Matt

IP.Board 3.0.0 and 2.3.6 cannot share a database. You either leave it as-is for 2.3.6 or upgrade it for 3.0.0. You can copy your database and upgrade the copy. The subscriptions manager: Yes, it's a separate download. Yes, it's being worked on. I hope to have a beta / RC out early next week.

Just a quick note on small update releases. We do plan to make them fairly frequently but you don't have to upgrade if you've not encountered any bugs. You can wait until 3.0.2 or later if you wish.

If you're interested, I'm making a start on the hooks and last few bits today.

You guys make my head hurt.

You can click the green plus on a tab to go full screen, if that helps.

Remember there is the option to create skin specific 'master' folders so you can work outside the ACP in your favourite editor. No matter how fancy a JS editor gets, it's never going to replace what you use on your desktop so we figured why even try?

Regarding the subscriptions manager. I'd like to point out that all of us at IPS take it very seriously and no one here wants to see it neglected. We realise that many of you monitize your forums using the SM so we have no plans to remove it or make it a poor cousin. Our immediate plans are: - Refactor the subscriptions manager from 2.3.6 to 3.0.0 [Done] - Tidy up the ACP interface [Done] - Fix reported issues with expiring subscriptions, upgrades and PayPal [Done] - Write hooks and add settings [To Do] - Release to QA [To Do] - Release Beta / RC [To Do] I spent most of last week and identified and fixed several issues. Most of which can be back-ported to 2.3.6. More details to follow.

Regarding the subscriptions manager: it is a separate application and can be released independently of IP.Board 3. Think of it this way: you could wait for two weeks for IPB 3 + SM or we could release IPB 3 and SM two weeks after. Either way, those that need the subscriptions manager will have to wait two weeks. Note: "two weeks" is for illustrative purposes only and is not a written-in-stone release schedule. I have been working hard on the subscriptions manager and it's almost done and ready for testing.

It is, yes.

The only thing that was removed was the search index which was to do with how items were saved before searching. The actual form hasn't changed. It may appear basic most existing items from 2.x are still there but only available once you've searched.

Much better now it's not bogged down by our 5/6 other sites, etc.

Josh has completed 95% of the subscriptions manager. I'm going to sort out the gateways (PayPal, etc) early next week. Keep an eye on our blog as we'll be asking for testers next week to help debug the subscriptions manager.

A lot of people get a rush of nostalgia and proclaim they miss the IPB 2 skin. If we actually forced it upon them for a month, they'll be pining for the new skin. It's human nature.

Well, those were mostly due to bugs which we've fixed. Apart from the odd spike when a cron is running load is very stable and speed is moderate.

Generally we recommend that customers chmod their conf_global.php file so that it is only readable by PHP. This makes it difficult to make a tool to edit that file as you'll just get an error saying that it did not have permission to save it. How often are you going to toggle FURLs? Once a day? ;)

We're not rushing anything. Although we are labeling the last 3.0.0 release as 'final' it would be short sighted to expect it to be 'final' as in 'the definitive most perfect edition'. As with any software there are always going to be bugs and brand new major releases are rich pickings. We could either dawdle around with RCs for the next month chipping away at the dribs and drabs reported or just release it and open the flood gates. We'd never release software that was obviously and seriously flawed but a few minor issues in hand isn't going to do any harm. What ever happens, we're going to get hit hard when we release it as 'final' because it'll be used more.

I'm going to switch that debug bar off. People just obsess over it. Refresh. ooh x.xx secods. Refresh. ooh, faster. Refresh. Oh, a bit slower. Etc. Guys, we are on a VPS with other sites. We routinely have 500+ users online. We are being spidered heavily. Expecting loads of under 1 and load times of under 500ms is unrealistic on this forum. Just wait and see how it does on your own server. And if it's 100ms slower, don't worry about it there are worse things in the world.

How on earth would you create a sitemap file for a board that had thousands of topics? I know that most search engines accept multiple site maps gzipped but it's a lot of overhead considering that on a busy board you'd have to re-create the sitemap every few hours to make it relevant. We will include a sample robots.txt file with the next release.

Indeed we are. Josh and Brandon spotted some bugs where incorrectly formatted BBCode can cause the process to hang. Disabling BBCode seemed to stabilize the server.

I used to use AOL dial up on a 33k modem. 10 seconds to me is fast.

Yes, we added that in. I wrote a function called IPSLib::wait( $seconds ). We felt that life was moving too fast and wanted to slow it down a bit. We think that huge delays in serving up webpages will be a popular feature. Even Google will do the same one day. Just imagine, having to wait for 5 seconds for your search results. I know, it's a crazy concept but I think many will appreciate it. We suggest you think about a loved one or task to do in that 5 seconds. Maybe take a peek out of the nearest window. Your choice, really. We're giving you the gift of time, it's up to you how you spend it.

Thanks, fixed.

20k people online is a substantial amount. What server(s) are you running and what other software (memcache, etc)?

What Michael said. Deleting posts is incredibly expensive. Each post has to be examined for attachments, etc then the topic and forum then board stats updating. It's just not possible without refreshes or an ajax style interface (and even then the possibility of the user navigating away or losing their connection could lead to unpredictable results). Something for the future maybe.

I've been writing and releasing code for 10 years and some things never change. :) Please understand that we are all working very hard to fix all the reported issues found. We will not release RC 2 until we feel that it is stable. It may or may not still have bugs. The 'final' version will probably have minor niggles here and there that are currently undiscovered. IP.Board 3.0.0 is a new platform and contains mostly new code. We realise and understand that there are going to be bugs that need fixing even in final which is why we are planning to release small maintenance releases relatively frequently.