10900 0 true 2008-09-16T11:55:15-04:00 30519 78 deprecate-should_be_restful 27964 5807 committed should_be_restful 2009-01-14T14:44:13-05:00 17641 Sean Hussey Dan Croak Dan Croak http://thoughtbot.lighthouseapp.com/projects/5807/tickets/78 should_be_restful is currently the worst portion of Shoulda and should be deprecated. Its problems: * It dissuades the programmer from good TDD practice. It's working at too high level of abstraction and encourages the programmer to take too many big steps. * It is an attempt to be like Rails scaffolding, which is a good way to learn REST, but the syntax is so cryptic it doesn't even accomplish that goal. * It's near impossible to find the line in your test file that is actually failing or erroring. * You have no idea how many tests it generates or how they're implemented, which lulls the programmer into a false sense of security. * It's slow. "If not now, when? If not us, who?" should_be_restful is currently the worst portion of Shoulda and should be deprecated. Its problems: * It dissuades the programmer from good TDD practice. It's working at too high level of abstraction and encourages the programmer to take too many big steps. * It is an attempt to be like Rails scaffolding, which is a good way to learn REST, but the syntax is so cryptic it doesn't even accomplish that goal. * It's near impossible to find the line in your test file that is actually failing or erroring. * You have no idea how many tests it generates or how they're implemented, which lulls the programmer into a false sense of security. * It's slow. "If not now, when? If not us, who?" <div><p>should_be_restful is currently the worst portion of Shoulda and should be deprecated. Its problems:</p> <ul> <li>It dissuades the programmer from good TDD practice. It's working at too high level of abstraction and encourages the programmer to take too many big steps.</li> <li>It is an attempt to be like Rails scaffolding, which is a good way to learn REST, but the syntax is so cryptic it doesn't even accomplish that goal.</li> <li>It's near impossible to find the line in your test file that is actually failing or erroring.</li> <li>You have no idea how many tests it generates or how they're implemented, which lulls the programmer into a false sense of security.</li> <li>It's slow.</li> </ul> <p>"If not now, when? If not us, who?"</p></div> 13443 0 should_be_restful is currently the worst portion of Shoulda and should be deprecated. Its problems: * It dissuades the programmer from good TDD practice. It's working at too high level of abstraction and encourages the programmer to take too many big steps. * It is an attempt to be like Rails scaffolding, which is a good way to learn REST, but the syntax is so cryptic it doesn't even accomplish that goal. * It's near impossible to find the line in your test file that is actually failing or erroring. * You have no idea how many tests it generates or how they're implemented, which lulls the programmer into a false sense of security. * It's slow. "If not now, when? If not us, who?" <div><p>should_be_restful is currently the worst portion of Shoulda and should be deprecated. Its problems:</p> <ul> <li>It dissuades the programmer from good TDD practice. It's working at too high level of abstraction and encourages the programmer to take too many big steps.</li> <li>It is an attempt to be like Rails scaffolding, which is a good way to learn REST, but the syntax is so cryptic it doesn't even accomplish that goal.</li> <li>It's near impossible to find the line in your test file that is actually failing or erroring.</li> <li>You have no idea how many tests it generates or how they're implemented, which lulls the programmer into a false sense of security.</li> <li>It's slow.</li> </ul> <p>"If not now, when? If not us, who?"