Letter to the Community: LinuxMCE 0810 - The Cold Honest Reality

[wierdbeard65]

Well um..

You know those little footers at the bottom of each developer page on the wiki?

You know the ones that say Category:Programmers Guide?

If you click on that, you get....

http://wiki.linuxmce.org/index.php/Category:Programmer%27s_Guide

and while it's not perfect..and some of it isn't completely up to date.. It is relevant.

-Thom

Ok, I can admit when I'm wrong  . I didn't spot that link (I don't tend to navigate wiki's using the categories links at the bottom  ) I'll start reading and see where it takes me.

Roll on the really good Noob users's guide that's been promised

[MarcoZan]

Dude, if you're a developer and you can't read source code...

...you need to learn more before ever tackling a system like LinuxMCE.

hmm.

Give me a break, guys. Yes, this documentation assumes you are capable of inductive and deductive reasoning and being able to think for yourself.

Yes, port 3450 is mentioned.. but which? use a port scanning program, or lsof, or hell, just guess. It's either/or, isn't it? and when you figure it out, guess what? YOU PUT IT IN THE DOCS! IT'S A WIKI! OPEN COLLABORATION FOR ALL.

I do NOT have time for people who willingly refuse to learn, and I am sorry, spoon feeding is NOT learning. I have no time to spoon feed developers who MAY or MAY NOT contrbute later. You all feel as if you are ENTITLED to being lead by the hand. I have news for all of us. We are entitled to NOTHING in this life. Not even Living. Even that is a gift, so we should all consider ourselves lucky, and make the most of it.

I hate replying in this manner, because it paints me out to be the asshole. However, I need to make one thing exceedingly clear. I bend over backwards to those who try and put in effort and initiative. How do I determine those people? by the questions they ask. The more specific the question you ask, the more specific my answers will be. Those of you who have taken the time to learn know this to be true of me.

Stop being children. I am not your father figure. Grow up and take some initiative.

-Thom

Better clarify what I myself was meaning by "spoon feeding" as I suspect that it may differs from what you mean.

I was referring to a possible series of workshop/screencast/similar things on well defined areas known as though ones, that in my opinion could have helped to boost a bit the learning process and possibly to reduce the required time to be somehow ready to go.
In my places they say " when too hot and dry, also hail is welcome".
I simply thought that as no experienced devs heve been stepping up so far, a way to speed up novices could at least be considered of some utility.
It was not meant as a replacement of the usual learning path, but a way to help willing people to accelerate and find earlier their way around.
Am I wrong? Maybe you have better vision and elements to judge. If you say it's a bad idea, that's fine for me.

I took the liberty to insist as this is exactly what has already been done (with good results) with HADesigner, with GSD, and what probably will be done with Datagrids. Never thought this was "spoon feeding" in the worst sense of the word.

Anyway I'm sorry to see that this kind of proposal have been systematically interpreted as laziness/intention to "manage" things/whatever similar thing you may have thought instead.

It was simply a way (maybe naive) to try and counteract the cronical lack of devs.

Marco

[wierdbeard65]

* hari gets some popcorn.. (but not to feed any trolls)

ROFLMAO 

[davegravy]

This thread started as an expression of frustration due to the situation with resources. The devs seem to have indicated they are not open to a change in direction or organization. I think it must therefore be accepted that there will be no relief for the devs, since to expect a different result without a change in approach would be insane (as defined by Einstein).

With this in mind, are there examples of wildly successful open source projects that are as big as linuxmce, that rely exclusively on self-motivated developers, and that have 5 developers pulling everything together? I’m not aware of any, so please enlighten me if there are. If there are not, what makes the devs think that linuxmce will be an exception? I am assuming there is a desire to make linuxmce wildly successful.

The way I see it, the devs can continue on the current path and achieve 50 units of progress in the short term and 1,000 units of progress in the long term. Or, they can increase their documentation/training efforts in the short term (at the expense of development efforts) to achieve 5 units of progress in the short term and 10,000 units of progress in the long term.

It's an accessibility investment that is urgently needed. If the investment is made to make this project accessible to the 100s of people who have 1-10 spare hours a week,  the invested time will pay for itself over and over and over.

I urge the developers to open their minds to this.

[tschak909]

I would agree with this, if no documentation were being written.

But there is an unbelievable amount of documentation being written _WHILE_ development is being done, and even before, when Pluto was in control.

Why don't you actually look at what's on the wiki, in the programmer's guide, for example?

I am VERY tired of people shit talking the documentation, when they haven't even looked at it! (This much keeps being proven true!)

I know it's not perfect, but we're trying to make it better, and it's EVERYONE'S responsibility to do so, not just ours. This does not mean everybody sit on their ass while the devs stop to write all the documentation, or any similar gradients thereof.

