Related
Hi all,
I really appreciate the work of all ROM cookers, even if I don't use a ROM from XDA anymore. (I was using some german releases here for quite some time)
I don't know much about ROM cooking, since I'm a SAP and Web-Deveopment guy, but I always asking myself:
Why is ROM cooking not like open source stuff?
Why is everyone/every group cooking it's own little stuff?
Why not working together on few versions to provide a ultimate, stable, bugfree, nice documented and "perfect" ROM?
Newbies are totally overloaded with 1337 ROM releases, which they should take? Where to find help for your decision? Not everyone has the time/KnowHow to try out 5+ ROMs...
There are so many many advantages if they would do so:
+ bugfixes are done only one time, not many times for many different roms
+ bugfixes are always up to date
+ mistakes are not done multiple times
+ more manpower to test/optimize/develop roms (since everyone works together)
+ clean and clear buglist and release history
+ bugtracking (easier for community to post bugs via bugtracker, easier to develop for cookers)
+ better for newbies (a stronger/bigger community is maybe not a bad thing?) - people know what to download (stable/beta/nightly builds releases - maybe light and full ROMs)
+ many many more, maybe even better reasons
And you can still release ROMs weekly and in multiple version (nightly builds, betas, alphas) but you also can make rock solid milestone releases for the community.
My feeling is SOMETIMES (not in all cases!), that cookers try to compete against each other, than working together.
As I said, its just a feeling... nothing personal against anyone.
Of course there are also disadvantages in organizing such a "structure", but in the end if a cooker would stop developing and invest this time into organizing the rest in to a developer group, that would be a benefit everyone. Means, a bit less "development power", but a lot more efficient developing. The big picture counts.
Just wanted to start a discussion about this thing. It is not like I'm totally experienced in this topic.
Because to be honest... much of the work (which is great and I really appreciate) is done multiple multiple times.
And many many ROMs are more or less the same. Different languages and "feature" levels like Light/Full would be sufficient.
It does seem like a good idea. Certainly, it would allow the chefs to develop an amazing Manilla X1 ROM, and a non-Manilla ROM, 6.1 ROM and a 6.5 ROM, light/full versions and other languages quite effectively. However, I'm not sure how much the chefs would like it. It would however, be excellent for the community. Would be interesting to hear the chefs views on this.
rom are developed really fast. too much people, too much organizations will slow down development.
most part of developers has trusted betatester and bug solver, working in small groups. afaik, looking at whole forum productions, this model works pretty well
yes I have the same question
if we all put our knowledge / findings together @ the cheffs share centre thread it would be A high mile achievement
guap said:
rom are developed really fast. too much people, too much organizations will slow down development.
most part of developers has trusted betatester and bug solver, working in small groups. afaik, looking at whole forum productions, this model works pretty well
Click to expand...
Click to collapse
I'm not too sure if you can call this situation "well". I would say it "kinda works".
And yes, in the short run, the development is slower, but in the long run the development is faster, because it is more efficient and there is less work which is done multiple times. As I said, it is the big picture you have to see, in the long run this would be much better.
Additionally to that "new" cooks who maybe have a new bugfix and want to release that, they have to do many of the other fixes again to release a new ROM (additionally to that they lack experience and will make "new" bugs).
In a single project, they simply post their fix, and the main developers check and integrate the fix ... done. Next build it is included.
And that's much faster!! And it is double checked for mistakes for more stable releases.
---
For the user it is unthinkable to update the ROM every week. There is a need for some "stable"-like ROM for general users.
The level of final release ROMs here are more or less like "Firefox RC" or even "Beta" Versions.
this is not the 1st time someone like u come up with this......loads of ppl had the same request. even i saw atleast a couple of such thread in this x1 thread!!!
but they neva work in most cases.....i only saw one such successful project!!!
That's not a bad idea at all.
I have been active for sometime with XDA equivalent for AVM Fritz modems (ip-phone-forum), together they made a compiler tool (Freetz.org) for all avm fritz modems, users download one tool, launch and select what features they want in their router, and build a new firmware/ROM, and there is a huge forum for support.
well, i don't say to make such a tool (while it will be so interesting to have, at least one per smartphone model)
having a xda organized in the way mentioned, many bugs can be resolved easily.
in the current state, a newbie will have much difficulty to read a 200 page thread about some custom ROM to find that in post 1232 describes the issue he encountered and later has been fixed in post 1325....
i was going to ask about making a sub forum per each custom rom, with threads divided per subjects, so if someone encounter an SMS issue will have to read one thread completely about sms issues rather than a very long one about all issues in that custom rom.....
if something could be done, then it will become much easier for everybody !!!
great idea.
