Shop OBEX P1 Docs P2 Docs Learn Events
Parallax github repository! — Parallax Forums

Parallax github repository!

Comments

  • jac_goudsmitjac_goudsmit Posts: 418
    edited 2014-11-07 18:07
    Yes! Best news of the day!

    P1V is not in there yet but hopefully that's only a matter of time.

    ===Jac
  • jmgjmg Posts: 15,173
    edited 2014-11-07 20:23
    Propeller 2 Verilog

    is quietly on the list ?!
  • Duane DegnDuane Degn Posts: 10,588
    edited 2014-11-07 20:32
    Thanks Seairth.

    I've had a github account for a while but I have yet to use it. Time to watch Jessica's videos.
  • Heater.Heater. Posts: 21,230
    edited 2014-11-07 23:39
    jmg,
    Propeller 2 Verilog....is quietly on the list ?!
    Yes it is.

    There is the idea that Parallax take ownership of the P8X32A_Emulation repository that I put up here: https://github.com/ZiCog/P8X32A_Emulation

    That is still Chips original code with only white space fixes, TABs replaced with the correct spaces to make it look as Chip wrote it no matter what editor/viewer one has. Like on the github web pages themselves.

    Plus the addition of a readme file and some documentation.

    That repo has already been forked a few times and has a pull request on it so it would still be a good starting point for Parallax to take it over.

    Chip has been too busy to act on this idea but following this announcement now would be a good idea.

    Or perhaps Chip and Ken have decided against it and are planning their own thing.
  • Cluso99Cluso99 Posts: 18,069
    edited 2014-11-08 13:18
    This is fantastic news!
    Guess I am going to have to learn github - up till now I have resisted - "resistance if futile"
  • jac_goudsmitjac_goudsmit Posts: 418
    edited 2014-11-09 00:52
    Heater. wrote: »
    That repo has already been forked a few times and has a pull request on it so it would still be a good starting point for Parallax to take it over.

    Chip has been too busy to act on this idea but following this announcement now would be a good idea.

    Or perhaps Chip and Ken have decided against it and are planning their own thing.

    From what I understand in Ken's message, Jeff Martin is putting the ParallaxInc Github organization together. So I think you misunderstand about Ken and Chip changing their minds about taking over your repo. I'm pretty sure they're well aware of its existance, and Jeff is in the process of putting things together, and I expect he will contact you soon about taking over the repo.

    https://github.com/parallaxinc is an organization, not a user ID; that's why the web page looks different from your own page and other people's pages. An organization on Github allows many people to work together on multiple projects. For example, earlier this year, the Heatsynclabs hacker space here in Mesa AZ organized the HackPHX event, and some people from Github.com were there to help organize the source code that the teams brought together, so they set up an Organization too, and made everyone part of it. That made it possible for everyone in the teams to check in code without having to deal with branching and pull requests. People outside an organization can still use the usual Fork model. Setting ParallaxInc up as an Organization is a great way to bring all the Parallax related projects together; it gives a sense of authority and approval of Github as the Place To Go for these projects, and still allows anyone to join in even in the unlikely event that others (even Parallax) might some day lose interest, or worse. And this is exactly what we would want from Parallax. Awesome!

    ===Jac
  • pik33pik33 Posts: 2,366
    edited 2014-11-09 01:11
    Cluso99 wrote: »
    This is fantastic news!
    Guess I am going to have to learn github - up till now I have resisted - "resistance if futile"

    I am going to learn github.... up till now I have resisted....
  • Heater.Heater. Posts: 21,230
    edited 2014-11-09 01:33
    jac,

    Yes, this all looks very good. It's all happened since Chip contacted me last so I started to wonder what might happen next.
  • jac_goudsmitjac_goudsmit Posts: 418
    edited 2015-01-03 23:42
    It looks like the Propeller Verilog files have finally shown up in the parallaxinc github area at https://github.com/parallaxinc/Propeller_1_Design, but unfortunately it's a new repository, not a fork of Heater's repo.

    I wonder if this "reboot" was intentional or if Jeff Martin (or someone else who's working on Github for Parallax) wasn't aware that Heater's repo exists... It looks like many commits were done a long time ago.

    I admit some bad decisions may have been made during our development but I don't know if I want to start over. It looks like the files in the Parallax repo aren't even de-tabbed yet.

