If you were handling a switch to vBulletin, how long would it take you?

[Crono]

Originally posted by: sundev
Any place to read up on why the switch to vBulletin, etc?

Essentially, FuseTalk has broken features, is missing desired features, and the support for it hasn't been particularly satisfying.

According to DerekWilson:

Originally posted by: DerekWilson
Hi guys ... I wanted to mention this in the PFI thread because I think it spoke to the issue at hand and I wanted to provide some information about what we are doing to fix certain issues of which we are aware. Specifically, vB allows the automated tracking of moderator warnings and vacations etc, which will go a long way to helping cut down on the workload our moderators have (currently we manually track all warnings and violations) -- and this will help to improve the experience for members and moderators alike.

I didn't really want to talk about it publically until I had something more to show you on it though. Some have pointed out that vB has an importer for FT, but this is for FT2 and not FT3 -- in addition we have customized our FT package to the point where we can't rely on their import feature and we must write our own software.

This is currently in development but we do not have a timeline and it may be a long process.

We want to make many changes including refining the way we manage and moderate, and to that end, we required some tools that, at this point in time, FuseTalk does not include. There are also some security issues that I'm not willing to discuss at this point that have also prompted our desire to move to a different forum package.

As for look and feel, it is my goal to make the new forums as clean and professional looking as possible. We haven't even started working on the UI yet as this will happen after we know our data can make it successfully into vB ... if you read into that a bit you'll see I'm hedgeing a bit there -- we may hit a problem that prevents the move from happening, but at this point things look certain enough for me to be comfortable talking about the fact that we are working on this.

As for what you guys need to know, we will do our best to import all the threads we currently have, archived and otherwise, as well as member info and post counts. We will keep as many avatars as we are able to (some of the avatars we have aren't exclusive to FT). We will try to minimize downtime when we eventually do try to switch, but at some point before that we will have a semi-closed beta open by invitation only ...

anyway that's about it for now.

 

[K1052]

About a year.

First I'd proclaim the switch then follow that with months of near silence on the subject, eventually set up a limited test to give the illusion something is actually happening. After a long enough time has elapsed that almost everyone has forgotten about the promised move I'd then request suggestions on how to improve the user experience (none of which will never be implemented into whatever version actually ends up being run).

At some later point when my trained chimpanzees have finished flinging poo at each other I'll task them with the finer points of the migration and bring the entire board live without any sort of testing whatsoever.

I will then disappear to Thailand for 3 months to enjoy a much deserved rest assuming all has gone to plan and that any user issues encountered are trivial and can await my leisure to be addressed.

 

[Crono]

Originally posted by: K1052
About a year.

First I'd proclaim the switch then follow that with months of near silence on the subject, eventually set up a limited test to give the illusion something is actually happening. After a long enough time has elapsed that almost everyone has forgotten about the promised move I'd then request suggestions on how to improve the user experience (none of which will never be implemented into whatever version actually ends up being run).

At some later point when my trained chimpanzees have finished flinging poo at each other I'll task them with the finer points of the migration and bring the entire board live without any sort of testing whatsoever.

I will then disappear to Thailand for 3 months to enjoy a much deserved rest assuming all has gone to plan and that any user issues encountered are trivial and can await my leisure to be addressed.

Derek, is that you?

 

[KingGheedora]

I wish I could help out with the project. I have experience with every facet of developing and running a large data-driven website, including data center moves, and site re-launches.

 

[Platypus]

Originally posted by: K1052
About a year.

First I'd proclaim the switch then follow that with months of near silence on the subject, eventually set up a limited test to give the illusion something is actually happening. After a long enough time has elapsed that almost everyone has forgotten about the promised move I'd then request suggestions on how to improve the user experience (none of which will never be implemented into whatever version actually ends up being run).

At some later point when my trained chimpanzees have finished flinging poo at each other I'll task them with the finer points of the migration and bring the entire board live without any sort of testing whatsoever.

I will then disappear to Thailand for 3 months to enjoy a much deserved rest assuming all has gone to plan and that any user issues encountered are trivial and can await my leisure to be addressed.

I always knew you were into Thai guys

 

[silverpig]

So they want to make windows stuff work on linux? I'd just use WINE.

 

[Evadman]

<= 2 weeks if I was 100% dedicated to the switch, and didn't have to answer to anyone except the end user. That time assumes I am intimately familiar with the underlying database structure of both systems. If I am not, tack on an additional 2 weeks for mapping and data constraint testing between the 2 systems.

Honestly, anyone who says < 2 weeks is full of shit, and is just guessing based on some made up bullsh!t process in their head. Failure is not allowed on a website. If the move fails, I would expect a GIANT drop in web traffic for a long time.

 

[allisolm]

Originally posted by: sao123

For a changover like this, Id start with a fresh blank system and force everyone to reregister.

There's going to be enough of that anyway when all the members who don't have a valid email address can't log in.

