arthur wrote:

Could i make the suggestion that perhaps one of the problems that new users have on milliways is that they get dumped straight into the main mass of conversation, which can be extremly technical and highly opinionated.

perhaps it would be better if new users started out in a different channel (yes it has lots of channels or rooms) one where a few of the more "polite" members could hang out in order to give advice and help.

and leave the more raucus crowd in a seperate section of the talker, where new people are welcome to go should they wish, but that they wont be dropped into unprepared.

This is certainly a possibility.  However, you need to be very careful that you don't split the community further.  People who start in the polite channel may be more inclined to stay there even though they would otherwise have ventured into the shark pool of room 0

You would also need very active moderation to keep the polite channel polite and unopinionated.

I'm also warey of splitting the rooms since you end up excluding people from your conversation who would otherwise have valuable input (for example, discussions on coding often gets valuable input from people like you, rohan, etc who would probably not be in the polite channel.  Thus people would (without realising) not necessarilly get the best advice.

rollercow wrote:

Firefury isn't wrong, milliways is still where the technical part of the community has its presence, it has to accept that new members in much the same way newer members have to accept the technical community.

I think this is major point. While it is important to get new members, a lot of these new members will come and go in a few years, yet some of the long life members who have been here for a long time, will still be here and continue to be members of the society when these new members have left (while that doesn't have to be the case and certainly some will stay, you get the idea).

The impression I get from some people is that they think the society belongs to the newer generations and that they should necessarily have the biggest voice (as they are larger in numbers and more uni relevant at that point), but this belittles the contributions over the years of these long time members and the often large contributions in time, effort,money/equipment and more importantly experience that they still provide.

saya wrote:

Given the current active student membership votes in at least part of the committee and the rest of the committee is generally again part of the currently active student membership, there is more room for reprimanding them when they fail to treat the members who they are representing and who voted them in, fairly. At the moment on MW at least not even current committee members (or at least those who are still students) have much of a say in how new members who deviate from the tolerated behaviour are treated. I know there isn't a way to completely stop unfair treatment, but at least by having people who are elected by the members to "moderate" that makes someone accountable when they don't do the job properly. Although I guess you might get the "who's going to deal with them then if they turn bad", but that's hardly different to how it is now.
The main thing with MW is (at least this is how I feel it is from being on there now), that it is like FF said _their_ community. Which to me means it's not SUCS' talker, so SUCS doesn't really have control over how it's moderated, and we're just using it or passing through really so it's just tough cookies if things are bit pants there for any current members.

foshjedi2004 wrote:

The only way I see to properly sort this out is to either restrict admin/moderator access to Commitee members only thus removing Admins as they currently stand.  Or introduce a way so that only Admins have the ability to kick/Gag people.

No offense intended, but I think this is an incredibly bad idea. While the suggestion is that it gives people someone to blame, they already have someone to blame and things can be done about people abusing power (as has been said by others) and you can't really use the yearly voting to solve the problem as by the time voting comes around the problem is solved as those members most likely won't be in or eligible to be in the exec again.

Either way, I think the idea of giving the current committee/exec all the power is horrible as the whole idea of having active older members is to help steer the course of the society over the years. The effect of any single exec or even normal set of members, is certainly not insignificant but can be somewhat transient, especially with respect to intent and vision for the society and old exec members having privileges (on mw and off) is just one way of showing their importance in the society (I kind of imagine it like a tribe with a set of elders and a young chief, while the young chief makes big decisions, the elders are there to offer guidance and often chastise the chief for making rash decisions - although I'm not saying all the "old" members are actually old in terms of age )


Mainly, I think that while encouraging a lot of new members to join in is admirable and wanting to help them is admirable, I think trying to help them integrate with the society more is a better idea rather than trying to integrate the society with them. I'm not suggesting new ideas and ways of doing things are bad, but there are reasons some things are the way they are, and just deciding to do certain things because they are new and exciting isn't always the best reason to do them (not that they can't be tried out and played with).

