Re: Figure 0

2014-04-08 Thread Georg Baum
Jürgen Spitzmüller wrote:

 2014-04-04 0:19 GMT+02:00 Uwe Stöhr uwesto...@web.de:
 
 The time you were the manager for 1.5/6.x I also many times upload test
 installers and you were aware of that. That time the main problem was
 that I did not had Vista and could not see if a problem that was only
 reported in that platform was a OS specific bug or not. With the testing
 releases we could solve some problems. So fine, or did this ever harmed
 LyX? No, in contrary it pushed us forward.
 
 
 You can pick on me, if you like, but again, note that several members of
 the core team expressed the same request to you.

No need to pick on Jürgen. For the 2.0 release I also asked to label 
installers in such a way that the source they were built from can be 
identified. This is no formalism, it is simply required by our license.


Georg



Re: Figure 0

2014-04-08 Thread Uwe Stöhr

Am 04.04.2014 12:01, schrieb Jürgen Spitzmüller:


If you want to see something bad, you will always find something. It hurts
me a lot that you see it like that. Why are you now that formalistic?


Call it formalistic, if you wish. But please note that I am not the only
one asking you for this. Georg asked, Vincent asked, others asked. You can
call all of us formalistic, or you can come to the conclusion that this is
obviously important to us, even if it is not important to you.


I will be more patient to such things in future.

regards Uwe


Re: Figure 0

2014-04-08 Thread Georg Baum
Jürgen Spitzmüller wrote:

> 2014-04-04 0:19 GMT+02:00 Uwe Stöhr :
> 
>> The time you were the manager for 1.5/6.x I also many times upload test
>> installers and you were aware of that. That time the main problem was
>> that I did not had Vista and could not see if a problem that was only
>> reported in that platform was a OS specific bug or not. With the testing
>> releases we could solve some problems. So fine, or did this ever harmed
>> LyX? No, in contrary it pushed us forward.
> 
> 
> You can pick on me, if you like, but again, note that several members of
> the core team expressed the same request to you.

No need to pick on Jürgen. For the 2.0 release I also asked to label 
installers in such a way that the source they were built from can be 
identified. This is no formalism, it is simply required by our license.


Georg



Re: Figure 0

2014-04-08 Thread Uwe Stöhr

Am 04.04.2014 12:01, schrieb Jürgen Spitzmüller:


If you want to see something bad, you will always find something. It hurts
me a lot that you see it like that. Why are you now that formalistic?


Call it formalistic, if you wish. But please note that I am not the only
one asking you for this. Georg asked, Vincent asked, others asked. You can
call all of us formalistic, or you can come to the conclusion that this is
obviously important to us, even if it is not important to you.


I will be more patient to such things in future.

regards Uwe


Re: Figure 0

2014-04-04 Thread Jürgen Spitzmüller
2014-04-04 0:19 GMT+02:00 Uwe Stöhr uwesto...@web.de:

  The problem is the naming of your binary.

 Oh, I simply increased the counter in the installer. I must give it a name
 and I really did not thought about the file name.


Exactly.


 However, the lyx.exe said it is a dev version, not the RC.


Good.


 If you want to see something bad, you will always find something. It hurts
 me a lot that you see it like that. Why are you now that formalistic?


Call it formalistic, if you wish. But please note that I am not the only
one asking you for this. Georg asked, Vincent asked, others asked. You can
call all of us formalistic, or you can come to the conclusion that this is
obviously important to us, even if it is not important to you.

You can stick with what seems important to you, or you can say: OK, if
these guys make such a big deal about this, I'll do it and make them happy.

In any case, as long as you think we do this in order to harass you,
something is going in the wrong direction.



 The time you were the manager for 1.5/6.x I also many times upload test
 installers and you were aware of that. That time the main problem was that
 I did not had Vista and could not see if a problem that was only reported
 in that platform was a OS specific bug or not. With the testing releases we
 could solve some problems. So fine, or did this ever harmed LyX? No, in
 contrary it pushed us forward.


You can pick on me, if you like, but again, note that several members of
the core team expressed the same request to you. And note that you seem to
be the only one who does not care about version naming. Again, you can
persist on your view, or you can go with what the majority of people
involved in this discussion prefer.



  This is not the point. This binary is in the wild.

 I already deleted it yesterday as I got the feedback I needed. This nasty
 bug cost me so much time and it had very strange side effects that I still
 don't understand. As I also got some mail from friends about this while
 others said, they cannot see the problem, so just wanted to be sure.
 I could have sent the test installer link by private mail as I did in the
 past so that the link is not in the Internet. But just 2 days ago JMarc
 encouraged me to bring more on the list what i normally do by private mail.
 So obviously how I do it, I do it wrong :-(


No. If you give the binary an appropriate name, you do it right.



 The problem I have is that you say, I violated a rule but we never voted
 and discussed this. It is not OK that some say +1 and that is then a
 rule. If we want to be more formalistic then we should do it right. Take
 for example the coding style rules - a file visible for everybody with
 precise rules we once discussed and also voted for. I cannot remember that
 we discussed the state of test builds for particular problems. As other GPL
 projects sent me test builds it doesn't seem a very uncommon thing and does
 not violate a license.


Again: Look at the list of people who asked you to care for binary naming
and conformance to the sources. Look at who is included: The people who are
responsible for the release process. Do you really want to make a poll?
Discuss this issue yet another bunch of weeks? I mean, we are talking about
binary naming, not about changing your life for the sake of LyX.


 I invested a lot of time the last weeks. The user feedback encourages me
 while your feedback frustrates me. This is not good and I am tired to have
 to explain myself for every bit. I swear I do my best to get a stable and
 well tested release. It seems to me that it is for you more important to
 fulfill certain rules.


If this is your impression, I will have to live with that.

Jürgen


Re: Figure 0

2014-04-04 Thread Jürgen Spitzmüller
2014-04-04 0:19 GMT+02:00 Uwe Stöhr :

> > The problem is the naming of your binary.
>
> Oh, I simply increased the counter in the installer. I must give it a name
> and I really did not thought about the file name.


Exactly.


> However, the lyx.exe said it is a dev version, not the RC.


Good.


> If you want to see something bad, you will always find something. It hurts
> me a lot that you see it like that. Why are you now that formalistic?


