| 2014-09-05 |
| ← brandonsavage left | 01:01 |
| → brandonsavage joined | 01:11 |
| ← beakman left | 04:58 |
| → beakman joined | 05:14 |
| ← beakman left | 05:24 |
| → devosc joined | 05:46 |
|
devosc
| hi guys, my name is greg and I've been working on php 5.5 "whatever/system", I've noticed that the system actually doesn't need to know anything about the actual interfaces of Request and Response objects | 05:48 |
|
| Stupid me never mind ... thanks for making ResponseSender callable. | 05:49 |
| ← rafi left | 07:07 |
| → rafi joined | 07:12 |
| ← wdna_ left | 07:36 |
|
devosc
| I wanted to see if I could remove my 'Response\Sender' and use Web\ResponseSender instead; I would then need to extend Web\Request so as to apply my own interface. | 07:49 |
|
| Since the request class has no interface methods I can't use it .. oh boy | 07:51 |
|
| Here is the DI Config I ended up with the Web\Request https://gist.github.com/devosc/0c821e506c4c85ce5c2d | 07:51 |
|
| I would of liked to get host name from the Request object with 'Request.host' where dot indicates it is a method call. | 07:53 |
| → wdna joined | 08:25 |
| → joelclermont joined | 12:44 |
| ← brandonsavage left | 13:00 |
| → brandonsavage joined | 13:02 |
| ← brandonsavage left | 13:03 |
| ← wdna left | 14:18 |
| → harikt joined | 14:38 |
|
harikt
| Hey good day to all | 14:39 |
|
| Good day pmjones . | 14:39 |
|
| You are early here :) | 14:39 |
|
pmjones
| hey there harikt | 14:39 |
|
harikt
| :) | 14:39 |
|
pmjones
| yeah i've been leaving IRC on overnight lately ;-) | 14:39 |
|
| i see your PRs and will address | 14:39 |
|
| harikt nods | 14:39 |
|
pmjones
| kernel and project-level stuff has to be dealt with delicately so we don't bust dependencies for too long | 14:40 |
|
| and regarding using `phpunit` instead of `./phpunit.sh` -- yes, i see your point | 14:40 |
|
| i wanted to make it "all tests all run the same way all the time" | 14:40 |
|
harikt
| I understand. | 14:41 |
|
pmjones
| but having the library-only unit tests go with just `phpunit` seems fine | 14:41 |
|
harikt
| great! | 14:41 |
|
| What about writing the testing as a seprate page ? | 14:41 |
|
pmjones
| the container, kernel, and project tests though -- those will stay ./phpunit.sh | 14:41 |
|
harikt
| so we can link to the page | 14:41 |
|
pmjones
| harikt: separate page? | 14:41 |
|
| oh, no, it's not that big a deal | 14:42 |
|
harikt
| ya for testing . | 14:42 |
|
| ok. | 14:42 |
|
pmjones
| update the README in each repo, and done | 14:42 |
|
harikt
| :) you need to update 20 repos Paul . | 14:42 |
|
pmjones
| if you like, though, you can see if the auraphp.com/contributing page needs touching up | 14:42 |
|
| 20 repos? that's what scripts are for ;-) | 14:42 |
|
harikt
| hm. you are doing with scripts | 14:42 |
|
| aaaha | 14:42 |
|
pmjones
| mostly | 14:42 |
|
harikt
| nice then! | 14:42 |
|
| not by me. | 14:43 |
|
pmjones
| the script does the heavy lifting, then i go back and tweak the results, and commit/push | 14:43 |
|
harikt
| hand made | 14:43 |
|
pmjones
| yeah | 14:43 |
|
harikt
| pmjones, I wonder then you have not created a auraphp console tool which can create package / library classes | 14:44 |
|
| like the artisian | 14:44 |
|
| pmjones shrugs | 14:44 |
|
pmjones
| when the time comes, there'll be one for creating ADR dirs | 14:44 |
|
| but for creating libraries, it's not that big a deal | 14:44 |
|
| oh, for the testing updates, the script was this: https://gist.github.com/pmjones/7b9a44749c4b838ac475 | 14:44 |
|
harikt
| :) | 14:44 |
|
pmjones
| ./newtest.sh Filter | 14:44 |
|
| but it has to be in a directory with all the other packages in it | 14:45 |
|
| blah blah blah | 14:45 |
|
harikt
| mm. | 14:46 |
|
| this script is nice. | 14:46 |
|
| I was wondering how you do for the readme.md | 14:46 |
|
| anyway may be the git helps to see a diff | 14:47 |
|
| :) | 14:47 |
|
| by the way pmjones I gave the link to http://voicesoftheelephpant.com/2012/05/01/fig-fud-fomo/ so DG can hear why PSR exists | 14:48 |
|
pmjones