I don't know what I'm really trying to say, so I guess I'll shut up.

foshjedi2004 wrote:

While I am not condoning such a course of action I have noticed that there are some people who do abuse their priviledges, from certain members whining when they lose there priviledges (alas I am one of these people), to other members who gag other members and/ or kick them from MW.

All of the "lost privs" incidents I've seen have either been a single SU deciding that another SU has done wrong and revoking his privs, or a "secret vote" held by a select few (i.e. they are selected since it is known which way they will vote).

If you want to revoke someone's priv and make it stick, you have to do it for a good reason and through open consultation.  Otherwise those revoking privs are no better than those alledgedly abusing them.

There have been a lot of "he mrodded me so I'm revoking all his SU privs" incidents over the years, and in those cases I think the person who's privs have been revoked has a right to complain.

The only way I see to properly sort this out is to either restrict admin/moderator access to Commitee members only thus removing Admins as they currently stand.  Or introduce a way so that only Admins have the ability to kick/Gag people.

This is a bad idea for a number of reasons:
1. Traditionally the committee has not always been technically competent enough to be admins.  Giving people superuser abilities when they are not competent enough to use them is a Bad Idea
2. The committee is not always around.  If noone else can moderate then you run a serious risk of chaos ensuing when the committee aren't around.

welshbyte wrote:

I've seen this working in action on many IRC channels. Like with Ubuntu, there's an #ubuntu-offtopic which is chock full of newbies and devs alike, conversing informally and at the other end of the scale there's #ubuntu-devel where only serious technical talk goes on.

I think there's a significant difference between separating subject areas (which is what Ubuntu and many other projects do) and separating behaviours (which is really what you're trying to do with Milliways).

I think splitting Milliways into 2 channels, with one of them being the current "room 0" and the other being a "polite" channel is potentially a bad idea since both rooms will carry similar subject matter.  You may even end up forcing people to participate in both channels (and put up with any badness that would not be tollerated in their preferred room) because there will be so much cross referencing of conversations between the rooms.  To some extent we already have this problem with frequent references to conversations happening on the CS forums (to which many people don't have access).

It's not really about separating users, it's about separating themes of conversation and allowing users to choose the kinds of conversations they engage in.

I don't think that's the purpose at all - noone is complaining about the themes of conversation on Milliways.  The complaints being cited are about behaviour.  I think you are fundamentally trying to solve a different problem (which may not even exist).

Another example is if I said something silly in #ubuntu-devel it'd be expected for them to kick me out but in #ubuntu-offtopic it would be left alone

I'm not sure you should expect to be kicked out of *any* Milliways room for just saying something silly (at least, not a proper kick - at the moment I would consider ZOD to not be an especially serious kick and is used as a bit of a minor slapping.  Maybe this is bad and ZOD should be reserved for people who've done something really bad.  Although notably, I tend to summon people back to the room after a ZOD if they don't immediately return themselves).

There are times when people are being persistently (and knowingly) annoying and in those cases a kick is probably appropriate no matter what channel they are in.

Also, I take exception to scripts which automatically abuse people based on keyword matching (which some users have set up on various occasions) - if someone needs to be kicked for being annoying then that's one thing, but if you feel the need to kick them when they say something you might find annoying, even when you're not reading the talker then there's something very wrong.

b) defaults like low timeouts don't allow users to lurk and acclimatise themselves to the conversation.

Low timeouts are designed to encourage participation.  I think the current default is probably too low (I think it's 30 minutes or something) but I'm wary of killing off timeouts altogether.

Maybe a better system would be to start off with a reasonably high timeout and gradually readjust it (automagically) based on the amount of participation.  I.e. you get to sit there lurking when you're a newbie, but if you've been lurking for a few weeks and still not really saying much then your timeout drops in an effort to prompt you to participate.

