[BUG] WIFI ESP3D file sd file upload MIssing #393
Reference in New Issue
Block a user
No description provided.
Delete Branch "%!s()"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
Upload file from web (ESP3D)
Expected behavior: file uploaded on the SD
Actual behavior: it's not possible to load files with extension .gco if I use .gco it's not visible on display of the printer
Same as https://github.com/knutwurst/Marlin-2-0-x-Anycubic-i3-MEGA-S/issues/330 but also with 1.5 beta firmware the problem persist...
Can someone tell me how can i download and use the Marlin bug fix updates
P.S. I have this display:
Anycubic 1.0 display
Bye
Stefano
I have added this fix in the configuration.h but i have this error:
https://github.com/MarlinFirmware/Marlin/pull/24752
#error "The ANYCUBIC LCD requires LCD_SERIAL_PORT to be defined."
The referenced pull request has nothing to do with the Knutwurst firmware at all. That's kind of a backport Marlin's default LCD implementation. And it only enables display of special menu entries and does not affect file loading.
The LCD firmware also filters entries by extension. Unfortunately this filter cannot be modified by the Marlin firmware.
By default ".gcode" and on Mega P display also ".bmp" are displayed. ".GCO" is just the shorthand for the legacy 8.3 names, so "longfilename.gcode" is represented as "LONGFI~1.GCO". Some displays do not filter, but at least the DGUS clone on the Mega P does and it does not show ".gco".
@stklcode,
Thanks a lot for your reply.
I know that the pull request is from another project but it's Marlin so i think share some stuff.
From what i remember with old firmware 1.3.1 i can see and print .gco with this new fimrware the printer won't see .gco anymore.
No, the code is 100% irrelevant for this project.
The Knutwurwt Firmware has its own Anycubic LCD implementation. The Marlin PR (which was made by myself, so let’s assume I know what it does) ports the Knutwurst behavior back to the Marlin core, so the Marlin implementation supports the Mega P display.
And, as already said, the change does not affect the files that are displayed. It’s only about the special menu pseudo-elements.
Which printer or which display are you using? I cannot reproduce it myself, as my display has a built-in filter and never supported .gco files.
Is there any reason why you want to use .gco files and not .gcode ? Pretty uncommon I’d say…
Only filter for SD card file display is here (this is unmodified Marlin code):
ac7c272403/Marlin/src/sd/cardreader.cpp (L224-L229)This has slightly changed since 1.3.1, but the relevant part is still the same:
21d062c285/Marlin/src/sd/cardreader.cpp (L173-L177)So from that I’d say the .gco extension should be fine from the Marlin side…
Additional question:
Are the files actually copied to the SD card and just not displayed on the printer or are they not copied at all?
Thanks a lot again for your reply!!
sorry for my delay… very very busy days at work… :(
The files are copied to SD card from Wi-Fi by the ESP01-S but for this I must use the 8.3 format…
I have also copied file from my PC to my SD card directly from PC in 8.3 format but unfortunately the same problem happen… The display of my printer can’t see the files in .gco format… :(
The problem is that .gco file il filtered by the display.
If you have a long file name you should see it because the extension is override by:
So I fix it doing more ore less the same thing
`
#if ENABLED(KNUTWURST_DGUS2_TFT)
if (fileNameLen >= MAX_PRINTABLE_FILENAME_LEN) {
fileNameWasCut = true;
fileNameLen = MAX_PRINTABLE_FILENAME_LEN;
}
`
Wouldn’t it be simpler to copy the filename to a local variable first place, append “de“ if it ends with “.gco“ and keep the sanitization routine in place as it is? (yes, that comes with some memory overhead)
The solution is not perfect though, think of two files named “test.gco“ and “test.gcode“. But maybe still better than not displaying the files. Same can be applied to
<longname>123~.gcodeand<longname>1234.gcodeyet, because of the truncation and apparently we can live with that fuzziness.I personally don’t get the problem at all, is there any reason beside “i want to“ to not just use the supported .gcode extension? Any software that can only generate .gco or something?
