Add-on deployment TTW

Development
21

I agree with @gyst and @zopyx on this; there is a lot of wishful thinking on some of the arguments presented above by the supporters of this new feature; seems sometimes we don't learn from our past mistakes.

you are presenting reason 530 on why Plone is losing market share and you think this can be magically fixed by installing add-ons TTW; the reality will be different as has been demonstrated by the TTW content type story: is not that easy.

now I'm going to give you some facts about the internet:

yes, the CMS market is probably going to grow in the next years and the reason is obvious: there are around 900.000.000 web sites worldwide and only around 25.000.000 are running under a CMS, that's way far from the 50% mentioned above.

Plone represents around than 0.1% of that market share on a global basis (a debatable figure, indeed) but you know what? that means nothing, because the websites we run are orders of magnitude more important than our grandma's blog.

CMS market worth an estimated $3.5 billion USD in 2015, but ECM worth $24.6 billion USD; so, where do you want to be? can we have an updated roadmap with a clear path based on solutions and not only on technology? can we deliver?

we have to focus our efforts and limited resources first, on fixing the features that are still broken on Plone 5 (toolbar, resource registriesā€¦); second, on helping integrators in porting their solutions to Plone 5 (add-ons, content migrationā€¦); only then we can start experimenting with new features on a larger scale (in the form of add-ons before integrating them into the core, as we have done with Mosaic) and learn from mistakes of early adopters.

if core developers think that delivering working functionalities is not fun, we, integrators, think that fixing half baked functionalities is neither. all of the above is causing waste of time, resources and money and is delaying adoption of Plone 5.

and no, don't blame Buildout as reason 531 for our own failures: Buildout works perfectly if you know what you are doing; if you don't, ping me on IRC and I will teach you in 10 minutes.

1 Like
22

You can set your buildout to display unpinned eggs:

[buildout]
show-picked-versions = true

And then run the buildout with the "non-newest" option (keeping current versions):

bin/buildout -N

After that you can easily copy the output of version pins and add them to your versions.cfg and presto you have a fully pinned buildout.

Obviously one of the tricks of not being scared of buildout is running everything on a local dev copy until you've got complete control.

23

exactly, but better you add the following to your production configuration so you don't need to remember to run in non-newest mode all the time:

[buildout]
newest = false

done! less than 10 minutes indeed.

24

IMO newest = false and bin/buildout -N are just temporary stopgaps until you've figured out your pins. In production I never allow unpinned eggs.

[buildout]
allow-picked-versions = false
3 Likes
25

Exactly, I even keep that for development, no bin/buildout should run successfully without all versions pinned. This way, not even during development you will have version jumping around, nor will you get surprised when deploying that a pin is missing.

1 Like
26

there's always something new to learn; thanks! :slight_smile:

27

Elsewhere I commented on the nice Odoo add-on "shop"/gallery... I think it would be ok to show and have nice looking front-facing functionality add-ons displayed and installable from the Site Setup Add-ons panel.

28

One of our keynote speakers in Boston will be Annette Lewis; she's the one who has done some serious wizardry using Plone entirely TTW. She is not alone; that was what we had to do as much as possible at UW Oshkosh simply because there are never enough file system devs to do everything.

https://2016.ploneconf.org/keynotes-and-talks

THE SUPER INTEGRATOR - Annette Lewis, who has created dozens of departmental and group Plone sites at Penn State University, will walk us through the process she uses to go from design comps to finished site in 2 weeks - with everything done through the web.

2 Likes
29

That seems to be common case for Plone at universities. Also we depend on our power users being able to implement and enhance features TTW. Of course, that often adds technical dept, but usually things work out well enough.

I would have no issues for Plone having two separate add-on stories (TTW add-ons / filesystem extensions) with the infamous Z-shaped learning curve between them. We just need clear brand/terminology separation between them. And I would not need any new TTW features, just a unified way to easily use all the existing features just by uploading a zip file or editing uploaded file TTW similarly to theming.

1 Like