</p></div> false 2008-09-16T11:55:15-04:00 30519 --- {} 78 deprecate-should_be_restful 0 5807 new should_be_restful 2008-09-16T11:55:20-04:00 30519 Dan Croak Dan Croak Tammer Saleh http://thoughtbot.lighthouseapp.com/projects/5807/tickets/78 13443 0 Other than the slow part, I agree completely. If we were going to deprecate it, how would we go about it? Should we just print a message when running that macro? Flag it in the docs? <div><p>Other than the slow part, I agree completely. If we were going to deprecate it, how would we go about it? Should we just print a message when running that macro? Flag it in the docs?</p></div> false 2008-09-16T12:00:07-04:00 30519 --- {} 78 deprecate-should_be_restful 0 5807 new should_be_restful 2008-09-16T12:00:10-04:00 13443 Tammer Saleh Dan Croak Tammer Saleh http://thoughtbot.lighthouseapp.com/projects/5807/tickets/78 13443 0 Both, I think, would be good. <div><p>Both, I think, would be good.</p></div> false 2008-09-16T12:01:57-04:00 30519 --- {} 78 deprecate-should_be_restful 0 5807 new should_be_restful 2008-09-16T12:02:02-04:00 10614 Jon Yurek Dan Croak Tammer Saleh http://thoughtbot.lighthouseapp.com/projects/5807/tickets/78 13443 0 Thank you for bringing this up. I've been trying to convince others to avoid using it, with no luck. Agreed, deprecate with a printed message and document. One concern: I think a lot of people are addicted to it. There will be push back. Does it make sense to start a shoulda-ext project for this kind of experimental stuff? <div><p>Thank you for bringing this up. I've been trying to convince others to avoid using it, with no luck.</p> <p>Agreed, deprecate with a printed message and document.</p> <p>One concern: I think a lot of people are addicted to it. There will be push back. Does it make sense to start a shoulda-ext project for this kind of experimental stuff?</p></div> false 2008-09-16T12:09:55-04:00 30519 --- {} 78 deprecate-should_be_restful 0 5807 new should_be_restful 2008-09-16T12:10:00-04:00 12160 Ryan McGeary Dan Croak Tammer Saleh http://thoughtbot.lighthouseapp.com/projects/5807/tickets/78 13443 0 I think that if there were enough demand for should_be_restful after it was removed, somebody would just move it to its own project. <div><p>I think that if there were enough demand for should_be_restful after it was removed, somebody would just move it to its own project.</p></div> false 2008-09-16T12:12:29-04:00 30519 --- {} 78 deprecate-should_be_restful 0 5807 new should_be_restful 2008-09-16T12:12:33-04:00 17223 Joe Ferris Dan Croak Tammer Saleh http://thoughtbot.lighthouseapp.com/projects/5807/tickets/78 10900 0 I have removed should_be_restful from the shoulda code base. The work was done in my remove_should_be_restful branch: http://github.com/dancroak/shoulda/tree/remove_should_be_restful Tammer - what do you mean by printing a message when running the macro or flagging it in the docs? If that is wanted, it still needs to be done. Josh is going to put should_be_restful in woulda: http://github.com/technicalpickles/woulda I suggest once he gets that together, we point the docs and message to his gem. Josh - take a look at these two commits to get a sense of what you'll need to put in woulda for should_be_restful: http://github.com/thoughtbot/shoulda/commit/10303b28f38c76c83bf99037eb9b2cb05ffec737 http://github.com/thoughtbot/shoulda/commit/f018d0844403ccdf4d89758966a2914b4d317c55 I am leaving this ticket open until should_be_restful finds a good new home and we update the docs. <div><p>I have removed should_be_restful from the shoulda code base. The work was done in my remove_should_be_restful branch:</p> <p><a href="http://github.com/dancroak/shoulda/tree/remove_should_be_restful">http://github.com/dancroak/shoul...</a></p> <p>Tammer - what do you mean by printing a message when running the macro or flagging it in the docs? If that is wanted, it still needs to be done.</p> <p>Josh is going to put should_be_restful in woulda:</p> <p><a href="http://github.com/technicalpickles/woulda">http://github.com/technicalpickl...