Uppps... cut and paste error. I've repaired the link for the time being, I'll set up a public git repo later.
Sweet! One minor thing: please omit the part of the patch that modifies build.sh, it's been obsolete since commit 38b320f9 (please update your repo).
Update: please commit your changes in atomic, topic-based commits. This way people can then cherry-pick the changes they want and ignore others (e.g. the RA/DEC label change).
One minor thing: please omit the part of the patch that modifies build.sh, it's been obsolete since commit 38b320f9
Done, but actually I liked the idea of specifying the exact revision/date of mingw (probably also boinc) to download in the build.sh, makes the build process much more reproducible and will avoid repetition of the problem where changes in the downloaded tools/libs will break the patches that are included in the starsphere repos. Just like it's done in the Einstein@Home build script for the apps wrt. LAL and BOINC.
Update: please commit your changes in atomic, topic-based commits. This way people can then cherry-pick the changes they want and ignore others (e.g. the RA/DEC label change).
Done, but actually I liked the idea of specifying the exact revision/date of mingw (probably also boinc) to download in the build.sh, makes the build process much more reproducible and will avoid repetition of the problem where changes in the downloaded tools/libs will break the patches that are included in the starsphere repos. Just like it's done in the Einstein@Home build script for the apps wrt. LAL and BOINC.
I agree with you as long as we're talking about projects that do have a working release/maintenance process :-) In general the GFX build script only uses fixed release tarballs or (even better) branches, hence only bug-fixes should be added to those branches which, in general, shouldn't break stuff. In case of BOINC and LAL the release processes have some considerable deficiencies which sometimes lead to problems unfortunately. However, using fixed date-based checkouts means that you don't get those bugfixes unless you monitor all 3rd party packages and update your build script on a regular basis. In addition to that, LAL is a very specific case and, as it isn't used by the GFX framework, we don't have to make provisions for correct/reproducible builds of it...
WRT MinGW, the problems we've experienced over the past few weeks were either caused by internal issues of the MinGW project (using date tags wouldn't have prevented that) or by their preparations of the 1.0 xscripts release. As soon as that's done I'm going to use their release branch. By the way, the package versions of MinGW as such have been, in fact, constant all the time - we simply used fresh copies of the MinGW cross-compile harness (xscripts).
If I minimize the starsphere graphics window, most of the time the mouse pointer gets corrupted (it's set to a random pattern), and stays that way.
Just to be sure, what exactly did you do? Do you mean minimizing a given starsphere window or changing it from fullscreen to windowed mode? The latter action involves changing the mouse pointer's visibility setting (hide/show) but none of the actions mentioned above change its appearance... Strange...
Update: I tried it myself using a Win XP Pro SP3 VM image, nothing unsusual though. However, I noticed that SDL seems to change/invert the mouse pointer's appearance as soon as one hovers the window's client area, but I don't see how this could be related to the issue you observed when minimizing the window.
Hi folks!
How are you doing?
I have a question related to (I think so anyway) the open source characteristic attached to BOINC software and projects:
I use sometimes cyrillic fonts (Russian alphabet) for some of my posts, since I am a Russian language learner.
I have realized some months ago that there´s no way to read and write correctly using russian fonts, because some little funny looking signs appear and make reading texts quite annoying. I bet you have found now and then somewhere those signs when viewing pages of other not so common visited countries/webpages.
I thought maybe some changes have been made in some folders related to those fonts.
This might be a question for programmers and webpage developers.
What do you think? Any experiences with unusual signs on your screen before?
Thanks for your help.
Cheers!
Hi Bikeman, RE: ...
)
Hi Bikeman,
The link to your patch is broken (local URL)... Use a public repo ;-)
Best,
Oliver
Einstein@Home Project
RE: Hi Bikeman, RE: ...
)
Uppps... cut and paste error. I've repaired the link for the time being, I'll set up a public git repo later.
RE: Uppps... cut and paste
)
Sweet! One minor thing: please omit the part of the patch that modifies build.sh, it's been obsolete since commit 38b320f9 (please update your repo).
Update: please commit your changes in atomic, topic-based commits. This way people can then cherry-pick the changes they want and ignore others (e.g. the RA/DEC label change).
Oliver
Einstein@Home Project
RE: One minor thing:
)
Done, but actually I liked the idea of specifying the exact revision/date of mingw (probably also boinc) to download in the build.sh, makes the build process much more reproducible and will avoid repetition of the problem where changes in the downloaded tools/libs will break the patches that are included in the starsphere repos. Just like it's done in the Einstein@Home build script for the apps wrt. LAL and BOINC.
RE: Update: please commit
)
Yeah, unlike BOINC! :)
RE: Done, but actually I
)
I agree with you as long as we're talking about projects that do have a working release/maintenance process :-) In general the GFX build script only uses fixed release tarballs or (even better) branches, hence only bug-fixes should be added to those branches which, in general, shouldn't break stuff. In case of BOINC and LAL the release processes have some considerable deficiencies which sometimes lead to problems unfortunately. However, using fixed date-based checkouts means that you don't get those bugfixes unless you monitor all 3rd party packages and update your build script on a regular basis. In addition to that, LAL is a very specific case and, as it isn't used by the GFX framework, we don't have to make provisions for correct/reproducible builds of it...
WRT MinGW, the problems we've experienced over the past few weeks were either caused by internal issues of the MinGW project (using date tags wouldn't have prevented that) or by their preparations of the 1.0 xscripts release. As soon as that's done I'm going to use their release branch. By the way, the package versions of MinGW as such have been, in fact, constant all the time - we simply used fresh copies of the MinGW cross-compile harness (xscripts).
Cheers,
Oliver
Einstein@Home Project
Hi! I'm encountering a
)
Hi!
I'm encountering a strange problem under Windows (Windows XP Home, SP 2).
If I minimize the starsphere graphics window, most of the time the mouse pointer gets corrupted (it's set to a random pattern), and stays that way.
Anybody else seeing this?
CU
Bikeman
RE: If I minimize the
)
Just to be sure, what exactly did you do? Do you mean minimizing a given starsphere window or changing it from fullscreen to windowed mode? The latter action involves changing the mouse pointer's visibility setting (hide/show) but none of the actions mentioned above change its appearance... Strange...
Update: I tried it myself using a Win XP Pro SP3 VM image, nothing unsusual though. However, I noticed that SDL seems to change/invert the mouse pointer's appearance as soon as one hovers the window's client area, but I don't see how this could be related to the issue you observed when minimizing the window.
Oliver
Einstein@Home Project
RE: Just to be sure, what
)
The former.
I've updated to SP3 now and will try again.
EDIT: My modifications are now hosted here, as suggested by Oliver:
http://wiki.github.com/Bikeman/eah_gfx_ext
CU
Bikeman
Hi folks! How are you
)
Hi folks!
How are you doing?
I have a question related to (I think so anyway) the open source characteristic attached to BOINC software and projects:
I use sometimes cyrillic fonts (Russian alphabet) for some of my posts, since I am a Russian language learner.
I have realized some months ago that there´s no way to read and write correctly using russian fonts, because some little funny looking signs appear and make reading texts quite annoying. I bet you have found now and then somewhere those signs when viewing pages of other not so common visited countries/webpages.
I thought maybe some changes have been made in some folders related to those fonts.
This might be a question for programmers and webpage developers.
What do you think? Any experiences with unusual signs on your screen before?
Thanks for your help.
Cheers!