OESF Portables Forum
Model Specific Forums => Sharp Zaurus => Zaurus - pdaXrom => Topic started by: chuckr on June 17, 2006, 09:01:17 pm
-
Well, in the last couple years, i looked at many projects that are available for the Zaurus, but none of them really were just "rigth on the money" for me. I won't even try to make this list complete, but here are some of my complaints:
OZ, GPE, Opie: All seemed to be based upon impenetrable Python scripts. Last time I looked (and it might have changed, I admit) they were very, very unstable; what built right now might be broken in 5 more minutes. The size of the tools were so huge, my chances of actually getting my head completely around it were slim.
pdaXrom: I like it was based upon GNU Make macros (I could understand these, and repair them myself), but when I tried to contribute, the guy who ran it, Sashz, wouldn't accept any software; it was a case of "say yes, but do no". I tried complaining to oesf members, but got no response, when I offered to shoftware host an archive for our own development. OK, they're out ...
OpenBSD: probably the single most stable, but the problem is, their compatibility level, software-wise, to the GNU tools, is very shakey, and since the QT and GNU tools are important to my end goals, unfortunately, this is also out. Too bad, they were otherwise the best.
There is a development tool that it seems everyone is ignoring, but which seems to be really, really interesting: it's the Pengutronix folks from Germany. This is the stuff that's at the center of the pdaXrom development. in fact, it seems to be more than 85% of what pdaXrom entirely is. this means, if I can get a system based upon ptxdist (Pengutronix's tool) to work, I will be able to re-gnerate the largest piece of what Sashz has been blocking me out of all these many months.
OK, so maybe I have a possible goal. I have already used the crosstools to make myself a arm/softfloat compiler, but I thin I will try to fall back on the one provided by ptxdist. If anyone wants to join me in this, it's still quite early, but send me mail, I will freeze no one out, I ask for no money at all. Tell me please, right up front, if you want me to let yo contribute directly, and (if so) how much you know about ssh and cvs, and whether you would like me to coach you on this. I will do it as fast as I can, and as fast as events allow.
Well, to goals. I am wiling to listen here, so I'm putting out my own goals, because THOSE I know. I want to have two parallel graphics, both available at the same time, and switchable via hotkeys. One of the graphics to be X11: not sure about the GUI: coiuld be GPE, could be KDE, or even something like WindowMaker. The other to be baed upon QTopia, the frame buffer, so as to allow all the stuff originally delivered with the Zaurus.
I have ptxdist happily building, but it's huge, needs some chanes, and I'm not reallyy ready to even commit very much, but if I get some folks together, then, yes, I will commit to an svn archive, and start things off. I only want to get some folks to come along. I won't even freeze Sashz out, if he wants in, but if he tries to strangle development, well, I will keep him out of it then. No one here can talk me out of that, And I'm really disappointed that folks here allow this to go on. I'm going to be ready to begin commit within a week maximum, even if I get no offers of help. Maybe less. I have a lot more questions to get answered.
Why do another dev system? Because this one is very simple to modify, so we can all make our own contributions, without having to die over it. On top of that, I have been a little miffed at some folks who are a lilttle too impressed with themselves, and i will try to keep this as open as possible, consistent with good quality control (I used to be a release engineer).
-
I will be the first reply to this thread...
If I could see how well built the ptxdist is built... and how it is stuctured.... then yes... I would be willing to help out the cause...
I going to do a lil research on this Pengutronix... will get back to you... if I could get a working ROM for the Z... I would do alot more building from the Z it self... Cross compiling isnt my thing...
Late
-
Not sure why another flavor is being discussed here.
Sharprom works fine. Not sure how long sharp will stay with the cause though and the next step would be openzaurus on an Axim?
I say stick with the guns and build software for the sharprom.
-
Not sure why another flavor is being discussed here.
I agree Cresho... If it is fine with Chuckr... I vote to move this to the general topic forum....
Late
-
Not sure why another flavor is being discussed here.
I agree Cresho... If it is fine with Chuckr... I vote to move this to the general topic forum....
Late
[div align=\"right\"][a href=\"index.php?act=findpost&pid=131568\"][{POST_SNAPBACK}][/a][/div]
it doesn't bother me, but maybe let this wait at least one day, see what folks who DO their own software say. If you are relying on Sharp, then it's extremely obvious that you don't do any software yourself, because Sharp hasn't opened up their software, they've only allowed you to sell your apps alongside theirs. You can't contribute to those who won't take it (which is why I couldn;'t get along with Cacko (pdaXrom), cause he wouldn't actually take patches.
But as far as moving it, please, I'm not as good as I could be at this (using oesf yet) so I maay have made a mistake in placment, and I wouldn't mind getting it moved, IF it wasn't hidden.
-
As far as the principle of linux is concerned, I guess it's not a bad thing to have more openness ... but I was just wondering if channelling new coding efforts to OZ/GPE or OZ/OPIE would be a better alternative than forking pdaXrom?
OZ/OPIE <-> SHARP/Cacko
OZ/GPE <-> pdaXrom
Is the above a good generally idea of the GUI match?