</a></p> <p>I suggest once he gets that together, we point the docs and message to his gem.</p> <p>Josh - take a look at these two commits to get a sense of what you'll need to put in woulda for should_be_restful:</p> <p><a href="http://github.com/thoughtbot/shoulda/commit/10303b28f38c76c83bf99037eb9b2cb05ffec737"> http://github.com/thoughtbot/sho...</a></p> <p><a href="http://github.com/thoughtbot/shoulda/commit/f018d0844403ccdf4d89758966a2914b4d317c55"> http://github.com/thoughtbot/sho...</a></p> <p>I am leaving this ticket open until should_be_restful finds a good new home and we update the docs.</p></div> false 2008-09-20T19:13:08-04:00 30519 --- :assigned_user: 13443 78 deprecate-should_be_restful 0 5807 new should_be_restful 2008-09-20T19:13:11-04:00 10900 Dan Croak Dan Croak Dan Croak http://thoughtbot.lighthouseapp.com/projects/5807/tickets/78 10900 0 Dan, What happened to deprecating it first? I think we should first just add a `warn` message to the method explaining that it is deprecated and that it is replaced by woulda. The rdoc for the should_be_restful method should also explain that it's deprecated. I think that's what Tammer and Jon had in mind. Unfortunately, I think the current implementation should stick around through at least one minor release (2.1.x) of the shoulda gem so people will understand that it's deprecated. Removing it can happen in the next minor release. <div><p>Dan, What happened to deprecating it first?</p> <p>I think we should first just add a <code>warn</code> message to the method explaining that it is deprecated and that it is replaced by woulda. The rdoc for the should_be_restful method should also explain that it's deprecated. I think that's what Tammer and Jon had in mind.</p> <p>Unfortunately, I think the current implementation should stick around through at least one minor release (2.1.x) of the shoulda gem so people will understand that it's deprecated. Removing it can happen in the next minor release.</p></div> false 2008-09-20T19:25:13-04:00 30519 --- {} 78 deprecate-should_be_restful 0 5807 new should_be_restful 2008-09-20T19:25:15-04:00 12160 Ryan McGeary Dan Croak Dan Croak http://thoughtbot.lighthouseapp.com/projects/5807/tickets/78 10900 0 I reverted the commits that removed should_be_restful. Dan - We agreed to deprecate should_be_restful for a while before removing it. I think there's no harm in keeping it around for a few months as long as the documentation and printed warning make it clear that it's not considered best practice, and that it will be removed in a future version. <div><p>I reverted the commits that removed should_be_restful.</p> <p>Dan - We agreed to deprecate should_be_restful for a while before removing it. I think there's no harm in keeping it around for a few months as long as the documentation and printed warning make it clear that it's not considered best practice, and that it will be removed in a future version.</p></div> false 2008-09-20T22:10:43-04:00 30519 --- {} 78 deprecate-should_be_restful 0 5807 new should_be_restful 2008-09-20T22:10:45-04:00 13443 Tammer Saleh Dan Croak Dan Croak http://thoughtbot.lighthouseapp.com/projects/5807/tickets/78 10900 0 GitHub didn't like the gemspec versioning switcheroo. It kept serving v2.0.1 even after the revert. So, to fix: `should_be_restful` is now [deprecated with this commit](http://github.com/thoughtbot/shoulda/commit/8c081d7739e9abf69977f17cde4af814119482e3). v2.0.2 of the gem is now [published with this commit](http://github.com/thoughtbot/shoulda/commit/3b5f2642733d93e9d3219bfdffdb213f601a1be2). Note: Even though, the core shoulda tests now print deprecation warnings and this should be resolved, I think it's beneficial to keep the should_be_restful tests around for a little while longer. <div><p>GitHub didn't like the gemspec versioning switcheroo. It kept serving v2.0.1 even after the revert. So, to fix:</p> <p><code>should_be_restful</code> is now <a href="http://github.com/thoughtbot/shoulda/commit/8c081d7739e9abf69977f17cde4af814119482e3"> deprecated with this commit</a>.