There is one issue though: people have strong personal flavors of ROMs. Some might want a fast and lean ROM while others might want a mighty one.
I guess some kind of sophisticated branching is needed to accommodate these requests
yes, different tastes, but same bugs
I am down on a common knowledge database, willing to share and learn
It would be great indeed to have a common project but all chefs would need to aim to the same release that would take time to come to a common design and implementation.
Hope this can happen though!
Maybe a cooperation of some chefs would help to reach something here.
mercuriussan said:
And yes, in the short run, the development is slower, but in the long run the development is faster, because it is more efficient and there is less work which is done multiple times. As I said, it is the big picture you have to see, in the long run this would be much better.
Click to expand...
Click to collapse
The problem is there is no long run for rom cooking.
Most of us (users or cook) seldom use a single phone longer than a year and we moved to a next device and start everything from the ground up again.
Even you keep the same device for a long time, newer WM build or other components comes very frequently and there's always a tradeoff between "new features" vs "stability". And every cook has their own perference and it natural to see varies roms among cooks. Therefore, it will never have an easy solution for rom user.
For cooks, we could open threads for individual components and work together. I did open a threads for TF3D v2 (landscape), and HD Camera as a project base, and Itje already create a chef discussion thread and I think this model works pretty well for cooks.
agree with jackleung in some points
I've created a Topic for cooking with windows mobile 6.5, too.
Everyone was invited to discuss their bugs and help each other.
But none like it really.
Or none got bugs
Basically what everyone wants is a ROM with PURE Windows Mobile as Microsoft gives it to manufacturers(No HTC, Sonyericsson stuff), cabs/packages of different applications and a program that combines all of them together according to individual needs.
I'm aware that I should probably post this is the Dev area, but I'm not a Dev yet lol and don't want to clutter up their precious 'new release' space with what I consider a n00b question.
However, I'm a budding software developer who was very interested in developing apps for android - all of my ideas have been taken up by people and though I contribute a little towards those with suggestions I would like to jump in at the deep end instead.
So i figure, why not start a thread with links to relevant places to learn how to mod the Android OS - whichever version. Not like the dev section where its pretty hardcore, but introductory links:
Everything from removing / adding bits and pieces to an established custom / stock rom, to compiling your own kernel and porting versions could be covered.
It might also offer an incentive to the best devs on the android asteroid to continue knocking out tutorials even after they've moved on to newer versions?
Anyway, just an idea, what do you guys think? Personally at the moment I'm trawling through craploads of official docs trying to get a handle on where to start!
What your looking for is the stickies. They are at the top of each forum subsection. Happy reading. Props on putting it in q&a section though.
doofah said:
I'm aware that I should probably post this is the Dev area, but I'm not a Dev yet lol and don't want to clutter up their precious 'new release' space with what I consider a n00b question.
Click to expand...
Click to collapse
DING DING DING! We have a winner! You sir, deserve a trophy for actually thinking!
Check out the "getting ADB up and running thread" It will point you to the android sdk and on that page you can download eclipse. Those should help you get on your way.
Resources for moddingo
Lol... Yay!! My brain was a worthwhile investment, must remember to send a letter of recommendation to my supplier!
Thanks guys I'll keep digging
My poor G1 is in for a ROUGH ride over the next few months I fear... maybe I should save for another just to be safe lol
There are several other threads that cover this. Search for those and they have links and documents relating to this.
You'll have to have knowledge on linux, java, c++, and someother stuff.
Hi,
Firstly, I don't want to sound unrespectful to developers. It is just a few complains but I admire the time each one spend for creating new roms. There are lots, lots of roms/kernels in the android development section. It's a complete jungle! For a non technical user, I find it really difficult to choose a rom because:
There are no standard table stating what is different with the other. The list described by each developer is not always clear. Especially when he or she sees "aroma", "CRT", "ram hack", "zipalign" without explication. Wouldn't one nice and small sentence be sufficient to explain these terms?
The issues for each rom are not tracked. I think there should be a tracking system like JIRA so that we can confirm if a bug is already described and if it is still present in future release.
It should be clear if a rom is based from stock or AOSP rom (edit: it's not always obvious).
There is no centralized topic explaining the differences. But I have seen there is topic listing roms.
To my mind, there should be in each rom description something like a history, a short sentence like "I have based this rom on this other rom because I feel it lacks this and that." (not always the case), instead of just throwing "hi, [screenshot], how to install this rom?". ^^
I believe what happen is that one user will check the 3 or 4 first roms and will get confused.
Generally speaking, I feel that when you read a rom description from the author, it is addressed to technical people. I think a centralized JIRA (each rom = one project) idea would be very helpful to compare but I don't know if it is easy to implement.
