[OT] TC development

[icfu]

IF CPU usage goes to 100% you may face concequences with scripts catching key presses. It can also be that the game itself uses techniques that conflict with AutoHotkey. As I don't know what you mean with "didn't work as it should" I don't know what's the problem here.

I suppose the better solution of the subj problem would be a new "wincmd.ini" file parameter AlwaysCopyMoveInQueue. What do you think?

This is obvious, every internal solution is probably better but I use AutoHotkey because I have no time and no desires to wait years till basic stuff like that is implemented.
Small glitches like that should be solved instantly on request and a new TC version with small enhancements should be released every week. If eMule or Bittorrent would be used for spreading these intermediate versions, bandwidth would also be no problem.
So, this is how I would do it if I was the developer.

Icfu

[mod]This thread has been split from AlwaysCopyInBackground and queue.

Hacker (Moderator)[/mod]

[SanskritFritz]

So, this is how I would do it if I was the developer.

I'm sure you have read the Cathedral and the Bazaar

[Lefteous]

2icfu

So, this is how I would do it if I was the developer.

Developers are just a group of many in the process of producing interactive products. Do you really think developers should (and actually) decide how often a new version should be released?

[icfu]

I'm sure you have read the Cathedral and the Bazaar

Nope, what is that? I'll check later.

Developers are just a group of many in the process of producing interactive products. Do you really think developers should (and actually) decide how often a new version should be released?

I am talking about TC which has only one developer and regarding my knowledge he is the only one deciding if a new version is released or not. Is that wrong, what do you think or know?

It's for sure not the users who decide, right?

Icfu

[Lefteous]

2icfu

what do you think or know?

I thought your statement above was more general not just about TC. But even in case of TC there are really many people involved but of course only one decides when a new version is released. If a betatester finds a serious bug a new release will be delayed. There are more factors which influence release dates.

Releasing a new version each week is quite difficult but maybe you have more experience?
Releasing just bugfix versions often is relatively easy. The problem is to work on a new bugfix version and a new major version at the same time. That means you have two branches were you have to watch that bug fixes are included in both versions but new features that require for exampe translations are only included in the major branch.
The overhead to develop this way is quite huge. If you don't have a versions management system it's even more work.
Actually there are two branches of TC at the moment as you probably know.

[SanskritFritz]

The problem is to work on a new bugfix version and a new major version at the same time.

Hmm, doesnt this require this:

Actually there are two branches of TC at the moment as you probably know.

So where is the problem if this is not true:

If you don't have a versions management system

I personally use Subversion, but code merging and project branching are techniques as old as TC itself. I quess Christian uses one.

[Lefteous]

2SanskritFritz

where is the problem

I just wrote it's more work especially on projects larger than a auto hotkey script.

[icfu]

But even in case of TC there are really many people involved but of course only one decides when a new version is released.

This is what I mean.

Releasing a new version each week is quite difficult but maybe you have more experience?

Is that you, Lefteous? I smell subversive attacks here so I won't comment on that.
Well, in case you mean my general experience with other developers: Yes, I have more experience. I prefer playing around with intermediate versions, public alpha testing for those who are interested in finding bugs and testing new features to prevent situations like in TC that features are developerd that are found out to be crap after release, like the content plugin configuration.

Releasing just bugfix versions often is relatively easy. The problem is to work on a new bugfix version and a new major version at the same time. That means you have two branches were you have to watch that bug fixes are included in both versions but new features that require for exampe translations are only included in the major branch.
The overhead to develop this way is quite huge. If you don't have a versions management system it's even more work.
Actually there are two branches of TC at the moment as you probably know.

Regarding that 6.54 is a bugfix version this is obvious. The problem is: Why are there no 6.53.x versions? Is it really the right decision to let the people live with annoying glitches for nearly a year? I don't think so.

A final release has to be intensively tested, this is what I like about ghisler, he doesn't release instable crap. But all the fun in between I have when playing around with other software is missing.

Icfu

[Lefteous]

2icfu

Well, in case you mean my general experience with other developers

No I ment managing software projects.

Is it really the right decision to let the people live with annoying glitches for nearly a year?

I don't think that is a decision at all - there is no other way if you are low on human resources.

ike the content plugin configuration

Did you read the thread about the problem in Shelldetails where a user had column shell extensions which installed more than 700 fields? It's really a pleasure to select a field from such a menu? Here is a little movie:
http://www.lefteous.de/tc/videos/damn_menu.sqx
( ~ 1:30 )

all the fun in between I have when playing around with other software is missing

Well I guess you know the answer.

[icfu]

I don't think that is a decision at all - there is no other way if you are low on human resources.

If the bugfix is done, he can release it as an intermediate version for interested users, where is the problem?
In small doses updates are much easier to eat, but when you get bombed with the changes from a whole year you are stressed.
I could go on and on but it leads to nowhere so I just stop it here.

Well I guess you know the answer.

As there was no question I don't know it, sorry.

Edit:

Did you read the thread about the problem in Shelldetails where a user had column shell extensions which installed more than 700 fields? It's really a pleasure to select a field from such a menu? Here is a little movie:
http://www.lefteous.de/tc/videos/damn_menu.sqx
( ~ 1:30 )

Well, if ghisler would tell his users what he plans to do (by releasing test versions), such situations could be prevented and would probably not arise. Everyone can see that this dialog doesn't fit the powerful concept of content plugins. As a standalone app I would have instantly dumped it to the bin. Funny vid, haha.

Icfu

[Sheepdog]

Well I guess you know the answer.

43, of course. Didn't you know?

sheepdog

[icfu]

You are right, of course.

Icfu

[icfu]

@SanskritFritz:
Excellent article, I know someone who should read it.

Icfu

[Lefteous]

2icfu

If the bugfix is done, he can release it as an intermediate version for interested users, where is the problem?

Let me answer with a question. Do you think 6.54 has delayed 7.0 yes or no? What would happen to new major releases when he would do that constantly?

In small doses updates are much easier to eat, but when you get bombed with the changes from a whole year you are stressed.

On one hand it's good to have always the latest bugfixes on the other hand installing updates all the time is just nasty - but this is just the user perspective.
The best would be a software completely without bugs (no bugfix releases necessary) but only major updates. Honestly TC is not so far away from that.

As there was no question I don't know it, sorry.

Well if you don't want to be a betatester.

[icfu]

Let me answer with a question. Do you think 6.54 has delayed 7.0 yes or no? What would happen to new major releases when he would do that constantly?

This question is strange because 6.54 is a bugfix release and 7.0 will bring new features, new bugs and includes the bugfixes.

As parallel development is the only way anyway, I really don't care which one delays what, but I definetely know that fixes that are hold back for one year is the wrong way.

Give the users two versions to play with: one stable one with fixes only and call the other one a technical preview or whatever. They will be more happy than now, at least this is true for me.

On one hand it's good to have always the latest bugfixes on the other hand installing updates all the time is just nasty - but this is just the user perspective.

If it's too nasty for YOU, don't install them. I LIKE updating, I am a software junkie.

The best would be a software completely without bugs (no bugfix releases necessary) but only major updates. Honestly TC is not so far away from that.

This is true, but it's not really a compliment because it effectively means that development has stopped.

Well if you don't want to be a betatester.

I have no interest to wait six months or even longer till I am allowed to find the bugs in features I never wanted (at least THAT way) and to know that my influence on the features of the final release is nearly at 0K(elvin).
I don't like being exploited and as soon as a developer just needs me for bug testing he should ask someone else, not me.

Icfu