Edit:
I see the issue with filename length 0.
Yes I can do and instead of append "de" its best to ovveride alse GCO just to avoid to have something like: "filename.GCOde"
yes, if you use ESP3D the file must be 8.3 format (i don't know if it's the same on last version)
It's possible to write better. I just test if it works because there was also an error in the original code:
this will never happen because
for (unsigned char i = 0; i <= fileNameLen; i++)Anyway it's possible to write it better and I think its better to see the files with manipolated name instead of wite line on the display.
OK, wasn't aware of that restriction, never used that combination.
Could you check whether the issue is still present with the latest master (since #398 has been merged now)? My Mega P does show a "beep.gco" file again, but I cannot test on any other printer (and don't know which pinter(s) you are actually using)
Sorry I take a look on:
So it's possible to see .GCO file on the screen
I will try to just use the correct fileName. If card.longFilename size == 0 just use card.filename.
It will not be necessary to manipolate it becasuse if it's long it works also with .GCO
I will try it later
The same idea is used in
card.longest_filename(), likely for the same reason (first use was some Ultra LCD implementation):8877ad534d/Marlin/src/sd/cardreader.h (L146)So we now use this display filename to start off, too.
Thanks.
I can not test now but I think theris something to fix
And last remove this code :
```
if (i >= fileNameLen) {
outputString[i] = ' ';
}
No idea what you're talking about, sorry....
In the last commit I see this:
` // THe longname may not be filed, so we use the built-in fallback here.
char* fileName = card.longest_filename();
int fileNameLen = strlen(fileName);
bool fileNameWasCut = false;
at line 1494 you have this:
Then at line 1203 you change the fileNameLen
so this
should be
ORIGINAL_FILE_SIZE is > fileNameLen
Otherwise i will never be >= fileNameLen because the loop is until <=fileNameLen
if (i >= fileNameLen) {
outputString[i] = ' ';
}
I will write the code that is best
See the code.
I hope it's clear if i'm not wrong
`
// The longname may not be filed, so we use the built-in fallback here.
char* fileName = card.longest_filename();
int fileNameLen = strlen(fileName);
bool fileNameWasCut = false;
`
Please stop posting full quotes of the code. It's really hard to get the single change from these answers.
That can't be 100% correct either.
We initialize:
Using the proposed loop:
We might overfill the character array.
Imagine
fileName = "1234567890". So we havefileNameLength = 10.We not truncate the name, so we have
fileNameLength = 5and initialize the output as an array of 5 characters:
outputString[5]The loop then iterates over the array and fills
outputString[0] = '1'outputString[1] = '2'outputString[4] = '5'outputString[5] = ' 'from here on we are out of the array bounds!outputString[9] = ' 'PS: I built a small demo here: https://www.sololearn.com/compiler-playground/c5a0mbpxc6xV
With NULL-Termination the output is exactly the same, because we terminate the string after
fileNameLenanyway, so there is no need to fill anything up with spaces here. However, we must not write data out of the array bounds, to not corrupt whatever is stored in the memory beyond the array memory.Ok
That's true but i will never be >= fileNameLen
The example works perfect but there's unecessary code
Agree. The minimal loop without unnecessary if-clauses should be correct and sufficient here.
Also
i <= fileNameLenshould bei < fileNameLen, because we setoutputString[fileNameLen] = '\0';after the loop, no need to copy the original NULL-Byte.That is what I meaning from the begining. Sorry if was not clear
I change the for cycle and test it on anycubic mega x. It works
Hello,
thanks a lot to everyone for the help, unfortunately i'm really busy at work this days, as soon as i can i will test the firmware in my printer :-D
Best Regards
Stefano
Hello to everyone!!!
I have updated the fimrware and work perfectly again now!!!
Thanks a lot!!!
Stefano
This issue has been automatically locked since there has not been any recent activity after it was closed. Please open a new issue for related bugs.