http://www.perlmonks.org?node_id=1214711

It has been announced that http://search.cpan.org is due to retire soon, redirecting to https://metacpan.org. So long and thanks for all the fish.

Replies are listed 'Best First'.
Re: Goodbye search.cpan.org
by pryrt (Prior) on May 17, 2018 at 13:04 UTC

    Are the monastery syntax [mod://...] and related Shortcuts going to change where they point, or is it going to rely on the automatic redirection at the other end?

      While I'm sad to see search.cpan.org go, as it's my preferred site, I expect that we will change the mod:// and dist:// shortcuts to the metacpan site.

        Why not wait and see how they do the redirect?

      This is a good question, and seems like the smart thing to do. I'll raise the issue. Even if they didn't it sounds like the redirect will cater for this, but where possible it'd make a lot of sense to alter the code here for [mod://] to do the same as [metamod://].

        With so many good Perl programmers here...

        Can Perlmonks not provide an extract of all the links in every post ever made?

        Turn it into some kind of Hackathon, best algorithm that can replace all links to the new site wins 1000 experience points :P

      I've been using

      [cpan://...]

      for my module references. Whoops!

Re: Goodbye search.cpan.org
by cavac (Curate) on May 18, 2018 at 09:14 UTC

    That's sad. The old search.cpan.org has a much cleaner interface, that is much easier too read for someone like me with a bit of color blindness.

    MetaCPAN might or might not allow to change the color scheme after logging in. I haven't tried, because i hate the fact that you can't just use username/password, but instead have to use a third party login provider. Why the hell would i allow Google or f...ing Facebook to sign in as me to another site? As a rule, i don't use those services for anything that is really relevant/important to me, because this is a security nightmare waiting to happen.

    But since there seems to be no alternative, i'll test over the next few weeks if this would work with running an OpenID API on my own server. If not, then i'll have to think about alternative solutions, like restart my webpage-modifying proxy project to filter out all the clutter and make metacpan more readable.

    "For me, programming in Perl is like my cooking. The result may not always taste nice, but it's quick, painless and it get's food on the table."

      "Why the hell would i allow Google or f...ing Facebook to sign in as me to another site?"

      Some people are unaware of the problems or just pain don't care. You don't need to use those, you can use github or openid (which as you know you can host yourself).

      "because this is a security nightmare waiting to happen."

      You are behind the times, it's already happened :P

      "If not, then i'll have to think about alternative solutions, like restart my webpage-modifying proxy project to filter out all the clutter and make metacpan more readable."

      It'd make more sense to work with the existing codebase and provide a PR for the functionality you discuss, then others could collaborate/get the benefits.

      Can you clarify what parts of the interface is a problem for people with color blindness? We definitely haven't put a lot of effort into accessibility features, but we also haven't gotten any feedback of where there are problems. Please feel free to file issues on GitHub for any accessibility problems you come across.

      MetaCPAN doesn't have any user-based UI preferences. Any UI customizations we do have are stored with in your browser directly using localStorage. The only purpose logging in serves is to allow using ++ on modules, and customizing a few things on your author page if you have a PAUSE account. We're unlikely to add local accounts, although we are considering adding support for AuthO.

Re: Goodbye search.cpan.org
by shmem (Chancellor) on May 18, 2018 at 11:00 UTC

    Sad news, and I frown upon the reasons for its going away. From the announcement:

    In recent years maintenance has become a burden. Most of the site is running 2005 era Perl code.

    As if old code would rot away. And below in the comments, this (emphasis mine):

    Nigel HorneMay 17, 2018 at 1:00 PM

    That's a shame. I prefer the clean interface of search.cpan.org over the metacpan site.

    Where's the source code? May be kinda fun to play around with.
    DanMay 17, 2018 at 1:29 PM

    https://github.com/metacpan/metacpan-web for the interface, https://github.com/metacpan/metacpan-api for the search/index backend


    DanMay 17, 2018 at 1:37 PM

    If you meant the source code of cpansearch, it has never been published.

    Wait, what? It has never been published? And then maintenance becomes a burden? Whilst I am deeply thankful for Graham for having it written in the first place, and all maintainers who helped to keep it running, I have to say:

    Congrats, search.cpan.org maintainers for not leveraging support from the community which this wonderful tool helped to build.

    Perl itself is in git for a long time now. Why didn't that happen with the code for that site?

    perl -le'print map{pack c,($-++?1:13)+ord}split//,ESEL'

      You make a good point, sadly I don't have any answers for you. Please reply if you do find the authoritative answers.

        No authoritative answers yet, sorry for that. But from the facts and the responses I got so far it looks like Graham Barr just didn't release the code under the terms of either the GNU General Public License or the Artistic License or such and doesn't mean to do so for whatever reason. I am not going to stalk GBARR to do so, and nobody should. If he does release in perl terms, fine.