Due to the general declining interest in C coding in the society since mw began (possibly due to comp sci not teaching it in much depth these days, more "shiny" languages appearing and other reasons), changes are few and far between so a rejig is unlikely to happen, as proven by the overall apathy towards Marvin.

I think to some extent the apethy towards Marvin is largely because noone's seeing it used on a daily basis (and indeed, it is in no state to do so) so don't feel that changes they make are of immediate benefit.

Whilest it's true that both students and the department seem to see C coding as unimportant, it *is* still very important in industry and if you can code C then you're probably going to get quite a good job (not just C coding jobs - if you can code C then you have a good grasp of a lot of important fundamentals that many academics don't seem to see as important these days).  On the other hand, the job market is flooded with people who can code the latest language du jour because a lot of a time the only significant reason for coding in those languages is because managers have heard a buzzword and want to jump on the bandwagon, even if it's completely inappropriate for the job.

A such, I think there is a significant benefit in keeping C coding going within the society, but we somehow need to encourage people to participate in the projects.

Frequently the comment comes up on Milliways "I would help with Marvin/MW3 but I don't know C" and those people show no interest in learning.  Participating in the society should be about learning, not just repetetively reusing your existing knowledge and rejecting anything you don't already know as "too hard".  What's more, there is really know excuse since there are a fair number of professional C coders on Milliways who would be more than happy to help and critique students' code.

Getting these improvements into mw (somehow) would be great but I understand that some members are investigating alternatives to mw that provide these features as standard. Although it may be seen by some as a threat to mw, or just being done because of some (imaginary) power struggle, and a waste of all the years of mw development, the points in the previous paragraph seem to justify the experiment.

There have been 2 attempts to move everyone over from Milliways over the past few years (notably both instigated by the same person) and they have both failed.  I think the reasons they have failed are down to one or more for the following reasons:
1. Maybe they don't offer significantly better features
2. Maybe they don't offer some significant features of Milliways
3. The person/people pushing the new system haven't clearly explained *why* they think the replacement system is better.  (No, "it's newer" is not a good reason - stuff isn't automatically better because it's new.  If you believe otherwise then you are perfect management material

The latest attempt basically boiled down to:
  "I think we should move everyone over to SILC"
  "Why?"
  "Because it's better"
  "How is it better?"
  "Uhhhh...  I dunno, it just is... and it's newer"
That sort of reasoning just doesn't cut it - if you want to run a replacement you can't just take a couple of people and go off on your own and then complain when noone else joins you - you have to convince the current Milliways population that it's a good idea - so far noone has managed to do that.

Times do change and if a new idea that really can offer SUCS a big improvement across the board comes along, maybe we should entertain the idea.

Absolutely.  But so far noone has actually said what the alternatives offer over the existing system and I'm sure noone sane wants to entertain the idea of ripping out the existing system and replacing it just for the sake of getting something new which is no better than what we already have.

Similarly, the people who are pushing for a new system can't just discount people stating important features that it doesn't have which Milliways does.  Remember, just because a feature isn't important to you doesn't make it unimportant to everyone.

There is effort involved in changing system.  If existing users don't see a reason to change (or worse: if they see the new system as sucky compared to the existing one) then they won't change.  And you shouldn't force them either - your idea has to stand on it's own merit, rather than succeeding through dictatorship.

Implementing something new will be very hard, of course. It's next to impossible to please everyone with something new, especially when they're comfortable with the status quo.

I think SUCS's needs would best be served by improving what we have at the moment (adding features, removing stuff we nolonger think is appropriate, and reengineering stuff).  Sadly it often seems that the most vocal group are the ones that are least interested in actually _doing_ stuff and just want to push everyone onto a pre-packaged solution even if it is inappropriate.

After doing that, rational discussions can be had about the advantages and disadvantages of each idea and eventually a shared idea of what everyone wants will form.

I think a key way to having a rational discussion is to document the ideas and reasons for doing stuff on a wiki so that people can actually take time to make an informed response.