Call it formalistic, if you wish. But please note that I am not the only
one asking you for this. Georg asked, Vincent asked, others asked. You can
call all of us formalistic, or you can come to the conclusion that this is
obviously important to us, even if it is not important to you.

You can stick with what seems important to you, or you can say: "OK, if
these guys make such a big deal about this, I'll do it and make them happy".

In any case, as long as you think we do this in order to harass you,
something is going in the wrong direction.



> The time you were the manager for 1.5/6.x I also many times upload test
> installers and you were aware of that. That time the main problem was that
> I did not had Vista and could not see if a problem that was only reported
> in that platform was a OS specific bug or not. With the testing releases we
> could solve some problems. So fine, or did this ever harmed LyX? No, in
> contrary it pushed us forward.


You can pick on me, if you like, but again, note that several members of
the core team expressed the same request to you. And note that you seem to
be the only one who does not care about version naming. Again, you can
persist on your view, or you can go with what the majority of people
involved in this discussion prefer.



> > This is not the point. This binary is in the wild.
>
> I already deleted it yesterday as I got the feedback I needed. This nasty
> bug cost me so much time and it had very strange side effects that I still
> don't understand. As I also got some mail from friends about this while
> others said, they cannot see the problem, so just wanted to be sure.
> I could have sent the test installer link by private mail as I did in the
> past so that the link is not in the Internet. But just 2 days ago JMarc
> encouraged me to bring more on the list what i normally do by private mail.
> So obviously how I do it, I do it wrong :-(


No. If you give the binary an appropriate name, you do it right.



> The problem I have is that you say, I violated a rule but we never voted
> and discussed this. It is not OK that some say "+1" and that is then a
> rule. If we want to be more formalistic then we should do it right. Take
> for example the coding style rules - a file visible for everybody with
> precise rules we once discussed and also voted for. I cannot remember that
> we discussed the state of test builds for particular problems. As other GPL
> projects sent me test builds it doesn't seem a very uncommon thing and does
> not violate a license.
>

Again: Look at the list of people who asked you to care for binary naming
and conformance to the sources. Look at who is included: The people who are
responsible for the release process. Do you really want to make a poll?
Discuss this issue yet another bunch of weeks? I mean, we are talking about
binary naming, not about changing your life for the sake of LyX.


> I invested a lot of time the last weeks. The user feedback encourages me
> while your feedback frustrates me. This is not good and I am tired to have
> to explain myself for every bit. I swear I do my best to get a stable and
> well tested release. It seems to me that it is for you more important to
> fulfill certain rules.


If this is your impression, I will have to live with that.

Jürgen


Re: Figure 0

2014-04-03 Thread Jürgen Spitzmüller
2014-04-03 1:13 GMT+02:00 Uwe Stöhr uwesto...@web.de:

 I did not release anything. A bug was fixed and I wanted to be sure and
 thus wanted to have user feedback. Why? Because I spent about 4 hours with
 this bug and I saw strange things and some only on some PCS. I noticed that
 it was not enough to replace the lyx.exe because also a registry key needed
 to be recreated to see that it is fixed. So I just upload the lyx.exe with
 an installer that does this automatically in order to get feedback that
 also on the PC of the initial bug reporter the bug is really fixed. So what
 is the problem, we now have the feedback that we need so the task fulfilled.


The problem is the naming of your binary.


 And besides this, did I announce my release?


Yes. In this very thread.


 When starting the lyx of this release does it state it is RC1? No.


Yes you did. The binary is named
LyX-210RC1-Installer-3.exehttps://sourceforge.net/projects/lyxwininstaller/files/TestVersions/LyX-210RC1-Installer-3.exe/download
 This is _not_ RC1.


 Does it appear on any official LyX server? No.


This is not the point. This binary is in the wild.


 Why do you have to criticize me for every single thing I do?


Why do you feel offended every time we ask you to stick with our
rules/agreements?


 For a change, why not seeing the good part of my work - in this case the
 benefit of user feedback for a tricky problem that was not easy to locate
 and not easy to verify that it was fixed? We are in a testing period and
 the aim if to get user feedback. And once again, why not trusting me that I
 don't want to harm LyX or whatever you might think from me. I am tired that
 I have to explain each and every bit nowadays.


Look, Uwe, this is a misconception. We say Uwe, please do not patch the
binaries and you understand Uwe, bad work. These is a significant gap
here in intended and received meaning.

We asked you more than once to not patch releases and to give them a
different name, and you just keep doing it. What do you expect us to do?
Just give up and let you do whatever you want?

This has nothing to do with the service itself you provide. Everybody
appreciates that. But _please_ conform to the group practice. For instance,
you can publish snapshots from GIT as often as you want, but then, name
these LyX-210Dev-76e66aaea890458 or something (appending the revision hash).



 Note, when I have a problem I often got special test builds from
 developers, for example recently from TortoiseGIT and in the past from
 ImageMagick, MiKTeX and JabRef. There is nothing spcial about that. Were
 these releases? No, just a test build for me.


How are these called? I am sure they give it special names.


 I provide test builds for years now. They are often only a few days on my
 SourceForge page in the test section because I can delete them  when the
 issue was discussed by private mail and the issue is reolved. For some bugs
 I let the test builds live a bit longer, especially when they were on the
 list. My SourceForge section is called test because that is what is it
 for - made it possible to get user feedback for particular problems.
 Sometimes one simply needs others to test things. I just cannot test on all
 Windows versions available and with all configurations.


OK, but then (again) give the binaries appropriate names _and_ document all
changes you have done.



 Sorry, I don't get your logic how I harmed LyX.


Then think twice. And read again what I wrote about the licensing issues.

Jürgen



 Uwe



Re: Figure 0

2014-04-03 Thread Vincent van Ravesteijn


 For a change, why not seeing the good part of my work - in this case the
 benefit of user feedback for a tricky problem that was not easy to locate
 and not easy to verify that it was fixed? We are in a testing period and
 the aim if to get user feedback. And once again, why not trusting me that I
 don't want to harm LyX or whatever you might think from me. I am tired that
 I have to explain each and every bit nowadays.


