[Author Prev][Author Next][Thread Prev][Thread Next][Author Index][Thread Index]

Re: [pygame] New pygame.org website



Can't we just make it so only a few known-good users have access to merge to master and then use pull requests for edits to the site from the community?  this is a common way to do it both for sites and otherwise - Ionic does their tutorial websites this way. 

The only real issue I would see is that you wouldn't really want a single repo with binaries or even of source of all games submitted.  That would clutter up the repo for the website with essentially backend data edits.

I would suggest making a template repository for users to fork on GitHub that would include an editable readme.md with a good starting framework.  They could update the readme to include descriptions of their games and these could be pulled into a list.

Regarding static vs dynamic sites and games lists, we could preprocess the games list into one big json file and then just implement the game list view as an Angular application.  That would allow sorting/filtering/searching of content with no dynamic processing on the server side whatsoever - just when a new project is accepted, that json file would be updated to include the link and some tags about each game.

On Fri, Dec 23, 2016 at 12:49 PM, Thomas Kluyver <takowl@xxxxxxxxx> wrote:
Hi Miriam,

Looking at ibiblio.org, it's talking about an application process to host a 'collection' there ("A decision will be made and you will be notified."). This sounds rather controlled for a simple static site, and I'm concerned that it won't be as easy as we want to update the site and for multiple people to work on it. Clicking around the distros section, it looks like most distros host downloads there but have a website somewhere else (so far, Fat Dog is the only one I've found with web pages there). This suggests to me that it's not ideal for hosting a website.

It's hard to put my finger on what's wrong with it, because clearly you *can* host static web pages there. But looking around the information pages and the Q&A site, I get a very strong feeling that hosting there would involve a lot of time and hassle that I don't want.

Similarly for archive.org, I think it's intended as... well, an archive! That's valuable, but it's not the problem we're trying to solve.

Github pages is a very convenient way to host a website, and to manage multiple people making changes to it. I know a lot of open source projects whose websites are hosted this way. It does require using git to update the site, and I'm sorry if that's difficult for you, but it's a system that works well for a lot of people. None of the arguments I've read so far provide a compelling reason not to host a site on Github, and the only option that I see as a reasonable alternative is Gitlab, which presumably works in a very similar way to Github.

> Is it possible to have something that is safe, but easy for newbies to add games to a site?

I think it is, if we accept that it just needs to be 'safe enough'. Automatically adding 'nofollow' on user provided links reduces its value to spammers a lot. I believe Google's recaptcha tool does a reasonable job of stopping bots. And if it's still not enough, we can authenticate users and require an admin to approve a user's first submission.

Thomas

On 23 December 2016 at 01:45, Miriam English <mim@xxxxxxxxxxxxxxxxxx> wrote:
The two repositories I mentioned, ibiblio.org and archive.org (I'm sure there are more) have, as far as I know, no limits on storage.

Have a look, for instance at the amount stored for Puppy Linux at ibiblio:
http://distro.ibiblio.org/puppylinux/

Each of those subdirectories you see relates to a different kind of Puppy.

The packages directories contain programs specifically tailored for a particular distribution of Puppy. The pet_packages-lucid directory alone contains more than 10 Gigabytes of programs, and there are more than 30 Puppy distros with associated package collections.

The puppy-528 directory contains a couple of ISO CD images for Puppy Lucid 528. It also contains an explanatory webpage:
http://distro.ibiblio.org/puppylinux/puppy-5.2.8/release-Lucid-528.htm

Most of the main distro directories contain such a page.

There is lots more on ibiblio, as a quick wander around http://distro.ibiblio.org/ will show.
They also have multiple mirrors around the world, so if one set of servers has a problem others are available.

Cheers,

    - Miriam


Charles Cossé wrote:
Hi,

On Thu, Dec 22, 2016 at 2:02 PM, Miriam English <mim@xxxxxxxxxxxxxxxxxx <mailto:mim@xxxxxxxxxxxxxxxxxx>> wrote:

    I don't see the point of using github for the web pages and
    keeping the content elsewhere. I don't have a lot of experience
    using github (I find it a pain actually). Github is intended as a
    versioning system. That has no utility for a pygame repository, as
    far as I can see -- or at least no advantage over an ordinary
    repository built purely with that purpose in mind.

    Wouldn't it be simpler to keep the whole thing in a repository? I
    mentioned 2 earlier: archive.org <http://archive.org> and
    ibiblio.org <http://ibiblio.org>, both of which are free and very

    secure.