-
so what you are proposing here is to create a fork of pdaXrom. Do you really think it is wise to do so? The Zaurus community is quite small and already quite fragmented. Adding more fragmentation surely will only add more confusion and user anguish?
but I can see where you are coming from. the current status of pdaXrom isn't very pretty. one could say it is out of control and at a risk to just vapourise but it is not just that bad yet.
maybe we should all take a little breath and try again to reconcile our perceived differences and try to work something out that will help everyone.
pdaXrom has lots of things going for it which is what attracted me to it initially. but lately, those things are vapourising and OZ is picking up trying to imitate some of them.
what I liked about pdaXrom is that it was based on stable packages, not bleeding edge stuff but that has changed...
the other things good about it was the native dev tools which made compiling on the Z itself easy as pie so I could compile stuff anywhere I and my Z were without needing to be at my PC.
also the vmware image and iso image for X86 was great (cause all my PCs run windows) so I could try out pdaX on a PC before putting it on my Z and also cross compile from the pdaX version.
pdaXrom also had a great community with lots of people contributing and generally being a cheerful bunch of non cynical users.
it was also using an almost compatible set of libraries and kernel to the Sharp ROM.
lots of games were being compiled for pdaXrom (I personally dont play much but its good to have some around when you feel like a game or two so its a good thing)
lots of other useful contributions were made by other users and pgas enthusiastically collected them all and put them on the contrib feeds.
the community worked together for coming up with better ways of doing things, ie the new package announcement thread.
beta1 was looking good, things were relatively stable and further betas should eventually resolve all the bugs and there was a notion that a stable 1.1 release was near.
then things started to go down the hill. rivalry with oz team, but that's over now thank goodness.
the pdaxrom site got revamped and links kept changing. it was bad, but everyone knew it was only going to be temporarily and that was accepted without much fuss.
then beta4 and new bootloader hit us.
there was a lot of chaos and a lot of people felt abandoned.
the 2.6 kernel stuff should had been done later or branched off to another release such as 1.1.1 or 1.2. would had made things less confusing
a release based on the 2.4 kernel should had been finalised first before playing with the 2.6 stuff or at least created two branches.
on the communication site, nobody really knew what was going on. the pdaxrom wiki was being worked on by volunteers from the pdaxrom community and blogs were made available. however, the blogs were mainly about the latest hacks, fixes to the latest beta4 but not much was communicated about the state of other things like, access to the contribution and user community section, version control system being worked on, etc... so people thought nothing was happening.
i was supposed to get access to the system to submit packages and code but it never happened and I lost interest while waiting...
-
my .02 U$
when I bought my 1st Z, a pxa250 5600, } thoughtthat dev on Z should concentrate on kernel/hardware stuff, and that pert should be sent back up to debian in hopes of making a real native debian zaurus distro (probably based on ssarge.) the pocketworkstation tools to lighten up debian stuff seeemd like a good idea. I was dissuaded by the lightweight ipk concept. pdax or oz has a deb to ipk conversion tool, I think. If we are gong to do some other damned project (like I do any work on any of this stuff lol) why not work to incorporate the Z into an extant system--- like debian. At this point, I doubt the Z is the most obscure arm computer around, and debian already buikds for arm....what...netwinders????
storage is now big and cheap. the current zs run on an entirely open 2.6. isn't native debian now conceivable? If you don't want to work on OZ, and can't work on pdax, and view sharprom as antique, and dead- then bring the z into a larger world, not another splinter of one. If DEBIan won' do it' keep trying distros-- ubuntu might, or novell might see cmmercial value innn a suse based Z. PLEASE! not another release of very whizzy sounding compilers 8 versions of aplha software and all the useful stuff broken--all based on some obscure script-fu that exists nowhere else but your project. In other words, I fear this will turn out to be a pdaxish son-of-OE. and we already have both. Then again, if you can get it going.......
-
my .02 U$
when I bought my 1st Z, a pxa250 5600, } thoughtthat dev on Z should concentrate on kernel/hardware stuff, and that pert should be sent back up to debian in hopes of making a real native debian zaurus distro (probably based on ssarge.) the pocketworkstation tools to lighten up debian stuff seeemd like a good idea.  I was dissuaded by the lightweight ipk concept. pdax or oz has a deb to ipk conversion tool, I think. If we are gong to do some other damned project (like I do any work on any of this stuff lol) why not work to incorporate the Z into an extant system--- like debian. At this point, I doubt the Z is the most obscure arm computer around, and debian already buikds for arm....what...netwinders????
storage is now big and cheap.  the current zs run on an entirely open 2.6. isn't native debian now conceivable? If you don't want to work on OZ, and can't work on pdax, and view sharprom as antique, and dead- then bring the z into a larger world, not another splinter of one. If DEBIan won' do it' keep trying distros-- ubuntu might, or novell might see cmmercial value innn a suse based Z. PLEASE! not another release of very whizzy sounding compilers 8 versions of aplha software and all the useful stuff broken--all based on some obscure script-fu that exists nowhere else but your project. In other words, I fear this will turn out to be a pdaxish son-of-OE. and we already have both. Then again, if you can get it going.......
[div align=\"right\"][a href=\"index.php?act=findpost&pid=131600\"][{POST_SNAPBACK}][/a][/div]
Wow, for me it will be the most amazing thing to be able to do on Z. Real and complete Debian!
But then the question is: What are the (technical) obstacles?
I've also heard that Debian is arm-friendly, and actually we can get some of the .deb packages to run directly on Sharp ROM by dropping the application and the required lib in the right paths. But how can it be done to make Z boot into Debian natively and directly? That's my dummy question.
-
NetBSD is pretty friendly on PDA/Handheld devices as well ... I remember running HPCArm port on a Jornada 680e.
Need to check the status on the Arm port though ...
-
There is a development tool that it seems everyone is ignoring, but which seems to be really, really interesting: it's the Pengutronix folks from Germany. This is the stuff that's at the center of the pdaXrom development. in fact, it seems to be more than 85% of what pdaXrom entirely is. this means, if I can get a system based upon ptxdist (Pengutronix's tool) to work, I will be able to re-gnerate the largest piece of what Sashz has been blocking me out of all these many months.
[div align=\"right\"][a href=\"index.php?act=findpost&pid=131551\"][{POST_SNAPBACK}][/a][/div]
I neither compile the software I use on the zaurus nor do I develop for the zaurus so I'm probably not the intended audience. Nevertheless, I'm not sure if more of the same is what "we" need -- although I admit that a pdaxrom-like x11-based distro with proper user accounts & vim7 & windowmaker & working suspend etc. would be extremely useful.
A debian-based distro would be nice too of course although this wouldn't fit into the internal ram I think -- not so cool.
