What happens with all those Feature Requests anyway?

[Kubist]

There really isn't a structured path visible between Feature Requests and Development. Or did I miss something? As I feel it, in becomes less and less attractive to post Feature Request, because nobody knows if and when a feature will be released. This in turn caused good ideas never even to make it onto the Feature Request forum. I think we should give the creative ones at least some concrete feedback as to whether and when their features will make it into a release.

I think It would be good if features were assigned some status and developed accordingly by the development team. Possible pathway:

1. Some vague idea: Somebody has a great idea and describes it.
2. Idea had some feedback, becomes more specific. Also reactions will make clear which idea is chanceless from the start and which ideas are good ones.
3. Developer assigns himself / herself to the idea. The idea gets a name.
4. Idea to be implemented in release n
5. Implemented.

Something like that.

[TB2]

Well I don't expect that even 10% of the requests are going to be implemented, but what I really would appreciate is if some of the devs would drop a quick note on how probable or possible it is for a certain request to be implemented. Just as a feedback, even if its negative, there would be a feeling that someone cares. And it's not such a workload, there are maybe 5 requests per day, that looked at pretty fast...

[Deciare]

Quite a few of the plugins available in Compiz Fusion owe their existence to developers who responded to earlier feature requests made in earlier forums, so someone was listening, certainly. ^^; As to whether anyone is listening... I hope so!

We should keep in mind that many feature requests propose ideas that may not be practical or feasible without advancements in hardware or software projects that Compiz Fusion's developers don't directly influence. Some features may be unpopular, some features may be uninteresting to the developers, and others still may have nothing to do with Compiz or its plugins.

And some, like features from Beryl that are still missing in Compiz... That's a sore spot in this project's history. Beryl's core was governed by far less stringent policies.

[maniac]

Well, I can only speak for myself, but I try to at least read all of the suggestions and put interesting stuff on my to-do list. Unfortunately my time is pretty limited, so it's hard to read all of them, let alone answer all of them or even code all of them ;)
We clearly don't have enough manpower to be able to fulfill many of the requests - a lot of things are said pretty easily, but are very difficult to code. We also have to keep in mind that coding it is only a part of the total effort - maintaining the code (and we already have a lot of code to maintain!) is also a big part.
So please bear with us if you don't get too many feature request answers - the problem just is a lack of time :(


And some, like features from Beryl that are still missing in Compiz... That's a sore spot in this project's history. Beryl's core was governed by far less stringent policies.
Yeah, but these policies ensure the code is maintainable in the long run. There are a number of examples for loss of maintainability in Beryl's code, starting from the OOo problem (legacy fullscreen support built into core) until the complete mess in rotate.c's code.

[delfick]

agreed, there isn't much organisation when it comes to pushing feasable feature requests into the spotlight.

hopefully that will be fixed as the project matures....

......
(and when we get input redirection :D)

[Kubist]

@maniac: In the first place: Great to hear this from you! I think you and your co-developers do a very good thing. But how do you prevent that something you put on your to-do list, doesn't end up on someone other's to-do list as well? I guess you guys developers do have some agenda sync in place? Wouldn't it be easy to share this 'agenda' (if it really is such a tangible thing, which i fear it isn't ;) ).

That would give bystanders some insight in your work in progress, and something to look forward to. I think maintainable code is an important enough reason not to rush any feature into a release. I think many non-developers would agree with that: The last thing a (semi)-newbie wants is to stare at a broken X without any clue about what went wrong. I don't think its the lack of patience that makes the waiting so hard to bear, but not the 'not-knowing-what-will-be-in-the-next-release'.

But again: Coming from the closed source world - as a lot of us are - this is by far the fastest evolving and most involving project - community-wise - I've encountered until now. I think a lot of people see compiz as the 'doing it right' where vista went the wrong way, only stalling its release to deliver bloated and non-functional (or even counter-functional) eyecandy on top of not so much at all.

