Time to re-evaluate one-click offerings?

apps

#1

Seems to me DH is doing the user base a dis-service by still offering dotProject. It hasn’t been updated in years (according to the website). There hasn’t been a single request for info here in a year or so. The web2Project app, a spinoff from dotProject, has had a few updates within the last year, but even there it looks like developers have largely lost interest.

As much as I like the OpenVBX software, I’d have to say the same about that one. It’s simply not maintained and it’s becoming tougher to get answers about basic usage. There’s almost no one still looking at the source anymore and most of the plugins are not maintained.

I wonder how many other apps are in this condition.

This sort of thing simply happens to FOSS projects. Some one individual gets a brilliant idea and spends every free moment to create excellent software in a fit of passion that only another developer could appreciate. They convey their passion to others and with any luck a Team developers. With luck, the software gets a lot of updates over time, the issue tracker is packed with open and closed enhancement requests as users become enamored and dependent on the software. But over time developers find other things to occupy their free time. They grow weary of the abuse of angry and demanding users who contribute nothing. They have families, jobs, and other things to do. The projects lose developers and eventually go stale. New developers aren’t interested in old software written by someone else - they have their own new and brilliant dreams to pursue. The user community is left with thousands of dead and under-maintained FOSS packages, killed off sometimes by over-popularity and people who don’t “get” the thing about Liberty versus Beer.

With all of the PHP developers in the world, and the popularity of the LAMP stack, you’d think this sort of thing wouldn’t happen. There would seem to be an endless supply of people to keep these popular packages alive. But the lack of stability of a “quid pro quo” model in the FOSS model has created this inherently volatile situation. More plainly stated, developers simply aren’t motivated to fix code for someone else’s benefit, especially when almost everyone expects their software to be “free” as in Beer.

We see this same thing happening with FOSS components as well as main projects. Drupal is hugely popular, but with every release we find modules that no longer work, and no one wants to maintain them. Part of the unspoken agreement that people make when they get a shared host and load it with FOSS, is that the software cannot be guaranteed to have the same features from one year to the next. You will upgrade, modules will break, you may never get them working again, you will have to decide to let some of those modules go and to adopt new ones which have a similar life cycle. The alternative of course is to compensate developers to help maintain software that you find of value. But really, how many people do this? Rather than paying someone to fix FOSS, most people would rather dump a module, and like a locust just move on to the next one. This is the “free Beer” issue playing out.

Well, how many of the One-Click packages have followed this pattern? I don’t know. Anyone? Is it now a dis-service to saddle DH users with software that has no future, something that they will come to regret later - and perhaps blame their host for the time lost? How many DH users realize that they’re installing a dead package when they do it? How many site managers are savvy enough to realize a package is dead and not use it? … Can DH see that reflected in decreasing popularity of specific One-Click installs?

Is it time for DH to review these packages? Stop offering new downloads for dead apps? Encourage users to start looking elsewhere? Start looking for alternatives?

Or might DH consider becoming the savior of the FOSS world, taking in these orphan packages, compensating developers for their time, making a real business out of getting people to understand the model, and getting them to pay for keeping their software of choice alive?

Is it time for me to shut up? Yeah, I can answer That one. :slight_smile:

Thanks for your time.


#2

Big heaps of them man. Big bloody heaps of them, sadly.


#3

I should have clarified there “I wonder how many other One-Click apps hosted at DH are in this condition”.

Yeah, on the world stage SourceForge, CodeProject, GoogleCode, GitHub, and other repositories are mass graveyards of dead code - as well as nurseries for new apps and lively homes for actively maintained apps. None of them that I am aware of provide a mechanism for auto-deprecation and obsolescence. In other words, when software hasn’t been updated in 8 months, it can no longer be downloaded and update routines return a warning of pending obsolescence. When not updated in 14 months, forums get closed - that will get people’s attention. When not updated in 18 months, trackers get closed and projects are marked for removal. And when not updated in 24 months, projects are completely closed.

Of course that’s brutal and incomplete, but it gives the world a basis for argument about where our tolerance is for dying and dead software, where right now we have none. And it gives companies like DH a metric for removing software from their own list of recommendations and offerings.


#4

Sadly, we don’t have magic pixie dust here at DreamHost which we can sprinkle on projects to make their maintainers keep working on them. (Or, if we do, nobody’s told me where we keep it.)

We try to pick applications which are under active development, and which seem to have a bright future ahead of them. We can’t call them all, though, and sometimes we pick some real stinkers. If you feel that there’s some applications on our installer which shouldn’t be — or applications that aren’t that should be! — please go ahead and let us know with a post in the Suggestions subforum.

As far as DotProject goes, I think you may be misreading their site. The project certainly isn’t bursting with activity, but they released two versions (2.1.6 and 2.1.7) in the last few months.


#5

Thanks for the response, Andrew. At least we’re at that point where we can discuss the grey areas. Until the discussion is opened and the packages re-evaluated, people are unknowingly going to be installing dead and dying software. That’s the problem I’d like to address.

As one more example, DH still offers Gallery2, but that’s fairly dead as well, being replaced by Gallery3.
Having tried both of those, I decided to look at Piwigo and I’m simply awestruck at the higher quality of Piwigo.
(And BTW, I don’t use One-Click except rarely to get a quick impression of how software works. This is for others…)

We can argue about quality and preferences, but when the package being offered is on the slab, we need a new discussion.

I’ve over-stayed my welcome in this thread. Thanks for following…


#6

My optimistic side likes to think that some of the dead projects are “complete” :smiley:

That’s another crazy thing that would make a selection process difficult - some of those projects are actually finished to the desired scope (final product) and just don’t need or want to become bloatware by adding on superfluous gimmicks in order to remain “active”. Remember when WP would run on 2MB and Joomla! would actually display a page in one second flat without requiring a minimum 4 core dedicated server? Ahh those were the days!


#7

@sXi, that’s funny n all, but it just doesn’t fly.

Consider another example - I’ve recently dumped Drupal and started to use the ModX CMS. The One-Click install is a very old version (Evolution) that’s not supported anymore. So someone installs that software, has issues, and they go to a forum where everyone tells them to upgrade (to Revolution).

No, better for new sites to get something current and actively supported.

BTW, ModX is great software but the community is a little disorganized - look for great things from them in 2013.