Anyway, I somehow agree that the sharp rom still is usability-wise the best solution at the moment -- in the sense of being able to do the most with it as an end user. I think one also has to think of what is an appropriate use for such a device. Things I'd like to do are: run vim(7), run ruby 1.8 programs (preferably with tk or gtk gui), run zbedic, read html files, read pdf files. The bottom line is, I really don't care about technicalities -- except that I prefer a distro with proper user accounts/roles (I don't want to log in as root) and that I'd prefer a proper package management system with proper dependency management (i.e. something other than ipk). So, what I'm currently looking for and what I personally see a need for is much rather one of these than developing the next most-likely-never-to-be-finished-polished-and-shining distro:
- improving the terminal/console for qpe-based roms (i.e. improve qkonsole so that the full screen mode works properly)
- develop a good pdf viewer (qpdf(2) is nice due to the lack of an alternative for qpe-based roms; xpdf isn't much better IMHO)
- create an easy to use & easy to setup cross-compiling sdk for the current cygwin environment
I still follow the pdaxrom-related discussion because I think that a x11-based distro would have enormous potential and because I had the impression that pdaxrom was the closest to be actually useful. If you think you can do it, please do so.
-
If you are relying on Sharp, then it's extremely obvious that you don't do any software yourself, because Sharp hasn't opened up their software, they've only allowed you to sell your apps alongside theirs. You can't contribute to those who won't take it (which is why I couldn;'t get along with Cacko (pdaXrom), cause he wouldn't actually take patches.
yep! your damn right! Nurses do not know C++, GNU development tools, java, python.............................................and I feel for you when you have people like sharp making stuff so difficult along with trolltech. and last but not least, people not accepting your fixes. I tried both pdaxrom and openzaurus. Both are great.....but have big minuses and this is understandible.
In your quest, you would need gangs of support and not only that, how many versions of zauruses are floating around? Even after building and getting a proper environment for the z which would be supported accross all sl's and c's platform, how about software???!!!! your going to be so busy like sashz and drown. Not only that.....only around less than 10000 registered users on the forums only which I see around 2000 active and about 1000 c users in the forums. now start carving the pie on how many people actually flashed another rom and it starts to get very small and it's getting more fragmented.
Need gangs of support.,,,,...something like sashz support. and no job too just to finish projects.
If we can get debian on the z something like pocketworkstation as the main operating system, then we can see alot of stuff come together like a snap because of the mountains of software available already. The only problem with that, the idea of pocketworkstation seems to be droped and no updates since......Even then, where are you going to get the drivers.....just hack away. Last i flashed pdaxrom, my z kept suspending and the light would not work. Only trickery of odd fingers made it come back to life and disabling suspend. If sashz knew how to fix the problem, it would of been done long ago but he is hacking stuff! sharp will not release hardware info for developers? We are stuck i guess huh.
consider these for the os-
how stable can you make it?
do you have the speed to launch mozilla in 10 seconds?
can i put the rom image on the microdisk instead of the 128mb?
how easy is it to develop or port over other apps?
how about a decent website where i can read usefull information on the rom, developer's tools download the applications and not juggle through light years of links?
how easy would it be to reflash the z to standard spec?
The way I see it, they own the software and the hardware. If they choose, they can totally stop the zaurus line and leave us in the dark. At the moment, zaurus in japan is pretty big and as long as that is around, we have the zaurus. I like the idea of a stable environment in my hand and Meanie(sorry guy but i have to preach you) Made and effort to bring pdaxrom on the sharprom. Seriously you can run alot of good applications on the sharprom.
anyway, I tried squeezing everything I know in this small paragraph. These are my hopes as of the moment and hope it will help your quest.
-
Why do another dev system? Because this one is very simple to modify, so we can all make our own contributions, without having to die over it. On top of that, I have been a little miffed at some folks who are a lilttle too impressed with themselves, and i will try to keep this as open as possible, consistent with good quality control (I used to be a release engineer).
It may be not what you want to hear, but I don't buy this logic. The system is simple to modify and easy to understand while it's small -- e.g. while you target only one or few models and have a monolithic system that is not configurable.
Sooner or later you will add more targets, more software, more configurability, and all the attractiveness through simplicity will vanish. The system will get streched and you will lock out all users who didn't grow up with it. The OpenZaurus guys have been there and done that. That's why we abandonded our Make-based system (buildroot).
Take a fresh look at OpenEmbedded, it's been designed from the ground up to prevent this problem. It may need you to learn a bit more at the beginning, but this learning curve is not exponential. Once you have learned the OE basics, there is not much more to learn, since it is based on object oriented concepts and simplicity at the heart of the .bb language - it's really a flat learning curve after the first steps. Most beginners who were motivated to learn something new are saying that now.
As for the size of the tools, well, BitBake is like 2.2Mb, Monotone is like 5Mb, OpenEmbedded repository is 122Mb. This is containing all our patches to applications and all revisions we ever did. Now you only need the additional sources, but this requirement is shared with all other tools...
Oh last but not least, if only you need a SDK, you can use OE to build one. It will be like 80Mb or so.
-
Take a fresh look at OpenEmbedded
I second this. I am not even a programmer, but I think OE is the best choice. Easy to understand, lots of momentum, helpful people.
My impression is that people "unknowledgeable" like me can make good contribution to OE since we find the absurdest bugs. Getting them out of OE is happening fast and I think one of the most valuable contributions to OE at this point in time.
I have very special requirements (http://gakusei.sf.net), but even these have been fulfilled as of late and I was able in contributing to make it happen and I have no clue about C and only know the rough concepts of python. I could not even write Hello World in python (although I wanted to learn for a long time how to program in python). There are more packages than you can chew in OE.
PS: I do agree that the first steps of even the latest versions were a bit rough. I suggest using a loop file on either CF or SD which you can back up from time to time and thus backup and recreate your complete system in 15 seconds by just copying a single file.
-
Well, I guess it's about time I replied.