So keep it up!

This year will be the year of ...

well...

Let's not spoil the fun.

[delfick]

I don't think its the lack of patience that makes the waiting so hard to bear, but not the 'not-knowing-what-will-be-in-the-next-release'.

after the amount of time it took for us to finally get rid of beryl, and the features that were made fairly public but not released for ages (for good reasons, we didn't want to release untill compiz-fusion (or compcomm as it was known back then) was in a decent state), I can say that impatience starts to take control :D

This is by far the fastest evolving and most involving project

it certainly is
in like a year we have gone so very far
so many things have changed
(for example, try out an early version of compiz-quinnstorm and compare to compiz-fusion :D

- community-wise -

6 different forums in a year definitely says community has changed fairly rapidly :D

This year will be the year of ...

well...

Let's not spoil the fun.

lol

[Deciare]

But again: Coming from the closed source world - as a lot of us are - this is by far the fastest evolving and most involving project - community-wise - I've encountered until now.
And I can't be happier about that! :D Being with so many excited, generous, interesting, inventive, and nice people is truly exhilarating throughout the day, and something to be thankful for at the end of the day. It's so depressing when nobody seems to care or listen to the same problems that crop up over and over again... So it's great to see that such depression has no place here!

This year will be the year of ...
Year of the Pig? XD

6 different forums in a year definitely says community has changed fairly rapidly :D
I'll bet you were there through most of it, too. :p That makes you one of our wisest and eldest. *bows in reverence*

[delfick]

I'll bet you were there through most of it, too. :p

1500 posts on compiz.net (this was my favourite of all our forums so far, such a tradegy when we lost it)
1000 posts on beryl-project.org
didn't really go on compiz-go.org
and from opencompositing.org, compiz.org and this compiz-fusion.org, my current post count

That makes you one of our wisest and eldest.

no, just one of the oldest :D lol

*bows in reverence*

:D

[Kubist]

But, to return to the initial question: Would it be doable to have some kind of feature roadmap online, so that we - the mere users - can track progress on them? Like this one:
http://developer.kde.org/development-versions/kde-3.0-features.html
or a little less structured, like this:
https://wiki.kubuntu.org/KubuntuFutureIdeas

[SmSpillaz]

But, to return to the initial question: Would it be doable to have some kind of feature roadmap online, so that we - the mere users - can track progress on them? Like this one:
http://developer.kde.org/development-versions/kde-3.0-features.html
or a little less structured, like this:
https://wiki.kubuntu.org/KubuntuFutureIdeas

Thats a nice idea, but unfortunately, there isn't really much of a roadmap for Compiz Fusion (as far as I know) and the project doesn't really have much planned for a BIG release.

It could work however. I like it!

[SmSpillaz]

Just for my 2c, I think that the feature requests board is not really given much attention (Not by the developers, I'm sure they're always looking for interesting stuff ;-)), but really by everyone else.

While I somewhat enjoy shooting down completely unrealistic suggestions, I feel the that amount of activity on the Feature Requests board is WAYYY lower than what it used to be in the Beryl and Compiz-Quinn days.

Is there any chance that the board can be moved 'up in the hiearchy' ?

SmSpillaz

[delfick]

While I somewhat enjoy shooting down completely unrealistic suggestions, I feel the that amount of activity on the Feature Requests board is WAYYY lower than what it used to be in the Beryl and Compiz-Quinn days.

i agree

especially compared to compiz.net days

Is there any chance that the board can be moved 'up in the hiearchy' ?

personally I don't see how that could help too much....

what we need is some way to highlight realistic ideas for the devs so they can set aside some time to implement some of them.....
(without having too much spend the time finding the ideas)

maybe a system where the mods can mark certain threads as unrealistic or realistic...
though that could be troublesome if a thread is tagged wrong.....
depends...

this one needs some thought :D

[Deciare]

Well, this subforum is equipped with the same kind of Thread Prefix (Forum Tools->Manage Thread Prefixes) features as what's already used in the Support forums, so we could use that...