What do you think?
ergosum said:
Hi,
Firstly, I don't want to sound unrespectful to developers. It is just a few complains but I admire the time each one spend for creating new roms. There are lots, lots of roms/kernels in the android development section. It's a complete jungle! For a non technical user, I find it really difficult to choose a rom because:
There are no standard table stating what is different with the other. The list described by each developer is not always clear. Especially when he or she sees "aroma", "CRT", "ram hack", "zipalign" without explication. Wouldn't one nice and small sentence be sufficient to explain these terms?
The issues for each rom are not tracked. I think there should be a tracking system like JIRA so that we can confirm if a bug is already described and if it is still present in future release.
It should be clear if a rom is based from stock or AOSP rom (not always the case).
There is no centralized topic explaining the differences. But I have seen there is topic listing roms.
To my mind, there should be in each rom description something like a history, a short sentence like "I have based this rom on this other rom because I feel it lacks this and that." (not always the case), instead of just throwing "hi, [screenshot], how to install this rom?". ^^
I believe what happen is that one user will check the 3 or 4 first roms and will get confused.
Generally speaking, I feel that when you read a rom description from the author, it is addressed to technical people. I think a centralized JIRA (each rom = one project) idea would be very helpful to compare but I don't know if it is easy to implement.
What do you think?
Click to expand...
Click to collapse
Actually always or it is baseed on stock or aosp or cyanogenmod it's pretty much in every thread i see. it's always described in the information they write.
and a centralized topic where is everything written well. it's for everybody else some like Lg stock ui and some like more clean android
Issue tracking jira i dont really understand i only know that developers are just wanting a catlog if there is a problem so they can see from what the problem is coming from or what it is related.
and that from a short sentence well. every rom misses something it's for everybody different and in almost all roms they are describing where it is based on and they have permission to use other work (the base) but some people are not asking it like the miui team who just took kowalski kernel without permission and they don't even give credit or something like that.
this post is just my opinion how i think it is
ergosum said:
...
Generally speaking, I feel that when you read a rom description from the author, it is addressed to technical people.
Click to expand...
Click to collapse
I do support the idea of asking the developers for more detailed info of their ROM / kernel features, however keep in mind that this is an Android development site - members are expected to have some level of knowledge and / or the will to learn stuff by themselves.
On the other hand, I've seen developers who care enough to spend time writing detailed documentation about their work and yet you'll find people asking questions or posting stuff without even reading the OP or using the search function. This obviously discourages the developer. Who will want to spend hours or even days writing how their stuff works or what are its main features if people won't take the time to read it?
DrummerMuppet said:
I do support the idea of asking the developers for more detailed info of their ROM / kernel features, however keep in mind that this is an Android development site - members are expected to have some level of knowledge and / or the will to learn stuff by themselves.
On the other hand, I've seen developers who care enough to spend time writing detailed documentation about their work and yet you'll find people asking questions or posting stuff without even reading the OP or using the search function. This obviously discourages the developer. Who will want to spend hours or even days writing how their stuff works or what are its main features if people won't take the time to read it?
Click to expand...
Click to collapse
I understand, but I don't ask for hours of well descriptions. One sum-up sentence and one link to a topic where other people has already spent the time should be enough. Example: an developer says "here are the 3 kernels [links]" and the only difference you can tell when looking at the links was "32rh", "16rh", "0rh". I had to guess that "rh" is ram hack and search for what it is. While with just "the ram hack is taking memory from GPU for the CPU", it's a clean and quick explanation. Otherwise people are bound to ask questions when they see this.
As for the JIRA system, it's just more obvious to see, for a newcomer:
how much issue a rom has at the moment
if a troubling issue important for you is described and prevent you from using the rom (so you don't waste time)
help the developer to see easyly what is wrong (instead of navigating through all the post)
help the developer by centralizing all the relevant information for one issue in one page, instead of having these spread everywhere
I'd be interested to see a list of the main developers involved in this
So far I've seen
(excuse my spelling)
Pulsar_g2
Chainfire
Eespen
Entropy512
XpLoDWilD
Rebellos
Apologies to the people I missed, give me some names below to add to the list!
Who are the rest?
t0ph0id said:
I'd be interested to see a list of the main developers involved in this
So far I've seen
(excuse my spelling)
Pulsar_g2
Chainfire
Eespen
Entropy512
XpLoDWilD
Who are the rest?
Click to expand...
Click to collapse
There isn't a list at present, for the simple reason that this is a community project, and it's more important to have community involvement than to celebrate some big names that are involved.