I can say a little bit that might help until someone with more knowledge has time to reply ... With GitHub pages your website is "just another" repo.  That's the main thing I wanted to point out.  There are no storage limits, and I'm pretty sure that GitHub would be happy to help pygame accomodation-wise if pygame needed anything special (within reason).   I also know that there is a 4 gigabyte file limit on GitHub.  (I know this because I once wanted to host an 8G SD card image and had to get it down to 4G in order to be housed on GitHub).

FWIW, I have also managed to run webapps on GitHub via GitHub pages.  For example http://asymptopia.github.io/TuxMathScrabble-2015/.

And, not trying to direct traffic to my site or anything, but here's my own site using GitHub pages: https://asymptopia.github.io/

Best regards,
Charles

    Cheers,

        - Miriam


    Thomas Kluyver wrote:

        Thanks everyone for your input. In the interests of making
        progress, I'd like to propose:

        - The informational site will be hosted on Github pages; I've
        used this for a number of websites before, it's reliable, we
        can point an external domain to it, and I imagine that most of
        the likely contributors have Github accounts already.
        - The pages will be generated by a Python static site
        generator. There doesn't seem to be a strong feeling between
        Sphinx/Nikola/Pelican, so it will likely depend on who is most
        excited to start building it.
        - The game feed will also be generated from content in Github,
        so /at first/ developers will need to submit a PR to add a
        game. Once that's working, we can build a simpler submission
        interface on Heroku/Appengine/similar which can push content
        to Github. Ideally the data will be in a format which would
        could move elsewhere later if necessary.

        I like the concept of drawing the game feed from an external
        source, but I don't think any of the sources proposed match
        what we want closely enough.

        Does anybody object to any of those proposals?

        Thanks,
        Thomas

        On 18 December 2016 at 20:18, Miriam English
        <mim@xxxxxxxxxxxxxxxxxx <mailto:mim@xxxxxxxxxxxxxxxxxx>
        <mailto:mim@xxxxxxxxxxxxxxxxxx

        <mailto:mim@xxxxxxxxxxxxxxxxxx>>> wrote:

        http://ibiblio.org is an enormous, free repository that also lets
            you have static webpages. Many of the Linux distros are hosted
            from there as well as much else too. I don't know how
        you'd set up
            a comments system there. It may be possible.

        http://archive.org is another gigantic free repository. They
            already have a comments system built into their pages. I don't
            know how it works. It might be worth checking out.

            Both these organisations are free and are aimed at helping
        make
            content available to the community which might otherwise
        be lost.
            You have complete control over the look of webpages at
        ibiblio.org <http://ibiblio.org>
        <http://ibiblio.org> because you simply upload static pages.

            I don't know how much control over the look archive.org
        <http://archive.org>
        <http://archive.org> provides because everything is dynamically

            served from xml data, I think. It might be possible to add
        static
            content, I don't know.

            But both are free, permanently available, and have
        excellent security.

            Cheers,

                - Miriam



            Peter Shinners wrote:

                Gitlab also has great static site support for free,
        and you
                can use custom domains. They also make it easy to run most
                static generation tools as a CI job. Although part of me
                thinks just pushing the static content is easiest. It
        sounds
                to me like there's a list of acceptable hosting
        choices that
                won't cost anything.

                Keeping the games list as a feed from other service sounds
                like it has the best chance of working.


                On 12/17/2016 10:51 PM, Lenard Lindstrom wrote:

                    Bitbucket also has static web site support. I set
        one up
                    for the Pygame docs awhile ago, but have not
        maintained it:

        http://pygame.bitbucket.org/docs/pygame/
        <http://pygame.bitbucket.org/docs/pygame/>
        <http://pygame.bitbucket.org/docs/pygame/
        <http://pygame.bitbucket.org/docs/pygame/>>

                    The repository is here:

        https://bitbucket.org/pygame/pygame.bitbucket.org
        <https://bitbucket.org/pygame/pygame.bitbucket.org>
        <https://bitbucket.org/pygame/pygame.bitbucket.org
        <https://bitbucket.org/pygame/pygame.bitbucket.org>>

                    Lenard Lindstrom

                    On 16-12-17 09:16 PM, Daniel Foerster wrote:

                        You know, I suppose we could just use GitHub
        pages.