But as you noted, feature requests are highly subjective things and different people will have different ideas about what's practical and what isn't... Which may again be different from the developers' perspective of what's feasible with consideration to technical and time constraints. Not to mention what's interesting, which can make a great difference. ;)

[Kubist]

@Deciare,

so could you, being moderator, introduce prefixes like [Totally unrealistic] , [Will be developed by....] , [To be included in 0.7.0]

..On second thought, this would mean a LOT of different prefixes. Maybe this isn't the best way. What if each developer had his own prefix?

[delfick]

[Will be developed by....] , [To be included in 0.7.0]

hmm, that would be going too far with the prefixes :D
if such an idea would be implemented, I'd imagine it'd just be simple realistic/unrealistic otherwise that'd be a real pain to maintain....

(also, as mods we all have the power to change any thread title, but I'm talking more of a way to mark a thread in a more invisible way to the user, with some mechanism in place for the devs to see a list of all just all the ideas marked as "realistic")...
edit: on second thoughts, having as invisible sort of removes the point of that being an answer to this thread :p, maybe have that invisble mark auto add to the thread prefix as either realistic or unrealistic, or something in between (I'm not a dev, so I'm not sure how the development process works or how realistic/unrealistic this idea itself is :D)

[SmSpillaz]

No, I was thinking the board needs to get more attention by users.

Maybe we should have an 'Idea of the week contest'

[delfick]

Maybe we should have an 'Idea of the week contest'

oh god..
past says this community + poll = flamewar
lol

(though maybe a fortnightly highlight of decent ideas in the blog......
though that could backfire with a false impression of what devs are actually implementing....
all depends)

[Deciare]

oh god..
past says this community + poll = flamewar
lol
... I liked Coral. *hides*

As ordinary users, let's not try to decide what's best for the developers. While it would be helpful and encouraging for the development team to somehow communicate to the users what kinds of ideas they'd like to see fleshed out and what's being worked on, I don't see anything wrong with the model where a developer--even presently unknown developers who aren't and won't be on the team over the long term--sees something interesting and spontaneously decides to go with it. Compiz's plugin-focused architecture is very conductive to that model, even.

What would belong on formal roadmaps are major architectural changes and system-wide implementations of new X server features (or something of that nature), and I'm sure the Compiz-core team has that well in hand. They're famously stringent.

[SmSpillaz]

I guess it would be nice, maybe if the mods choose something and the person had the opportunity to change their title to 'Idea of the week winner'

[delfick]

I guess it would be nice, maybe if the mods choose something and the person had the opportunity to change their title to 'Idea of the week winner'

not a good idea

then people will start doing that to their threads even if it isn't true......

and anyway, we don't get in enough ideas for such a thing .......

[SmSpillaz]

Meh, we need something to promote ideas :P

[delfick]

true.....

personally, methinks the following

a) we need a dev to comment on the direction of this thread
b) better organisation of the feature requests section (after friday my exams finish (one more left !!!! and uni exams are great, unlike that crap in year 12 last year :p), I'll have a think about that a bit more after then :D)

[SmSpillaz]

true.....

personally, methinks the following

a) we need a dev to comment on the direction of this thread
b) better organisation of the feature requests section


Maybe it should be brought out a bit more. We don't just have to have Ideas of the Week, we could have artwork of the week, config profile of the week - etc :)

Maybe with the other ones it could be *of the month*.


(after friday my exams finish (one more left !!!! and uni exams are great, unlike that crap in year 12 last year :p), I'll have a think about that a bit more after then :D)
Heh, same here - I had my geography and french exam yesterday... Year 10 exams seriously suck. I've only got an english now - which I am fairly OK at :)

[Kubist]

@Delfick and SMSpillaz: glad I'm not in Australia. Had my only exam last week: it was in Process & Data modeling.

