-
-
Notifications
You must be signed in to change notification settings - Fork 19.3k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Adds versioning for Configuration.h & Configuration_adv.h #3609
Adds versioning for Configuration.h & Configuration_adv.h #3609
Conversation
Actually if you try to use a 1.0.x configuration with Marlin now it will always bark. But this is a good possible additional measure. |
* the configuration files. | ||
*/ | ||
#define REQUIRED_CONFIGURATION_H_VERSION 1 | ||
#define REQUIRED_CONFIGURATION_ADV_H_VERSION 1 |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I suppose we could also use 010100 (i.e., version 1.1.0) etc. Two digits for each part of the version, still an integer that can be compared with < == >. A thought…
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
The problem I see is that we could have multiple config updates during a tag release.. I leave it up to you to decide which is best.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I think we should only put version numbers the configurations in real release versions. This might be version 1, and the configs with 1.1.1 might be version 2 if they change very much. But in between they should just be considered works-in-progress with options that come and go as we figure things out.
Consider this as a safe net under |
+1 from me! Luckily, I tested with one hand on the "off" switch of the printer... Demanding a minimum config version could avoid such issues. |
@thinkyhead one more use case. |
@tkurbad I'm surprised this didn't catch your error: /**
* Endstops
*/
#if DISABLED(USE_XMIN_PLUG) && DISABLED(USE_XMAX_PLUG) && !(ENABLED(Z_DUAL_ENDSTOPS) && Z2_USE_ENDSTOP >= _XMAX_ && Z2_USE_ENDSTOP <= _XMIN_)
#error You must enable USE_XMIN_PLUG or USE_XMAX_PLUG
#elif DISABLED(USE_YMIN_PLUG) && DISABLED(USE_YMAX_PLUG) && !(ENABLED(Z_DUAL_ENDSTOPS) && Z2_USE_ENDSTOP >= _YMAX_ && Z2_USE_ENDSTOP <= _YMIN_)
#error You must enable USE_YMIN_PLUG or USE_YMAX_PLUG
#elif DISABLED(USE_ZMIN_PLUG) && DISABLED(USE_ZMAX_PLUG) && !(ENABLED(Z_DUAL_ENDSTOPS) && Z2_USE_ENDSTOP >= _ZMAX_ && Z2_USE_ENDSTOP <= _ZMIN_)
#error You must enable USE_ZMIN_PLUG or USE_ZMAX_PLUG
#elif ENABLED(Z_DUAL_ENDSTOPS) && !Z2_USE_ENDSTOP
#error You must set Z2_USE_ENDSTOP with Z_DUAL_ENDSTOPS
#endif |
* will force a minimum config file revision, otherwise Marlin will not build. | ||
*/ | ||
#if ! defined(CONFIGURATION_H_VERSION) || CONFIGURATION_H_VERSION < REQUIRED_CONFIGURATION_H_VERSION | ||
#error You are using an old Configuration.h file, updated it before building Marlin. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
"update" here instead of "updated"
I'm still hesitant to incorporate this because I know we will get a huge raft of complaints about how hard it is to update configurations and how we don't provide any tools to make it easier. Users would prefer to just drop in their old configurations and have them work. Or, barring that, they want some helpful message to tell them what needs changing. Although this prevents the use of old configuration files (which is the safest route) it doesn't give any help with updating. We ought to compare the config options from 1.0.2-1 to 1.1.0 soon and list all the changes in the documentation someplace. Ultimately, people want to drop their old config files onto a web page and have it give them shiny new configs. We should look at that possibility too. @MagoKimbra has done some of this before. |
@thinkyhead your js configuration is working with all the conf options or does it need any updating ? |
@tinkyhead Well, what keeps us from checking the current config version A of the user vs. the one (B) needed for the state of Marlin he/she wants to compile? One could error out with all the changes necessary to transition from A to B. The only downside of this approach would be that we would have to handle unversioned configs (which may be from yesterday or from the stone age) somehow and find the necessary diff or just hint the user to read the documentation... Perhaps a combination of the current SanityCheck approach (for older configs) and versioned configs (from now on) would be feasible. |
@jbrazio I wouldn't use my configurator as a starting point for a simple converter. It's still in the experimental phase and needs updates for the reformatted configuration files. It was never an "expert system" but so far built to parse the configurations and put all the settings into a structured format, capture dependencies, etc. For an "upgrade tool" I would do things differently. First I would just scrape the configs from older versions of Marlin (1.0.2-1 for example) and capture all the configuration options in a raw form – just get their names basically, and then capture their values. (e.g., For commented-out options the value is "off.") Then I would scrape the newest configurations and get a full list of all new settings that have been added since the older config. Then from that we would make a spreadsheet with all the old and new settings and how they relate. For most they will just move over literally. But there will be a few needing code to rearrange or rename them. From that information we could build a drag-and-drop tool to convert old configs into the new form. |
Wouldn't something like repetier's online configuration generator be feasible for Marlin? I guess that would mean a lot of work, though... |
@tkurbad We have a great wealth of excellent ideas. |
Great @thinkyhead .... |
@thinkyhead ... as it is with all community driven open source projects I know. sigh |
One more for reference: #3573 |
@jbrazio I guess this needs a rebase, there have been so many changes lately. Whenever you find the time…. Maybe it can get merged soon…? |
@thinkyhead overall time this week has been short, will rebase Friday or so. |
9a2daaf
to
04bdec8
Compare
@thinkyhead rebased |
I guess you can see already… Failing Travis CI. |
04bdec8
to
8ccd97e
Compare
@thinkyhead It must be related with another thing, travis is green and I've forced travis to run again without issues. https://travis-ci.org/MarlinFirmware/Marlin/builds/136990199 |
Delete mysterious .swp files from #3609
…esio UI Support) ・Add implimantation of PR MarlinFirmware#3609 to configuration files of Cartesio ・Standardize macro names
Update current FW version
We try our best to include sanity checks for all the changes configuration directives because people have a tendency to use outdated config files with the bleding edge source code, but sometimes this is not enough. This PR will force a minimum config file revision, otherwise Marlin will not build.
PS: It also includes a much improved
generate_version_header_for_marlin
.