 Look, Uwe, this is a misconception. We say Uwe, please do not patch the
 binaries and you understand Uwe, bad work. These is a significant gap
 here in intended and received meaning.


That's what I indeed meant to say. Nothing more, nothing less. And if we
wouldn't have had this discussion already a few days ago, I might have said
it in a nicer way.

Vincent


Re: Figure 0

2014-04-03 Thread Georg Baum
Uwe Stöhr wrote:

 Am 02.04.2014 09:26, schrieb Vincent van Ravesteijn:
 
 This is the third time we have to tell you to not create and release an
 installer for RC1 if it is not exactly what I released as RC1.
 
 What is going on nowadays?

This is going on: You keep ignoring all attempts to explain to you why a 
binary release must be be labelled in such a way that interested parties can 
see which source code was used to build it. You assume that we want to 
stop you from producing updated installers. This is wrong. As you say there 
are very good reasons for producing an updated installer. The only 
requirement we have is that these updated installers need to be named 
different than RC1 if they were not produced from the RC1 tarball.

Others have explained the reasons why this is a bad practice better than I 
could do, so I'll only add an example of packagers who do it correctly: Most 
linux distributions apply small patches to their LyX packages for various 
reasons, but they always do it in such a way that it is reproducible. On 
https://packages.debian.org/sid/lyx you see the page of the 'lyx' package in 
debian. It contains a link to the archive of the unmodified source which was 
used (currently  
http://ftp.debian.org/debian/pool/main/l/lyx/lyx_2.0.6.orig.tar.xz), and a 
link to an archive with the applied changes (currently 
http://ftp.debian.org/debian/pool/main/l/lyx/lyx_2.0.6-1.debian.tar.gz). The 
latter also uses the suffix which is applied to the original version number 
and used for the resulting binary package. In this case it is simply -1, 
but sometimes it is something more complicated.

With this information everybody can study the source of the binary package, 
which is not possible for the updated installers you provided after the RC1 
release.

If you do not use a simple version suffix, but a git hash as Jürgen 
suggested, you would not even need to provide a patch file, since everybody 
could just fetch the particular git revision. This would be _zero_ extra 
work for you, and everyboy would be happy.


Georg



Re: Figure 0

2014-04-03 Thread Uwe Stöhr

Am 03.04.2014 08:43, schrieb Jürgen Spitzmüller:

 The problem is the naming of your binary.

Oh, I simply increased the counter in the installer. I must give it a name and I really did not 
thought about the file name. However, the lyx.exe said it is a dev version, not the RC.


 And besides this, did I announce my release?

 Yes. In this very thread.

If you want to see something bad, you will always find something. It hurts me a lot that you see it 
like that. Why are you now that formalistic? The time you were the manager for 1.5/6.x I also many 
times upload test installers and you were aware of that. That time the main problem was that I did 
not had Vista and could not see if a problem that was only reported in that platform was a OS 
specific bug or not. With the testing releases we could solve some problems. So fine, or did this 
ever harmed LyX? No, in contrary it pushed us forward.


 This is not the point. This binary is in the wild.

I already deleted it yesterday as I got the feedback I needed. This nasty bug cost me so much time 
and it had very strange side effects that I still don't understand. As I also got some mail from 
friends about this while others said, they cannot see the problem, so just wanted to be sure.
I could have sent the test installer link by private mail as I did in the past so that the link is 
not in the Internet. But just 2 days ago JMarc encouraged me to bring more on the list what i 
normally do by private mail. So obviously how I do it, I do it wrong :-(




Why do you have to criticize me for every single thing I do?


Why do you feel offended every time we ask you to stick with our
rules/agreements?


The problem I have is that you say, I violated a rule but we never voted and discussed this. It is 
not OK that some say +1 and that is then a rule. If we want to be more formalistic then we should 
do it right. Take for example the coding style rules - a file visible for everybody with precise 
rules we once discussed and also voted for. I cannot remember that we discussed the state of test 
builds for particular problems. As other GPL projects sent me test builds it doesn't seem a very 
uncommon thing and does not violate a license.
For some Windows issues I will always need the help of users and sending them a special lyx.exe is 
often not enough (as in this case).


However, I have more points I wanted to say, please have a look about my mew thread of the LyX 2.1 
release. There are so many things in the air that lead to confusions and I don't thick that it is 
right what we are doing right now for LyX 2.1.


I invested a lot of time the last weeks. The user feedback encourages me while your feedback 
frustrates me. This is not good and I am tired to have to explain myself for every bit. I swear I do 
my best to get a stable and well tested release. It seems to me that it is for you more important to 
fulfill certain rules. LyX 2.1 is in my opinion not yet ready for a release and I would like to get 
more feedback and more time. Both are in my opinion very important at that state!


--

I am currently very busy and will be at a business trip. So it might be that I cannot respond until 
Sunday and I did also not yet had the time to respond to the other posts. Please don't be offended.


regards Uwe


Re: Figure 0

2014-04-03 Thread Uwe Stöhr

Am 03.04.2014 08:43, schrieb Jürgen Spitzmüller:

 his has nothing to do with the service itself you provide. Everybody
 appreciates that. But _please_ conform to the group practice. For instance,
 you can publish snapshots from GIT as often as you want, but then, name
 these LyX-210Dev-76e66aaea890458 or something (appending the revision hash).

That is of course possible.

Nevertheless, please trust me that it was not my intention to release anything nor so say it is an 
RC. I don't think about these things that much. (It was very late and it cost me some time until I 
realized that replacing the lyx.exe was not enough so that I need to build an installer). However, 
for me formalism is not that important as you know ;-) and I think the tester cared about my wrong 
name. I wrote him what this build is about.



Note, when I have a problem I often got special test builds from
developers, for example recently from TortoiseGIT and in the past from
ImageMagick, MiKTeX and JabRef. There is nothing spcial about that. Were
these releases? No, just a test build for me.


How are these called? I am sure they give it special names.