    ===Jac
  • David BetzDavid Betz Posts: 14,516
    edited 2015-01-04 05:45
    P1 is Chip's design and I'd like to see the Parallax git repository reflect only changes that have been accepted by Chip as something that would likely be part of new P1 silicon should Parallax ever decide to fabricate a new chip. It seems like extensions done by forum participants should be in separate repositories and pulled into the official Parallax repository only when reviewed and accepted by Chip or someone else to whom he grants "push" privileges. Of course, I'd like to see the tab-to-space conversion adopted by Parallax but that is Chip's decision.
  • Heater.Heater. Posts: 21,230
    edited 2015-01-04 06:25
    Not sure what happened there. Last time Chip contacted me he was getting all set to take ownership of my Propeller emulation repository. Never did hear back. Oh well.

    It's a shame because that repo has been forked and used by others already and it would have been good to pull that along. Now that is all lost and people will have to do all that again.

    David,

    For sure all changes should be acceptable to Chip. Or someone he really trusts I guess.

    Contributions by others would come from other repos. That is how git works, you fork the upstream repo into your own new repo, make changes, then you can request that the upstream maintainers pull the changes. Which they may or may not do of course.

    My changes were strictly limited to de-tabbing, which Chip seemed OK with, and adding some docs.

    I guess I'll have to fork the new repo and try to get the de-tabbing pulled in.



    Who is actually running that repo? The changes are already months old and as far as I know Chip has had nothing to do with git in that time.
  • David BetzDavid Betz Posts: 14,516
    edited 2015-01-04 06:36
    Heater. wrote: »
    Not sure what happened there. Last time Chip contacted me he was getting all set to take ownership of my Propeller emulation repository. Never did hear back. Oh well.

    It's a shame because that repo has been forked and used by others already and it would have been good to pull that along. Now that is all lost and people will have to do all that again.

    David,

    For sure all changes should be acceptable to Chip. Or someone he really trusts I guess.

    Contributions by others would come from other repos. That is how git works, you fork the upstream repo into your own new repo, make changes, then you can request that the upstream maintainers pull the changes. Which they may or may not do of course.

    My changes were strictly limited to de-tabbing, which Chip seemed OK with, and adding some docs.

    I guess I'll have to fork the new repo and try to get the de-tabbing pulled in.



    Who is actually running that repo? The changes are already months old and as far as I know Chip has had nothing to do with git in that time.
    I have no idea who is running the P1V repository. It is probably just a snapshot of what Parallax released at the start. Couldn't they still merge in your changes? If all you've done is detab and add some documentation that doesn't seem like it would require any special blessing from Chip except maybe to verify that the docs are correct.
  • Heater.Heater. Posts: 21,230
    edited 2015-01-04 07:06
    Yes, they can still merge the changes. But first I have to fork their new repo, copy my changed files into my fork and commit them, then issue a pull request. Perhaps raise an issue on their repo to go with all that.

    I'm too tired this evening to think about any of that and tomorrow is a busy day. We'll see.
  • jac_goudsmitjac_goudsmit Posts: 418
    edited 2015-01-04 07:57
    David Betz wrote: »
    P1 is Chip's design and I'd like to see the Parallax git repository reflect only changes that have been accepted by Chip as something that would likely be part of new P1 silicon should Parallax ever decide to fabricate a new chip. It seems like extensions done by forum participants should be in separate repositories and pulled into the official Parallax repository only when reviewed and accepted by Chip or someone else to whom he grants "push" privileges. Of course, I'd like to see the tab-to-space conversion adopted by Parallax but that is Chip's decision.

    I agree that the Parallax repo should contain the code that replicates the P1 as accurately as possible. Anyone who wants to start with the code as Chip intended, should be able to just clone/fork the master branch from the Parallax repo.

    But I also think that there are a few minor but important problems that should be fixed in the Parallax repo:
    • All known bug fixes should be implemented; I haven't verified yet if that is the case. Also I don't know if Chip has confirmed each of the bugs that were found in this forum and I don't know if that's necessary. I expect that fixes for real bugs will be accepted by Parallax as pull requests, by Chip or by someone who's in control of the repo.
    • Tabs should be converted to spaces in all text files, since there is always some disagreement and therefore ambiguity on how many spaces are represented by a tab. With computer memory sizes and storage sizes of the 21st century they should have no reason to exist.
    • The three directories should be combined so that improvements don't have to be manually merged between targets. Alternatively, there could be three branches instead of directories to make it easier to merge changes between targets. Git is good at merging between branches but merging changes between directories in the same branch is basically not supported.

    Furthermore, I think it would be a good idea to convert the AHDL files to Verilog so that beginners only have to learn one language. And apparently it also helps with the Xilinx port to have all files in Verilog.

