Eclipse Community Forums: EPF » Source for (uncorrupted) templates?

Help

Home

Home » Archived » EPF » Source for (uncorrupted) templates?

Show: Today's Messages :: Show Polls :: Message Navigator

Source for (uncorrupted) templates? [message #15475]

Sat, 15 July 2006 00:18

Eclipse User

Originally posted by: christopher.fuhrman.etsmtl.ca

Hello,

The latest build of EPF seems to have again the problem of corrupted
templates for the artifacts (Vision document, SRS, etc.). We like to use
these templates for our courses at our university, since we teach UP.

Is there an alternate place where we can point our students to these
templates?

I'd be more than willing to do some binary diffs to try to spot where
the corruption is occurring (I have seen the bug logged in bugzilla as
https://bugs.eclipse.org/bugs/show_bug.cgi?id=147945).

Cheers,

Cris

--
Prof. Christopher Fuhrman
Department of Software and IT Engineering
University of Quebec - École de technologie supérieure (ETS)
http://profs.logti.etsmtl.ca/cfuhrman/index.shtml?en
+1 (514) 396 8638

Report message to a moderator

Re: Source for (uncorrupted) templates? [message #15508 is a reply to message #15475]

Mon, 17 July 2006 21:12

Ricardo Balduino

Messages: 191
Registered: July 2009

Senior Member

Christopher, the problem is still there and we are investigating. If you can
help finding where the corruption is, that would be great.
For now, I'm attaching a working version of these templates in CVS in, you
can donwload from there:

http://dev.eclipse.org/viewcvs/index.cgi/org.eclipse.epf/lib raries/resources.zip?cvsroot=Technology_Project

I hope that it helps.
Cheers,

Ricardo Balduino
IBM / EPF Committer

"Christopher Fuhrman" <christopher.fuhrman@etsmtl.ca> wrote in message
news:e99c9d$30u$1@utils.eclipse.org...
Hello,

The latest build of EPF seems to have again the problem of corrupted
templates for the artifacts (Vision document, SRS, etc.). We like to use
these templates for our courses at our university, since we teach UP.

Is there an alternate place where we can point our students to these
templates?

I'd be more than willing to do some binary diffs to try to spot where
the corruption is occurring (I have seen the bug logged in bugzilla as
https://bugs.eclipse.org/bugs/show_bug.cgi?id=147945).

Cheers,

Cris

--
Prof. Christopher Fuhrman
Department of Software and IT Engineering
University of Quebec -

Report message to a moderator

Re: Source for (uncorrupted) templates? [message #15539 is a reply to message #15508]

Tue, 18 July 2006 01:30

Eclipse User

Originally posted by: christopher.fuhrman.etsmtl.ca

Hi Ricardo,

Ricardo Balduino wrote:
> Christopher, the problem is still there and we are investigating. If you can
> help finding where the corruption is, that would be great.

The symptom of the problem seems to be (at least at the start of the
corruption) a carriage return (0x0D) that is getting removed from some
(not all!) of the files coming from CVS.

I downloaded the "bad" versions of the a few of the .dot files from CVS,
for example, and I hope this doesn't wrap:

http://dev.eclipse.org/viewcvs/index.cgi/org.eclipse.epf/lib raries/OpenUP/openup_basic/guidances/templates/resources/sup porting_req.dot?cvsroot=Technology_Project

Then I file-compared (using "fc /b" in windows cmd.exe) the "good"
version with the "bad", and here's the output (which I've annotated) for
two of the files that have problems (I didn't have time to do all of them):

---------
Comparing files RESOURCES_GOOD\supporting_req.dot and
RESOURCES_BAD\SUPPORTING_REQ.DOT
000083B9: 0D 0A <--- first carriage return that gets removed
000083BA: 0A 1A
000083BB: 1A 0A
000083BC: 0A 00
000083BF: 00 0D
000083C0: 0D 49
000083C1: 49 48
000083C2: 48 44
.
.
.
---------
Comparing files RESOURCES_GOOD\glossary.dot and RESOURCES_BAD\GLOSSARY.DOT
00001680: 0D 0A <--- first carriage return that gets removed
00001681: 0A 00
00001683: 00 0E
00001684: 0E 0A
00001685: 0A 00
00001687: 00 18
00001688: 18 0A
.
.
.
---------

I'm far from a CVS expert, but it almost seems like these files weren't
all added to indicate they're "binary" files in CVS (-kb is the command
option I remember on the add). I quickly spotted with Google the
following advice for setting binary files in CVS using cvswrappers
(perhaps including .dot as an extension for binary files):

http://www.cs.utexas.edu/~almstrum/cs373/general/cvsguide.ht ml

> For now, I'm attaching a working version of these templates in CVS in, you
> can donwload from there:
>
> http://dev.eclipse.org/viewcvs/index.cgi/org.eclipse.epf/lib raries/resources.zip?cvsroot=Technology_Project
>
> I hope that it helps.

Thanks for the zip file with the "good" templates, and I hope this
partial diagnosis helps!