First, I think I better say, if you don't do any software yourself, then I don't believe you have any right to even comment upon this. the phrase "Dutch Uncle" comes to mind here, so if you don't do software, I don't think you have a right to comment, in any way, upon the doing of those who do. I'm not trying to insult you, but those are my feelings.
Second, I've had some comments says that the OPIE/GPE tools are simple. That's untrue. Try understanding the workings of all those python scripts. You actually CAN do that with the ptxdist stuff, the scripts I saw made sense, and while their documentation level is very, very low, their support level is very very right (maybe as high as the Opie/GPE folks all put together! they have mailing lists, irc channels, and lots of willingness to help.
So, which you might (after a lot of work) be able to understand WHAT the python scrupts are doing, I rather doubt you will get to the level of fixing the python scripts themselves. They are FAR more complicated than the Pengutronix stuff is.
Those of you who suggested staying with either pdaXrom or Cacko: are you aware that pdaXrom IS Cacko? They are one and the same, Cacko was the earlier name, is all. Past that, Sashz controls all of the code with an iron hand, and you CAN'T COMMIT to it, he won't open his archive. If it were open, I would have conributed to it a long time back, when I first tried to teach Sashz the workings of cvs and ssh. If he won't allow it, all I could do is sit aroudn with you folks, but I want to contribute, not just talk about contributing. I suspect those who suggested staying wiht Cacko must have been one of the folks who doesn't themselves write software. Like I said, those are Dutch Uncles, they don't qualify to comment on a development project if they aren't developers.
-
Second, I've had some comments says that the OPIE/GPE tools are simple. That's untrue. Try understanding the workings of all those python scripts. You actually CAN do that with the ptxdist stuff, the scripts I saw made sense, and while their documentation level is very, very low, their support level is very very right (maybe as high as the Opie/GPE folks all put together! they have mailing lists, irc channels, and lots of willingness to help.
So, which you might (after a lot of work) be able to understand WHAT the python scrupts are doing, I rather doubt you will get to the level of fixing the python scripts themselves. They are FAR more complicated than the Pengutronix stuff is.
Do they need fixing? Is this a requirement for developing using them? To draw an admittedly absurd parallel, should I have to understand how the Linux virtual memory paging system is implemented to be able to write a "Hello World!" binary in C?
Those of you who suggested staying with either pdaXrom or Cacko: are you aware that pdaXrom IS Cacko? They are one and the same, Cacko was the earlier name, is all.
No. Cacko is based on Sharp's Qtopia, pdaXrom is an x11 based system, this makes them different from a number of points of view (including the build system afaik).
Past that, Sashz controls all of the code with an iron hand, and you CAN'T COMMIT to it, he won't open his archive. If it were open, I would have conributed to it a long time back, when I first tried to teach Sashz the workings of cvs and ssh. If he won't allow it, all I could do is sit aroudn with you folks, but I want to contribute, not just talk about contributing.
You can contribute to OpenEmbedded, we're always looking for more contributors, and not just for code - bug fixing, documentation writing (also for OpenZaurus), there's lots of opportunity for everyone to get involved.
Cheers,
Si
-
Chuckr,
Cacko and pdaxrom are not one and the same. As I recall, back in the pre c7x0 days, Maslovsky and Sashz worked together to produce enhancements to the Sharp ROMs available then. In those days, rather than the Sashz/OE wars, we had the Sashz/tkcROM wars, with accusations from Sashz that Proto has ripped off Cacko for tkcROM.
Some time after that, Sashz split off on his own to produce pdaXrom. Maslovsky continued work on Cacko and as far as I know still does, although things hasve been quiet recently.
So although there are historical ties, Cacko and pdaxrom are not the same.
Maybe I'm wrong, but you post suggests that you won't go ahead with this unless you get some people to join you in your efforts. Is that the case? I'd like to encourage you to just get on with it. I think people are tired of the "this system is better than that" bickering that we've had here for a long time. Show us what can be done that makes this system better, and people will follow you.
I'd like to suggest you speak the the OE guys in more depth, though. I think they made the choices they have for good reasons. I suggest it is worthwhile taking some more time to understand their reasons, so that even if you decide in the end that OE is not the way forward for you, you are at least prepared for the issues that you may run up against down the road.
Mike.
-
that pretty much leaves you all by yourself, there, chuck, since apparently neither the developers or the users meet your high standards. soo....go ahead, blow us away with a new system. I'll apologize for my cynicism if it appears i na useble form. in fact I'll even set up a build system and port some software.
In the meantime, OE makes progress, sashz is still around, and sharp and cacko are pretty useful.
-
First, I think I better say, if you don't do any software yourself, then I don't believe you have any right to even comment upon this. the phrase "Dutch Uncle" comes to mind here, so if you don't do software, I don't think you have a right to comment, in any way, upon the doing of those who do. I'm not trying to insult you, but those are my feelings.
In other words: "Fsck off users - I only want skilled devels and I do not care at all about users".
Users are most important part of Zaurus world - if you do not get them then your work is useless. Zaurus world is split into few parts and if something new came then it has to describe users why it need to be choosen instead of one of already existing solutions.
Second, I've had some comments says that the OPIE/GPE tools are simple. That's untrue. Try understanding the workings of all those python scripts. You actually CAN do that with the ptxdist stuff, the scripts I saw made sense, and while their documentation level is very, very low, their support level is very very right (maybe as high as the Opie/GPE folks all put together! they have mailing lists, irc channels, and lots of willingness to help.
As many know I'm one of core OpenEmbedded developers. I use it for over 2 years now and do not try to understand why some stuff is written in a way like they are written - it works and thats enough for me. If I want to get knowledge about some bbclasses then I ask other developers and get answer. I wrote many recipes during those 2 years and know that adding new stuff can take less then 5 minutes (plus buildtime) - and that some newbies are able to reach that too during their first week of using OE.
And OE is not "OPIE/GPE tools" but one of method to build those enviroments. Many people are building OPIE/GPE without even trying OE and it works for them.
