On 7/7/20 12:47 PM, Wolfgang Hartmann via Webobjects-dev wrote:
Although I am not working anymore with webobjects but with more recent
technologies I was still the one who did the last updates on the
wolips-project. And therefore here is my opinion about the discussion
of work to be done on the wolips-plugin:
As you can see in the commit-history for my account "Wolfy42"
(https://github.com/wocommunity/wolips/graphs/contributors) there was
some annual necessary. The reason for that was that the wolips-plugin
is based on the completely outdated eclipse e3-platform. Since around
2012 all new versions of eclipse are based on the e4-plaform. Eclipse
knowns that not all plugins are actively maintained and the upgrade to
the e4-platform can be a lot of work and therefore there are a lot of
compatibility-layers which could be applied so that also old plugins
are capable of running in new releases of the platform. That was the
annual work I was doing in the wolips-project.
I was not even aware of the e3 vs e4 issues, so this information is
appreciated.
But this compatibility-layer is also the reason why the plugin is
rather slow and that it is possible that there is more annual work
necessary. Every year there is a new version of eclipse with the
possibility to break a lot of old plugins ...
I think it would be relevant to break down the parts of wolips and
consider which part should be maintained or updated. I think the major
parts of the wolips-project are:
* The Component-Editor with the Binding-Validation: I think this is
the most relevant part and should be completely updated with a current
HTML-Editor and it should be upgraded to the e4-platform. And the
race-conditions in the binding-validator should be fixed. If you look
at the amount of "synchronized"-keywords in the
binding-validation-code it should be consider to rewrite it from
scratch. The ruleset how the bindings have to be validator can still
be reused
Agreed.
* Hot-Code-Replacement (=automatic component-class-reloading) after
changing the components. This is a very relevant part and should be
automatically available without the need of third-party-technologies
like JRebel
* The EO-Modeler: This is still a relevant part when working with
existing projects but there are other good
database-domainmodel-modeling-tools out there. It should be considered
if one of them should be extended to just generate the .plist-files
and .java-files necessary. Maybe even the cayenne-modeler could help
here because I consider cayenne to be a modern version of EOF
There probably are better tools. But...
* The EO-Reverse-Engineer-Tool: This should be completely ignored
because no new project should be started with EOF. If someone still
wants to use webobjects than at least the database-layer should be
based on cayenne or JPA-Hibernate.
I strongly disagree with this. I can create a database and tables, shove
some data in, create a Wonder project, reverse-engineer, add a wo:loop
and a wo:str to a main page and see data in a browser in under 5
minutes. Which other tools can do this?
* All the "wizards": The only wizard necessary is the one to generate
new Components (all the .wod, .woo, .api and .java-files necessary).
All other wizards are irrelevant
* All the "refactorings": The only one relevant is the one to rename a
component (=rename all the files/folder in one step). The other ones
are not relevant.
Although webobjects is long dead and not maintained for 12 years now
there are still some great and maintained projects built with
webobjects out there and therefore it is relevant to have a a good
development-environment. Although sometimes I had the feeling that my
eclipse was hating me it is still a good IDE.
I absolutely bow down to Wolfy for his wolips work. I certainly have not
managed to contribute, though I have set up the wolips dev environment
several times. Getting started doing this dev work is just a huge pill
to swallow all at once.
But let's not get rid of things that are still very usuable and very useful.
And I agree with another's comment that we should make sure the issues
are clear and reproducible before we point someone at them.
cheers - ray
Best Regards,
Wolfy
------------------------------------------------------------------------
*Von:* Paul Hoadley via Webobjects-dev <webobjects-dev@lists.apple.com>
*Gesendet:* Freitag, 3. Juli 2020 11:33
*An:* Hugi Thordarson <h...@karlmenn.is>
*Cc:* WebObjects-Dev <webobjects-dev@lists.apple.com>
*Betreff:* Re: WOLips bugfixes and new features planning
On 2 Jul 2020, at 20:25, Hugi Thordarson via Webobjects-dev
<webobjects-dev@lists.apple.com
<mailto:webobjects-dev@lists.apple.com>> wrote:
Many of the issue reporters are still with us. I suggest announcing a
short grace period where people can look at their own issues (and
other issues, of course) and add the "keep" label to issues they want
to keep. When that grace period expires, all unlabeled issues are
closed. Worst case scenario: We have to re-open a closed issue. Most
users probably get an e-mail notification when their issues are
closed anyway so they can complain at that time :).
I support this approach. Issues that aren't *actionable* for any of
the reasons you describe should be closed.
Regarding funding, I would be very willing to operate on some sort of
a "per feature/issue" basis. I.e. I'd dedicate a fixed amount of
money to the resolution of an issue. Perhaps that also solves the
problem of prioritization? I'm guessing the issues most valuable to
the community will end up being the ones with most funding attached
to them.
My preference would be prioritised list + pool of money → work on the
list until the money runs out (just because it's simpler). But I'd
participate in your system too if you want to run it.
Regarding specific issues, there's one issue I'm *really* interested
in: I've attempted to do WOLips development on some occasions but
always gave up since I didn't get everything to work (the docs are
kind of convoluted/outdated).
Perhaps I'm just stupid, but if not; I believe we would benefit
greatly from having a functional, up-to-date step-by-step guide for
how to do development on WOLips. Teach a man to fish and all that :).
This would be good.
--
Paul Hoadley
https://logicsquad.net/
https://www.linkedin.com/company/logic-squad/
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
https://lists.apple.com/mailman/options/webobjects-dev/ray%40ganymede.org
This email sent to r...@ganymede.org
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
https://lists.apple.com/mailman/options/webobjects-dev/archive%40mail-archive.com
This email sent to arch...@mail-archive.com
