R2-Beta - Priorities (All Users)

Return to Index Page
Most recent messages (500 max) are listed first.

#UserMessageDate
191RobertIMO we should try to get a 2.7.7 release done as Carl pointed out in his Blog.8-Apr-09 7:07
190RobertIIRC there are some GUI bugs that we can fix and some mezzanine bugs.8-Apr-09 7:07
189RobertI have lost track about the 2.7.7 priorities.8-Apr-09 7:07
188GreggPARSE: I think where it gave up makes sense. Not sure what else it would do.

CALL/SHOW: What is wrong is that the refinement should have been /no-show (or something). The change reversed the existing behavior and broke code. Since my last complaint, I've made an expansive sweep and changed a lot of code to work with it (since I didn't think it would be fixed at this point). So, ironically, I don't think I want you to change it now. Just make it a lesson about changing existing behavior.

26-Mar-09 0:34
187CarlSo, we need a release manager.24-Mar-09 22:21
186Carlhttp://www.rebol.com/article/0403.html24-Mar-09 22:21
185CarlRobert, there is not a lot of time for R2, however, releases can be made if everyone else wants to help with it, and all I do is the build, or any internal changes.24-Mar-09 21:59
184CarlCALL/SHOW changes: can be fixed, but we need a precise test to show what is wrong, and to confirm it is fixed.24-Mar-09 21:58
183CarlG: PARSE backtracks a lot, so what kind of indication do you want, just the stream point where it gave up?24-Mar-09 21:57
182RobertAny real chance to get a 2.7.7 or 2.8.0 release?15-Mar-09 11:18
181GreggI asked on 26-Dec, and I'll ask again (having just found another script broken by it). Is there any plan, or any hope for CALL/SHOW to be changed, so the default behavior is the same as < 2.7.6 releases?9-Mar-09 5:20
180GreggI've asked about this before, but I'll ask again. Otherwise I have to roll my own18-Feb-09 6:17
179GreggHow hard would it be to have parse set a marker so, if it fails, you can see *where* it failed (how far along in the input it got)?18-Feb-09 6:16
178BrianHAt the very least, I would want a 2.7.7 to have a version that fixes post-2.7.6 mezzanine bugs, and 2.7 series regressions vs. 2.6.3.8-Jan-09 22:20
177BrianHWe've been backporting R3's new natives as mezzanines when possible, which means when they don't require R3's changes to the basic semantic model. We can hope that's enough.8-Jan-09 22:18
176BrianHFor me, the big question is what kind of release we will be doing: - 2.7.7: Patching glaring bugs in a few natives, VID fixes, and continuing the backports and mezzanine fixes. - 2.8.0: Backporting some of the R3 native changes (function, not infrastructre), and the above. I think that the decision a long time ago was to focus on R3 as a priority, and just patch up R2 as necessary.8-Jan-09 22:13
175BrianHYeah, we never really had the thorough discussion of FORMAT that so many of the other new functions got. I still consider it a work in process in R3 - that's why I didn't backport it yet.8-Jan-09 22:07
174GreggI would still like to see a discussion about what the standard FORMAT func should do. The one from R3, IIRC, simply pads values for columar output. That seems a very specific case for a very general word.2-Jan-09 17:47
173RobertI would like to see a 2.7.7 too :-)2-Jan-09 16:35
172BTiffinI'd like to see a 2.7.7 soon, just to give people a sense that RT cares about the public folder pollution on Windows. Or, (and I'm a fan of conspiracies) is it a secret plan to gently shove rebols towards a real operating system? Well, that and I'd like to see the FORMAT mezzanine released to the general population out of the box.1-Jan-09 9:09
171GrahamI use call in 2.7.6 as it works better for me than previous versions ie. it didn't work on Linux.27-Dec-08 2:23
170GrahamIs 2.7.7 still on the table??27-Dec-08 2:23
169GreggI think I've brought this up before, but I'll mention it again. Is there any plan to change the 2.7.6 call/show refinement to call/no-show? i.e. the default behavior would be the same as for <= 2.7.5, with the option to not show the window via the refinement.