I cannot tell as I don't have it anymore, I received it by private mail after 
filing this issue:
http://code.google.com/p/tortoisegit/issues/detail?id=1788can=1q=reporter%3Auwestoehr

The bug I found was on Win XP 64bit (aka Win 2003)-only  and the developers didn't had access to 
that Windows version. I got a build (full working installer) for this version:

http://code.google.com/p/tortoisegit/source/detail?r=1c72523e0345

And that is what I also did in the past and will have to do in future as well.

regards Uwe


Re: Figure 0

2014-04-03 Thread Jürgen Spitzmüller
2014-04-03 1:13 GMT+02:00 Uwe Stöhr :

> I did not release anything. A bug was fixed and I wanted to be sure and
> thus wanted to have user feedback. Why? Because I spent about 4 hours with
> this bug and I saw strange things and some only on some PCS. I noticed that
> it was not enough to replace the lyx.exe because also a registry key needed
> to be recreated to see that it is fixed. So I just upload the lyx.exe with
> an installer that does this automatically in order to get feedback that
> also on the PC of the initial bug reporter the bug is really fixed. So what
> is the problem, we now have the feedback that we need so the task fulfilled.
>

The problem is the naming of your binary.


> And besides this, did I announce my "release"?


Yes. In this very thread.


> When starting the lyx of this "release" does it state it is RC1? No.


Yes you did. The binary is named
"LyX-210RC1-Installer-3.exe"
 This is _not_ RC1.


> Does it appear on any official LyX server? No.
>

This is not the point. This binary is in the wild.


> Why do you have to criticize me for every single thing I do?


Why do you feel offended every time we ask you to stick with our
rules/agreements?


> For a change, why not seeing the good part of my work - in this case the
> benefit of user feedback for a tricky problem that was not easy to locate
> and not easy to verify that it was fixed? We are in a testing period and
> the aim if to get user feedback. And once again, why not trusting me that I
> don't want to harm LyX or whatever you might think from me. I am tired that
> I have to explain each and every bit nowadays.
>

Look, Uwe, this is a misconception. We say "Uwe, please do not patch the
binaries" and you understand "Uwe, bad work". These is a significant gap
here in intended and received meaning.

We asked you more than once to not patch releases and to give them a
different name, and you just keep doing it. What do you expect us to do?
Just give up and let you do whatever you want?

This has nothing to do with the service itself you provide. Everybody
appreciates that. But _please_ conform to the group practice. For instance,
you can publish snapshots from GIT as often as you want, but then, name
these LyX-210Dev-76e66aaea890458 or something (appending the revision hash).



> Note, when I have a problem I often got special test builds from
> developers, for example recently from TortoiseGIT and in the past from
> ImageMagick, MiKTeX and JabRef. There is nothing spcial about that. Were
> these "releases"? No, just a test build for me.
>

How are these called? I am sure they give it special names.


> I provide test builds for years now. They are often only a few days on my
> SourceForge page in the test section because I can delete them  when the
> issue was discussed by private mail and the issue is reolved. For some bugs
> I let the test builds live a bit longer, especially when they were on the
> list. My SourceForge section is called "test" because that is what is it
> for - made it possible to get user feedback for particular problems.
> Sometimes one simply needs others to test things. I just cannot test on all
> Windows versions available and with all configurations.
>

OK, but then (again) give the binaries appropriate names _and_ document all
changes you have done.


>
> Sorry, I don't get your logic how I harmed LyX.
>

Then think twice. And read again what I wrote about the licensing issues.

Jürgen


>
> Uwe
>


Re: Figure 0

2014-04-03 Thread Vincent van Ravesteijn
>
>
>> For a change, why not seeing the good part of my work - in this case the
>> benefit of user feedback for a tricky problem that was not easy to locate
>> and not easy to verify that it was fixed? We are in a testing period and
>> the aim if to get user feedback. And once again, why not trusting me that I
>> don't want to harm LyX or whatever you might think from me. I am tired that
>> I have to explain each and every bit nowadays.
>>
>
> Look, Uwe, this is a misconception. We say "Uwe, please do not patch the
> binaries" and you understand "Uwe, bad work". These is a significant gap
> here in intended and received meaning.
>

That's what I indeed meant to say. Nothing more, nothing less. And if we
wouldn't have had this discussion already a few days ago, I might have said
it in a nicer way.

Vincent


Re: Figure 0

2014-04-03 Thread Georg Baum
Uwe Stöhr wrote:

> Am 02.04.2014 09:26, schrieb Vincent van Ravesteijn:
> 
>> This is the third time we have to tell you to not create and release an
>> installer for RC1 if it is not exactly what I released as RC1.
> 
> What is going on nowadays?

This is going on: You keep ignoring all attempts to explain to you why a 
binary release must be be labelled in such a way that interested parties can 
see which source code was used to build it. You assume that "we" want to 
stop you from producing updated installers. This is wrong. As you say there 
are very good reasons for producing an updated installer. The only 
requirement "we" have is that these updated installers need to be named 
different than RC1 if they were not produced from the RC1 tarball.

Others have explained the reasons why this is a bad practice better than I 
could do, so I'll only add an example of packagers who do it correctly: Most 
linux distributions apply small patches to their LyX packages for various 
reasons, but they always do it in such a way that it is reproducible. On 
https://packages.debian.org/sid/lyx you see the page of the 'lyx' package in 
debian. It contains a link to the archive of the unmodified source which was 
used (currently  
http://ftp.debian.org/debian/pool/main/l/lyx/lyx_2.0.6.orig.tar.xz), and a 
link to an archive with the applied changes (currently 
http://ftp.debian.org/debian/pool/main/l/lyx/lyx_2.0.6-1.debian.tar.gz). The 
latter also uses the suffix which is applied to the original version number 
and used for the resulting binary package. In this case it is simply "-1", 
but sometimes it is something more complicated.

With this information everybody can study the source of the binary package, 
which is not possible for the updated installers you provided after the RC1 
release.

If you do not use a simple version suffix, but a git hash as Jürgen 
suggested, you would not even need to provide a patch file, since everybody 
could just fetch the particular git revision. This would be _zero_ extra 
work for you, and everyboy would be happy.


Georg



Re: Figure 0

2014-04-03 Thread Uwe Stöhr

Am 03.04.2014 08:43, schrieb Jürgen Spitzmüller:

> The problem is the naming of your binary.

Oh, I simply increased the counter in the installer. I must give it a name and I really did not 
thought about the file name. However, the lyx.exe said it is a dev version, not the RC.


