > > Tilman, SORRY! I actually though I had already applied this! Busy with RubyConf etc. My apologies. I've just applied it. Thanks for this! Works great! Chad From tilman at code-monkey.de Mon Oct 31 15:20:57 2005 From: tilman at code-monkey.de (Tilman Sauerbeck) Date: Mon, 31 Oct 2005 21:20:57 +0100 Subject: [Rubygems-developers] [PATCH] Support extensions built with rake In-Reply-To: References: <20050812135812.GA5828@code-monkey.de> <3FF6741F-6FD5-4B26-B0E6-622CF62F772E@chadfowler.com> <20050830215620.22432409.tilman@code-monkey.de> <20050830230621.4fe3add9.tilman@code-monkey.de> <20050907211432.2b2007fd.tilman@code-monkey.de> <20050925081121.GA5808@code-monkey.de> <20051031185917.GA5631@code-monkey.de> Message-ID: <20051031202047.GA6023@code-monkey.de> Chad Fowler [2005-10-31 14:09]: > > Tilman Sauerbeck [2005-09-25 10:11]: > > > > > >> [patch] > >> > > > > I know it's only ~5 seeks since I sent the last patch in but it would > > be really awesome if I could get some kind of feedback on this. > > Of course if that's a problem I'll happily wait a few more months. > > > > > > Tilman, SORRY! I actually though I had already applied this! Busy > with RubyConf etc. My apologies. > > I've just applied it. Thanks for this! Works great! Cool, thank you :) Regards, Tilman -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: not available Url : http://rubyforge.org/pipermail/rubygems-developers/attachments/20051031/38573911/attachment.bin