I don't know about anyone else, but this change keeps me from using 2.7.6 in production because it breaks things.

26-Dec-08 17:43
168Grahamis OpenSolaris gaining any traction ?30-May-08 7:47
167TomChttp://www.rebol.com/platforms.html27-May-08 6:53
166TomCyes27-May-08 6:52
165GrahamWas there ever a view version for Solaris?24-May-08 8:23
164CarlNext is Linux. And will attempt solaris at same time.14-Mar-08 23:08
163TomCyes please9-Mar-08 17:35
162CarlCore?9-Mar-08 17:20
161TomCcan we discuss a Solaris version. If need be I could work on building it.9-Mar-08 6:02
160CarlI am typing this message from AltME 16 running on 2.7.6 on Linux. A cool thing.9-Mar-08 3:55
159BrianHI have to do some family stuff this afternoon, so I will be able to make any mezzanine fixes this evening. I am not aware of any errors in them though, so make test code if you find any and post the code in the Mezzanines group.8-Mar-08 17:45
158CarlGoal is to release later today, but we can delay if you want.8-Mar-08 17:41
157CarlUpdate on status...

1. Brent is coordinating 2.7.6 testing and docs, with various developers helping out.

2. BrainH is making a few last minute fixes to mezz.

3. I'm working on Linux and OSX bugs. No time estimate known yet (debugging can be more of a hassle using gdb ;)

8-Mar-08 17:41
156Maarten(sorrty wrong group)7-Mar-08 21:50
155MaartenRight now I'm enbaseing, shifting strings and back again.7-Mar-08 21:50
154MaartenBasically, whjat we know from << and >> as operators on binaries in other languages. Not polymorphic in REBOL, but one of the few things really missing.7-Mar-08 21:49
153MaartenOnly for binaries: bitshift: func [ b [binary!] {Shifts the binary by 1 bit to the right} /left {Shift to the left}7-Mar-08 21:48
152MaartenTracker # 37.7-Mar-08 21:47
151BrentI think 4263 is it? http://www.rebol.net/cgi-bin/rambo.r?id=4263&7-Mar-08 21:46
150CarlIf not, can someone post small piece of test code for it. Such things save me just a bit of time.7-Mar-08 21:44
149CarlIs there a RAMBO for #20?7-Mar-08 21:43
148CarlNo, must be done after they are merged here. Otherwise, I'll not be able to know I have them done.7-Mar-08 21:31
147BrianHDo I mark the ones that have been implemented as built, or do you after you have accepted the DevBase submissions?7-Mar-08 21:30
146Carl(Tracker changes can superset RAMBO)7-Mar-08 21:29
145CarlBrianH: RAMBO entries are not required for those.7-Mar-08 21:29
144CarlMaarten: can you spec them in the tracker here?7-Mar-08 21:28
143BrianHCan you do them today? If not, perhaps you should ask later, after the 2.7.6 release. Put it in the tracker.7-Mar-08 21:25
142Maarten....and give us something to really test ;-)7-Mar-08 21:24
141MaartenTheyw ere there in rebcode, they are highly usable (think protocol implementations) and they would add a lot of value to REBOL.7-Mar-08 21:24
140MaartenI have a priority I know a lot of people share: getting bitt shift and rotate operators for binary! types.7-Mar-08 21:23
139BrianHNone of these have RAMBO entries, as the ideas either came from Carl's blog, DevBase source or the R3 alpha list.7-Mar-08 21:01
138BrianHDone so far: - Tracker #s 25 through 33 are done already and in DevBase - see the response section for references. - Tracker #s 34 through 36 have not been done yet, mostly because they are better done as natives. If you want mezzes, say so.7-Mar-08 20:59
137Brentyou want me to go through something in DevBase and add tracker items?7-Mar-08 19:31
136CarlAlso, maybe you can have someone else help you on it?7-Mar-08 19:31
135CarlOk, great.7-Mar-08 19:30
134BrianHI've been busy. I'm listing them now.7-Mar-08 19:30
133CarlAll changes must be listed in Tracker - even mezz changes.7-Mar-08 19:30
132CarlNP. So, there are no mezz update requests?7-Mar-08 19:29
131Brentoops read priority # instead of bug #. Sorry.7-Mar-08 19:29
130CarlOk thanks for correction. Also, that's actually bug #5.7-Mar-08 19:28
129CarlAll submissions reviewed.7-Mar-08 19:25
128Brent#11 should be RAMBO #4192, not #4129. not sure if this gives you more specifics like you need: http://www.rebol.net/cgi-bin/rambo.r?id=4192&7-Mar-08 19:25
127Carl#11 is non specific. I agree with the general idea, but I want to know specifically what change is requsted.7-Mar-08 19:22
126CarlStill updating it.7-Mar-08 19:21
125CarlRather than post each one here... just look at Tracker, and select Active Priorities sort order.7-Mar-08 19:21
124Carl#8 also.7-Mar-08 19:16
123Carl#20 - Pending.7-Mar-08 19:15
122Carl#18 - PENDING - scheduled.7-Mar-08 19:14
121Carl#19 - vectorial fonts in Linux - changed to WAITING, here's why:

