Most players also didn't download from WAU every day. And you're right, I missed the difference between Beta and Development, but I'd make the case that Beta is more likely to be the version that, under WAU, would have been checked into the trunk, while development would (should) have been checked into other branches not exposed to the public. Again, Curse's implementation is not fundamentally different from WAU's in that regard.

While that is certainly possible, most authors on wowace just checked their changes in as they made them and let god sort 'em out. When they broke something they checked in fixes. They didn't pull branches off the trunk and deal with release methodologies or any of that stuff because ace didn't really support that, because wowaceupdater wasn't originally meant for enduser use. Curse enforces the dev/beta/release methodology, because mods not tagged "release" won't make it to their website. If mod authors decide to circumvent this by just tagging everything release, well, that'll be a real shame.

SVN updates and Curse file mirroir is not the same thing. Even if you set Downloader to alpha version it won't pull every single update.

Are you saying that the intention was for Ace Addon developers to always develop on a branch, and then only merge 'release' versions to the trunk? Seems kind of a strange way to use SVN to me, but then I'm not really that experienced when it comes to version control.

Do you really think that the Ace Addon developers (whatever that means) are knowing much about SVN and serious usage of svn ? I am an addon developer, part of the Wowace community. I am also a professional developer, where I use (in a somewhat more sane way) SVN, but coding addon is a hobby, not a job.

A lot of people are asking me to update my addons right now. I didn't update most of them, or did not release the update, because I think I have the right to enjoy patch day like everyone else. I have taken my characters in different instances, I have done a few achievements, started to level inscription on a reroll, tried to fix my UI, set up a raid on the Calendar for today, ... pretty much the same thing everyone else has done on patch day.

I, as a developer, completely support the wowace change, even if it is a burden for me and not yet as easy to use as the previous one. I am glad for the tremendous service that wowace (funded by curse) as given me and continue to give me and I have absolutely no issue with curse trying to make a living out of it. I never wanted to make a profit of the addons I wrote, and I understand that providing the hosting service has a cost.

Maybe the final user should, too, be a little more supportive of the hosting provider and correctly value the service offered instead of just thinking about the author.

What?

"Knowledge of SVN" doesn't mean you don't check in small changes. It doesn't mean you don't do active development in trunk, either. Your understanding of how the majority of wowace developers used wowace SVN is a little bit naive (read: incorrect) as well.

I have to laugh as I read this posted over and over again. Wowace had a paypal donate button on its site for a long time. How many of you actually donated? I'll go on record right now that 80% of the people who say they did are lying.

There are a lot of generous people out there, but there are a lot more who figure someone else will take care of it. After the fact, everyone is the generous sort who would donate to get back what they now miss. Funny how the urge to donate lessens once the rubber hits the road though.

As someone else stated before me: if you think it's easy to put together a site which distributes lots of addons, has forums, is user friendly, and attracts talented developers and great addons, get started. I'm sure everyone here is just waiting for you to open your doors. But don't try to make any money while you're doing it, even if it's not covering your costs. That means you're anti-open source!

I've yet to really see anyone griping about having to pay except for who they'd be paying too and that's really a perception issue with Curse.

If the announcement would have been "ok our bills are stacking up and we're crying uncle so now you have to pay to download using WAU or get your mods elsewhere" I think the community reaction would have been a lot different than what we had with Curse, which already had a bit of a shady view within the community, coming out and basically saying "not only is our new client not open source and therefore not open to peer review, but it's going to sit in your task tray doing mysterious stuff and here's a privacy/terms of use policy that's broad enough we can essentially snoop around all we want and sell your information to 3rd parties".

The reality is that the client probably isn't really ever going to be doing anything particularly "bad" and since we can still monitor it with stuff like Process Monitor and other tools, if it ever did, they'd likely get called on it in a publicly embarrassing way. But the marketing and community management (or lack thereof) around this whole thing hasn't exactly been giving people any warm fuzzies.

As an aside, in case someone from Curse actually reads this thread anymore, I'd like to humbly request that you start digitally signing your client and installer.

There are as many different ways of using SVN (or any source code repository) as there are users. At a minimum, there are people who believe you should do all dev in trunk and use branches for stability releases, and those who believe in using branches for dev and keep trunk always stable, merging down only for releases. Neither one is "right", they are basically mirror images.

Your second set of points are the heart of the matter though. Yes, everyone agrees that if only devs all properly labeled their check-ins and builds, and if only WAU had a way to set what quality of check-ins you wanted to get, and if only, if only, if only...

There is probably a parallel universe out there where this is exactly what happened, and somehow all the multitudes of SVN contributors were arm-twisted into doing the right thing, even though they were doing it for free and had no real leverage held against them. It takes an endless supply of enthusiasm, time, and energy to wrangle and lead a divergent group of resources you don't have an employer-employee relationship with.