>> And besides this, did I announce my "release"?
>
> Yes. In this very thread.

If you want to see something bad, you will always find something. It hurts me a lot that you see it 
like that. Why are you now that formalistic? The time you were the manager for 1.5/6.x I also many 
times upload test installers and you were aware of that. That time the main problem was that I did 
not had Vista and could not see if a problem that was only reported in that platform was a OS 
specific bug or not. With the testing releases we could solve some problems. So fine, or did this 
ever harmed LyX? No, in contrary it pushed us forward.


> This is not the point. This binary is in the wild.

I already deleted it yesterday as I got the feedback I needed. This nasty bug cost me so much time 
and it had very strange side effects that I still don't understand. As I also got some mail from 
friends about this while others said, they cannot see the problem, so just wanted to be sure.
I could have sent the test installer link by private mail as I did in the past so that the link is 
not in the Internet. But just 2 days ago JMarc encouraged me to bring more on the list what i 
normally do by private mail. So obviously how I do it, I do it wrong :-(




Why do you have to criticize me for every single thing I do?


Why do you feel offended every time we ask you to stick with our
rules/agreements?


The problem I have is that you say, I violated a rule but we never voted and discussed this. It is 
not OK that some say "+1" and that is then a rule. If we want to be more formalistic then we should 
do it right. Take for example the coding style rules - a file visible for everybody with precise 
rules we once discussed and also voted for. I cannot remember that we discussed the state of test 
builds for particular problems. As other GPL projects sent me test builds it doesn't seem a very 
uncommon thing and does not violate a license.
For some Windows issues I will always need the help of users and sending them a special lyx.exe is 
often not enough (as in this case).


However, I have more points I wanted to say, please have a look about my mew thread of the LyX 2.1 
release. There are so many things in the air that lead to confusions and I don't thick that it is 
right what we are doing right now for LyX 2.1.


I invested a lot of time the last weeks. The user feedback encourages me while your feedback 
frustrates me. This is not good and I am tired to have to explain myself for every bit. I swear I do 
my best to get a stable and well tested release. It seems to me that it is for you more important to 
fulfill certain rules. LyX 2.1 is in my opinion not yet ready for a release and I would like to get 
more feedback and more time. Both are in my opinion very important at that state!


--

I am currently very busy and will be at a business trip. So it might be that I cannot respond until 
Sunday and I did also not yet had the time to respond to the other posts. Please don't be offended.


regards Uwe


Re: Figure 0

2014-04-03 Thread Uwe Stöhr

Am 03.04.2014 08:43, schrieb Jürgen Spitzmüller:

> his has nothing to do with the service itself you provide. Everybody
> appreciates that. But _please_ conform to the group practice. For instance,
> you can publish snapshots from GIT as often as you want, but then, name
> these LyX-210Dev-76e66aaea890458 or something (appending the revision hash).

That is of course possible.

Nevertheless, please trust me that it was not my intention to release anything nor so say it is an 
RC. I don't think about these things that much. (It was very late and it cost me some time until I 
realized that replacing the lyx.exe was not enough so that I need to build an installer). However, 
for me formalism is not that important as you know ;-) and I think the tester cared about my wrong 
name. I wrote him what this build is about.



Note, when I have a problem I often got special test builds from
developers, for example recently from TortoiseGIT and in the past from
ImageMagick, MiKTeX and JabRef. There is nothing spcial about that. Were
these "releases"? No, just a test build for me.


How are these called? I am sure they give it special names.


I cannot tell as I don't have it anymore, I received it by private mail after 
filing this issue:
http://code.google.com/p/tortoisegit/issues/detail?id=1788=1=reporter%3Auwestoehr

The bug I found was on Win XP 64bit (aka Win 2003)-only  and the developers didn't had access to 
that Windows version. I got a build (full working installer) for this version:

http://code.google.com/p/tortoisegit/source/detail?r=1c72523e0345

And that is what I also did in the past and will have to do in future as well.

regards Uwe


Re: Figure 0

2014-04-02 Thread Vincent van Ravesteijn
On Wed, Apr 2, 2014 at 1:39 AM, Uwe Stöhr uwesto...@web.de wrote:

 Am 31.03.2014 22:50, schrieb Uwe Stöhr:


  I found now a recipe to reproduce, see
 http://www.lyx.org/trac/ticket/9070


 This bug is now fixed and I prepared an installer for this:
 https://sourceforge.net/projects/lyxwininstaller/
 files/TestVersions/LyX-210RC1-Installer-3.exe/download

 Please give it a try and report back (you must probably uninstall LyX
 2.1RC1 before).

 thanks and regards
 Uwe


Uwe,

This is the third time we have to tell you to not create and release an
installer for RC1 if it is not exactly what I released as RC1.

I hope you take notice of this now.

Vincent


Re: Figure 0

2014-04-02 Thread Jürgen Spitzmüller
2014-04-02 9:26 GMT+02:00 Vincent van Ravesteijn:

 This is the third time we have to tell you to not create and release an
 installer for RC1 if it is not exactly what I released as RC1.

 I hope you take notice of this now.


I may add that this also probably causes license problems. We violate our
own license.

The GPL requires that

*[§] 2.* You may modify your copy or copies of the Program or any portion
of it, thus forming a work based on the Program, and copy and distribute
such modifications or work under the terms of Section 1 above, provided
that you also meet all of these conditions: *a)* You must cause the
modified files to carry prominent notices stating that you changed the
files and the date of any change.

and

[§] *3.* You may copy and distribute the Program (or a work based on it,
under Section 2) in object code or executable form under the terms of
Sections 1 and 2 above provided that you also do one of the following:

*a)* Accompany it with the complete corresponding machine-readable source
code, which must be distributed under the terms of Sections 1 and 2 above
on a medium customarily used for software interchange; or,
*b)* Accompany it with a written offer, valid for at least three years, to
give any third party, for a charge no more than your cost of physically
performing source distribution, a complete machine-readable copy of the
corresponding source code, to be distributed under the terms of Sections 1
and 2 above on a medium customarily used for software interchange; or,
*c)* Accompany it with the information you received as to the offer to
distribute corresponding source code. (This alternative is allowed only for
noncommercial distribution and only if you received the program in object
code or executable form with such an offer, in accord with Subsection b
above.

