[Catalyst] installing catalyst on mac os

Ashley apv at sedition.com
Wed Dec 24 23:14:08 GMT 2008


On Dec 24, 2008, at 2:07 PM, Tomas Doran wrote:
>
> On 24 Dec 2008, at 18:27, Ashley wrote:
>
>> On Dec 24, 2008, at 2:08 AM, Tomas Doran wrote:
>>> On 24 Dec 2008, at 00:25, Ashley wrote:
>>>> I love OS X but the Perl it has historically shipped with is  
>>>> historically no so hot. I would strongly recommend you install  
>>>> the latest 5.8 (NOT over your system perl but beside it) or 5.10.
>>>
>>> I don't consider this good advice, especially for a beginner.
>>
>> He didn't say he was a Perl beginner. :)
>
> Yeah, sorry - total misreading on my part there..
>
>> He said he was new to Catalyst. The perl + libs OS X ships with  
>> have caused me many problems (nothing with Cat springs to mind,  
>> more C-based things like Storable) in the past from the public  
>> beta to 10.4.
>
> I've never had an issue with Apple's perl myself. I wouldn't use it  
> for production (where I build my own from source), but I've always  
> found it fine for development..
>
> It's not bad advice, generally - but readers could have got the  
> impression that Catalyst itself is unlikely to work on the perl  
> which Apple ship, which isn't true..

I've used every single OS X from the PB and I've had problems with  
all of them except 10.5 but I didn't keep it long enough to know if  
it would cause problems, I upgraded Perl as soon as I got my new  
machine. Maybe I do more odd things than you, but many system  
installs are not particularly hot. The vendors make goofy, subjective  
choices and releases get frozen with bugs. Red Hat Perl has also had  
big problems, which clobbered DBIC among other libs, just as another  
example. You should cook your own if you know how.

>> This might be the problems with the temporarily out of sync  
>> content decoding layers in LWP v Mech. Which are all fixed in the  
>> newest versions. So installing LWP and WWW::Mechanize before  
>> Catalyst might be all that's needed.
>
> I think that is just idle speculation...

I think that's just a cliche to fill space... There was an ongoing  
problem between Mech and LWP regarding decoding layers it was  
discussed on those lists, Perl Monks, and here too as it was messing  
with testing Cat apps. It was fixed. The last time I updated was  
about then and everything went fine. Without looking at test output  
it was a perfectly reasonable guess.

> I can confirm that right now for me, a freshly installed 5.8.8,  
> pulling everything (including latest WWW::Mech, and latest LWP)  
> from CPAN, WWW::Mech fails tests for me, as per:


You're right. It's bad live tests. They're checking live sites which  
is prone to falling down, obviously, because sites change content.

All the unit tests pass so a force install in this case is completely  
reasonable.

jinx at jasper[124]~/build/WWW-Mechanize-1.52>prove -l lib t
t/00-load.................1/2 # Testing WWW::Mechanize 1.52, with LWP  
5.822, Perl 5.010000, /usr/bin/perl
t/00-load.................ok
t/add_header..............ok
t/aliases.................ok
t/area_link...............ok
t/autocheck...............ok
t/clone...................ok
t/cookies.................ok
t/credentials-api.........ok
t/credentials.............ok
t/die.....................ok
t/field...................ok
t/find_frame..............ok
t/find_image..............ok
t/find_inputs.............ok
t/find_link-warnings......ok
t/find_link...............ok
t/find_link_id............ok
t/form-parsing............ok
t/frames..................ok
t/image-new...............ok
t/image-parse.............ok
t/link-base...............ok
t/link-relative...........ok
t/link....................ok
t/new.....................ok
t/pod-coverage............ok
t/pod.....................ok
t/regex-error.............ok
t/save_content............ok
t/select..................ok
t/taint...................ok
t/tick....................ok
t/upload..................ok
t/warn....................ok
t/warnings................ok
All tests successful.
Files=35, Tests=339, 20 wallclock secs ( 0.45 usr  0.41 sys + 14.26  
cusr  2.88 csys = 18.00 CPU)
Result: PASS




More information about the Catalyst mailing list