But back to the subject, I think that contests have one big drawback: It causes people to think less freely: because of the pressure of the contest they might apply some - unnneeded - self-censoring to their ideas, which causes those ideas to die before they are even born. Which would be a sad thing.

What about splitting the feature requests forum into two pieces: The one that's already there, with al kinds of totally pointless ideas, but also some unpolished diamonds. This forum can be posted at/in/on/to (<-- please advise as to which word to use here) by anyone. There will be hardly any interference by the moderators (no more than there is now). But:

There will be another forum, called "Feasible Requests" or something like that. Threads from the original forum can be moved to this forum - by a DEV only. Discussing the feature will still be possible, but the topmost post (being sticky, if posts can be sticky, I know threads can be) will always reflect the current state of the project.
The thread subject in this forum will be prefixed with the name of the dev, like

[DevX]Combining snow and window miniatures
Post #0 (or sticky post or something)
2007-11-14: Discontinued for apparent reasons (see below)
2007-11-12: Good idea, but might consume a LOT of GPU/CPU
2007-11-11: Easy to implement, but can't think of any functional rationale for it.

Post #1
The original post about it, before moving it to another forum:
"Man I think it would be way coooool to have the snow plugin use windows instead of snowflakes..."
......
..
etcetera

[Deciare]

I kind of like your most recent idea. Especially the part about allowing only Developers to move a thread into the to-do/status forum. :) I also love the idea of a first post that summarises the current state of the effort to get that feature implemented. And the fact that discussion can continue in a thread after it's been moved, so the idea can continue to be refined as it's worked on. Actually, I like your idea more than "kind of". :D

An additional benefit to a dedicated to-do/status forum is that it'll be easy for everyone to see exactly which ideas have been attempted in the past, and why they were successful or unsuccessful. It will give us all a good idea of what's doable, what's interesting, and how to take ideas in a direction that will help them become a reality.

[delfick]

@Delfick and SMSpillaz: glad I'm not in Australia.

no your not :D
the heat wave of the past few days has ended, damn it's really nice outside :D

But back to the subject, I think that contests have one big drawback: It causes people to think less freely: because of the pressure of the contest they might apply some - unnneeded - self-censoring to their ideas, which causes those ideas to die before they are even born. Which would be a sad thing.

definitely agreed
we need more imagination, not less :D

What about splitting the feature requests forum into two pieces: The one that's already there, with al kinds of totally pointless ideas, but also some unpolished diamonds. This forum can be posted at/in/on/to (<-- please advise as to which word to use here) by anyone. There will be hardly any interference by the moderators (no more than there is now). But:

the problem comes when the second forum becomes as full as the current one and pretty much removes the point of it :D

maybe a better structure would be

*Feature requests section
--Initial requests = Please post any idea you have, no matter how wacky
--Possible requests = This thread is for ideas considered possible, you can discuss but not initiate ideas here
--Graduated requests = requests that have been implemented, you can continue to discuss how this idea could be improved.

Though, there are issues with this

a) devs may not see all threads (i'm not sure how much they read, so I could be wrong there) also ideas evolve, devs may not read entire threads (especially the ones where every post is half the screen long, like the one we once had about making all this into a DE), and so it would be up to the mods (who collectively should read everything) to report decent ideas, but that could lead to confusion to do with what we may perceive possible or not.....

b) It might not work when ideas are lost because the devs don't have time to keep watch on all the threads in the graduated and possible sections
also ideas in those threads might evolve back to not possible.....

c) I had a third problem, but it took me too long to write the last two problems in words that make sense (only just woke up) and so I forget the third issue.....