If changes are committed to the released sources, then there is strictly
speaking no accompanying source code to the released binary.

So there is also a legal reason why modified binaries contain detailed
notes about the changes to the original source and a different name. It
might be nitpicking, but it could be that we buy problems by releasing
patched binaries without documentation and source (at least in for of
diffs). We might weaken our own position in the case someone willingly
infringes the license if it can be proves that we do not follow the license
ourselves.

I suggest that we put a note to the relevant document for binaries
packagers stating something along this line:

Please note that binaries must be build from the released sources. Please
do not build from the GIT repository. Please do not patch the sources for
building. If a patch is absolutely necessary, the patch must be included in
your package as well, and you must give the binary a different version in
order to make clear that this is not 100% identical with the released
sources.

Jürgen



 Vincent




Re: Figure 0

2014-04-02 Thread Stephan Witt
Am 02.04.2014 um 12:37 schrieb Jürgen Spitzmüller sp...@lyx.org:

 2014-04-02 9:26 GMT+02:00 Vincent van Ravesteijn:
 This is the third time we have to tell you to not create and release an 
 installer for RC1 if it is not exactly what I released as RC1.
  
 I hope you take notice of this now.
 
 I may add that this also probably causes license problems. We violate our own 
 license.
 
 
...
 So there is also a legal reason why modified binaries contain detailed notes 
 about the changes to the original source and a different name. It might be 
 nitpicking, but it could be that we buy problems by releasing patched 
 binaries without documentation and source (at least in for of diffs). We 
 might weaken our own position in the case someone willingly infringes the 
 license if it can be proves that we do not follow the license ourselves.
 
 I suggest that we put a note to the relevant document for binaries packagers 
 stating something along this line:
 
 Please note that binaries must be build from the released sources. Please do 
 not build from the GIT repository. Please do not patch the sources for 
 building. If a patch is absolutely necessary, the patch must be included in 
 your package as well, and you must give the binary a different version in 
 order to make clear that this is not 100% identical with the released 
 sources.
 Jürgen

I agree.

Stephan

Re: Figure 0

2014-04-02 Thread Uwe Stöhr

Am 02.04.2014 02:19, schrieb Jürgen Lange:


Report: I can confirm, that the bug is now fixed. Thanks to all participating.


Many thanks for testing! Good to hear that it works for you.

regards Uwe


Re: Figure 0

2014-04-02 Thread Uwe Stöhr

Am 02.04.2014 09:26, schrieb Vincent van Ravesteijn:


This is the third time we have to tell you to not create and release an
installer for RC1 if it is not exactly what I released as RC1.


What is going on nowadays?

I did not release anything. A bug was fixed and I wanted to be sure and thus wanted to have user 
feedback. Why? Because I spent about 4 hours with this bug and I saw strange things and some only on 
some PCS. I noticed that it was not enough to replace the lyx.exe because also a registry key needed 
to be recreated to see that it is fixed. So I just upload the lyx.exe with an installer that does 
this automatically in order to get feedback that also on the PC of the initial bug reporter the bug 
is really fixed. So what is the problem, we now have the feedback that we need so the task fulfilled.


And besides this, did I announce my release? When starting the lyx of this release does it state 
it is RC1? No. Does it appear on any official LyX server? No.
Why do you have to criticize me for every single thing I do? For a change, why not seeing the good 
part of my work - in this case the benefit of user feedback for a tricky problem that was not easy 
to locate and not easy to verify that it was fixed? We are in a testing period and the aim if to get 
user feedback. And once again, why not trusting me that I don't want to harm LyX or whatever you 
might think from me. I am tired that I have to explain each and every bit nowadays.


Note, when I have a problem I often got special test builds from developers, for example recently 
from TortoiseGIT and in the past from ImageMagick, MiKTeX and JabRef. There is nothing spcial about 
that. Were these releases? No, just a test build for me.
I provide test builds for years now. They are often only a few days on my SourceForge page in the 
test section because I can delete them  when the issue was discussed by private mail and the issue 
is reolved. For some bugs I let the test builds live a bit longer, especially when they were on the 
list. My SourceForge section is called test because that is what is it for - made it possible to 
get user feedback for particular problems. Sometimes one simply needs others to test things. I just 
cannot test on all Windows versions available and with all configurations.


Sorry, I don't get your logic how I harmed LyX.

Uwe


Re: Figure 0

2014-04-02 Thread Vincent van Ravesteijn
On Wed, Apr 2, 2014 at 1:39 AM, Uwe Stöhr  wrote:

> Am 31.03.2014 22:50, schrieb Uwe Stöhr:
>
>
>  I found now a recipe to reproduce, see
>> http://www.lyx.org/trac/ticket/9070
>>
>
> This bug is now fixed and I prepared an installer for this:
> https://sourceforge.net/projects/lyxwininstaller/
> files/TestVersions/LyX-210RC1-Installer-3.exe/download
>
> Please give it a try and report back (you must probably uninstall LyX
> 2.1RC1 before).
>
> thanks and regards
> Uwe
>

Uwe,

This is the third time we have to tell you to not create and release an
installer for RC1 if it is not exactly what I released as RC1.

I hope you take notice of this now.

Vincent


Re: Figure 0

2014-04-02 Thread Jürgen Spitzmüller
2014-04-02 9:26 GMT+02:00 Vincent van Ravesteijn:

> This is the third time we have to tell you to not create and release an
> installer for RC1 if it is not exactly what I released as RC1.
>
> I hope you take notice of this now.
>

I may add that this also probably causes license problems. We violate our
own license.

The GPL requires that

"*[§] 2.* You may modify your copy or copies of the Program or any portion
of it, thus forming a work based on the Program, and copy and distribute
such modifications or work under the terms of Section 1 above, provided
that you also meet all of these conditions: *a)* You must cause the
modified files to carry prominent notices stating that you changed the
files and the date of any change."

and

"[§] *3.* You may copy and distribute the Program (or a work based on it,
under Section 2) in object code or executable form under the terms of
Sections 1 and 2 above provided that you also do one of the following:

*a)* Accompany it with the complete corresponding machine-readable source
code, which must be distributed under the terms of Sections 1 and 2 above
on a medium customarily used for software interchange; or,
*b)* Accompany it with a written offer, valid for at least three years, to
give any third party, for a charge no more than your cost of physically
performing source distribution, a complete machine-readable copy of the
corresponding source code, to be distributed under the terms of Sections 1
and 2 above on a medium customarily used for software interchange; or,
*c)* Accompany it with the information you received as to the offer to
distribute corresponding source code. (This alternative is allowed only for
noncommercial distribution and only if you received the program in object
code or executable form with such an offer, in accord with Subsection b
above."