I'm not 100% sure why this doesn't work. So, requires Cyphre to do it.

7-Mar-08 19:12
120CarlDitto #227-Mar-08 19:10
119CarlStatus of #4 changed to PENDING - scheduled.7-Mar-08 19:09
118CarlOk. But, move the discussion to Mezzanines group. Thanks.7-Mar-08 19:09
117BrianHThe reason I would add potential backports to the tracker is so we could decide which would be good ideas.7-Mar-08 19:08
116Carl(meaning - it is scheduled for 2.7.6 release)7-Mar-08 19:08
115CarlStatus of #10 changed to PENDING.7-Mar-08 19:08
114CarlStatus of #9 changed to WAITING.7-Mar-08 19:06
113CarlAt one point, a couple years ago, Cyphre even took a look at it.

My guess is the fix is going to require substantial work from someone of that level. It cannot be done today.

To get it fixed, I would suggest we make a bounty for it, and pay Cyphre (or equivalent) to do the work.

7-Mar-08 19:05
112BrianHMost of the backports so far are just changing words to get-words to make more of the built-in functions safe to use with function values. I tested them with R2 as well when I was making the changes for R3. Any that are unknown I haven't ported.7-Mar-08 19:04
111Carl(Let me refer to this RAMBO #4274, #4161 - View / right-align bug as Bug #9 - the tracker #.)7-Mar-08 19:03
110CarlBrianH: only if they are fully tested in R2.7-Mar-08 19:01
109CarlTop of list shows: RAMBO #4274, #4161 - View / right-align bug

This bug fix is of a high level of complexity. I think Jim originally wrote this code. I've posted the source a few times for others to look at it, and it is not simple.

The root problem comes from the fact centering and right-align is algorithmic, but the reverse map is static. To explain:

To get from an xy to a char position you need to find the char line and the char offset. Char lines (Y maps) are cached. But, the first char offsets (X offsets) are not cached. Therefore, to find the X position, we will either need to store those X offsets, or we need to compute them again, on demand.