[DevX]Combining snow and window miniatures
Post #0 (or sticky post or something)
2007-11-14: Discontinued for apparent reasons (see below)
2007-11-12: Good idea, but might consume a LOT of GPU/CPU
2007-11-11: Easy to implement, but can't think of any functional rationale for it.

agreed...

that would be good

though once again it comes down to how much time the devs have to maintain such things for many threads....


we really need some devs to comment on their thread viewing habits....... :D

[Deciare]

the problem comes when the second forum becomes as full as the current one and pretty much removes the point of it :D
The second forum is populated only by ideas that have already been attempted or completed, so it will be much shorter than the unabridged feature requests forum by design. Since it's been established that every thread in that forum has at least some immediate merit, any contribution to those threads should be less likely to go unseen than random contributions to random threads.

maybe a better structure would be

*Feature requests section
--Initial requests = Please post any idea you have, no matter how wacky
--Possible requests = This thread is for ideas considered possible, you can discuss but not initiate ideas here
--Graduated requests = requests that have been implemented, you can continue to discuss how this idea could be improved.
That's an interesting adaptation, but I don't think a third forum is necessary. Rather, I think that every idea that is being actively developed (i.e. code has been written within the past 48 hours) should be stickied in the second forum whereas any idea that is on hiatus or finished should be unstickied. Moderators can help by ensuring that the proper threads are un/stickied. This keeps ideas of greatest interest near the top of the list without moving ideas of lesser interest out of sight and out of mind.

[delfick]

The second forum is populated only by ideas that have already been attempted or completed, so it will be much shorter than the unabridged feature requests forum by design. Since it's been established that every thread in that forum has at least some immediate merit, any contribution to those threads should be less likely to go unseen than random contributions to random threads.

true.....

That's an interesting adaptation, but I don't think a third forum is necessary. Rather, I think that every idea that is being actively developed (i.e. code has been written within the past 48 hours) should be stickied in the second forum whereas any idea that is on hiatus or finished should be unstickied. Moderators can help by ensuring that the proper threads are un/stickied. This keeps ideas of greatest interest near the top of the list without moving ideas of lesser interest out of sight and out of mind.

then we get the situation we currently have with the packages forum where we have to scroll the window to get past the sticky threads....

not a good outcome :)


also what happens to those ideas that are possible, but not without upstream changes
for example any idea that requires input redirection.....
if they are put into the possible section then it's contradictive....
but if they are left in the first section, they'd get lost.....

[Deciare]

then we get the situation we currently have with the packages forum where we have to scroll the window to get past the sticky threads....

not a good outcome :)
With the notable exception that all threads that are sticky, are sticky for a reason. ;) Since the decision of whether a thread should be stickied or not is directly dependent on whether a developer is working on that idea "right now", I think we can expect that we'll have different stickies every few days, so the list of threads at the top of the list won't become as stale as those in the packages forum.


also what happens to those ideas that are possible, but not without upstream changes
for example any idea that requires input redirection.....
if they are put into the possible section then it's contradictive....
but if they are left in the first section, they'd get lost.....
That's a tricky question, yes, and one for which I don't yet have a good answer. But I'm not thinking of the second forum as one for ideas that are possible. I'm thinking of it as a place for developers to keep track of ideas that they will have a very high probability of implementing, even if not immediately.

[delfick]

With the notable exception that all threads that are sticky, are sticky for a reason. ;) Since the decision of whether a thread should be stickied or not is directly dependent on whether a developer is working on that idea "right now", I think we can expect that we'll have different stickies every few days, so the list of threads at the top of the list won't become as stale as those in the packages forum.

oops, my bad, didn't interpret correctly :)

then that's a good idea...
(if devs choose to maintain that though, it might get a little tedious for them to constantly have to turn sticky on and off.....)

That's a tricky question, yes, and one for which I don't yet have a good answer. But I'm not thinking of the second forum as one for ideas that are possible. I'm thinking of it as a place for developers to keep track of ideas that they will implement, even if not immediately.

hmm, yep, that'd be better description of such a section......

