Jump to content



  • Content Count

  • Joined

  • Last visited

Community Reputation

0 Обычный

About Submin

  • Rank

Recent Profile Visitors

The recent visitors block is disabled and is not being shown to other users.

  1. @dxp Submodules are dictating the file tree and therefore is not an option.
  2. I'm sorry @Raven I used all the words I know. The scheme with dependencies requires tuning and precision order of actions or, as I named it, tracking - that what such tool is for. Not GUI nor CLI but tracking. As the example consider keeping the commit SHA of nested repo within the upper layer repo as its part. Is it possible do manually - yes for sure. Would it convenient to do so - obviously not. Will such solution error prune - no doubt. If my explanation is insufficient that I'm sorry again. Thank you for trying to help.
  3. @Raven I'm not sure that we're on the same page. I have an independent repo. GIT takes care about it. There is another independent repo which depends on the first one. So in the second I can use a certain commit. Because both repos are independent, I can clone the first one wherever I need, modify it and commit. Then in the second all I need to do is to change the commit. If the first repo was updated outside my project, I can test it with the latest commit or may ignore it. Is it clear? If this idea is understood and seems to be matching my requirements, it there any tool that can be configured for such independent repo dependencies and keep track of of the changes?
  4. @Raven, Well, I don't think that anything but GIT and Hg are supporting submodules/subrepos. GIT submodule is not acceptable because it requires it's on file tree(does not meet the requirements) and some other people on the project dislike Hg... I am not aware of any other distributed SVC. Does anyone know some kind of a system which can manage INDEPENDENT GIT repos, tracked and managed by this system externally in relation to GIT?
  5. @Raven, I'd appreciate if you elaborate your idea in more details - I'm not sure that understood it. As I mentioned in the very first post, I'm not strong with SVC and GIT as a tool. What I do know, that GIT indeed will create the folder structures as you said but this is NOT suitable for the project. If it is possible to fool git - then it is great, doing the same with the actual project is unacceptable: tool is used by the project not the inverse.
  6. Take it as simple as just they can't. At least one reason: GIT sub-modules are creating/demanding their one file tree structure that is NOT suitable for the project. Using independent sub-repos does not require such a hassle. If you have something to offer then please do if not then thank you for trying to help.
  7. This is an example of dependencies. For example Repo1 depends on Repos 3,5,6 and so on...
  8. Submodule the way it is implemented in GIT does not work. Distributed repos expected to do the job. Is there any tool to manage multiple and nested repos?
  9. So after long internal discussions we're going to try using a project as a main repo and clone other independent repos to whatever location/folders tree structure required by this particular project. The challenge will be how to track all those folders and repos(or actually subrepos in relation to the main project) and their relations. Is my explanation clear? Do you have any suggestions and/or tools which can help? Thanks.
  10. I don't have a Russian keyboard so I will use English. Fill free to reply in Russian. We were using Hg for a while for our small team version control. The order is to migrate to something else. It is not really important is it free or commercial. What is important to have a distributed cloud-based platform with the ability to support sub-repositories. If you want to say GIT is the one then I don't know how to apply it correctly. The reason is that what is called submodules is being cloned within the project subfolder and thus creates a bunch of copies of the master repo. When you have quite a few projects on the same PC managing all of them becomes a nightmare at least for me. I'm not a Git guru so maybe it is my problem, not GITs. I do recall that about 10 years ago GIT could do the trick of having a single repo linked to a bunch of projects but it has changed a lot since. Could you guys please advise what should we look at or what is the correct technology to use for that workflow? Here is a very brief example: Let's say there is some HAL. Then there are several example projects showing how to use apply this HAL. Of course, with the time HAL is changing but the requirement still has old example projects linked to the old version of HAL, not a copy but exactly the reference to the repo. As I mentioned, I'm not really qualified to discuss Version control tools and asking you to excuse me and or I can serve as a translator between you and our team member responsible for the VC system. Thank you in advance.
  11. To be short: I have the working code but trying to better understand the quite ambiguous documentation. For that purpose, I intentionally set up the argument for ACMD41 command for the voltage range below acceptable for my SDHD card. In my case, it was 2.0 - 2.1v. What happened next is the subject of my question: After that action, I'm not able to send ANY command to the card. No matter what I'm doing, I'm getting the CMD time out. I tried to remove the card, turn the power off/on - nothing helps. But if I plug the same card into PC, which finds it, does something, what I'd like to know, and then if I put it back, it works as expected. Of course, if I do not send erroneous command. What initialization command I'm missing. My guess is that somehow I need to reset the OCR. The sequence of commands I'm applying is Go Idle State(0) and Send If Condition(8), ACMD41... and so on. And I can basically, lock the card for my code, as described above. Thanks in advance. PS: I can easily read Russian but can't type in Russian from my office. Sorry about that.
  12. We have 2 separate projects for CY8PROTO-063 BLE: one is for central another is for the peripheral device. The challenge is to understand, document and merge those two projects into one which will be convertible on a fly(for example via GPIO) to either Central or peripheral. Another challenge that those two are hardcoded to be connected. We need to be able to configure such bonding/pairing at will. If you are capable of performing such work, please contact me for further conversation. Thanks.
  13. The price makes this part out of consideration.
  14. I never said that we're going to use STEVAL-SPIN3201 as is. Our goal is to find the alternative to the existing and working on our product TI C2K solution. One of the reasons is to use Cortex-M MCU. The first step is to EVALUATE the FOC software, its performance, the process of tuning it to a custom motor and make a decision would it worth to try redesigning our TI based motor-controller. Unfortunately, I wasn't given the permission to publish the picture but the working board dimensions are 60*40*16mm