Currently a lot of this stuff has just been randomly discussed in real time - real time communication will get you a lot of comments and ideas fast, but they are rarely well thought through.  Worthwhile ideas and comments get lost in the discussion and people who weren't there at the time lose out on the ability to help direct the process.

P.S. I've already seen a good change in how people are being treated in mw since the AGM and since this thread started. It is in fact good to talk™ and it does make things better. Good stuff, and thanks

I agree entirely here.  And whilest this thread prompted a lot of discussion I think it's important for people outside the admin circle to understand that there have been big discussions amoungst the admins both before and after this thread started in an effort to improve things.  It's an ongoing process and we're all always learning.

frosty wrote:

You keep mentioning significant features of milliways that a new system would be missing. Could you explain what those would be? I'm afraid (in my opinion, at least), you're equally guilty of not providing reasons - you say Milliways is better

I haven't said Milliways is better - indeed I haven't really looked into any alternative systems.  However, until someone's provided good reason's to switch and a suggestion as to what to switch to, I'm not sure there's a lot of point in anyone spending the time deciding what Milliways does that another, undecided, system doesn't do.

If you want people to switch to a new system, it is your job to convince people that it's a good idea.  The bulk of the Milliways users are not going to change to a whole new system on the basis that there's no reason _not_ to unless there is a good reason to do it.  Change is effort - why spend the effort if there's nothing significant to gain?

and refactoring what we currently have is better, but you don't really explain *why*.

Because with Milliways we can always tune it to do what we want.  If we use a prepackaged solution I think that's going to be much harder.  Whilest another Free solution could be modified to work the way we want, any code changes are going to make handling upgrades an almighty pain in the arse unless we can get the changes accepted upstream.

In the end, if we use a prepackaged solution I think we'll end up with just another talker with absolutely no unique and defining features.  Whilest it may "work" just fine, I feel that making all our systems run nothing but generic software somehow removes part of the society's soul.

cmckenna wrote:

However, what is below are what I consider to be the key features, for me of milliways (in no particular order).

[*]emote
[*]raw

Personally, I quite like the feature found in some talker systems where the name of the person who says something is prepended to their message when it appears on your screen It's quite disconcerting when messages just appear out of nowhere, or you can't be quite sure who a message really came from.

I don't see why people should have to play a guessing game, nor need to have a colours script/spy privilege/raw detection script to allow this. Most users don't want to participate in what feels like a mw-script arms race (or at least the fallout from one).

cmckenna wrote:

[*]multiple rooms (or equivalent)
[*]private conversations (eg .sayto/.whisper)

The multiple rooms in MW at the moment are quite a pain, as welshbyte highlighted earlier. I wouldn't really call .sayto good for private conversations. Odd private messages, sure.

You can quickly lose the conversation in the sea of other text in the room, and it's all too easy to screw up and say something to the wrong person or the entire room.

You might say that Jabber is a better bet for having private conversations. Maybe it is, but only once you've messed around adding people to your roster. Newer members probably don't have many SUCS people on IM. Allowing similar though the context of MW/whatever could really help to draw people into the society in my opinion.

cmckenna wrote:

[*]customisability - mw does this partly through scripts - that allow people to view it how they want. For example mw allows me to have all the lines coloured, using colour schemes that I can read on a black background, whereas (AIUI) Pyschodom uses a white background with black text with only people's names in colour.
[*]timestamps

Customisability is good, but I believe things should work pretty well without having to customise them too. Currently new/bbs MW users don't get any colours at all.

cmckenna wrote:

[*]zod/mrod (or eqivalents to remove misbehaving users from the conversation)
[*]the ability to run on any operating system with minimal overheads (no/small programs needed, not eating huge amounts of system resources) - inlcuding on handheld devices

I'm not sure that zod/mrod make it into my list of essentials, at least not for dealing with misbehaving people. If it's someone we really would rather not have around then sure, kick and ban them. Otherwise, the ability to temporarily silence people while calmly explaining what they've done wrong sounds a far more useful solution.