Cris

> Cheers,
>
> Ricardo Balduino
> IBM / EPF Committer
>
>
> "Christopher Fuhrman" <christopher.fuhrman@etsmtl.ca> wrote in message
> news:e99c9d$30u$1@utils.eclipse.org...
> Hello,
>
> The latest build of EPF seems to have again the problem of corrupted
> templates for the artifacts (Vision document, SRS, etc.). We like to use
> these templates for our courses at our university, since we teach UP.
>
> Is there an alternate place where we can point our students to these
> templates?
>
> I'd be more than willing to do some binary diffs to try to spot where
> the corruption is occurring (I have seen the bug logged in bugzilla as
> https://bugs.eclipse.org/bugs/show_bug.cgi?id=147945).
>
> Cheers,
>
> Cris
>

--
Prof. Christopher Fuhrman
Department of Software and IT Engineering
University of Quebec - École de technologie supérieure (ETS)
http://profs.logti.etsmtl.ca/cfuhrman/index.shtml?en
+1 (514) 396 8638

Report message to a moderator

Re: Source for (uncorrupted) templates? [message #15571 is a reply to message #15539]

Tue, 25 July 2006 03:24

Ricardo Balduino

Messages: 191
Registered: July 2009

Senior Member

It seems the problem was added when we commited the files to CVS as ASCII,
instead of Binary. CVS seems to modify the file as you described.

In order to fix that, I tried this:
- I added 'good' files to CVS
- Changed type to binary (in my local workspace) and committed those files,
but apparently they are converted back to ASCII by CVS server when someone
updates their local workspace.

I wonder if there's any CVS expert that could give us some help on this.

For now I can see the files as binary in my workspace. Please let me know if
the latest weekly build of OpenUP works fine and let's keep an eye in the
next few weekly builds.

Thanks,

Ricardo.

"Christopher Fuhrman" <christopher.fuhrman@etsmtl.ca> wrote in message
news:e9hdk1$o3$1@utils.eclipse.org...
Hi Ricardo,

Ricardo Balduino wrote:
> Christopher, the problem is still there and we are investigating. If you
> can help finding where the corruption is, that would be great.

The symptom of the problem seems to be (at least at the start of the
corruption) a carriage return (0x0D) that is getting removed from some
(not all!) of the files coming from CVS.

I downloaded the "bad" versions of the a few of the .dot files from CVS,
for example, and I hope this doesn't wrap:

http://dev.eclipse.org/viewcvs/index.cgi/org.eclipse.epf/lib raries/OpenUP/openup_basic/guidances/templates/resources/sup porting_req.dot?cvsroot=Technology_Project

Then I file-compared (using "fc /b" in windows cmd.exe) the "good"
version with the "bad", and here's the output (which I've annotated) for
two of the files that have problems (I didn't have time to do all of them):

---------
Comparing files RESOURCES_GOOD\supporting_req.dot and
RESOURCES_BAD\SUPPORTING_REQ.DOT
000083B9: 0D 0A <--- first carriage return that gets removed
000083BA: 0A 1A
000083BB: 1A 0A
000083BC: 0A 00
000083BF: 00 0D
000083C0: 0D 49
000083C1: 49 48
000083C2: 48 44
..
..
..
---------
Comparing files RESOURCES_GOOD\glossary.dot and RESOURCES_BAD\GLOSSARY.DOT
00001680: 0D 0A <--- first carriage return that gets removed
00001681: 0A 00
00001683: 00 0E
00001684: 0E 0A
00001685: 0A 00
00001687: 00 18
00001688: 18 0A
..
..
..
---------

I'm far from a CVS expert, but it almost seems like these files weren't
all added to indicate they're "binary" files in CVS (-kb is the command
option I remember on the add). I quickly spotted with Google the
following advice for setting binary files in CVS using cvswrappers
(perhaps including .dot as an extension for binary files):

http://www.cs.utexas.edu/~almstrum/cs373/general/cvsguide.ht ml

> For now, I'm attaching a working version of these templates in CVS in, you
> can donwload from there:
>
> http://dev.eclipse.org/viewcvs/index.cgi/org.eclipse.epf/lib raries/resources.zip?cvsroot=Technology_Project
>
> I hope that it helps.

Thanks for the zip file with the "good" templates, and I hope this
partial diagnosis helps!

Cris

> Cheers,
>
> Ricardo Balduino
> IBM / EPF Committer
>
>
> "Christopher Fuhrman" <christopher.fuhrman@etsmtl.ca> wrote in message
> news:e99c9d$30u$1@utils.eclipse.org...
> Hello,
>
> The latest build of EPF seems to have again the problem of corrupted
> templates for the artifacts (Vision document, SRS, etc.). We like to use
> these templates for our courses at our university, since we teach UP.
>
> Is there an alternate place where we can point our students to these
> templates?
>
> I'd be more than willing to do some binary diffs to try to spot where
> the corruption is occurring (I have seen the bug logged in bugzilla as
> https://bugs.eclipse.org/bugs/show_bug.cgi?id=147945).
>
> Cheers,
>
> Cris
>

--
Prof. Christopher Fuhrman
Department of Software and IT Engineering
University of Quebec -

Report message to a moderator