So, which you might (after a lot of work) be able to understand WHAT the python scrupts are doing, I rather doubt you will get to the level of fixing the python scripts themselves. They are FAR more complicated than the Pengutronix stuff is.
Getting started with PTXdist need you to know make syntax (which has some caveats). To add something into OE you need to write "recipe" for it. For many targets simplest version contain two lines. For example any application which use autotools (common "./configure;make;make install" thing) is:
SRC_URI = "ftp://ftp.example.com/helloworld.tar.gz"
inherit autotools
I suspect those who suggested staying wiht Cacko must have been one of the folks who doesn't themselves write software. Like I said, those are Dutch Uncles, they don't qualify to comment on a development project if they aren't developers.[div align=\"right\"][a href=\"index.php?act=findpost&pid=131668\"][{POST_SNAPBACK}][/a][/div]
Like I said - you ignore users.
Anyway I wish you good luck on getting another distro for Zaurus working. I will look into your repository to check how some things get done and import some of them into OE and can offer help (on IRC) on some stuff (but it depends on free time from my free time).
-
A debian-based distro would be nice too of course although this wouldn't fit into the internal ram I think -- not so cool.
I was under the vague impression that pdaXrom was a hair's breadth away from being somewhat debian compatible if some of the innards were updated. If that happened and more debian packages worked on pdaXrom wouldn't that foster greater enthusiasm for it?
If someone were to fork pdaXrom or any other Z distribution. Making it compatible with a more mainstream system, so that it was a micro laptop, rather than a pda with a good screen and keyboard, would be the way to go.
However, I don't think pdaXrom needs to be forked. I just think it needs to be opened up, updated, cleaned up, debugged, given a direction and a clear roadmap for all to see where it's going, how it's going to get there and when.
-
However, I don't think pdaXrom needs to be forked. I just think it needs to be opened up, updated, cleaned up, debugged, given a direction and a clear roadmap for all to see where it's going, how it's going to get there and when.
[div align=\"right\"][a href=\"index.php?act=findpost&pid=131821\"][{POST_SNAPBACK}][/a][/div]
That' s what i' m telling Sashz since time but didn' t have a valid answer.
I' m tired and there to leave PdaXrom.
Happy Luck to you everyone.
Hope that something change, as i would like that the very little i can do and later improvements, will be for Zauri comunity benefits.
BYEZ!
P.S. Sorry guys, i have only one Zauri and it will make me do Distro choices time by time.
-
Second, I've had some comments says that the OPIE/GPE tools are simple. That's untrue. Try understanding the workings of all those python scripts. You actually CAN do that with the ptxdist stuff, the scripts I saw made sense, and while their documentation level is very, very low, their support level is very very right (maybe as high as the Opie/GPE folks all put together! they have mailing lists, irc channels, and lots of willingness to help.
So, which you might (after a lot of work) be able to understand WHAT the python scrupts are doing, I rather doubt you will get to the level of fixing the python scripts themselves. They are FAR more complicated than the Pengutronix stuff is.
I don't see the point here. You don't have to understand anything of either bitbake or the python backend of OpenEmbedded to be a productive developer and customize it for your liking. Even I don't understand everything of bitbake or the more advanced build classes -- that didn't stop me from being productive with OE
-
I was under the vague impression that pdaXrom was a hair's breadth away from being somewhat debian compatible if some of the innards were updated. If that happened and more debian packages worked on pdaXrom wouldn't that foster greater enthusiasm for it?
Afaik, debian don't use soft-float (yet) so neither pdaX nor OpenZaurus are compatible.
OpenZaurus used to be compatible (back in the days of OZ3.2) but in reality debian isn't optimised to work on these small screen, memory and storage constrained devices so the choice was made to move on and try to make things better.
I can't see that the lack of software is a real issue for either OZ or pdaX, or is it?
Si
-
However, I don't think pdaXrom needs to be forked. I just think it needs to be opened up, updated, cleaned up, debugged, given a direction and a clear roadmap for all to see where it's going, how it's going to get there and when.
[div align=\"right\"][a href=\"index.php?act=findpost&pid=131821\"][{POST_SNAPBACK}][/a][/div]
I agree with this statement.
-
I was under the vague impression that pdaXrom was a hair's breadth away from being somewhat debian compatible if some of the innards were updated. If that happened and more debian packages worked on pdaXrom wouldn't that foster greater enthusiasm for it?
Afaik, debian don't use soft-float (yet) so neither pdaX nor OpenZaurus are compatible.
OpenZaurus used to be compatible (back in the days of OZ3.2) but in reality debian isn't optimised to work on these small screen, memory and storage constrained devices so the choice was made to move on and try to make things better.
I can't see that the lack of software is a real issue for either OZ or pdaX, or is it?
Si
[div align=\"right\"][a href=\"index.php?act=findpost&pid=131846\"][{POST_SNAPBACK}][/a][/div]
hmmm ... how about NetBSD then? NetBSD seemed to be ported to all kinds of hardware. It supports X11 and its a full desktop OS. comments?
PS: I'm not trying to undermine Cacko, OZ or pdaXrom ... they are all great ... but on the topic of having a full desktop OS that can work off a device like a Zaurus, NetBSD comes to mind.
-
Second, I've had some comments says that the OPIE/GPE tools are simple. That's untrue. Try understanding the workings of all those python scripts. You actually CAN do that with the ptxdist stuff, the scripts I saw made sense, and while their documentation level is very, very low, their support level is very very right (maybe as high as the Opie/GPE folks all put together! they have mailing lists, irc channels, and lots of willingness to help.
So, which you might (after a lot of work) be able to understand WHAT the python scrupts are doing, I rather doubt you will get to the level of fixing the python scripts themselves. They are FAR more complicated than the Pengutronix stuff is.
Do they need fixing? Is this a requirement for developing using them? To draw an admittedly absurd parallel, should I have to understand how the Linux virtual memory paging system is implemented to be able to write a "Hello World!" binary in C?
