rails-devel

[Rails-devel] Progress on 1880 merge into Rails 2.0, but still open issues

[Stefan F. <ste...@we...>]

I have made some more progress to remove bugs that still exist in 1880 
from the merge into Rails 2.0. Most of them were related to the fact 
that executed actions are checked against allowed actions.

However it seems that the code base used to merge 1880 into Rails 2.0 
did not include everything up-to Rails 1.8.7:

Especially some svg tiles are missing and the xml definitions deviate.
I do not know exactly how this could happen, as they should have been 
identical at the start of the merge.

In addition there are a lot of keys missing in LocalisedText.properties
for 1880, which is not the case in Rails 1.8.7.

An example what is missing are the svg-files for tiles 455,-9021,-9031.

Those belonged to commit 3f9305 from two years ago, which was based on 
a patch from Martin commited by Erik.

However I cannot simply cherry-pick this commit, as it also contains 
many changes to the java classes, which later on have changed again.

Could anyone of you double-check what happened and give an indication 
which patches/components are still missing?

Otherwise I will simply add those files that are obviously missing (svg) 
and will compare the *.xml definition files and LocalisedText properties 
and edit them manually.

===> Imporant remark:
Please keep in mind, that I will only do this once 1880. I will not do 
this for any future game that is developed from now on under Rails 1.x.

This is a lot of work to do to narrow down all those little bits and now 
that Rails 2.0 is really close to finalize, there is no reason to keep 
working on Rails 1.x.

And I for sure do not want to repeat this experience. ;-)

To make things easier, I will give more information what are best 
practices to implement a new game in Rails 2.0. However I want to have 
Rails 2.0 out of the door, before I seriously start writing about it.
In the meantime best thing to do, is simply asking.

Stefan

Re: [Rails-devel] Progress on 1880 merge into Rails 2.0, but still open issues

[Martin B. <dr....@t-...>]

Hi Stefan,

i'll try later or tomorrow to fix the svg and xml discrepancies as well
as the Localized Text.

Kind Regards,
Martin

Am 22.03.2014 11:27, schrieb Stefan Frey:
> I have made some more progress to remove bugs that still exist in 1880 
> from the merge into Rails 2.0. Most of them were related to the fact 
> that executed actions are checked against allowed actions.
>
> However it seems that the code base used to merge 1880 into Rails 2.0 
> did not include everything up-to Rails 1.8.7:
>
> Especially some svg tiles are missing and the xml definitions deviate.
> I do not know exactly how this could happen, as they should have been 
> identical at the start of the merge.
>
> In addition there are a lot of keys missing in LocalisedText.properties
> for 1880, which is not the case in Rails 1.8.7.
>
> An example what is missing are the svg-files for tiles 455,-9021,-9031.
>
> Those belonged to commit 3f9305 from two years ago, which was based on 
> a patch from Martin commited by Erik.
>
> However I cannot simply cherry-pick this commit, as it also contains 
> many changes to the java classes, which later on have changed again.
>
> Could anyone of you double-check what happened and give an indication 
> which patches/components are still missing?
>
> Otherwise I will simply add those files that are obviously missing (svg) 
> and will compare the *.xml definition files and LocalisedText properties 
> and edit them manually.
>
> ===> Imporant remark:
> Please keep in mind, that I will only do this once 1880. I will not do 
> this for any future game that is developed from now on under Rails 1.x.
>
> This is a lot of work to do to narrow down all those little bits and now 
> that Rails 2.0 is really close to finalize, there is no reason to keep 
> working on Rails 1.x.
>
> And I for sure do not want to repeat this experience. ;-)
>
> To make things easier, I will give more information what are best 
> practices to implement a new game in Rails 2.0. However I want to have 
> Rails 2.0 out of the door, before I seriously start writing about it.
> In the meantime best thing to do, is simply asking.
>
> Stefan
>

Re: [Rails-devel] Progress on 1880 merge into Rails 2.0, but still open issues

[Stefan F. <ste...@we...>]

Okay, perfect. The most current update of my code is in sfy_1880.
It runs your test game until the end, however it fails to build the map 
as tile 455 is missing.
Thanks,
Stefan