There are a few more people than that though
I guess we can get together a list later on, but I'll have to make sure I don't miss anyone - this is why people lists are tricky to maintain
Going forward though, omni is intended to be something anyone can contribute into, be it translating, testing (once builds are being made officially), developing features, giving devices support, or helping out in forums.
pulser_g2 said:
There isn't a list at present, for the simple reason that this is a community project, and it's more important to have community involvement than to celebrate some big names that are involved.
There are a few more people than that though
I guess we can get together a list later on, but I'll have to make sure I don't miss anyone - this is why people lists are tricky to maintain
Going forward though, omni is intended to be something anyone can contribute into, be it translating, testing (once builds are being made officially), developing features, giving devices support, or helping out in forums.
Click to expand...
Click to collapse
Thanks for the response, I should have known something like that was coming.
:cyclops:
t0ph0id said:
I'd be interested to see a list of the main developers involved in this
So far I've seen
(excuse my spelling)
Pulsar_g2
Chainfire
Eespen
Entropy512
XpLoDWilD
Who are the rest?
Click to expand...
Click to collapse
That list gives me goosebumps.
You can add Rebellos to the list!
Just names
And as pulser said the term "lead developers" is misleading. We see ourselves not as that.
Sent from my Find 5 using xda app-developers app
maxwen said:
Just names
And as pulser said the term "lead developers" is misleading. We see ourselves not as that.
Sent from my Find 5 using xda app-developers app
Click to expand...
Click to collapse
Nice to see you, and the others, here
I think there will probably always be a need for leads/senior developers (e.g. those with global +2) just to prevent something from descending to chaos - however, I want to try and minimize the distinction between leads and others, and only have overriding happen when there's a really good reason to do so, as opposed to something getting -2ed because a maintainer just doesn't like it.
Most importantly, if something is rejected, it's going to have to have a detailed explanation from multiple people of why.
pulser_g2 said:
There isn't a list at present, for the simple reason that this is a community project, and it's more important to have community involvement than to celebrate some big names that are involved.
There are a few more people than that though
I guess we can get together a list later on, but I'll have to make sure I don't miss anyone - this is why people lists are tricky to maintain
Going forward though, omni is intended to be something anyone can contribute into, be it translating, testing (once builds are being made officially), developing features, giving devices support, or helping out in forums.
Click to expand...
Click to collapse
Don't forget me.. like you forgot to add my ssh key when you switched domains (btw, can you fix that?)
Entropy512 said:
I think there will probably always be a need for leads/senior developers (e.g. those with global +2) just to prevent something from descending to chaos - however, I want to try and minimize the distinction between leads and others, and only have overriding happen when there's a really good reason to do so, as opposed to something getting -2ed because a maintainer just doesn't like it.
Most importantly, if something is rejected, it's going to have to have a detailed explanation from multiple people of why.
Click to expand...
Click to collapse
Maybe the term "project overseers" is more appropriate
maxwen said:
Maybe the term "project overseers" is more appropriate
Click to expand...
Click to collapse
I would prefer "dark overlords" personally.
Entropy512 said:
I would prefer "dark overlords" personally.
Click to expand...
Click to collapse
Master Blaster
Sent from my Find 5 using xda app-developers app
I assume the primary initiators of this rom have had this discussion but one thing I'd like to see is "this is the place where we talk about things."
Taking cues from other major rom projects you'll get a decent user base in using several forums or social services and some places get overwhelming or spread to thin.
Examples: a rom site having their own forum which doesn't get laid out properly and never gets used anyways. then an xda subforum along with a device-specific thread in each device's subforum. Then an omnirom wiki, an xda wiki, an irc channel, Google+/Twitter/Facebook trifecta, androidcentral,rootzwiki and a few others, jira, gerritt, mailing list, hangout...
You get the idea.
Social media sites aren't going away- personally Twitter and Facebook for a rom seem frivolous.
Mailing lists and irc are antiquated unless developers want a quiet place to talk. The many forum thing just seems onerous and duplicated. I dunno. I like a wiki, this subforum but broken down as necessary, jira, gerritt, and whoever wants to tackle the social trifecta can do that.
Basically, you can see what things I'm saying could get dropped or consolidated. I think this rom has that in mind and I could see it growing fastest knowing where to go.
And, personally, after using Arch Linux, their wiki is freakin' impressive. I've rarely had to turn anywhere else.
If omnirom broke the mold and had an awesome wiki that developers and active users maintained for device specific statuses, feature descriptions and statuses, etc. it would eliminate clutter.
Just my thoughts but it needs support. I guess you'd have to use Arch Linux to even know how impressive it is. Cyanogenmod did a pretty good job starting their wiki but it sort of slowed down. I think their wiki actually helped them get to being the most popular rom because of the wiki because of how much it offered for the new tinkerer.