Alas, in this case Kaubel, ckk, and the gang instead went a different route which is more fulfilling to them, or for one of a hundred different possible reasons.

As I mentioned, if people think they did wrong, and there is a better way out there, the market is ready, willing, and able to see if they can deliver. I'm no curse apologist, I just think people are far too willing to complain that all the free shit they got isn't available any more.

"I just want an updater that is easy to use, works out of the box, is fast, does everything i want and nothing I don't, and is free, though I'll say I'm open to donating as long as I don't actually have to enter in any account numbers." Yeah, right. And I want a pet unicorn. If there's a will and a way, someone will provide it. If it can't sustain itself, I'll see you back here in this thread when it implodes under the unfunded weight of its success 12 months from now.

That is my point. People aren't griping about having to pay because there is a market force at work which prevents it, and no consequences if they don't actually follow through on their alleged promises. As soon as they are back to getting what they want for free, you better believe the vast majority will find excuses never to click that donate button, just like they did all those months when they thought WA was doing fine and they were happy as clams getting their free downloads.

I've witnessed great success of voluntary donations funding a project before. I'm associated with one now. I don't know how or why it works - perhaps it's just a perfect storm. But I know that it rarely works as well as it does in the few success stories.

Just my opinions. YMMV.

I have no problems with curse making money through ads or even charging for value-added services. More power to em. I just wish they waited until the curse updater had matured to a point where it was a good replacement for wowaceupdater before forcing everybody to switch.

I wish the same. I'm not them, so I can't say for sure why they switched now, but my (hopefully reasonable) guess is that they wanted to stop hemorrhaging money paying for bandwidth, and we all know that the 3.0 major patch is going to be responsible for a huge, huge, huge amount of addon downloading. Ready or not, the CC was getting thrown out for public consumption, and WA was getting locked.

Since maybe my post came accross a bit wrong, I'd like to clarify that I wasn't trying to imply anything about Addon Developers, Ace or otherwise, I was merely replying to Adoriele, who (as far as I can tell), implied that unstable dev versions "(should) have been checked into other branches not exposed to the public.", and to me this was a strange way of using SVN (using branches for release, trunk for unstable).

As Dinesh has since pointed out, there are many different ways of using SVN, none of them wrong, I just thought that this one seemed quite strange to me. As mentioned though, I'm still fairly knew to managing projects in version control.

Anyway, sorry for the offtopic about SVN usage!

This works beautifully for you and your setup, I agree.

But what about a company that has multiple versions and releases all in the wild at the same time? A real world example might be, I dunno, Oracle (8i, 9i, 10g, 11g are all still working somewhere, and Oracle supports maintenance releases on more than one platform at a time) or Microsoft (XP and Vista both get releases now). A WoW example might (in a different parallel universe) be Ace, Ace2, and Ace3.

Or what about a product that has simultaneous dev, beta, and release builds? Beta is intended to be stable. It doesn't get new features, just bug fixes.

To various extents, these scenarios don't play well with "everyone knows to go to trunk for the stable release" model.

I posted this because it's a subject I enjoy bullshitting about, but we're pretty far off-topic now, so I'll probably just shut up about it going forward.

*Shrug* It's still on-topic if something mentioned here gets incorporated into however Curse decides to run their development platform (not that there's a large chance of that happening), or someone takes some of these ideas and runs with them to create a "new" WoWAce community for those who hate Curse. Maybe something like "Best Practices: Running an addon development and distribution community", where we jaw on and on about how to set things up in a way that avoids the downfall the WoWAce saw.

And yes, there are definitely other ways to set up SVN that work differently in other situations, your Oracle example being a very good one. The reason I mention how I would set it up is that the WAU knew of only one way to grab addons for people (devs using actual SVN access excluded from this discussion for obvious reasons): grab the trunk. In light of that, it seems fairly obvious to me that, in order to minimize bandwidth costs and make updating easier on the end user, there should have been rules set up that the trunk should always be a functional version of the addon. How the developer made that work was up to them: develop in branches, merge to trunk; only commit changes after thorough testing in your working copy; whatever. Such rules could easily be implemented using SVN folder properties and such which, while the addon authors may not have any knowledge of, the WoWAce leads sure should have been able to do.

Your example of ACE, ACE2, ACE3, counters this, but those were also being written (if I'm not mistaken) by the same people who hosted/led/etc the WoWAce site and community, and it could have been set up to enforce the rules by default, while allowing people who knew what they were doing more unrestricted access to the SVN.

Sorry to get off topic but I thought if people are reading this thread looking for a Wowace replacement as far as updating your mods with one click and not wanting to use the Curse client. I have started using Wowmatrix located at Wowmatrix and it has worked great so far.