Those of you who suggested staying with either pdaXrom or Cacko: are you aware that pdaXrom IS Cacko? They are one and the same, Cacko was the earlier name, is all.
No. Cacko is based on Sharp's Qtopia, pdaXrom is an x11 based system, this makes them different from a number of points of view (including the build system afaik).
Past that, Sashz controls all of the code with an iron hand, and you CAN'T COMMIT to it, he won't open his archive. If it were open, I would have conributed to it a long time back, when I first tried to teach Sashz the workings of cvs and ssh. If he won't allow it, all I could do is sit aroudn with you folks, but I want to contribute, not just talk about contributing.
You can contribute to OpenEmbedded, we're always looking for more contributors, and not just for code - bug fixing, documentation writing (also for OpenZaurus), there's lots of opportunity for everyone to get involved.
Cheers,
Si
[div align=\"right\"][a href=\"index.php?act=findpost&pid=131672\"][{POST_SNAPBACK}][/a][/div]
I said that the Cacko development was closed, and you replied that the OpenEmbedded isn't closed, don't you see your own bias there? I never said, and certainly never meant, thatn the OpenEmbedded was closed, my comments were ALL about Sashz and his closed shop, and how so many of you oesf folks appear not to mind being frozen out of all possible development.
Just, don't get me erong, you;'ll have the entier world thinking I said that about OpenEmbedded. My complaint about OpenEmbedded waas about the size of their tools, and the stability of them, but never about their openness.
-
However, I don't think pdaXrom needs to be forked. I just think it needs to be opened up, updated, cleaned up, debugged, given a direction and a clear roadmap for all to see where it's going, how it's going to get there and when.
[div align=\"right\"][a href=\"index.php?act=findpost&pid=131821\"][{POST_SNAPBACK}][/a][/div]
I agree with this statement.
[div align=\"right\"][a href=\"index.php?act=findpost&pid=131885\"][{POST_SNAPBACK}][/a][/div]
So what can be done to further this particular aim?
-
So what can be done to further this particular aim?
Please bear in mind that I only started using a Zaurus since Oct 2005 and pdaXrom shortly after that and thus don't know much about the trials and tribulations of how pdaXrom came to be what it is today ...
pdaXrom is Sashz's baby and as such it is understandable that he would want to retain tight control over its direction and development. However what is puzzling is why he seemingly/allegedly fails to incorporate patches for bug fixes etc. Perhaps it is due to the contributions regime. Until recently Sashz has been relying upon the contributions for his day to day expenses. Maybe Sashz is worried that if there were more contributors (of development work not cash) then he doesn't deserve or would feel guilty in claiming all the (cash) contributions. And the other thing is recognition, more developers may mean that Sashz's hardwork and contributions gets diluted (in terms of recognition).
So if we can reassure Sashz that he gets to keep all the cash contributions, and gets suitable recognition for all the work he has put into pdaXrom so far then maybe he would reconsider opening up pdaXrom a bit more?
Just my tuppence.
-- cheers
-
I'd like to re-examine the subtitle for this thread in a calm and non judgemental manner, ignoring a lot of the venom which seems to have infected this thread. We have to remember that english isn't everyone's first language and that people can read a piece of text and colour it however their brain wants to read it, without actually picking up on the poster's original intent. Please read this without emotion and just think logically. Thanks! :-)
The subtitle for this thread is:
"Re-examining what's done, could be done"
Without getting way too technical, because I'm not, what "could be done"? What would people like to see done?
Whatever happens to pdaXrom (or an alternative should someone wish to build it) this information is useful because it allows everyone to see what's wanted and possible.
My own opinion is this:
- I want a lean, fast and clean ROM for my Z.
- I would like suspend and resume to work as flawlessly as it does in the Sharp ROM.
- I would like battery life to be the same as in the Sharp ROM.
- I don't want lots of pre-bundled apps and tools I don't even know how to use or de-install.
- I like the ubuntu approach to linux, one decent app per category, but the ability to add whatever you want after installation would be a good approach.
- I think a number of images with different pre-installed apps might be an idea, a base (nothing pre-installed), basic user (word processor, spreadsheet, PIM, web browser, pdf viewer and mp3 player) and advanced (the developer centric tools and apps).
- I really want apps to open and fit the screen so that I'm not forever resizing something before I can even use it.
- I really want all the keys to work first time, correctly as expected. I actually like the sticky shift and function approach used in the Sharp ROM.
- Personally, I know I've said this repeatedly, I would like Open Office 2 to run on my Z. This is purely because my work uses it and it would be great not to have to take my laptop everywhere or to have to save files down into word/excel formats just to view them on my Z.
I've been using OZ/GPE for about a week and a bit now. I feel that this meets more of my needs than pdaXrom does in it's current form.
I'd enjoy hearing other people's thoughts.
-
I'd like to re-examine the subtitle for this thread in a calm and non judgemental manner, ignoring a lot of the venom which seems to have infected this thread. We have to remember that english isn't everyone's first language and that people can read a piece of text and colour it however their brain wants to read it, without actually picking up on the poster's original intent. Please read this without emotion and just think logically. Thanks! :-)
The subtitle for this thread is:
"Re-examining what's done, could be done"
Without getting way too technical, because I'm not, what "could be done"? What would people like to see done?
Whatever happens to pdaXrom (or an alternative should someone wish to build it) this information is useful because it allows everyone to see what's wanted and possible.
My own opinion is this:
- I want a lean, fast and clean ROM for my Z.
- I would like suspend and resume to work as flawlessly as it does in the Sharp ROM.
- I would like battery life to be the same as in the Sharp ROM.
- I don't want lots of pre-bundled apps and tools I don't even know how to use or de-install.
- I like the ubuntu approach to linux, one decent app per category, but the ability to add whatever you want after installation would be a good approach.
- I think a number of images with different pre-installed apps might be an idea, a base (nothing pre-installed), basic user (word processor, spreadsheet, PIM, web browser, pdf viewer and mp3 player) and advanced (the developer centric tools and apps).
