submitqc4 bugs

[bmarkus]

Found following bugs.

...-KERNEL is not handled in .dep and creates missing dep file warning, like

'ax25-KERNEL.tcz is missing in the repo as a dependency of libax25.tcz.dep'

In .info recently introduced Tags: field not handled yet:

'Tags: field is not valid in libax25.tcz.info.'

In case of empty /usr/local/tce.installed directory error message is wrong, misleading:

'... has more than one /usr/local/tce.installed script'

Just a cosmetic; when maintainer and submitter is the same there is an advice to talk to myself

In few cases I get zero length empty difflist files.

Also there are diff lists where the same file is reported as both missing and extra when submitqc4 is applied on the same set of files as in the repo so they are the same.

[Jason W]

KERNEL is replaced with the current TC kernel version and considers the 64 bit one, but I will look again at it, as upgrading kernels during a TC release cycle as well as perhaps other kernels existing will cause error.
 
Tags checking was added shortly after they were added to the repo.  It checks for existence of the Tags: field.

I will revisit the checking of tce.installed.

I guess I could make it only echo maintainer when they are different, but I hesitate to add complexity as I am in the mood to make the script in general simpler and more maintainable.

File lists are always going to differ on an update, and since some start as usr/ and others start as ./usr, in that case every file will differ.  In other words file list differences are of little value if any at this point, as app or library version updates will always mean differently named files.  I think I will drop the list file diffs.

[bmarkus]

KERNEL is replaced with the current TC kernel version and considers the 64 bit one, but I will look again at it, as upgrading kernels during a TC release cycle as well as perhaps other kernels existing will cause error.
 
Tags checking was added shortly after they were added to the repo.  It checks for existence of the Tags: field.

I will revisit the checking of tce.installed.

I guess I could make it only echo maintainer when they are different, but I hesitate to add complexity as I am in the mood to make the script in general simpler and more maintainable.

File lists are always going to differ on an update, and since some start as usr/ and others start as ./usr, in that case every file will differ.  In other words file list differences are of little value if any at this point, as app or library version updates will always mean differently named files.  I think I will drop the list file diffs.

Jason,

KERNEL is not substituted in .dep, see

ax25-KERNEL.tcz is missing in the repo as a dependency of libax25.tcz.dep

ABout tags in .info:

Tags: field is not valid in libax25.tcz.info.

which indicates that tags: line presence is not checked.

File list difference can be omitted, I do not see benefits. In my case i used it against the sames .info files locally as in the repo and still got wearnings. Just drop this feature.

Maintainer vs. submitter warning is OK as it is, we can have some fun

[Jason W]

Before the kernel upgrade the KERNEL variable substitution worked, but I will change the variable to uname -r, though that will not allow auditing 32 bit extensions on a 64 bit machine, or vice versa.

The Tags check works here on that extension, it it were not checked it would not produce an error.  But I get no error message  when the Tags line is present, only if it is spelled wrong or is missing.  .

I will change the text of the maintainer check, check with maintainer if you and he are not the same. 

[Jason W]

The kernel can be a different one than uname -r now, use the

submitqc4 --kernel=3.0.21-tinycore

if different than uname -r.

Also the startup script check is fixed.

[aus9]

Hi Jason W

I am not a coder so forgive me for silly questions

does this look like a bug.....report of good news and then reporting an error?

2) and is there supposed to be something in that big space between the #s?

Code: [Select]

lilyterm.tcz has proper startup script name lilyterm
 
lilyterm.tcz is maintained by  blofsy. You are  aus9.  If you are not the maintainer, please make sure the maintainer is aware you are updating his extension.
 
lilyterm.tcz  has some similarly named extensions in the repo.  Make sure yours does not    collide with these:
 
lilyterm.tcz
 
The following errors are found in lilyterm.tcz.info. No news is good news:
Change-log: field is not valid in lilyterm.tcz.info.
lilyterm.tcz is a valid tcz file. Checking MD5 for lilyterm.tcz:
lilyterm.tcz: OK
 
##############################################
##############################################
 
 
 
##############################################
##############################################
 
SOME OF THESE EXTENSIONS HAVE CORRUPT INFO FILES.  PLEASE REVIEW /tmp/submitqc/wronginfofile

and I know the next bit is not a bug but thought I better show the contents to wronginfofile in case it helps

Change-log: field is not valid in lilyterm.tcz.info.

and in case it helps because I am stumped on what exactly my bad field is, I hope you don't mind if I post my intended info file? EDIT as attachment


trivia ...I may add my stuff to the tce.installed script but then I have to tell them where it is.
....space is truncated.....but I went back into the info file in mc, and originally I had used leafpad and pressed the TAB key....this time I used the space bar, if that helps

[Jason W]

Hi aus9,
The spaces are put there on purpose to make the output easier to read.

You have named the field "Changelog:", which is incorrect, should be "Change-log:", so the error message is correct.  Syntax must be exact. 

And sorry, I am not sure what you are asking ot the tce.installed script.

Thanks,
JW

[aus9]

thanks for quick reply

don't worry about the tce.installed bit I know what I mean and don't want to waste too much of your time which is why I called it trivia

The only reason I posted was because if that was an error, even a silly one from me, I was hoping it was not say good news when it meant bad news

sorry to have troubled you

sneaks away to correct info files

[Jason W]

Its all good, feedback is always welcome.

[aus9]

great

so I found the reason for the gap because I did not stick any depends in /tmp when running your app the first time....well it wasn't really the first time, giggles

Code: [Select]

##############################################
##############################################
 
 
Dep files seem to have no detectable errors.
 
##############################################
##############################################

Thanks again for your app