There's a fair amount of controversy with wowmatrix; they leech bandwidth from sites like wowinterface and curse without showing their ads, they don't give developers proper attribution, and so on. Google if you're interested. Even if you don't care about that stuff (and I don't blame you if you don't), many prominent authors have convinced wowmatrix to remove their addons, so their selection has suffered. If it has your addons it does work well, but personally I'm staying away from it and suffering with the curse updater.

Or hell, even before having to take that step, just limiting the downloads per IP to a certain metered amount per session. Being able to use WAU only once a week for under 5 MB of downloads or whatever would still be vastly superior to what we have now. (or only letting non-developers download whatevers tagged as "release" and not beta)

Also most of the code files were text and very small on their own, perhaps something should have been done about mod users inflating their archives with silly fonts, statusbar textures, variants of SharedMedia-3.0, and so on. Lossless TGA textures doubled or tripled the size of every download for many mods for nothing but aesthetic benefit (and even there benefit is a relative term, as 99% of SM statusbar textures were abominations I had no use for and dreaded having to go through and delete so they wouldn't clog up my small, orderly bar selection after a major update)

This thread/announcement was started in July, and the patch didn't come until October. Shut it down before the next major revamp if they have to, but 3 months should be adequate time (sylvanaar earlier quoted spending 1500 man-hours or about 2 months as an unpaid hobbyist to make WAU what it was) to build a mature little utility app better than what the Curse Client is now (esp. with real corporate expectations and the nature of having total closed-source control over it and not having to muck through the open contributions.) If it wasn't ready for Windows/Mac/Linux platforms with no-lib updating (still a problem a week after the patch with MANUAL updating, as in I don't see an obvious link on the Curse site itself for 3.0-ready no-lib mods,) etc. then it shouldn't have been talked up as a viable solution that would be ready on patch day (and this is assuming the WoWAce guys were only sure about the transfer in July when they announced it, the Curse Client has been in development since May)

It's just last version is so much better . Current version is pretty stable, although it still have some problems with identifying addons and basically you need to reinstall every addon without libs.

I was actively against the Curse client in this thread and others around the world...of warcraft, so I thought I would give some updated impressions.

After getting tired of checking for updates on the curse site using a favorites list, I broke down about a week after 3.0.2 came out, and installed the client. The client has certainly become much more stable than initial builds. It installed some different versions of addons I had already, but otherwise did quite well from the start. I minimized it to my tray where it's been sitting ever since. I've never had any addon issues at all since then, with quite a large number of addons installed. (albeit, every one of them from curse)

I get the urge to check up on it every few days, and make sure my addons are up-to-date. At first I was confused as to why clicking the "check for addon updates" button wasn't doing anything. I realized it was just automatically checking for updates on whatever schedule it has, and my addons are always up-to-date to at least that day. It appears to be getting stable versions of all these addons, nothing is giving me any grief in game at least.

I'm sure it's reporting back to the servers with my name, credit card numbers, and favorite types of porn, but I suppose I can live with that for the convenience factor. Dare I say it's better than WAU ever was now?

So I suppose once this little free trial expires, they've probably got my money! Seriously impressed with the progress made on this thing.

I've had similar success keeping my addons stable and in working order, but by keeping a simple spreadsheet, where each row has a column for the name of the addon, the version, last updated date, and a URL I can click on to go to the addon's web page.

I found that despite my old addiction to WAU, very, very few addons need to be updated regularly, and I rather enjoy knowing when a new version of an addon hits my machine in many cases, so I know what new features or changed features to look for.

No doubt, WAU was an addiction, but I'm finding that the alternative to running an application in the background of my computer and paying money to do so is really no big deal. I run around 50-60 addons, by the way, not all found at curse, and after the initial week or two after the 3.0.2 patch I have spent very little time on addons, perhaps an average of 30 minutes a week.

For people who liked WAU, but don't like the Curse Client, I think the main argument is "installation preview", or rather, a lack thereof. The Curse Client is getting there (I've been following some of the discussions on CurseForge and #wowuidev on IRC), but WAU had a really nice listing to show what mods would be updated, and to what version, when you clicked update. If you were on a Mac, WAU had a native interface as opposed to a cross-platform Java interface (meaning that various Mac shortcuts worked as expected).

That said, this mattered more when you consider the way SVN was used on wowace (see the October posts in this thread), where some developers were checking in very buggy versions that WAU would happily download. The Curse Client will only grab Beta or Release packaged versions of mods. As a mod developer, I can check in as often as I want, and those builds will get packaged up as alphas on CurseForge, but until I tag my mod as beta/release it won't show up on Curse or be available to the client.

My sense is that right now the Ace -> Curse migration has been mostly about ironing out the hosting costs, and making development easier; the improved usability will come soon. For now, I think the Curse Client is not as good as WAU, but still does its job.

The curse client supports that methodology. Just set it to release or beta and turn off autoupdate. I believe that matches the non-subscriber featureset anyway. It's missing a few features, and it's unpolished and slow, but if you're careful and leave automatic processes disabled it's quite usable now.