On 03/22/2014 04:03 PM, Martin Brumm wrote:
> Hi Stefan,
>
> i'll try later or tomorrow to fix the svg and xml discrepancies as well
> as the Localized Text.
>
> Kind Regards,
> Martin
>
> Am 22.03.2014 11:27, schrieb Stefan Frey:
>> I have made some more progress to remove bugs that still exist in 1880
>> from the merge into Rails 2.0. Most of them were related to the fact
>> that executed actions are checked against allowed actions.
>>
>> However it seems that the code base used to merge 1880 into Rails 2.0
>> did not include everything up-to Rails 1.8.7:
>>
>> Especially some svg tiles are missing and the xml definitions deviate.
>> I do not know exactly how this could happen, as they should have been
>> identical at the start of the merge.
>>
>> In addition there are a lot of keys missing in LocalisedText.properties
>> for 1880, which is not the case in Rails 1.8.7.
>>
>> An example what is missing are the svg-files for tiles 455,-9021,-9031.
>>
>> Those belonged to commit 3f9305 from two years ago, which was based on
>> a patch from Martin commited by Erik.
>>
>> However I cannot simply cherry-pick this commit, as it also contains
>> many changes to the java classes, which later on have changed again.
>>
>> Could anyone of you double-check what happened and give an indication
>> which patches/components are still missing?
>>
>> Otherwise I will simply add those files that are obviously missing (svg)
>> and will compare the *.xml definition files and LocalisedText properties
>> and edit them manually.
>>
>> ===> Imporant remark:
>> Please keep in mind, that I will only do this once 1880. I will not do
>> this for any future game that is developed from now on under Rails 1.x.
>>
>> This is a lot of work to do to narrow down all those little bits and now
>> that Rails 2.0 is really close to finalize, there is no reason to keep
>> working on Rails 1.x.
>>
>> And I for sure do not want to repeat this experience. ;-)
>>
>> To make things easier, I will give more information what are best
>> practices to implement a new game in Rails 2.0. However I want to have
>> Rails 2.0 out of the door, before I seriously start writing about it.
>> In the meantime best thing to do, is simply asking.
>>
>> Stefan
>>
>
>
>
> ------------------------------------------------------------------------------
> Learn Graph Databases - Download FREE O'Reilly Book
> "Graph Databases" is the definitive new guide to graph databases and their
> applications. Written by three acclaimed leaders in the field,
> this first edition is now available. Download your free book today!
> http://p.sf.net/sfu/13534_NeoTech
>
>
>
> _______________________________________________
> Rails-devel mailing list
> Rai...@li...
> https://lists.sourceforge.net/lists/listinfo/rails-devel
>

Re: [Rails-devel] Progress on 1880 merge into Rails 2.0, but still open issues

[Martin B. <dr....@t-...>]

Am 22.03.2014 16:03, schrieb Martin Brumm:
Hi Stefan,

the svg files should now be upto date.

I couldnt find differences in the xml-files though ?

REgards,
Martin

Re: [Rails-devel] Progress on 1880 merge into Rails 2.0, but still open issues

[Stefan F. <ste...@we...>]

Martin:
perfect, thanks.
There are only a few changes in map.xml, that were missing in sfy_1880.
All other diffs are due to my changes.
I will take care of those myself tomorrow.
Stefan


On 04/04/2014 08:10 PM, Martin Brumm wrote:
> Am 22.03.2014 16:03, schrieb Martin Brumm:
> Hi Stefan,
>
> the svg files should now be upto date.
>
> I couldnt find differences in the xml-files though ?
>
> REgards,
> Martin
>
>
>
> ------------------------------------------------------------------------------
>
>
>
> _______________________________________________
> Rails-devel mailing list
> Rai...@li...
> https://lists.sourceforge.net/lists/listinfo/rails-devel
>

Re: [Rails-devel] Progress on 1880 merge into Rails 2.0, but still open issues

[Stefan F. <ste...@we...>]

I merged your changes to my sfy_1880 and also added the remaining 
changes from map.xml and LocalisedText in rails_1.8.x to sfy_1880.

However there are a few issues remaining:

* The StockMarket.xml and StockMarket implementation seems changed in 
rails_1.8.x. I have to take a look at this.

* What is the reason for all those city labels that start with "test" in 
map.xml?

* At least on the gray Bejing tile the tokens are all plotted in the 
middle, so only one token is displayed.

* The blue bars for water blocked hex sides appear to be wrong on hex E1.

As soon as those issues are fixed (assuming they are minor), an Alpha 5 
can be published.