I find the ability to access MW anywhere I can use SSH very useful. Most of the time though I run it on my own computer, where having a client which takes full advantage of this would be preferred.

cmckenna wrote:

[*]no/minimal wasted screen space
[*]idle time display
[*]copy and paste

The screen space that something takes up, or how many other gadgets apart from the pure text which show up, is another thing which should be customisable. People have different definitions of screen space "wasted".

Knowing how long someone has been idle is useful, but MW could do other things to tell me if I'm likely to get a reply. Especially since lots of people are permanently connected to MW, I think IM-style away states could be useful here.

Copy and paste... how about copy and paste which doesn't make you worry whether you have multiple lines in your clipboard when pasting? Copy and paste which isn't utterly broken-by-default for our Windows users?

firefury wrote:

Copy and paste which isn't utterly broken-by-default for our Windows users?

How is paste "utterly broken" for Windows users?  It seems to work for me (and yes, I have Milliways running under Windows at work - it's about the only useful thing I can use the Windows machine to do).

For clarification, this is just a snipe at PuTTY. Most incoming SUCS members are from a Windows background, and PuTTY is the software which we recommend to them to use with Milliways. I'm fed up with having to immediately explain to people that they shouldn't use their right mouse button in PuTTY (something which the Windows interface encourages a lot) and seeing people flounder with accidental pastes.

I know it's configurable to act in a more Windows-like way, but this isn't the first conversation I want to have with people new to the society.

chckens wrote:

Personally, I quite like the feature found in some talker systems where the name of the person who says something is prepended to their message when it appears on your screen It's quite disconcerting when messages just appear out of nowhere, or you can't be quite sure who a message really came from.

It has been discussed repeatedly recently that things like colours, and spy (shows who actually sent a message) etc etc should go in the default permissions for everyone.  The current state of these features isnt a reason to find an alternative, its a reason to get off your arse and do some coding to fix it.

chckens wrote:

The multiple rooms in MW at the moment are quite a pain, as welshbyte highlighted earlier. I wouldn't really call .sayto good for private conversations. Odd private messages, sure.

You can quickly lose the conversation in the sea of other text in the room, and it's all too easy to screw up and say something to the wrong person or the entire room.

You might say that Jabber is a better bet for having private conversations. Maybe it is, but only once you've messed around adding people to your roster. Newer members probably don't have many SUCS people on IM. Allowing similar though the context of MW/whatever could really help to draw people into the society in my opinion.

This is more a user interface issue. You cant easily have seperate windows for every room/privatechat in a text based user interface without much gymnastics and the loss of certain features.

someone could write a whizy curses based client, which gives some kind of tabs for each channel and for private message threads, but then you lose the normal scroll buffers of past conversation that you currently have.

perhaps the easiest way to make multi rooms work better would be to have your scripts set the background colour according to the room the message was in, and the foreground for who it was.

chckens wrote:

Knowing how long someone has been idle is useful, but MW could do other things to tell me if I'm likely to get a reply. Especially since lots of people are permanently connected to MW, I think IM-style away states could be useful here.

milliways knows how long it is since the user last typed something, and is what it sets as the 'idle' time in the who listing. there is no way for it to know anything else.  instant messenger programs are running native on your desktop, they can find out if youve moved the mouse or typed in another window.  ssh clients dont pass that on, so mw cant know.  this rather makes the idle system less than optimal, oh and before you suggest it auto away messages are utter pain in the arse. every irc channel ive ever hung out on would kick and ban anyone that used them.

chckens wrote:

Copy and paste... how about copy and paste which doesn't make you worry whether you have multiple lines in your clipboard when pasting? Copy and paste which isn't utterly broken-by-default for our Windows users?

absolutely nothing to do with mw, nothing we can do about it. what the users choose for the ssh client and how they configure it is their problem.