-Thom

[WhateverFits]

I do believe that things like Programming Challenge: DCE Device: Torrent Downloader PlugIn http://forum.linuxmce.org/index.php?topic=8442.0 are a good step in the right direction. That is a simple scope task that is moderately well defined (the rest of the definition is left to the consumer as a training exercise). I have been setting up my build system for 8.10 for the past few days and I intend on starting on that project simply so I can learn more about the entire system. I am a complete noob on LinuxMCE development and I expect to run into some walls on it but I also intend on documenting the daylights out the process. Once I'm done going through it I should be able to provide documentation on what I had a difficult time with. I think the first part is that I never knew where to start. The Programming Challenge is good for that.

[tmoore]

Despite it being mentioned repeatedly by some of the developers here, more developers is NOT the complete answer. It is certainly a big part of it, but the project is destined to fail (IMHO) if the focus is purely on random contributions to code.  I think the term "management" may have hit a nerve with a couple of people.  Maybe their bosses suck and have given the term a bad reputation?  So, let's put it another way.  This project needs to be organized differently to enable more people to help.  The number of people who have made that point in this thread proves it to be true, no matter what we each may think.

If the 5 devs neither have the time nor inclination to reorganize the project (and that is their right, as it is their free time after all), then the community should come together to form and agree a vision, a strategy, an organizational structure, and start coordinating the project in a way that promotes success.  And if the devs don't want any part of that, then that is fine too, but without the cooperation of those people then there is no alternative other than to fork the project.

Perhaps the best thing to do is to fork this project now?  Those who want the project to continue on its current path (whatever that is) can stick with LMCE, and those of us who feel we need a different approach can start a new project that has room for coders of all abilities and also people who feel that they can contribute in other ways.  I mean no animosity towards anyone, and I would hope that the two projects could work in parallel on good terms and share information/code.  There are different approaches to the same problem, and I think there is a fundamental split in this community.  Perhaps everyone will get along a lot better if those two styles of management/organization are separated into different teams.  Clearly there are a bunch of us who are very eager to contribute but who feel unable to, for one reason or another, as we currently stand.

What does everyone think?

[brake16]

*** hari, pass the popcorn!***

[Techstyle]

After I have eaten some of Hari's popcorn I would have to say that I think we should try where ever possible to not fork.  If you look at the Wiki you will see that Thom (for one) has made some changes to the development pages in an effort to help bridge the alledged/percieved documentation gap.  I couldn't tell you if there was a gap to start with because although there I don't understand it I have read the some of the programmers guide category.

Your comments about Project Management mimic mine from a couple of days ago but I still think task management within the current project could be helpful - adopting the project coordinator role not the project manager would be appropriate and map the tasks required from now to production.

I also hope that Thom's 'programming challenge' post is the first of many that will help the community and the developers with the goal of 0810 release by accessing community resources.  I just hope eventually I will be far enough up the learning curve to take one of these challenges myself.

[tschak909]

*breath*

well,

If you feel that way, nothing stopping you from forking.

The code is here in SVN:
http://svn.linuxmce.org/svn/branches/LinuxMCE-0810

I wish those of you who wish to fork the best of luck. Please be sure to comply with the Pluto Public License for code so labeled, and be sure to comply with the GNU Public License, where so labeled.

I do have a caveat that you may have difficulty maintaining a community around it without any developers.

-Thom

Despite it being mentioned repeatedly by some of the developers here, more developers is NOT the complete answer. It is certainly a big part of it, but the project is destined to fail (IMHO) if the focus is purely on random contributions to code.  I think the term "management" may have hit a nerve with a couple of people.  Maybe their bosses suck and have given the term a bad reputation?  So, let's put it another way.  This project needs to be organized differently to enable more people to help.  The number of people who have made that point in this thread proves it to be true, no matter what we each may think.

If the 5 devs neither have the time nor inclination to reorganize the project (and that is their right, as it is their free time after all), then the community should come together to form and agree a vision, a strategy, an organizational structure, and start coordinating the project in a way that promotes success.  And if the devs don't want any part of that, then that is fine too, but without the cooperation of those people then there is no alternative other than to fork the project.

Perhaps the best thing to do is to fork this project now?  Those who want the project to continue on its current path (whatever that is) can stick with LinuxMCE, and those of us who feel we need a different approach can start a new project that has room for coders of all abilities and also people who feel that they can contribute in other ways.  I mean no animosity towards anyone, and I would hope that the two projects could work in parallel on good terms and share information/code.  There are different approaches to the same problem, and I think there is a fundamental split in this community.  Perhaps everyone will get along a lot better if those two styles of management/organization are separated into different teams.  Clearly there are a bunch of us who are very eager to contribute but who feel unable to, for one reason or another, as we currently stand.