</p> <p>v2.0.2 of the gem is now <a href="http://github.com/thoughtbot/shoulda/commit/3b5f2642733d93e9d3219bfdffdb213f601a1be2"> published with this commit</a>.</p> <p>Note: Even though, the core shoulda tests now print deprecation warnings and this should be resolved, I think it's beneficial to keep the should_be_restful tests around for a little while longer.</p></div> true 2008-09-20T23:40:17-04:00 30519 --- :state: new 78 deprecate-should_be_restful 0 5807 committed should_be_restful 2008-09-20T23:42:49-04:00 12160 Ryan McGeary Dan Croak Dan Croak http://thoughtbot.lighthouseapp.com/projects/5807/tickets/78 10900 0 My misunderstanding, guys. Sorry for the confusion and extra work you had to do. <div><p>My misunderstanding, guys. Sorry for the confusion and extra work you had to do.</p></div> true 2008-09-21T11:44:22-04:00 30519 --- {} 78 deprecate-should_be_restful 0 5807 committed should_be_restful 2008-09-21T11:44:26-04:00 10900 Dan Croak Dan Croak Dan Croak http://thoughtbot.lighthouseapp.com/projects/5807/tickets/78 10900 0 Absolutely. We need to give a grace period before removing it completely. Dan - if you removed it from the main repo, can you please revert that change as soon as possible? Tammer <div><p>Absolutely. We need to give a grace period before removing it completely.</p> <p>Dan - if you removed it from the main repo, can you please revert that change as soon as possible?</p> <p>Tammer</p></div> true 2008-09-23T13:11:25-04:00 30519 --- {} 78 deprecate-should_be_restful 0 5807 committed should_be_restful 2008-09-23T13:11:31-04:00 13443 Tammer Saleh Dan Croak Dan Croak http://thoughtbot.lighthouseapp.com/projects/5807/tickets/78 10900 0 should_be_restful is back in the main repo. I believe Ryan committed it on Saturday night. <div><p>should_be_restful is back in the main repo. I believe Ryan committed it on Saturday night.</p></div> true 2008-09-23T19:32:15-04:00 30519 --- {} 78 deprecate-should_be_restful 0 5807 committed should_be_restful 2008-09-23T19:32:18-04:00 30519 Dan Croak Dan Croak Dan Croak http://thoughtbot.lighthouseapp.com/projects/5807/tickets/78 10900 0 I think some screwy time-line problem happened with Lighthouse here. My guess is that Tammer replied via email with his most "recent" message but did so a few days ago. Lighthouse hid it until now? Perhaps some background Lighthouse process went on the fritz? <div><p>I think some screwy time-line problem happened with Lighthouse here. My guess is that Tammer replied via email with his most "recent" message but did so a few days ago. Lighthouse hid it until now? Perhaps some background Lighthouse process went on the fritz?</p></div> true 2008-09-23T19:38:39-04:00 30519 --- {} 78 deprecate-should_be_restful 0 5807 committed should_be_restful 2008-09-23T19:38:41-04:00 12160 Ryan McGeary Dan Croak Dan Croak http://thoughtbot.lighthouseapp.com/projects/5807/tickets/78 10900 0 Might want to say something about deprecation on the tutorial page http://thoughtbot.com/projects/shoulda/tutorial/controllers <div><p>Might want to say something about deprecation on the tutorial page <a href="http://thoughtbot.com/projects/shoulda/tutorial/controllers">http://thoughtbot.com/projects/s...</a></p></div> true 2008-11-18T13:27:16-05:00 30519 --- {} 78 deprecate-should_be_restful 0 5807 committed should_be_restful 2008-11-18T13:27:20-05:00 23475 Matthew Platte Dan Croak Dan Croak http://thoughtbot.lighthouseapp.com/projects/5807/tickets/78 10900 0 Thanks Matthew - I've updated the tutorial. <div><p>Thanks Matthew - I've updated the tutorial.