| he knows why it exists :-) | 14:48 |
|
| he just doesn't like that other people think "there is only one set of standards" | 14:48 |
|
| and that the set of standards is not something he likes | 14:49 |
|
| pmjones shrugs | 14:49 |
|
pmjones
| he's changed the name and added attribution, so i think we're ok now | 14:49 |
|
harikt
| :) . | 14:49 |
|
| pmjones, here you go https://github.com/php-group/standards/issues/1#issuecomment-54634622 | 14:50 |
|
| deal ? | 14:51 |
| → beakman joined | 14:52 |
|
harikt
| Thanks Paul | 14:55 |
|
| :) | 14:55 |
|
pmjones
| yeah, we'll see if that actually goes anywhere or not :-) | 14:55 |
|
| and thank *you* for noting it | 14:55 |
|
harikt
| yup, we will see. | 14:56 |
| ← beakman left | 14:56 |
|
harikt
| and you are welcome. Always a pleasure to let you know something | 14:56 |
| → brandonsavage joined | 15:38 |
|
devosc
| hey paul, did you see my notes from earlier about the Web\Request ? | 15:38 |
|
harikt
| devosc, anything public so I can also have a look ? | 15:39 |
|
devosc
| http://colabti.org/irclogger/irclogger_log/auraphp?date=2014-09-05 | 15:39 |
|
| I ended up using SF. I was interested in having the separate Response and Responder classes. | 15:43 |
|
harikt
| devosc, not sure I understand. I was looking into it . | 15:46 |
|
| Looking at your github profile I think I have came accross your profile :) | 15:46 |
|
pmjones
| oh hey there devosc | 15:47 |
|
| pmjones scrolls up | 15:47 |
|
harikt
| devosc, what are you tring to do actually ? | 15:47 |
|
devosc
| The snag is the Request object doesn't have any interface methods. | 15:47 |
|
harikt
| ok, pmjones may be able to help with it . | 15:48 |
|
pmjones
| devosc: sorry, i don't think i understand your question | 15:48 |
|
devosc
| I'm trying to chain method calls to retrieve values, e.g 'Request.host' would resolve to $request->host() but there no methods | 15:48 |
|
pmjones
| yes, the Aura.Web Request object is basically a collection of properties, that are themselves objects, with the various functionality separated that way | 15:49 |
|
| rather than grouping all the various methods in a single object | 15:49 |
|
harikt
| devosc, have you had a look at https://github.com/auraphp/Aura.Web/blob/develop-2/README-REQUEST.md ? | 15:49 |
|
devosc
| And the object that has the actual value, doesn't have a matching interface method (since it is array access). | 15:49 |
|
pmjones
| magic property access iirc, but yeah | 15:50 |
|
| is that a problem? | 15:50 |
|
devosc
| It sort of prevents me from using the request object. I'm finding in DI that the 'code' really doesn't need the request object since all its data can be retrieved and injected .. but my poor mans method chain won't work | 15:51 |
|
| Even if the respective Param object had the methods, I could maybe target that obj directly and call the method. | 15:52 |
|
pmjones
| i'm not understanding why a method-chain is a requirement here -- $request->server->get('WHATEVER') is not an option? | 15:53 |
|
harikt
| yes. I am also wondering the same. | 15:53 |
|
devosc
| I'm using a config syntax and not runtime php. | 15:53 |
|
| new Call('Request.host') | 15:53 |
|
pmjones
| if i may be so bold, perhaps the problem is that the config syntax is underpowered | 15:54 |
|
harikt
| devosc, so the new Call should return request->host ? | 15:55 |
|
devosc
| :) Maybe. But method interfaces are important (and thats why its failing) :) | 15:55 |
|
| harikt, yeah | 15:56 |
|
pmjones
| i agree that interfaces are important for methods, but these are properties :-) | 15:56 |
|
| magic ones, to be sure, but still | 15:56 |
|
devosc
| interface object access doesn't care what the method actually does | 15:57 |
|
pmjones
| devosc: are you making an indirect feature request here, asking for method-based access to the same propeties? | 15:57 |
|
devosc
| I'm mentioning ... I'm seeing changes (atleast in my thinking) and DI + Interfaces are important. | 15:58 |
|
pmjones
| ok, i agree that interfaces are important, for as far as that goes | 15:58 |
|