 

[olds]

What makes anyone think it's not already up?

 

[Q]

Over 9000.

 

[K1052]

Originally posted by: oldsmoboat
What makes anyone think it's not already up?

FuseTalk Enterprise Edition - © 1999-2009 FuseTalk Inc. All rights reserved.

 

[olds]

Originally posted by: K1052

Originally posted by: oldsmoboat
What makes anyone think it's not already up?

FuseTalk Enterprise Edition - © 1999-2009 FuseTalk Inc. All rights reserved.

Powered by vBulletin® Version 3.7.3
Copyright ©2000 - 2009, Jelsoft Enterprises Ltd.

 

[flashbacck]

Originally posted by: K1052

Originally posted by: oldsmoboat
What makes anyone think it's not already up?

FuseTalk Enterprise Edition - © 1999-2009 FuseTalk Inc. All rights reserved.

:laugh:

 

[K1052]

Originally posted by: oldsmoboat

Originally posted by: K1052

Originally posted by: oldsmoboat
What makes anyone think it's not already up?

FuseTalk Enterprise Edition - © 1999-2009 FuseTalk Inc. All rights reserved.

Powered by vBulletin® Version 3.7.3
Copyright ©2000 - 2009, Jelsoft Enterprises Ltd.

Define "up"

 

[Crono]

Originally posted by: oldsmoboat

Originally posted by: K1052

Originally posted by: oldsmoboat
What makes anyone think it's not already up?

FuseTalk Enterprise Edition - © 1999-2009 FuseTalk Inc. All rights reserved.

Powered by vBulletin® Version 3.7.3
Copyright ©2000 - 2009, Jelsoft Enterprises Ltd.

Powered by DarkCentury? Version 15.2.1
Copyright ©1997-2030, Skynet, Inc.

 

[olds]

Originally posted by: K1052

Originally posted by: oldsmoboat

Originally posted by: K1052

Originally posted by: oldsmoboat
What makes anyone think it's not already up?

FuseTalk Enterprise Edition - © 1999-2009 FuseTalk Inc. All rights reserved.

Powered by vBulletin® Version 3.7.3
Copyright ©2000 - 2009, Jelsoft Enterprises Ltd.

Define "up"

Text

 

[xSauronx]

Originally posted by: Leros

Originally posted by: xSauronx

Originally posted by: yllus

Originally posted by: Leros
Why does size matter? Wouldn't the script be the same regardless of size?

Nope. A single server setup is no big deal, but when you have to scale the site over multiple servers (that is, multiple Web frontends and multiple database backends) things get complicated in a hurry.

Add to the fact that the AnandTech Forums currently run on a Windows environment and the people who administrate it probably have little in the way of experience with the Linux-based Apache, PHP and MySQL - it'll take time to do right.

and the database is huge. at least, thats what anand claimed was the biggest hold up when he was at the raleigh meet. they want everything and its a massive amount of stuff to move over

How big could anandtech possibly be? Its all text. I would imagine 1TB to be an extreme upper bound.

im just telling you what they told us. he didnt give a specific size but anand said "tens and tens of gigabytes of text" so even a terabyte is probably pushing it, they just made it sound like the conversion was a bitch to get done, and that they had wished to have it done before now.

so they were kinda vague but sounded like even they were frustrated that it was taking so long. dont know what else to tell you.

 

[TruePaige]

Hmm...~2 weeks of full dedication, no distraction to allow adequate time for testing.

Really is all about the DB though. If the systems are hard enough to adapt to each other could be a month.

I seriously doubt this site is spread across multiple DB servers, and if it is...what the heck...judging by the most concurrent online users I ever see a single server could handle the forums.

 

[mb]

Bout 3 hours 50 minutes.

 

[Hacp]

Do we get to keep our hard earned experience points? I spent 4 years grinding this shit out.

 

[ultimatebob]

Originally posted by: Evadman
<= 2 weeks if I was 100% dedicated to the switch, and didn't have to answer to anyone except the end user. That time assumes I am intimately familiar with the underlying database structure of both systems. If I am not, tack on an additional 2 weeks for mapping and data constraint testing between the 2 systems.

Honestly, anyone who says < 2 weeks is full of shit, and is just guessing based on some made up bullsh!t process in their head. Failure is not allowed on a website. If the move fails, I would expect a GIANT drop in web traffic for a long time.

It sounds like they're not moving the data, though, and deploying vB from scratch. That should only take a few days to configure a beta forums for people to test, and then a week or so to configure the live site.

 

[sao123]

Originally posted by: oldsmoboat

Originally posted by: sao123
15 minutes.

For a changover like this, Id start with a fresh blank system and force everyone to reregister.

That's the plan anyway.

NOOOOOOOOOOOOOOOOOOOOOOOOOOO
/vader



waves goodbye to postcount... almost made lifer too.
/cries

 

[Zolty]

20 minutes or so, but then I don't really care about people's data or posts.