On 04/04/2014 08:41 PM, Stefan Frey wrote:
> Martin:
> perfect, thanks.
> There are only a few changes in map.xml, that were missing in sfy_1880.
> All other diffs are due to my changes.
> I will take care of those myself tomorrow.
> Stefan
>
>
> On 04/04/2014 08:10 PM, Martin Brumm wrote:
>> Am 22.03.2014 16:03, schrieb Martin Brumm:
>> Hi Stefan,
>>
>> the svg files should now be upto date.
>>
>> I couldnt find differences in the xml-files though ?
>>
>> REgards,
>> Martin
>>
>>
>>
>> ------------------------------------------------------------------------------
>>
>>
>>
>> _______________________________________________
>> Rails-devel mailing list
>> Rai...@li...
>> https://lists.sourceforge.net/lists/listinfo/rails-devel
>>
>
> ------------------------------------------------------------------------------
> _______________________________________________
> Rails-devel mailing list
> Rai...@li...
> https://lists.sourceforge.net/lists/listinfo/rails-devel
>

Re: [Rails-devel] Progress on 1880 merge into Rails 2.0, but still open issues

[Martin B. <dr....@t-...>]

Am 09.04.2014 11:56, schrieb Stefan Frey:
> I merged your changes to my sfy_1880 and also added the remaining 
> changes from map.xml and LocalisedText in rails_1.8.x to sfy_1880.
>
> However there are a few issues remaining:
>
> * The StockMarket.xml and StockMarket implementation seems changed in 
> rails_1.8.x. I have to take a look at this.
>
> * What is the reason for all those city labels that start with "test" in 
> map.xml?
>
> * At least on the gray Bejing tile the tokens are all plotted in the 
> middle, so only one token is displayed.
>
> * The blue bars for water blocked hex sides appear to be wrong on hex E1.
>
> As soon as those issues are fixed (assuming they are minor), an Alpha 5 
> can be published.
>
>
> On 04/04/2014 08:41 PM, Stefan Frey wrote:
>> Martin:
>> perfect, thanks.
>> There are only a few changes in map.xml, that were missing in sfy_1880.
>> All other diffs are due to my changes.
>> I will take care of those myself tomorrow.
>> Stefan
>>
>>
>>
Hi Stefan,

we defined an additional field value to the StockspaceType. Any field
can have an Revenue value that will be used in 1880 by a dynamic Revenue
Modifier to add the value on the stockmarket to the revenue generated.

Relict will be fixed tonight in map.xml

Hm ? the svg tile for bejing looks correct ? at least in rails1.8.x its
the correct one. i have fixed that in my development branch..

It should be on the right side of E11 and on the left side of E13  ?
Theres no tile E1 on the map ?

One issue: the medium city large circle around a black dot are a
speciality of 1880 and can be upgrade into a large city or downsized to
a small stop. The game engine sometimes doesnt recognise the true
function of the halt/stop if its upgraded or downgraded.

Regards
Martin

Re: [Rails-devel] Progress on 1880 merge into Rails 2.0, but still open issues

[Stefan F. <ste...@we...>]

Martin,
Thanks for your replies, see my comments below.
I have attached screenshots of both rails 1/2 for comparison of the 
visual map effects.
Stefan

> we defined an additional field value to the StockspaceType. Any field
> can have an Revenue value that will be used in 1880 by a dynamic Revenue
> Modifier to add the value on the stockmarket to the revenue generated.

I did not refer to the revenue value, but to the number of slots that 
can be used to start used at a par price to start a company.
This seems to be not correct in rails 2.0, there are the variables 
parPlace_100 ... and they are not used anywhere. Maybe they are a 
relict. And there is the parslot attribute of startspace which is never 
read? It might be correct in 1.8.7, so it most have somehow got lost in 
during the merge.

>
> Relict will be fixed tonight in map.xml
>
> Hm ? the svg tile for bejing looks correct ? at least in rails1.8.x its
> the correct one. i have fixed that in my development branch..

The svg is correct, however the tokens are not displayed correctly.
Most likely there is code missing from rails1.8.7 that painted the 
tokens correctly.

>
> It should be on the right side of E11 and on the left side of E13  ?
> Theres no tile E1 on the map ?

Sorry I meant E11. It shows the blue bars left and right (the one to E9 
is incorrect). This might be related to the double-definition (in E11 
and both E13, this is not required), however it should not cause to show 
up incorrect.

>
> One issue: the medium city large circle around a black dot are a
> speciality of 1880 and can be upgrade into a large city or downsized to
> a small stop. The game engine sometimes doesnt recognise the true
> function of the halt/stop if its upgraded or downgraded.

What do you mean game engine? The revenue calculation engine? Or for 
token lays? In Rails 1.8.7 or in Rails 2?

