Shop OBEX P1 Docs P2 Docs Learn Events
New Parallax Website and new OBEX? Do it right, this time. — Parallax Forums

New Parallax Website and new OBEX? Do it right, this time.

msrobotsmsrobots Posts: 3,709
edited 2012-06-06 10:23 in General Discussion
Hi All,

in some thread a while ago I stumbled over someone asking why BUMP is so silent and someone else mentioned there will be a new WEBsite and (possible?) a different OBEX ?

Since finding things in the OBEX is somehow painfull, I wonder what is planned there and if there are any plans to integrate Objects via http-link in the source files, allowing repositries/librarys like in PERL or putting objects into some CODEPLEX/SOURCEFORGE-clone to view files before download for example?

the OBEX is one very good thing for Propeller. But we need more selections P1,P2 spin,forth ,basic or C, homespun/bst required and probably more.

webservice-interface? to use in dev-programs?
json interface to use from Prop?

this should be a opportunity for Parallax to get things right, after a while as usual, but putting all the brains together here we might be able to IMPROVE the OBEX to a mighty Tool. (even ZOG would approve this, feeling lonely as the mightiest one...)

Integrating the OBEX more could produce something like a 'cloud' of services usable with dev-tools and propeller-projecs, given spinneret and P2 doing ethernet.
(and most of the dev-computer anyways)

Hell, lurking here in those foras for a while I am confident that we all together can do WAY better than any PIC/ARM/Whatever other company is doing.

Because it is Parallax, and CHIP and shoot you all know them. They are used to do things different as anybody else when they are convinced that it is the better way.

And so am I. Lets show the world what makes Parallax and our community special.

So lets collect some thoughts and give some (unwanted) advice.

Enjoy!

Mike

Comments

  • Heater.Heater. Posts: 21,230
    edited 2012-06-01 23:24
    I can't help but think that the OBEX should be for Spin objects only rather than cluttering it up with C, Forth and whatever other code.

    It could perhaps do with a slicker way to find things, a versioning system, maybe things like the possibilty to include objects directly from the web into your projects etc etc. At the end of the day a tool for Spin programmers.

    C/C++ guys might be happier with a more regular repository like sorceforge and such.

    Forth guys will want something totally backward:)

    Zog is not so sad, still snoozing in his iceburg.
  • xanaduxanadu Posts: 3,347
    edited 2012-06-02 00:59
    For me the OBEX is like going to a grocery store to get chips and guacamole. The avocados and tortilla chips are many isles apart. Between lies an adventure.
  • LoopyBytelooseLoopyByteloose Posts: 12,537
    edited 2012-06-02 01:50
    My own experiences with OBEX are rather embarrassing. I had expected it to make learning easy and I became quite frustrated with trying to work into it as a whole. All was made worse by it being a period of rapid expansion and code was being changed or added daily.

    But after complaining and making myself upset, I found that the SPIN objects included with the original Propeller Tool are extensive and the most helpful for the learner. So I wasted quite a bit of time thinking that OBEX was blocking my learning.

    If you are a new learner - don't be too greedy. Work with one recent version of an IDE and the examples it offers until you have good reason to move on.

    Linux has several version revision applications that Parallax can use. Some are rather huge, too ambious, and unweidly; but others are small and work quite well. Since many of these have been ported over to Windows, I see no reason to reinvent the wheel and create a new one.

    About the only thing I would want in OBEX at this point is to have it clearly mention that the Propeller Tool comes with a complete set of learner materials. It would help with BST if it also mentioned that these can be migrated over for BST.

    Are far as C, Forth, and whatever - please don't mix everything together. What is to be gained by that.
  • Dave HeinDave Hein Posts: 6,347
    edited 2012-06-02 04:57
    I think the OBEX is fine the way it is. The only thing I would suggest is that people document their objects a bit better. A readme file that just contains an object map is mostly useless. It would be much better to put a description of the object in the readme file. I understand the idea of putting all the documentation in the source file, which I suppose is OK. However, it would be nice if the readme file also contained the same information, even if it is just the comments extracted from the source files.

    Also, all objects should have a demo program so that programmers can see how the object is used. A lot of applications are just modifications of the original demo program.

    EDIT: One other thing. When possible, avoid the use of the special Parallax characters. I prefer working with 8-bit ASCII files than the 16-bit files that contain the Parallax font.
  • prof_brainoprof_braino Posts: 4,313
    edited 2012-06-02 08:24
    It seems the original post is about FINDING STUFF.

    Specifically, getting the right information from the person that has it to the person that needs it in a timely manner.

    I can never find the right stuff the first time around here. The search doesn't work, the sort doesn't work, the stuff is not organized in a way suitable for a person looking for it. Like many places, stuff is arranged for the folks that already know where it is and generally are not the ones that need it.

    If the OBEX is limited to spin objects, fine. The real need is that we need to find those spin objects when we need them, and we need to find all the other stuff we need without relying on luck. Saying "you just have to learn where stuff is from experience" is that whole silos of knowledge thing that everybody is trying to fix, and is a poor cop-out.