| but stating that does not seem to resolve your problem | 15:58 |
|
| sooo ... ? | 15:58 |
|
devosc
| I'm not looking for immediate resolvement, but I thought it worth mentioning. | 15:59 |
|
| pmjones nods | 15:59 |
|
pmjones
| consider it mentioned | 15:59 |
|
harikt
| devosc, if you like PR ;) | 15:59 |
|
pmjones
| for the record, if you want something, ask for it directly :-) | 15:59 |
|
harikt
| pmjones, will look into it. | 16:00 |
|
| Here like me. | 16:00 |
|
| Blame Paul is wrong when you think he is wrong. | 16:00 |
|
| ;) | 16:00 |
|
pmjones
| :-) | 16:00 |
|
devosc
| :) | 16:00 |
|
harikt
| pmjones, great looking at your commit https://github.com/auraphp/Aura.Auth/commit/1a64f527775961d764e33ea05450f8dbe177412b | 16:01 |
|
| :) | 16:01 |
|
pmjones
| harikt: I BLAME YOU | 16:01 |
|
harikt
| I take the responsibility :) | 16:01 |
|
pmjones
| harikt: we'll see how travis and scrutinizer like it, then i can apply to all the others | 16:01 |
|
harikt
| why they don't like ? | 16:02 |
|
pmjones
| i don't know if they do or not -- waiting to see :-) | 16:03 |
|
| travis seems ok | 16:04 |
|
| oh hey harikt -- let drop the dravid grudl conversation | 16:04 |
|
| i can't tell if he's really serious or not | 16:04 |
|
| if he is, so much the better | 16:04 |
|
| if not, no amount of added conversation will sway him | 16:04 |
|
harikt
| ok pmjones . | 16:07 |
|
brandonsavage
| pmjones: Are you an original member of the FIG? | 16:07 |
|
harikt
| brandonsavage, yes | 16:07 |
|
| :) | 16:07 |
|
brandonsavage
| harikt: No offense, but I think I asked pmjones that question, not you. | 16:07 |
|
pmjones
| brandonsavage: yes i am | 16:07 |
|
| i was in the "smoke filled room" at the start of ti | 16:07 |
|
harikt
| brandonsavage, I heard Fig foma and food. | 16:08 |
|
brandonsavage
| pmjones: Aha. So when Crell insists that the room was smoke filled, he isn't lying. | 16:08 |
|
pmjones
| well, when the myth is flattering, publish the myth ;-) | 16:08 |
|
harikt
| and I will consider for next time when responding | 16:08 |
|
| :) | 16:08 |
|
brandonsavage
| "When the legend becomes fact, print the legend." | 16:09 |
|
pmjones
| that's what i was looking for | 16:09 |
|
| bravo sir | 16:09 |
|
brandonsavage
| The Man Who SHot Liberty Valance. | 16:09 |
|
pmjones
| great flick | 16:09 |
|
brandonsavage
| A corollary to "the victors make the history." | 16:09 |
|
pmjones
| harikt: all libs converted back to `phpunit -c tests/unit/` instead of `cd tests/unit; ./phpunit.sh` | 16:17 |
|
harikt
| Thanks pmjones | 16:21 |
|
| and you are fast | 16:21 |
|
pmjones
| thank you for keeping me honest ;-) | 16:22 |
|
brandonsavage
| What's the advantage of the conversion? | 16:22 |
|
pmjones
| brandonsavage: short version is that it's easier for windows folks to run the plain old unit tests. | 16:22 |
|
brandonsavage
| How many hours of work did you do for 0.02% of all Aura users? :-P | 16:23 |
|
pmjones
| two days ago i started changing to a ./phpunit.sh script to standardize across unit, integration, kernel, bundle, etc testing | 16:23 |
|
| brandonsavage: about 20 minutes ;-) | 16:23 |
|
brandonsavage
| Ah, well, small wins then. ;-) | 16:23 |
|
pmjones
| but harikt pointed out, i think rightly, that there's no benefit to that, and perhaps even a small drawback | 16:23 |
|
| so plain old library level unit tests are back they way they were, albeit in a different directory (tests/unit/ instead of tests/, so we can also have tests/container/ et al in the same tests dir) | 16:24 |
|
| brandonsavage nods | 16:25 |
|
brandonsavage
| I leave it to you to run the unit tests. | 16:26 |
|
| I don't bother running them myself. | 16:26 |
|
devosc
| what about giving the properties a configuration interface :) | 16:28 |
|
pmjones
| brandonsavage: ha! :-) | 16:30 |
|
| devosc: what do you mean? | 16:30 |
|
harikt
| pmjones, when you have a badge do we need to run unit tests on local system ;) | 16:31 |
|
| https://github.com/auraphp/Aura.Web#quality | 16:31 |
|
devosc
| Which object gives me the path from $_SERVER ? | 16:31 |
