Koozali.org: home of the SME Server
Legacy Forums => General Discussion (Legacy) => Topic started by: djhomeless on December 29, 2005, 10:04:37 PM
-
My New Years Feature Suggestion is an interesting one, how about working towards a set of requirements that would allow you to run SME in a hosted (ie co-lo) environment.
Hosted environments the world over are starved for a platform like SME. Take a look at any webhost, the best they can do is trot out a 5 year old version of RedHat on the old Cobalt platform, or some overpriced management system like Plesk or CPanel. Or, of course, Windows 2003.
Simply put, there are no good apps or platforms to manage individual servers for businesses big and small in the hosted world.
Case in point: My company. We have our own internal infrastructure but have a need to have an external server to be run by a small marketing department. They have their own developers who also want to manage the server. They would love to learn Linux, but as there is nothing out there that can help them through the learning curve, the ultimate decision was made to go with a Windows box (arrrgh!).
Years ago I got my feet wet running a Cobalt Linux box. Cobalt, for what it was, was great. I got a nice web interface, I didn't really had to touch the command line, and it was (for the most part), stable and easy to use.
I don't think there is a lot that would be needed really to make this happen. The upshot, I think, would be huge. Rolling SME out to a wider audience would bring more people to our community, and bring more people to help make SME better.
1. Give the admin the ability to upgrade SME via the network (either via YUM or the Server Manager).
2. Give admin's the ability to toggle password strength via the admin panel (which can be done today from the command line).
3. ...the ability to toggle between IMAP/POP or IMAPS/POPS.
4. SSH enabled by default.
5. Load balancing/failover of Ibays. ie Ibay data can be setup to be mirrored on another SME. So, if one server falls over, a switch somewhere can send traffic to the second.
6. Ruby on Rails support. Ok, so this is not really a co-lo requirement. But it still would be cool!
7. Oracle support. Oracle can install on CentOS 3.4, so I assume it wouldn't be to difficult for SME.
8. Web-based interface for the templating system. So, instead of a user having to create template fragments and then manually rebuild/expand that template, just manage it in the server manager by treating each template change as a versioned object (so you can rollback....)
9. Jabber server support (controlled IM is becoming big in the enterprise).
If you really think about it, not much from the above list actually deviates from the core SME mission imho. SME is a great product, run by a great community. I really would love to see the day when SME is being run in every respectable hoster in the land!
By the way, just so we are clear, this is not some backdoor way to gripe about the state of the distro. With the exception of number 3, I do not require any of the above for me personally. I am quite happy with my SME 7 thank you! :)
Happy New Year everyone,
Geoffrey
-
My New Years Feature Suggestion is an interesting one ...
"SME Server Future" in the bug tracker is the place to post New Feature Requests.
Your wish list seems rather long - unless some of the businesses benefitting from SME start investing in R&D, it's quite unlikely much of the list will be realised any time soon.
IIUC, your item 3 is already included in SME 7.
-
Hi Charlie,
I don't want to submit this as a FR unless there is some desire/interest in the community to pursue this. You guys have alot on your plate without being distracted with peripheral stuff like this.
Though, if I may, I really think it would not be that hard to do all of the above. Even if you took out the pipe-dream stuff, like load balancing, isn't the above somewhat achievable in a full development cycle?
In any case, if anyone else thinks there would be some value in this, I'd be happy to submit it. Otherwise, I'd rather leave the suggestion here instead of bogging you guys down.
Geoffrey
-
djhomeless, i understand your requests. and i think you have a point. there definitely is a need for product and a void in the market.
still, i think they are outside the scope of the sme project. sme can't be everything to everybody. doing so would lose focus. and if it isn't clear, it isn't helpful.
possibly a fork or a sister product. and again alot of work including manpower and commitment. i hope there's interest
----------------------------------------------------
my requests for sme would be organizational rather than features. possibly a yearly membership to a paid-for forum that weeds out alot of OT and worthless threads (ie sme9beta9, LOL).
the goal would be increasing service to moderate to serious users.