What does everyone think?

[darrenmason]

Tim,

You have certainly made some valid points - but your comments about the project "failing" are certainly subjective and may not necessarily align with everyone.

If the goal was to achieve complete market dominance in a fixed amount of time - then I think that then there is a good chance with the current setup there would be a failure.

If the goal is to have a codebase that is built on a solid architectural framework and can be used by whoever is inclined for whatever puprose they see fit then it is pretty close and is probably achieveable the way things are going.

The point is, it is only failing if the goals are clear and agreed by all and even then they might not be common goals so it might still be a success by others criteria.

I think getting a new 'release' out seems to be a common goal to most people that is having issues at the moment and perhaps there is an amount of management lacking.

I don't think that there is this big rift that you speak of. Certainly not one that would warrant a fork. There is constant stream of code and documentation updates (just check the logs) so in my view the project is not stale which to me would be a criteria for success/failure. I think that most people that are eager/keen enough find that they can contribute.
There are strong personalities aplenty and some "interesting" discussions but I am not seeing many peoples questions actually not getting answered (if there is an answer).

So if you  feel that you can add some value in the project managment aspect then step up and do it. Suggest some updates to the vision/strategy etc. If there is value to others then I think you will see adoption of it.

If you really feel that the developers investing all there time are not in alignment with your goals then there is room for you to seperately follow your goals independantly and still be involved here. I think that Andrews(totallymaxed) company is a good example. He has obviously thought out his own business model/plan and invested heavily to establish it. I assume as part of that plan he manages/limits his hard dependancies upon linuxmce but at the same time has also been one of the driving forces on giving the project continual momentum.

Just some thoughts

Darren

[tmoore]

Good points, and don't misunderstand, I don't want the project to fork either.  That would be a last resort, but others have raised the possibility to me privately.  It would be far better that everyone pulls together.  But we need to work better as a team.  People can and should be able contribute in different ways and at different levels of experience.  We should be accepting to all and do what we can to help them help the project.  Success is measured by everyone (that's everyone) enjoying the process and creating something good.

At the moment the 5 devs seem to hold the keys to this working in LMCE.  If the community can get the agreement from them that they are prepared to relinquish and/or share some responsibility, in order to help them and the project as a whole, then we will be on the right track.  We haven't heard this to date.  There have been too many comments along the line of "no, we're doing it this way", and not enough of "let's discuss the pros and cons and decide".  I'm not suggesting that every detail should be discussed, as nothing would be achieved, but fundamental issues such as how the project is run, whether it should have a commercial element etc. should be open for community discussion and decision.

I'll propose some changes to our general approach over the next few days, and hopefully we can discuss and agree a way forward.

Tim

[tschak909]

tmoore, why don't you actually SPEND SOME TIME STUDYING THE SYSTEM ARCHITECTURE, FIRST?

-Thom

[tschak909]

And before you say any sort of retort resorting to elitism...

* Those of us doing the work, tend to respect others who also dig in and do work too.

* Those of us doing the work write the code. He who writes the code HAS THE FINAL SAY.

This is because we're down in the trenches, and we choose to give away the fruits of our work for others. There is an unwritten rule here that those who recieve should also give in kind.

So, when you sit there on top of the mount and preach that "you should do this and this and this and this..." guess what? it sounds like you're preaching from the mount.

tmoore, it is THAT reason why I am ascerbic to you. You have done absolutely nothing to gain credibility in this community. If you want your incredibly assertive statements to carry any weight, you need to change this and get down here in the dirt with us.

And the sheer fact you just back-pedaled on your last thread really did nothing to help your credibility. Just a tip.

-Thom

[merkur2k]

See, you're getting confused tmoore. Its not "no, we're doing it this way", its "these decisions have already been made and we are sticking to them". You need to sit down and actually learn why these decisions were made the way they were and why the project would fragment and go backwards if such core goals were to be changed so late in the game.
I am also having a hard time understanding just exactly what it is that potential devs are expecting here. I myself came into this with a general software developement background (actually mostly web scripting stuff) and a good solid understanding of linux. Yes it has been difficult getting up to speed, but not at all impossible. It seems like people are expecting to get full training in c++ or whatever other language, which is simply not gonna happen. There are plenty of books and expensive college courses out there where you can learn that stuff.
Its all there in the wiki, or at least is mostly there. The rest is in the code. If someone is trying to work on a task but cant quite figure out the details of how some part of the system works, ask in the developers forum. And then add it to the wiki. Thats all that is being asked here.
I don't even know c++, but I have found plenty of projects to work on so far. I have submitted patches for the web admin, installer related shell scripts, and am currently working on a dvd-based installation method. I've never done alot of this stuff before, but i've got the wiki, the code, and google at my disposal.