So ultimately you are suggesting a solid wiki. Great. Help us do so by contributing to docs.omnirom.org
Sent from my SM-N900T using XDA Premium 4 mobile app
As a general Linux user the Arch wiki is still incredibly useful. I agree that wiki format is probably the best for conveying information about a ROM, the hard part is just getting people to keep it up to date and accurate. It'll be good if for instance each device had a wiki page that is kept up to date with known issues and workarounds. I feel though that the forum/discussion needs to be integrated with the wiki somehow for maximum effectiveness, because otherwise you often get a situation here on xda where a thread OP is filled with all sorts of useful, even up to date information, and yet people ignore it and ask questions that are covered there. So it's partly a social problem and partly a UI problem IMO. Anyway, just my random additional thoughts, agreed on the importance of a good wiki.
Well, I've taken it to heart and looking at the wiki more. Right now Last night I walked through the build instructions for the n4 but fell asleep during the repo sync. I've always wondered why projects have you download every old sdk. I guess it's in case you can't get them from Google? Seems unnecessary. I guess I can find out how to omit them from syncing. Those types of tips are interesting to me. I'm sure some things with development wouldn't seem quite as overwhelming if a repo was tiny.
jawz101 said:
Well, I've taken it to heart and looking at the wiki more. Right now Last night I walked through the build instructions for the n4 but fell asleep during the repo sync. I've always wondered why projects have you download every old sdk. I guess it's in case you can't get them from Google? Seems unnecessary. I guess I can find out how to omit them from syncing. Those types of tips are interesting to me. I'm sure some things with development wouldn't seem quite as overwhelming if a repo was tiny.
Click to expand...
Click to collapse
There is a flag for repo to only download the current branch. Might be worth using
jawz101 said:
I assume the primary initiators of this rom have had this discussion but one thing I'd like to see is "this is the place where we talk about things."
Taking cues from other major rom projects you'll get a decent user base in using several forums or social services and some places get overwhelming or spread to thin.
Examples: a rom site having their own forum which doesn't get laid out properly and never gets used anyways. then an xda subforum along with a device-specific thread in each device's subforum. Then an omnirom wiki, an xda wiki, an irc channel, Google+/Twitter/Facebook trifecta, androidcentral,rootzwiki and a few others, jira, gerritt, mailing list, hangout...
You get the idea.
Social media sites aren't going away- personally Twitter and Facebook for a rom seem frivolous.
Mailing lists and irc are antiquated unless developers want a quiet place to talk. The many forum thing just seems onerous and duplicated. I dunno. I like a wiki, this subforum but broken down as necessary, jira, gerritt, and whoever wants to tackle the social trifecta can do that.
Basically, you can see what things I'm saying could get dropped or consolidated. I think this rom has that in mind and I could see it growing fastest knowing where to go.
And, personally, after using Arch Linux, their wiki is freakin' impressive. I've rarely had to turn anywhere else.
If omnirom broke the mold and had an awesome wiki that developers and active users maintained for device specific statuses, feature descriptions and statuses, etc. it would eliminate clutter.
Just my thoughts but it needs support. I guess you'd have to use Arch Linux to even know how impressive it is. Cyanogenmod did a pretty good job starting their wiki but it sort of slowed down. I think their wiki actually helped them get to being the most popular rom because of the wiki because of how much it offered for the new tinkerer.
Click to expand...
Click to collapse
It is always an interesting questions of user dynamics, as to where people go.
I, for one, refuse to really use any of the social networks for this myself, as they're really not cut out for the task. They're also generally closed platforms that want to erode your right to privacy, and simply use you as a money making tool.
Aside from that though, social networks aren't ideal for hierarchical organisation. Right now there are a few things to consider.
There must be a code review platform for reviewing and approving code submissions. This will be separate from the others and most likely be gerrit. You can't replace that with a wiki or anything else.
You also need a proper bug tracking and project management area. That's jira. As omni is about having an open direction, it's important we can have road maps visible for comment and discussion and improvement. It is also important to ensure that the bug reporting and project management area has enough functionality long term. The only feasible way to do that is with dedicated software.
In terms of irc, it is most definitely not antiquated in the open source development community. If anything it is highly active, and often better than the modern alternatives, as it is an open standard allowing you to use any client software you please. It is also pleasantly simple and allows for real time group collaboration.
The big plan for omni is proper documentation though. That's where the wiki comes in. Not just of user information, but of developer information to help new developers.
delete post. post 9 is the real conversation
delete post
delete post. I'm creating another topic to change the subject