If changes are committed to the released sources, then there is strictly
speaking no accompanying source code to the released binary.

So there is also a legal reason why modified binaries contain detailed
notes about the changes to the original source and a different name. It
might be nitpicking, but it could be that we buy problems by releasing
patched binaries without documentation and source (at least in for of
diffs). We might weaken our own position in the case someone willingly
infringes the license if it can be proves that we do not follow the license
ourselves.

I suggest that we put a note to the relevant document for binaries
packagers stating something along this line:

"Please note that binaries must be build from the released sources. Please
do not build from the GIT repository. Please do not patch the sources for
building. If a patch is absolutely necessary, the patch must be included in
your package as well, and you must give the binary a different version in
order to make clear that this is not 100% identical with the released
sources."

Jürgen


>
> Vincent
>
>


Re: Figure 0

2014-04-02 Thread Stephan Witt
Am 02.04.2014 um 12:37 schrieb Jürgen Spitzmüller :

> 2014-04-02 9:26 GMT+02:00 Vincent van Ravesteijn:
> This is the third time we have to tell you to not create and release an 
> installer for RC1 if it is not exactly what I released as RC1.
>  
> I hope you take notice of this now.
> 
> I may add that this also probably causes license problems. We violate our own 
> license.
> 
> 
...
> So there is also a legal reason why modified binaries contain detailed notes 
> about the changes to the original source and a different name. It might be 
> nitpicking, but it could be that we buy problems by releasing patched 
> binaries without documentation and source (at least in for of diffs). We 
> might weaken our own position in the case someone willingly infringes the 
> license if it can be proves that we do not follow the license ourselves.
> 
> I suggest that we put a note to the relevant document for binaries packagers 
> stating something along this line:
> 
> "Please note that binaries must be build from the released sources. Please do 
> not build from the GIT repository. Please do not patch the sources for 
> building. If a patch is absolutely necessary, the patch must be included in 
> your package as well, and you must give the binary a different version in 
> order to make clear that this is not 100% identical with the released 
> sources."
> Jürgen

I agree.

Stephan

Re: Figure 0

2014-04-02 Thread Uwe Stöhr

Am 02.04.2014 02:19, schrieb Jürgen Lange:


Report: I can confirm, that the bug is now fixed. Thanks to all participating.


Many thanks for testing! Good to hear that it works for you.

regards Uwe


Re: Figure 0

2014-04-02 Thread Uwe Stöhr

Am 02.04.2014 09:26, schrieb Vincent van Ravesteijn:


This is the third time we have to tell you to not create and release an
installer for RC1 if it is not exactly what I released as RC1.


What is going on nowadays?

I did not release anything. A bug was fixed and I wanted to be sure and thus wanted to have user 
feedback. Why? Because I spent about 4 hours with this bug and I saw strange things and some only on 
some PCS. I noticed that it was not enough to replace the lyx.exe because also a registry key needed 
to be recreated to see that it is fixed. So I just upload the lyx.exe with an installer that does 
this automatically in order to get feedback that also on the PC of the initial bug reporter the bug 
is really fixed. So what is the problem, we now have the feedback that we need so the task fulfilled.


And besides this, did I announce my "release"? When starting the lyx of this "release" does it state 
it is RC1? No. Does it appear on any official LyX server? No.
Why do you have to criticize me for every single thing I do? For a change, why not seeing the good 
part of my work - in this case the benefit of user feedback for a tricky problem that was not easy 
to locate and not easy to verify that it was fixed? We are in a testing period and the aim if to get 
user feedback. And once again, why not trusting me that I don't want to harm LyX or whatever you 
might think from me. I am tired that I have to explain each and every bit nowadays.


Note, when I have a problem I often got special test builds from developers, for example recently 
from TortoiseGIT and in the past from ImageMagick, MiKTeX and JabRef. There is nothing spcial about 
that. Were these "releases"? No, just a test build for me.
I provide test builds for years now. They are often only a few days on my SourceForge page in the 
test section because I can delete them  when the issue was discussed by private mail and the issue 
is reolved. For some bugs I let the test builds live a bit longer, especially when they were on the 
list. My SourceForge section is called "test" because that is what is it for - made it possible to 
get user feedback for particular problems. Sometimes one simply needs others to test things. I just 
cannot test on all Windows versions available and with all configurations.


Sorry, I don't get your logic how I harmed LyX.

Uwe


Re: Figure 0

2014-04-01 Thread Uwe Stöhr

Am 31.03.2014 22:50, schrieb Uwe Stöhr:


I found now a recipe to reproduce, see
http://www.lyx.org/trac/ticket/9070


This bug is now fixed and I prepared an installer for this:
https://sourceforge.net/projects/lyxwininstaller/files/TestVersions/LyX-210RC1-Installer-3.exe/download

Please give it a try and report back (you must probably uninstall LyX 2.1RC1 
before).

thanks and regards
Uwe


Re: Figure 0

2014-04-01 Thread Jürgen Lange
Report: I can confirm, that the bug is now fixed. Thanks to all  
participating.


Re: Figure 0

2014-04-01 Thread Uwe Stöhr

Am 31.03.2014 22:50, schrieb Uwe Stöhr:


I found now a recipe to reproduce, see
http://www.lyx.org/trac/ticket/9070


This bug is now fixed and I prepared an installer for this:
https://sourceforge.net/projects/lyxwininstaller/files/TestVersions/LyX-210RC1-Installer-3.exe/download

Please give it a try and report back (you must probably uninstall LyX 2.1RC1 
before).

thanks and regards
Uwe


Re: Figure 0

2014-04-01 Thread Jürgen Lange
Report: I can confirm, that the bug is now fixed. Thanks to all  
participating.