    I believe the request is to address this. [/rant]
  • msrobotsmsrobots Posts: 3,709
    edited 2012-06-02 16:22
    @heater, versioning is a good example. But - i would really like to find ZOG there or CP/M or ... all those things in the foras. If you don't know you will never find them. Maybe its just me not beeing able to find things. Old threads vanish. So what I am asking for is some combination of forum and obex to better publish/find already existing information buried deep down there. Until now I am still ZOGless and without CP/M. :frown:

    @xanadu - yes. Looking for one thing finding another.

    @Loopy. Hm. OK I try not to be greedy. But I feel I am constantly reinventing stuff already there. Pullmolls VT100 for example. I do not complain about existing documentation needed for programming. All there and very,very good compaired to other Micros. I love the Propeller and am able to do most of what I need. It is just so that Parallax is redoing their website right now and IF we want to have some feature it is the perfect time to ask for.

    @Dave. Yes. different versions of allmost the same object - why to choose what version? Grouping them somehow ? maybe links to forum-threads? By now I download tons of them, read thru the source to find documentation - if any. Now I have some form of local copy. And use (just for example) an older version of KYEs SD-driver. Putting my stuff as Archiv into the obex and there we are.

    @prof_braino - exactly. And not just finding, keeping. things up to date like finding out about the new 4portserial or such.

    @all - I am just brainstorming here. I agree mixing Spin, C, Forth, etc. might not be a good move.

    Enjoy!

    Mike
  • SRLMSRLM Posts: 5,045
    edited 2012-06-02 22:59
    I think the most important change needed is to have a method of recommending the "best" drivers, along with a way of updating them. The over abundance of serial objects is a prime example of this. There are too many objects, both here and in the obex, that do the same thing. One is better than all the others, but how do we find it? And how can the community contribute to making it better?

    They can't, and that's why the Obex needs significant improvement.
  • jrjr.jrjr. Posts: 45
    edited 2012-06-03 20:42
    I agree that the OBEX is needing a rebuild, but I would hope that the end function would define it's location for all flavors for programs.

    It would make best sense to have a decent FILTERING system would be the key to segregating the desired queries.

    The current OBEX has the format to filter results if the attributes column was extended.

    My reasoning for this is that even if you are looking for a spin routine, it may be just as usable to find
    a PASM or C routine to study or reference during it's implementation.

    The filtering serves the purpose of limiting choices without scattering the database by program flavor at the expense of program function.

    We use the OBEX for both direct use as well as reference, it would be a shame to break that functionality.

    As to the best drivers, an OBEX implementation forum category might be a good choice here as well as a
    moderated voting method that reflects user confidence in implementing the routines.

    jack
  • Heater.Heater. Posts: 21,230
    edited 2012-06-04 05:29
    msrobots,

    ...i would really like to find ZOG there or CP/M or ... ... I am still ZOGless and without CP/M.

    That's terrible, everyone should see what CP/M can do on the Prop and everyone should have a giant Zog to guard over them:)

    Actually you might be the first person to request that these go into OBEX. I did think about it a long while ago but as they are more projects rather than general purpose objects I did not pursue the idea. Besides they have never really arrived at being finished:)


  • mindrobotsmindrobots Posts: 6,506
    edited 2012-06-04 06:20
    Heater. wrote: »

    ...they have never really arrived at being finished:)



    I tend to think of projects as a journey and not a destination. It makes stress and performance issues more bearable (this comment does not reflect the opinions of management.)

    msrobots,

    There was this discussion of rewriting/reinventing the OBEX from September of 2011: OBEX-for-non-MIT-licensed code Hopefully, some of these suggestions were taken to the dark recesses of Rocklin to inspire the Web Wranglers.
  • Heater.Heater. Posts: 21,230
    edited 2012-06-04 06:45
    msrobots,
    I tend to think of projects as a journey and not a destination.

    I love it. I'll suggest that to my boss at our next progress meeting:)
  • msrobotsmsrobots Posts: 3,709
    edited 2012-06-05 21:18
    Heater - yes it is terrible ...

    but you still have time to finish! I am reading myself thru ZiCogs threads an am still in 2009 so way to go !

    I already decided I need this to remind me WHY I like c# - any COBOL out there for CPM ? this Propeller stuff is so much fun.

    reading the other thread mindrobots mentioned.
    google-code or clones provide bug-reports and revisions but are very project-orientated by design. they also easy allow to work toger like duane design and tracy allen did on 4portserial.(ex tim moore ones).
    Problems with MIT. hm. Is that really important for 2K of PASM ? I am not a open Source guy but objects for microcontoller?

    Anyhow - as far as I can see consens would be at least

    1. better search and filtering
    2. viewing of source before download
    3. some sort of rating with explanation why better
    4.?

    Enjoy!

    Mike
  • doggiedocdoggiedoc Posts: 2,245
    edited 2012-06-06 10:23
    Heater. wrote: »
    ... I'll suggest that to my boss at our next progress meeting:)
    Heater - I don't know your boss, but I'm guessing that wouldn't go over so well. :D

    Paul
Sign In or Register to comment.