</p></div> true 2008-11-18T16:45:07-05:00 30519 --- {} 78 deprecate-should_be_restful 0 5807 committed should_be_restful 2008-11-18T16:45:09-05:00 13443 Tammer Saleh Dan Croak Dan Croak http://thoughtbot.lighthouseapp.com/projects/5807/tickets/78 10900 0 I'm surprised by this deprecation; should_be_restful seemed like the best part of shoulda to me. Of the five points in the original commit, all but the first seem to be issues with the implementation, and not the concept itself. As for the first point, I don't see how it "dissuades the programmer from good TDD practice". In my opinion, one of the great things about RESTful design is the consistency that should allow us to hit as-yet unseen levels of abstraction. The higher the level of abstraction, and the bigger the steps, the better, right? On the shoulders of giants and all that? I can see how it could provide a false sense of security, where the user is under the impression that should_be_restful is "testing their show action", whereas it's actually testing the default rails show action that they have since added code to. That also seems like an implementation detail to me though. Can anyone enlighten me? :) <div><p>I'm surprised by this deprecation; should_be_restful seemed like the best part of shoulda to me.</p> <p>Of the five points in the original commit, all but the first seem to be issues with the implementation, and not the concept itself.</p> <p>As for the first point, I don't see how it "dissuades the programmer from good TDD practice". In my opinion, one of the great things about RESTful design is the consistency that should allow us to hit as-yet unseen levels of abstraction. The higher the level of abstraction, and the bigger the steps, the better, right? On the shoulders of giants and all that?</p> <p>I can see how it could provide a false sense of security, where the user is under the impression that should_be_restful is "testing their show action", whereas it's actually testing the default rails show action that they have since added code to. That also seems like an implementation detail to me though.</p> <p>Can anyone enlighten me? :)</p></div> true 2008-11-25T07:53:29-05:00 30519 --- {} 78 deprecate-should_be_restful 0 5807 committed should_be_restful 2008-11-25T07:53:31-05:00 7971 Ben Hoskings Dan Croak Dan Croak http://thoughtbot.lighthouseapp.com/projects/5807/tickets/78 10900 0 Ummm did should_be_restful get spun off anywhere? I don't see it in woulda (http://github.com/seanhussey/woulda/tree/master). I understand the arguments against should_be_restful, but I have a number of 'helper' models where some automated testing of the restful controller pieces is helpful - and just about the right amount of attention for those actions. <div><p>Ummm did should_be_restful get spun off anywhere? I don't see it in woulda (<a href="http://github.com/seanhussey/woulda/tree/master)">http://github.com/seanhussey/wou...</a>.</p> <p>I understand the arguments against should_be_restful, but I have a number of 'helper' models where some automated testing of the restful controller pieces is helpful - and just about the right amount of attention for those actions.</p></div> true 2008-12-15T19:48:30-05:00 30519 --- {} 78 deprecate-should_be_restful 0 5807 committed should_be_restful 2008-12-15T19:48:32-05:00 40929 Cynthia Kiser Dan Croak Dan Croak http://thoughtbot.lighthouseapp.com/projects/5807/tickets/78 10900 0 I contemplated bringing should_be_restful into woulda, but after thinking it over, I agree with the reasons for the deprecation, so I didn't want to include or maintain it. Would Thoughtbot be ok with someone extracting should_be_restful into its own plugin? (I'm not volunteering...) <div><p>I contemplated bringing should_be_restful into woulda, but after thinking it over, I agree with the reasons for the deprecation, so I didn't want to include or maintain it.</p> <p>Would Thoughtbot be ok with someone extracting should_be_restful into its own plugin? (I'm not volunteering...)</p></div> true 2009-01-14T14:44:09-05:00 30519 --- {} 78 deprecate-should_be_restful 0 5807 committed should_be_restful 2009-01-14T14:44:13-05:00 17641 Sean Hussey Dan Croak Dan Croak http://thoughtbot.lighthouseapp.com/projects/5807/tickets/78