                        On Dec 17, 2016 17:32, "Charles Cossé"
        <ccosse@xxxxxxxxx <mailto:ccosse@xxxxxxxxx>
        <mailto:ccosse@xxxxxxxxx <mailto:ccosse@xxxxxxxxx>>
        <mailto:ccosse@xxxxxxxxx <mailto:ccosse@xxxxxxxxx>
        <mailto:ccosse@xxxxxxxxx <mailto:ccosse@xxxxxxxxx>>>>
                        wrote:



                            On Sat, Dec 17, 2016 at 4:12 PM, Daniel
        Foerster
        <pydsigner@xxxxxxxxx <mailto:pydsigner@xxxxxxxxx>
        <mailto:pydsigner@xxxxxxxxx <mailto:pydsigner@xxxxxxxxx>>
        <mailto:pydsigner@xxxxxxxxx <mailto:pydsigner@xxxxxxxxx>
        <mailto:pydsigner@xxxxxxxxx <mailto:pydsigner@xxxxxxxxx>>>> wrote:

                                Using S3/CloudFront is a lot cheaper
        than the
                        EC2 setup you're
                                imagining (and which a Django stack would
                        require).



                            I never said to use Amazon at all.  Just
        use the
                        current server,
                            whatever it is (unless it's Amazon).

                                On 12/17/2016 05:11 PM, Charles Cossé
        wrote:

                                    Yikes!  who's gonna pay the Amazon
        bill?

                                    On Sat, Dec 17, 2016 at 4:09 PM, Paul
                            Vincent Craven
        <paul@xxxxxxxxxxxxxxxx <mailto:paul@xxxxxxxxxxxxxxxx>
        <mailto:paul@xxxxxxxxxxxxxxxx <mailto:paul@xxxxxxxxxxxxxxxx>>
        <mailto:paul@xxxxxxxxxxxxxxxx <mailto:paul@xxxxxxxxxxxxxxxx>
        <mailto:paul@xxxxxxxxxxxxxxxx
        <mailto:paul@xxxxxxxxxxxxxxxx>>>> wrote:

                                        If most of the site is static,
        then I
                            think Django would
                                        be overkill. The static
        portion of the
                            site can easily be
                                        deployed via Amazon
        S3/CloudFront and
                            then we'd not have
                                        to maintain a server.

                                        Paul Vincent Craven

                                        On Sat, Dec 17, 2016 at 5:00 PM,
                            Charles Cossé
        <ccosse@xxxxxxxxx <mailto:ccosse@xxxxxxxxx>
        <mailto:ccosse@xxxxxxxxx <mailto:ccosse@xxxxxxxxx>>
        <mailto:ccosse@xxxxxxxxx <mailto:ccosse@xxxxxxxxx>
        <mailto:ccosse@xxxxxxxxx <mailto:ccosse@xxxxxxxxx>>>> wrote:


                                            On Sat, Dec 17, 2016 at
        3:26 PM,
                            Thomas Kluyver
        <takowl@xxxxxxxxx <mailto:takowl@xxxxxxxxx>
        <mailto:takowl@xxxxxxxxx <mailto:takowl@xxxxxxxxx>>
        <mailto:takowl@xxxxxxxxx <mailto:takowl@xxxxxxxxx>

        <mailto:takowl@xxxxxxxxx <mailto:takowl@xxxxxxxxx>>>> wrote:


                                                So far, I think the
        proposals
                            for the static
                                                information part of
        the site
                            are Nikola (a static
                                                site generator
        oriented around
                            blogs) and Sphinx
                                                (oriented around
        docs). Both
                            are written in
                                                Python. Does anyone
        want to
                            make the case for any
                                                other system?


                                            Can Django factor-in there? I
                            guess it would reside
                                            underneathe the other pkgs
        ... but
                            might as well run
                                            Python through-and-through
        imho.





                                    --
                                    Linkedin
        <https://www.linkedin.com/in/charles-cosse
        <https://www.linkedin.com/in/charles-cosse>
        <https://www.linkedin.com/in/charles-cosse
        <https://www.linkedin.com/in/charles-cosse>>> |
                                    E-Learning <http://www.asymptopia.org>






                            --
                            Linkedin
        <https://www.linkedin.com/in/charles-cosse
        <https://www.linkedin.com/in/charles-cosse>
        <https://www.linkedin.com/in/charles-cosse
        <https://www.linkedin.com/in/charles-cosse>>> | E-Learning
        <http://www.asymptopia.org>








            --     There are two wolves and they're always fighting.
            One is darkness and despair. The other is light and hope.
            Which wolf wins?
            Whichever one you feed.
             -- Casey in Brad Bird's movie "Tomorrowland"



    --     There are two wolves and they're always fighting.
    One is darkness and despair. The other is light and hope.
    Which wolf wins?
    Whichever one you feed.
     -- Casey in Brad Bird's movie "Tomorrowland"




--

Linkedin <https://www.linkedin.com/in/charles-cosse> | E-Learning <http://www.asymptopia.org>



--
There are two wolves and they're always fighting.
One is darkness and despair. The other is light and hope.
Which wolf wins?
Whichever one you feed.
 -- Casey in Brad Bird's movie "Tomorrowland"