>
> Regards
> Martin
>
>
>
> ------------------------------------------------------------------------------
> Put Bad Developers to Shame
> Dominate Development with Jenkins Continuous Integration
> Continuously Automate Build, Test & Deployment
> Start a new project now. Try Jenkins in the cloud.
> http://p.sf.net/sfu/13600_Cloudbees
>
>
>
> _______________________________________________
> Rails-devel mailing list
> Rai...@li...
> https://lists.sourceforge.net/lists/listinfo/rails-devel
>

Re: [Rails-devel] Progress on 1880 merge into Rails 2.0, but still open issues

[Martin B. <dr....@t-...>]

Am 10.04.2014 12:34, schrieb Stefan Frey:
> Martin,
> Thanks for your replies, see my comments below.
> I have attached screenshots of both rails 1/2 for comparison of the
> visual map effects.
> Stefan
>
Hi Stefan,
The list scratches attachments.
>> we defined an additional field value to the StockspaceType. Any field
>> can have an Revenue value that will be used in 1880 by a dynamic Revenue
>> Modifier to add the value on the stockmarket to the revenue generated.
>
> I did not refer to the revenue value, but to the number of slots that
> can be used to start used at a par price to start a company.
> This seems to be not correct in rails 2.0, there are the variables
> parPlace_100 ... and they are not used anywhere. Maybe they are a
> relict. And there is the parslot attribute of startspace which is
> never read? It might be correct in 1.8.7, so it most have somehow got
> lost in during the merge.
>

This has been changed by Mike, ParslotModel and ParSlotManager are used
to achieve that now. SO yes that code is a relict imho.

>>
>> Relict will be fixed tonight in map.xml
>>
>> Hm ? the svg tile for bejing looks correct ? at least in rails1.8.x its
>> the correct one. i have fixed that in my development branch..
>
> The svg is correct, however the tokens are not displayed correctly.
> Most likely there is code missing from rails1.8.7 that painted the
> tokens correctly.
>

Let me think again. The basic code didnt support 6 Token cities...

Commit IDs: 6744707 and 49f89e3

made changes to GuiHex.java to support 6 token cities.


>>
>> It should be on the right side of E11 and on the left side of E13  ?
>> Theres no tile E1 on the map ?
>
> Sorry I meant E11. It shows the blue bars left and right (the one to
> E9 is incorrect). This might be related to the double-definition (in
> E11 and both E13, this is not required), however it should not cause
> to show up incorrect.
>
>>
>> One issue: the medium city large circle around a black dot are a
>> speciality of 1880 and can be upgrade into a large city or downsized to
>> a small stop. The game engine sometimes doesnt recognise the true
>> function of the halt/stop if its upgraded or downgraded.
>
> What do you mean game engine? The revenue calculation engine? Or for
> token lays? In Rails 1.8.7 or in Rails 2?
>
The revenue engine for sure and sometimes of course a subsequent
sideeffect the code trying to find valid connections for the investors.
And i am not particularly sure that the Tile laying Code suggesting new
tiles was handling that correctly also all the time. (And yes thats in
Rails1.8.x)

>>
>> Regards
>> Martin
>>
>>
>>

Regards,
Martin

Re: [Rails-devel] Progress on 1880 merge into Rails 2.0, but still open issues

[Stefan F. <ste...@we...>]

Thanks again Martin, see further comments below.

> This has been changed by Mike, ParslotModel and ParSlotManager are used
> to achieve that now. SO yes that code is a relict imho.

OK I understand. On first glance I have to rewrite the ParSlotModel 
anyway to confirm the new Rails 2.0 approach (and move it to the 
specific_1880 package). And most likely not all of ParSlotManager is 
undo-proof.
However this will be done for the next alpha.

> Let me think again. The basic code didnt support 6 Token cities...
>
> Commit IDs: 6744707 and 49f89e3
>
> made changes to GuiHex.java to support 6 token cities.

So I manually copied the relevant code changes from a diff between 1.8.x 
and sfy_1880 into my branch. On first glance it seems to achieve the 
effects. Cherry-pick unfortunately does not work in this occasion.

> The revenue engine for sure and sometimes of course a subsequent
> sideeffect the code trying to find valid connections for the investors.
> And i am not particularly sure that the Tile laying Code suggesting new
> tiles was handling that correctly also all the time. (And yes thats in
> Rails1.8.x)

To find those problems I will need more details on concrete test cases.
Maybe you could check the alpha from today how it currently works in 
Rails 2.x.