[Deciare]

(if devs choose to maintain that though, it might get a little tedious for them to constantly have to turn sticky on and off.....)
I volunteer to help with the stickiness. ^_^

I know that my presence can be rather inconsistent, though, so it'd be nice if other mods could drop by gitweb now and then and adjust the threads as needed.



hmm, yep, that'd be better description of such a section......
Thank you, I'm glad you think so. ^_^

[delfick]

I know that my presence can be rather inconsistent, though, so it'd be nice if other mods could drop by gitweb now and then and adjust the threads as needed.

hmm, then we get the latency we got before all the mods were made global.....
once again it's hard to make judgments without knowing the habits of the devs, but I assume that wouldn't be too much of an issue....
I gotta make a smiley allowance for myself, I swear I've used them like atleast twice in every post I've made this morning, that's a lot of smileys, lol *resists use of smilies*

[SmSpillaz]

Heh, well the idea behind the contest was to expose the feature requests section a little more. However way that is done, I am fine.

It just seems as though the compiz-Fusion comunnity isn't as lively as the compiz-Quinn community :(

[delfick]

It just seems as though the compiz-Fusion comunnity isn't as lively as the compiz-Quinn community :(

hasn't been since we lost compiz.net
:(

[SmSpillaz]

Dammit. I would give anything to have something like the old compiz.net forums back.

Oh well, I think as soon as we get the theme in (Can you poke RYX about that? It's been supposedly ready for weeks :-|) things should look nicer

Also, we need more games to play on this forum. I think we should add wishlists for plugins like scale, animation etc.

The problem I've found with this merge is that it is now harder to get 'almost useless but kind of cool' features into the core :( Oh well.

[delfick]

Dammit. I would give anything to have something like the old compiz.net forums back.

I want the old (http://delfick755.googlepages.com/old_theme.png) theme (http://delfick755.googlepages.com/old_forum_theme_posts.png) back
but I doubt that'll ever happen :(
btw, when did we actually lose the forums, I can't find the announcement on beryl-project.org, though it seems the earliest post was November 28....
looking at that second screenshot (those are the only two screenshots I have of that forum), I had 1000 posts a month before we lost the forums, yet I had over 1500 before we lost the forums, that means I posted over 500 posts in a month...lol

Oh well, I think as soon as we get the theme in (Can you poke RYX about that? It's been supposedly ready for weeks :-|) things should look nicer

still need people to reply to http://lists.compiz-fusion.org/pipermail/web/2007-October/000523.html

Also, we need more games to play on this forum. I think we should add wishlists for plugins like scale, animation etc.

hmm, want games, go to ubuntuforums.org community cafe section :D
or even better, over here, http://ubuntuforums.org/forumdisplay.php?f=298

[...] I've found [...] this merge [...] cool
*cough*that quote may have been edited for the greater good*cough*
:D

[SmSpillaz]

LOL "Hopefully my glasses are correct :D"

[delfick]

LOL "Hopefully my glasses are correct :D"

there's a reason iXce gave me the "Crazy guy" title :)

[SmSpillaz]

Meh, we need more members. Or more activity at least.

*Is jealous of UbuntuForums :-|*

[sdlvx]

I think that even if you say things and you don't get a response, people still take it into consideration.

Back on the Beryl forums, I tried to make the snow plugin look like stars. It was close, but no matter what the stars still had a little jitter to them. Now, there is a stars plugin.

I also said in the Beryl forum that it would be nice to use the cube tops for displaying lookup charts and stuff (I.E. an image of an ASCII table), and low and behold, there's a plugin to scroll cube caps.

I never once heard anyone say "Oh, that's a great idea, I'm going to code it tomorrow.", but it did eventually get done.

They do listen, they're just too busy coding instead of replying to forum posts. I wouldn't get discouraged about making posts about new ideas, because you never know who is listening/reading.

[SmSpillaz]

sdlvx == Krissss ?

If so, please tell me as I have pretty much copy-pasted your work into snow and called it stars :P