    I guess the bottom line is that the Parallax repo is the "official" one, making all other ones "derivatives". With Parallax' excellent reputation of working with users, it makes me think that they didn't intend to discard all our work and basically make it irrelevant. I probably wouldn't mind starting over; my work so far has been mostly reorganizing the existing code to make it easier for others to add features. I know there are various problems in addition to the ones mentioned above and it would be a good opportunity to do things right this time. But others have invested a lot of time in their own forks of the various branches produced by Heater, Mindrobots, myself and others, and may not be so willing to start over.

    So, before I do a lot of work on the new Parallax repo, I'd like to know what their intentions are. What's the reason that there is now a second repo for the same code, with apparently more authoritative-ness because of the Parallax name? Was it a mistake and was Jeff (or whoever) unaware of Heater's depot, or was the creation of the new repo intentional? And if it was a mistake, is Parallax going to still take over Heater's repo, as he intended?

    ===Jac
  • Heater.Heater. Posts: 21,230
    edited 2015-01-04 08:29
    jac,

    I agree with all your bullet points.

    Looking at it I think it's probably too late to revert this and use my repo. My version is basically the initial tarball from Chip de-tabbed. What I see on the Parallax repo has some changes already. At least support for the BeMicroCV micro has been added.

    I think I'll try and find time to fork the Parallax repo and de-tab it again. Then make an issue and a pull request for that.

    Note: For all those who fuss about tabs and indent levels, please browse the verilog code in the Parallax github. It displays as a mess. All I want is to have it display nicely, that is as Chip saw it when he wrote it.

    As another bullet point, I don't like the new repo name "Propeller_1_Design". The word "design" is at least redundant. It's also not true. This is not what the Pi was built from !

    I called mine "P8X32A_Emulation" in recognition of that but I was not overly happy about that name either.
  • David BetzDavid Betz Posts: 14,516
    edited 2015-01-04 08:55
    Jac's bullet points all make sense to me. I had forgotten the duplicate files. It would be good to have all of the targets build from the same core set of files.
  • Duane DegnDuane Degn Posts: 10,588
    edited 2015-01-04 12:20
    When Ken proposed the Open Propeller Project #8 (Porting the Eddie frimware to the Activity Board), one of the things he said was:
    Ken Gracey wrote: »
    Parallax GitHub

    The source code repository for this project will be in the Parallax GitHub. Parallax is in the process (as of November 1, 2014) of moving all of our community open source code development efforts to this GitHub. All code will be open source under MIT license (copy and paste the license from other code into this one). The original Eddie code didn’t have the MIT license because it was created for Microsoft.

    We will create the GitHub repository for these files on Tuesday this week.

    I'm very new to using GitHub myself but I don't see any sign of the Eddie firmware on the Parallax repository. I have added my updates to my personal GitHub account.

    Am I correct in assuming I need some sort of permission to add projects to the Parallax GitHub account and the proper course of action is to wait for Parallax to "Fork" or "Pull" or whatever one does to copy code from one repository to another.

    BTW, Heater, you were right. GitHub does make sense even for a single developer. I like to think I had a pretty good version control system going but now that I'm getting used to GitHub, I can certainly see an advantage to using it.
  • Heater.Heater. Posts: 21,230
    edited 2015-01-04 12:48
    Duane,

    What is the license on the Eddie code? Ken's statement that is was created for MS makes me worry you are not even allowed to redistribute.

    I would image that only Parallax people will ever be able to add projects to their git hub account.

    That is "account" though, which will have many git repositories within it. Potentially they might invite a trusted few to collaborate on those repos and be given access.

    Generally though I would expect perhaps only one person at Parallax to be responsible for each repo and make use of the usual fork, update, pull request of the git work flow.

    Glad to here you are getting on with git and github. It's worth it all just for the peace of mind of not having to worry about backups all the time. Or which machine happens to have the correct versions of dev code on it. The freedom to just hack on code as an experiment and not have to worry about trampling over the last know working version. And so on...
  • Duane DegnDuane Degn Posts: 10,588
    edited 2015-01-04 13:15
    Heater. wrote: »
    Duane,

    What is the license on the Eddie code? Ken's statement that is was created for MS makes me worry you are not even allowed to redistribute.

    But he also provides the following instructions to whomever is modifying the Eddie code :
    Ken Gracey wrote: »
    All code will be open source under MIT license (copy and paste the license from other code into this one).

    I think the statement:
    Ken Gracey wrote: »
    The original Eddie code didn’t have the MIT license because it was created for Microsoft.

    Is to let us know why a MIT license wasn't already in the code.
Sign In or Register to comment.