Re: Figure 0

2014-03-31 Thread Jürgen Lange

Am 31.03.2014, 01:49 Uhr, schrieb Uwe Stöhr uwesto...@web.de:


Am 31.03.2014 00:55, schrieb Jürgen Lange:

In document view all figures and tables are labelled Figure 0 and Table  
0.
The resulting pdf is ok (Figure 1, Figure 2 etc.). I've this problem  
since I updated to 2.1.0rc1.


Does this problem disappear when you restart LyX?
If so, I can say that I get the same problem sometimes.

regards Uwe


The problem survived some restarts, but did disappear after a while . It
reappears then after a rather short time. Figure 0 problem seems to be  
relatively difficult to reproduce. (Win7, 1.3.0rc1)


regards
Jürgen


Re: Figure 0

2014-03-31 Thread Jürgen Lange

Am 31.03.2014, 08:49 Uhr, schrieb Jürgen Lange juergen.la...@unitybox.de:


Am 31.03.2014, 01:49 Uhr, schrieb Uwe Stöhr uwesto...@web.de:


Am 31.03.2014 00:55, schrieb Jürgen Lange:

In document view all figures and tables are labelled Figure 0 and  
Table 0.
The resulting pdf is ok (Figure 1, Figure 2 etc.). I've this problem  
since I updated to 2.1.0rc1.


Does this problem disappear when you restart LyX?
If so, I can say that I get the same problem sometimes.

regards Uwe


The problem survived some restarts, but did disappear after a while . It
reappears then after a rather short time. Figure 0 problem seems to be  
relatively difficult to reproduce. (Win7, 1.3.0rc1)


regards
Jürgen


Correction: (Win7, 2.1.0rc1)


Re: Figure 0

2014-03-31 Thread Jürgen Lange
Remark: Rebooting the system (win7) makes Figure 0 problem a lot more  
reproducible.


regards
Jürgen


Re: Figure 0

2014-03-31 Thread Uwe Stöhr

Am 31.03.2014 20:47, schrieb Jürgen Lange:


Remark: Rebooting the system (win7) makes Figure 0 problem a lot more 
reproducible.


I found now a recipe to reproduce, see
http://www.lyx.org/trac/ticket/9070

I must admit that I don't have any clue what exactly triggers the bug and why 
it is only on Windows 7.

regards Uwe


Re: Figure 0

2014-03-31 Thread Jürgen Lange

Am 31.03.2014, 01:49 Uhr, schrieb Uwe Stöhr <uwesto...@web.de>:


Am 31.03.2014 00:55, schrieb Jürgen Lange:

In document view all figures and tables are labelled Figure 0 and Table  
0.
The resulting pdf is ok (Figure 1, Figure 2 etc.). I've this problem  
since I updated to 2.1.0rc1.


Does this problem disappear when you restart LyX?
If so, I can say that I get the same problem sometimes.

regards Uwe


The problem survived some restarts, but did disappear after a while . It
reappears then after a rather short time. Figure 0 problem seems to be  
relatively difficult to reproduce. (Win7, 1.3.0rc1)


regards
Jürgen


Re: Figure 0

2014-03-31 Thread Jürgen Lange

Am 31.03.2014, 08:49 Uhr, schrieb Jürgen Lange <juergen.la...@unitybox.de>:


Am 31.03.2014, 01:49 Uhr, schrieb Uwe Stöhr <uwesto...@web.de>:


Am 31.03.2014 00:55, schrieb Jürgen Lange:

In document view all figures and tables are labelled Figure 0 and  
Table 0.
The resulting pdf is ok (Figure 1, Figure 2 etc.). I've this problem  
since I updated to 2.1.0rc1.


Does this problem disappear when you restart LyX?
If so, I can say that I get the same problem sometimes.

regards Uwe


The problem survived some restarts, but did disappear after a while . It
reappears then after a rather short time. Figure 0 problem seems to be  
relatively difficult to reproduce. (Win7, 1.3.0rc1)


regards
Jürgen


Correction: (Win7, 2.1.0rc1)


Re: Figure 0

2014-03-31 Thread Jürgen Lange
Remark: Rebooting the system (win7) makes Figure 0 problem a lot more  
reproducible.


regards
Jürgen


Re: Figure 0

2014-03-31 Thread Uwe Stöhr

Am 31.03.2014 20:47, schrieb Jürgen Lange:


Remark: Rebooting the system (win7) makes Figure 0 problem a lot more 
reproducible.


I found now a recipe to reproduce, see
http://www.lyx.org/trac/ticket/9070

I must admit that I don't have any clue what exactly triggers the bug and why 
it is only on Windows 7.

regards Uwe


Figure 0

2014-03-30 Thread Jürgen Lange

In document view all figures and tables are labelled Figure 0 and Table 0.
The resulting pdf is ok (Figure 1, Figure 2 etc.). I've this problem since  
I updated to 2.1.0rc1.

What's going wrong? Example file appended.

thanks and regards
Jürgen

Figure 0.lyx
Description: Binary data


Re: Figure 0

2014-03-30 Thread Uwe Stöhr

Am 31.03.2014 00:55, schrieb Jürgen Lange:


In document view all figures and tables are labelled Figure 0 and Table 0.
The resulting pdf is ok (Figure 1, Figure 2 etc.). I've this problem since I 
updated to 2.1.0rc1.


Does this problem disappear when you restart LyX?
If so, I can say that I get the same problem sometimes.

regards Uwe


Figure 0

2014-03-30 Thread Jürgen Lange

In document view all figures and tables are labelled Figure 0 and Table 0.
The resulting pdf is ok (Figure 1, Figure 2 etc.). I've this problem since  
I updated to 2.1.0rc1.

What's going wrong? Example file appended.

thanks and regards
Jürgen

Figure 0.lyx
Description: Binary data


Re: Figure 0

2014-03-30 Thread Uwe Stöhr

Am 31.03.2014 00:55, schrieb Jürgen Lange:


In document view all figures and tables are labelled Figure 0 and Table 0.
The resulting pdf is ok (Figure 1, Figure 2 etc.). I've this problem since I 
updated to 2.1.0rc1.


Does this problem disappear when you restart LyX?
If so, I can say that I get the same problem sometimes.

regards Uwe