- I really want apps to open and fit the screen so that I'm not forever resizing something before I can even use it.
- I really want all the keys to work first time, correctly as expected. I actually like the sticky shift and function approach used in the Sharp ROM.
- Personally, I know I've said this repeatedly, I would like Open Office 2 to run on my Z. This is purely because my work uses it and it would be great not to have to take my laptop everywhere or to have to save files down into word/excel formats just to view them on my Z.
I've been using OZ/GPE for about a week and a bit now. I feel that this meets more of my needs than pdaXrom does in it's current form.
I'd enjoy hearing other people's thoughts.
[div align=\"right\"][a href=\"index.php?act=findpost&pid=132003\"][{POST_SNAPBACK}][/a][/div]
As end user I totally agree with everything you said. I have not tried pdaXrom or OZ until now because every time I read about new releases some or most of the above is not working out of the box and I'm too lazy to mess with keyboard, window config files etc (ok, I do not consider openoffice to work out of box to be a requirement but firefox or some other full featured browser is a must - without waiting several minutes to start up, swap, etc - if opera can do it in cacko it's obviously possible).
-
Well, I've had my C1000 about 2 weeks now, and I'm completely new to the Zaurus! I've tried every ROM going (at least twice each!! ), loading applications, seeing how they work, etc., etc. In the end I've got so confused, I had to write myself a set of notes on each ROM and what worked with what. It's been very labour intensive, and I still don't know which ROM is most suitable for me, which GUI I like best, what apps I like, etc.
However what I do know, is that there is too much choice; each ROM with its own pros and cons - none do all that I want. As was mentioned earlier the community is probably too small to support the kind of diversity we have now - we should work together not split-apart, however that means making a choice and sticking to it, and then ensuring that we influence the development/features etc. of our choice - isn't this a better approach; rather than forking-off another development, let us choose one that is "looking pretty good", and then influence that so we get what we want in the future.
As a software developer (and Zaurus user), I would prefer to see a layered-model that allows more of a mix and match approach... We don't need loads of different kernels/boot loaders - just one that is stable, and includes all that the community needs. I would like to choose between different window managers on top of the kernel, and I would like to get any application to run on top of it. Furthermore I want any application to be able to run on the given window manager (or at least know it won't before I start!).
From my small amount of Zaurus knowledge I have got in the past 2 weeks, it seems that OE/OZ and the to-be-released Angstrom distribution is going the right way. Since I want to help the community I will soon have to choose which distribution to work with; pdaXrom is a non-starter since it is a closed development, and looks like it has stalled.
-
It's true, there is too much distro ... for instance i don't have make a choice because the all softwares i want are not in a distro feed.
I think people who buy a Zaurus just want to use it and choose a distro which they feel better with ( i'm talking about the design and the gui, qtopia like or X11 like).
But for now to make a choice, we must test the distro and see which distro offer which softwares ...
for example : OZ don't have KO/PI but PdaXrom and cacko have it.
So if you want OZ because you like his kernel improvement or something else, you must use another software. Or you must build it yourself, and most of Zaurus owner are not able to do that ( i think ... )
So, before thinking of splitting anything, why don't make a 2.6 standard kernel and a universal feed ?
perhaps it's simply impossible but i hope it's possible.
And then the users have just to choose the gui they prefers.
hope this help
ps : please stop child war, this is useless and it make the linux pda community not reliable
-
<snip>
for example : OZ don't have KO/PI but PdaXrom and cacko have it.
<snip>
Initially that's what I thought (and I was disappointed, as Ko/Pi is a very good PIM), but you can install it for OZ(OPIE) from the feed (can't recall the exact path as I can't access http://ewi546.ewi.utwente.nl/mirror/www.op...s.org/official/ (http://ewi546.ewi.utwente.nl/mirror/www.openzaurus.org/official/) for some reason, but it's under there).
This is the problem, when there are so many choices!
Anyways, I'm off now to watch England beat Sweden... ...
-
OZ/OPIE: ipkg install kopi == installed KO/PI
OZ had KO/PI before 3.5.1 was released ;)
-
oki that was not a good example ...
but i hope my idea was clear ( sorry i don't foudn another word to mean that i wish you understand me ... sorry for my english )
what do you think about working together and building a big feed for all the distro?
it is possible?
i think yes if there is a system like the debian depot ( don't think it's the right word ... )
-
baallrog: like we (OE/OZ devs) told many times:
OpenEmbedded can generate packages which will be usable also under pdaX. Other config allow to build packages compatible with SharpROM etc.
OE is powerfull thing, ptxdist (used by pdaX and who-knows-how-chuckr-will-call-his-distro) has some power too - someone who know both can work on integrating stuff from one in another. But I do not feel that one day all distros will unite - it will just does not happen.
-
I personally would also like to see WPA support built-in and multiple virtual consoles. It would also be nice if bvdd was safer because sometimes crashes hard on me. I've also had this bug where I have trouble connecting to any protected wireless network. I can rarely connect to WEP networks.
-
TJBK_TJB: OZ already support WPA (I use WPA-PSK at home). Multiple virtual consoles are easy to setup - about one minute in "vi /etc/inittab".
-
Busy, busy! - The sources are available and if anyone fixes anything i will personally make sure the fixes get implemented into the build - we are moving to SVN asap but until then.
Btw. the biggest problems arent in the kernel stuff which Sash is focused on - it so many "experts" are disliking the pdaXrom i can't believe why they just spend like 20-30 minutes looking at the preferences programs which are all written in python and make them better and maybe even in C. All the current python stuff were just quick workarounds until we had the time to code them in C - acutally Zumi is working on that but im sure he could use some help. http://mail.pdaxrom.org/soc/ (http://mail.pdaxrom.org/soc/)
Its just a matter of looking how the python scripts does stuff and then make that prettier and into C native apps.