|
pmjones
| devosc: which path? | 16:32 |
|
devosc
| I'm seeing AbstractValues which has a get method that I can work with (chained gets) .. | 16:32 |
|
| https://github.com/auraphp/Aura.Web/tree/develop-2/src/Request | 16:32 |
|
harikt
| devosc, https://github.com/auraphp/Aura.Web/blob/develop-2/README-REQUEST.md#superglobals | 16:32 |
|
devosc
| uri path | 16:32 |
|
pmjones
| devosc: where would you normally find that in $_SERVER? | 16:32 |
|
devosc
| isn't that the request_uri ? | 16:32 |
|
pmjones
| so then, $request->server->get('REQUEST_URI') | 16:33 |
|
devosc
| what pathInfo or similar does | 16:33 |
|
harikt
| devosc, you should have a look at https://github.com/auraphp/Aura.Web#instantiation | 16:33 |
|
| Rather than just the api. | 16:33 |
|
| read me points to all I guess | 16:33 |
|
| if you find something missing please do send a PR or create an issue ;) | 16:34 |
|
devosc
| I can do $request->get('server')->get('REQUEST_URI') ... since that would new Param('Request.server.REQUEST_URI') | 16:34 |
|
pmjones
| devosc: there's also https://github.com/auraphp/Aura.Web/blob/develop-2/README-REQUEST.md#url | 16:34 |
|
| $request->url->get(PHP_URL_PATH) | 16:34 |
|
devosc
| I need a chain of method calls | 16:36 |
|
| So I would need 'url' to be returned by a get method also. | 16:36 |
|
| pmjones nods | 16:36 |
|
pmjones
| i think __get() may do the trick for you, then | 16:36 |
|
| alternatively, you may wish to use a DI system that does not limit you in that way | 16:37 |
|
devosc
| Actually its the other way around, you'r suggesting that I have a bootstrap to make method calls -- that is exactly what I am trying to avoid and have direct DI :) | 16:38 |
|
pmjones
| mmm, 6 of one, half-dozen of the other | 16:39 |
|
| something has to do the setup work, and that means something is going to be executing code in a bootstrap | 16:39 |
|
devosc
| its about static and serialization | 16:39 |
|
| pmjones shrugs | 16:39 |
|
pmjones
| whatever floats your boat man | 16:39 |
|
devosc
| Another way of looking at is, is the configuration is the inversion of control of the DI. | 16:40 |
|
| Its potentially faster, and the config interface opens up more possibiliities | 16:41 |
|
pmjones
| yeah, maybe | 16:41 |
|
| good luck with it :-) | 16:41 |
|
devosc
| Its done :) | 16:41 |
|
pmjones
| well then :-) | 16:41 |
|
devosc
| I was trying to integrate with your components :) and couldn't :( | 16:42 |
|
brandonsavage
| pmjones: BTW, how's the new job? | 16:42 |
|
pmjones
| devosc: that's too bad. like i said, if you have specific things you want to see, ask for them directly. | 16:43 |
|
devosc
| I will thanks. What I'm seeing in Aura.DI is there could be few other 'convenient' type functions similar to lazyNew | 16:45 |
|
| pmjones nods | 16:46 |
|
pmjones
| folks, i'm afraid i have to step away | 16:47 |
|
| devosc: it's nice to see you here | 16:47 |
|
| harikt: sleep well, sir, when that time comes :-) | 16:47 |
|
devosc
| you too ... | 16:47 |
|
harikt
| pmjones, thank you. | 16:47 |
| pmjones → pmjones_away | 16:47 |
|
harikt
| will be leaving shortly. | 16:47 |
|
| devosc, in case you have trouble please mail to google groups, | 16:48 |
| → beakman joined | 16:48 |
|
harikt
| I hope you will find an answer . It is time for me to sleep. | 16:49 |
|
| I am leaving. Talk later ! | 16:49 |
| ← harikt left | 16:49 |
| → wdna joined | 17:06 |
| ← wdna left | 17:28 |
| ← beakman left | 17:49 |
| → beakman joined | 17:52 |
| ← joelclermont left | 17:53 |
| pmjones_away → pmjones | 19:08 |
| ← beakman left | 19:24 |
| → beakman joined | 19:35 |
| ← rafi left | 20:00 |
| ← beakman left | 20:37 |
| → beakman joined | 20:58 |
| ← jakefolio left | 21:30 |
| ← beakman left | 21:40 |
| ← brandonsavage left | 21:44 |
| → jakefolio joined | 21:53 |
| → wdna joined | 22:04 |
| → beakman joined | 22:17 |
| ← beakman left | 22:17 |
| → brandonsavage joined | 22:18 |
| → beakman joined | 22:18 |
| ← beakman left | 22:29 |
| ← wdna left | 22:35 |
| → cxj joined | 22:48 |
| → beakman joined | 22:52 |