        All Things Must Pass.

        perl -le'print map{pack c,($-++?1:13)+ord}split//,ESEL'
Re: Goodbye search.cpan.org
by hippo (Chancellor) on May 22, 2018 at 19:15 UTC
Re: Goodbye search.cpan.org
by Your Mother (Bishop) on May 18, 2018 at 16:51 UTC

    For my part I was supremely grateful for search.cpan.org ever since I discovered it almost two decades ago. Until metacpan.org launched, even with some initial bugs and index drop-outs; which were addressed quickly because of the open code and better reporting/social mechanisms. At that point I was only irritated every time I used a search engine and search.capn.org was the top result and metacpan.org wasn't even on the top page; or someone linked to the former and sent me to go find that module on the latter because I prefer it enough to do so. Excepting page load speed, metacpan.org is better and preferable to me on every level; except that they didn't choose my logo in the contest. :P Mine was better than the winner.

      "metacpan.org is better and preferable to me on every level;"

      I agree, and for me the page load times seem indistinguishable between the two systems.

      "they didn't choose my logo in the contest. :P"

      I warned you at the time that you'd regret getting that tattoo :P

        I'd like to implement search.cpan.org

        In fact, so much so, That I picked up an appropriately named domain, created an identical (search) page. Created a CPAN mirror ( soon to be official ). All that's left (and this is pretty important) is the search implementation.
        Time is running short. The 25th is drawing near. Anyone willing to help; suggest how I might best (re)create the search implementation. As I understand it, it's not open(source), or available. :-(

        IOW As I'd really like to get this right the first time. Any suggestions, pointers, ... greatly appreciated!

        λɐp ʇɑəɹ⅁ ɐ əʌɐɥ puɐ ʻꜱdləɥ ꜱᴉɥʇ ədoH

      > or someone linked to the former and sent me to go find that module on the latter because I prefer it enough to do so.

      Poor Friedrich ... ;p

      Distributions on cpan.org have forward links to metacpan.org in the upper right corner (but not the other way round AFAIK).

      e.g. compare Data::Dump vs Data::Dump

      Cheers Rolf
      (addicted to the Perl Programming Language and ☆☆☆☆ :)
      Wikisyntax for the Monastery

        Distributions on cpan.org have forward links to metacpan.org in the upper right corner…

        Yes, and I was grateful when it appeared somewhat recently. Fried-rich, a new style of sandwich as only I know how to deep fry it.

        Distributions on cpan.org have forward links to metacpan.org in the upper right corner (but not the other way round AFAIK).

        MetaCPAN has linked back to search.cpan.org for as long as I can recall. It's in the left nav under the heading "S.C.O". HTH (for the next month or so).

Re: Goodbye search.cpan.org
by VinsWorldcom (Prior) on May 18, 2018 at 15:19 UTC

    I see above the discussions about OpenID and / or using another site to authenticate. I tried with GitHub and it *seems* to work, but my "Profile" is blank prompting me to connect my PAUSE account. Clicking the link prompts for my PAUSE id, which I dutifully enter and then it says it sent an email to my PAUSE address (presumably PAUSEid@cpan.org, which I have setup in PAUSE to forward to my *real* email) but I never get it. Testing shows that emails sent to my PAUSEid@cpan.org do in fact get to me and I always receive the automated emails from PAUSE, i.e., uploads, site activity, RT, etc.

    I've opened a case with MetaCPAN.

    Anyone else got this to work?

Re: Goodbye search.cpan.org
by Anonymous Monk on May 17, 2018 at 18:27 UTC

    Note also that http://metacpan.org also redirects to https, thus the ability to search over vanilla http will be lost. But maybe this was the idea.

Re: Goodbye search.cpan.org
by Anonymous Monk on May 17, 2018 at 10:20 UTC
    Very sad... I wonder what the maintenance was ...
Re: Goodbye search.cpan.org
by Anonymous Monk on May 19, 2018 at 09:04 UTC
    Can something please be done to just stop this nonsense?

    As an active developer I use search.cpan.org every working day!

    My browser history shows an average of 20 visits per day in the past month.

    It's a PERFECT website. The form and function are impeccable. It must remain!


    Not only should we save it but really search.cpan.org or something exactly like it belongs in the Perl core so we could simply do things like:

    pydoc -g
    http://localhost:7464/

    gem server
    http://localhost:8808/

    php -S localhost:8080
    http://localhost:8080/

    Like search.cpan.org for local modules:

    perldoc -port:8888
    http://localhost:8888/


    Calling all Perl; Monks, Mongers and Poets!

        what's wrong with metacpan.org?

        Minor quibble: on /recent, (which is the most-used page for me) the line height is too large, so that less entries fit on a page (than is the case on the old website). I do see the advantages of metacpan.org but less lines on a page is still something that is a bit annoying. Perhaps it can be fixed.

        "So, what's wrong with metacpan.org?"

        (speaking for myself) It's not that I hate metacpan; it's that I don't like it.

        It's not that I like search.cpan; it's that I love it.

        Qualification of the above;
        I find metacpan too messy, and busy for my needs, and likes -- it get's in my way, and workflow.

        search.cpan OTOH has exactly what I need, and only what I need.

        That pretty much covers it, for me. :-)

        λɐp ʇɑəɹ⅁ ɐ əʌɐɥ puɐ ʻꜱdləɥ ꜱᴉɥʇ ədoH

      "It must remain!"

      This seems unlikely.

      "Not only should we save it but really search.cpan.org or something exactly like it belongs in the Perl core so we could simply do things like:"

      What are you doing to save search.cpan? Have you submitted a patch for the functionality you desire?

      Update: fixed typo.