-
Busy, busy! - The sources are available and if anyone fixes anything i will personally make sure the fixes get implemented into the build - we are moving to SVN asap but until then.
Btw. the biggest problems arent in the kernel stuff which Sash is focused on - it so many "experts" are disliking the pdaXrom i can't believe why they just spend like 20-30 minutes looking at the preferences programs which are all written in python and make them better and maybe even in C. All the current python stuff were just quick workarounds until we had the time to code them in C - acutally Zumi is working on that but im sure he could use some help. http://mail.pdaxrom.org/soc/ (http://mail.pdaxrom.org/soc/)
Its just a matter of looking how the python scripts does stuff and then make that prettier and into C native apps.
[div align=\"right\"][{POST_SNAPBACK}][/a][/div] (http://index.php?act=findpost&pid=132144\")
Thanks Laze for the info ... I, for one, am up to the invite. Checking out the links right now ...
EDIT: Read the link and ... how do I help? His email or contact?
This package is by him [a href=\"https://www.oesf.org/forums/index.php?showtopic=16649&st=25]https://www.oesf.org/forums/index.php?showtopic=16649&st=25[/url] so I suppose I should PM him?
EDIT: PMed him.
-
There's a solution for you who want to contribute!
The closed-ness of pdaXrom was a problem since I first saw it and started to like it. After some asking, Sashz started to put up sources of apps and build system with every release (and sometimes more often).
I don't think that Sashz kept the source tightly closed, so no one can contribute. They were up, he made some documentation to help developers getting started (Build your ROM (http://www.pdaxrom.org/index.php?showid=47&menuid=1)).
I made about 3 smaller contributions in the previous year and he accepted them all. Send diff files by email and comments, I don't think Sashz threw any of them away.
It wasn't really easy to contribute, that's for sure... make two copies of the source, edit one of them, make diff, email that. That's what I call hard.
So today I talked with Laze and we agreed that we need some versioning system. The idea is to setup an SVN server (objections, other ideas?) and give developer acc-s to you guys, who want to contribute. You can start a branch there, and Sashz or I will merge the things from it from time to time. So as the usual open source projects developments are going.
I hope you like this kind of open develop feeling (I know I do...) more.
What do you think?
Zumi
-
I think that is a great step!
Thanks
-
Is pdaXrom finally going to be openly developed? If so that is fantastic news for the Z community- I am glad sash and laze have seen the light. I never understood why it was closed in the first place- what was being gained? Its not like it was proprietary software they were making money out of. I suppose sash has more to brag about if its a one man show but he'll get more respect by co-operating and we'll all benefit from a faster developed, more stable OS- not that everyone didn't know this already.
-
Is pdaXrom finally going to be openly developed? If so that is fantastic news for the Z community- I am glad sash and laze have seen the light. I never understood why it was closed in the first place- what was being gained? Its not like it was proprietary software they were making money out of. I suppose sash has more to brag about if its a one man show but he'll get more respect by co-operating and we'll all benefit from a faster developed, more stable OS- not that everyone didn't know this already.
[div align=\"right\"][a href=\"index.php?act=findpost&pid=132237\"][{POST_SNAPBACK}][/a][/div]
OK, isn't good to talk in Sashz place but I think it was purely laziness based. He like to hack and develop pdaXrom in his own way. For example I think to move to 2.6 included a lot of hacking and it would take too much time to commit everything in a versioning system in the way, so it won't be broken from time to time. That requires a bit more strict developing fashion that doesn't really works with Sashz.
Other problem was that he doesn't always have and had good Internet connection (GPRS only at the moment, for example), so working with something like SVN would be hard too...
I don't know whether the sources on the SVN server will be the ones Sashz works with, but I will import his sources from time to time to the main trunk at least!
Zumi
-
Its really good to hear that a change is going to be made....
I'm just waiting to hear from Sashz...
-
Zumi: Right on the money - we never wanted to keep it closed but take these things into consideration:
1. Sash was working in Novosibirsk - he could max upload at 1-2 Kb/s - and it was very costly to do so.
2. All the compilation and development was/is done 100% by him - and most of the people who wanted to help often just turned out to be people wanting help with getting their hardware/sofware working.
3. He is a fulltime geek and spent all his time on his baby, pdaXrom, and he wanted to share but often to find that the stuff he shared was included in other people work without any credit. True that we also sometimes forgot to credit people but we tried to always keep it updated.
And now i need to go and spend time on somethings that actually give me money to pay the bills, which again keeps paying Sashs bill, and not spend countless hours on discussing with people a small group of people who always knows better. :-)
-
Thats very sad that sash has such a poor internet connection- he could well argue he's the greatest geek in the world- top 5 at least! We've got to consider people like RMS and Linus then. RMS is G E E K!
No really, I have the utmost respect for sashz work- I can only dream about having his coding skills! But still nothing you say justifies a closed devel model. pdaXrom would be developed faster that way- it cannot be denied.
-
And now i need to go and spend time on somethings that actually give me money to pay the bills, which again keeps paying Sashs bill, and not spend countless hours on discussing with people a small group of people who always knows better. :-)
[div align=\"right\"][a href=\"index.php?act=findpost&pid=132251\"][{POST_SNAPBACK}][/a][/div]
You really know how to butter people up don't you....
No really, I have the utmost respect for sashz work- I can only dream about having his coding skills! But still nothing you say justifies a closed devel model. pdaXrom would be developed faster that way- it cannot be denied.
[div align=\"right\"][a href=\"index.php?act=findpost&pid=132257\"][{POST_SNAPBACK}][/a][/div]
Yup, having an open development model would have reduced the amount of time he'd have had to spend fixing things and allowed him to concentrate on the things he's most keen on and delegated things to others. Plus things would have continued to progress in the hands of others on those occasions when time and internet connection eluded Sash.
-
Zumi, so as you mentioned, you will be ready timewise in 2 weeks time ... let us know when and I'll be up.
-
Is anything happening with this?
Just curious if anything is happening behind the scenes.