7-Mar-08 19:00
108BrianHShould I enter the R3 backports into the tracker?7-Mar-08 18:55
107CarlFor discussion related to tracker priorities.7-Mar-08 18:50
106CarlRenamed this group to Priorities.7-Mar-08 18:50
105BrianHI haven't overloaded the item names yet either.3-Mar-08 20:41
104BrianHNo, I can't upload. There are too many security fixes that haven't been accepted yet.3-Mar-08 20:41
103CarlOtherwise we'll conflict with R2 items.3-Mar-08 20:40
102CarlBut, remind me, did you find a way to overload item names?3-Mar-08 20:40
101CarlWell, let me diff to the SDK, but if no diffs, you can go ahead and upload the SDK R2 sources.3-Mar-08 20:40
100CarlSo, the team has to set the priorities, make the fixes, and test them. To degree possible.3-Mar-08 20:39
99BrianHRight. When are you putting the R2 code in DevBase where we can work on it? That way all you would have to do is review and accept.3-Mar-08 20:39
98CarlAlso, keep in mind: for me, I'm only about 20% on R2 project.3-Mar-08 20:38
97CarlLet's fix what we can between now and Friday. Prioritize well. But, not miss Friday deadline.3-Mar-08 20:38
96CarlLet me restate it differently....3-Mar-08 20:37
95BrianHHow about mezzanines equivalent to some of the new natives, at least the ones we don't port directly?3-Mar-08 20:37
94BrianHThat should be an easy fix. I just fixed a regression in ARRAY too.3-Mar-08 20:36
93Carlex: http://www.rebol.com/test/ --> http:/www.rebol.com/test/3-Mar-08 20:35
92CarlE.g. I tried to use R3 CLEAN-PATH recently with some URLs and found it problematic. ;)3-Mar-08 20:34
91BrianHMany of the changes and additions would be easy to port back - some originated as R2 code.3-Mar-08 20:34
90CarlAnd well TESTED.3-Mar-08 20:34
89CarlI would be inclined to do that for ones that are compatible and consistent with R2.3-Mar-08 20:33
88CarlSo, the question has been posed: Do we want to move some of the R3 mezz back into R2?3-Mar-08 20:33
87BrianHLet's hope that we don't see much discussion of bugs here, as there is a seperate group for that.3-Mar-08 20:33
86CarlCreated "chit chat" to replace purpose of ALL.3-Mar-08 20:32
85CarlRequirements.3-Mar-08 20:31
84CarlI'm becoming more adamant that we use group topics wisely... so with that in mind, renaming this group....3-Mar-08 20:31
83CarlSo, this ALL group has been "shanghaied" into requirements.3-Mar-08 20:31
82CarlAlso, one I want badly is D&D of files. But, not sure R2 event system can handle it.3-Mar-08 20:28
81CarlSome of those types of things will port quickly too.3-Mar-08 20:27
80BrianHNot in 2.7.6 is a good answer, though. It may be worth it for 2.8 though.3-Mar-08 20:27
79BrianHWould it be a problem to port some of the new natives to R2 as well, at least the ones that don't require the new architecture? I'm thinking of the speedup additions like FIRST+, TAKE, and APPLY.3-Mar-08 20:26
78CarlWell, as long as we don't need Unicode in R2.... we can fix some stuff. ;)3-Mar-08 20:22
77BrianH:(3-Mar-08 20:21
76CarlWow, this world really heated up fast. (Do I sense neglected developers?)3-Mar-08 20:21
75GrahamDoc, isn't there a bug using title in Encap when encapping Cheyenne or something?3-Mar-08 20:01
74GrahamRambo #3660 ( reported march 2005 )3-Mar-08 19:59
73GrahamIt's easily remedied in windows with a system call though.3-Mar-08 19:57
72GrahamIs anyone concerned that shell windows opened in Encap do not contain the correct window title?3-Mar-08 19:57
71Paul** Script Error: Feature not available in this REBOL3-Mar-08 18:26
70PaulDoc - about 'run. I forgot about 'run and I did hear that it was a wrapper before. We just need to get it enabled then. Currently, the feature is not supported:3-Mar-08 18:26
69DocKimbelCould we see somewhere a changelog of things that were already fixed or added to 2.7.5 ?3-Mar-08 15:42
68Henrikscreen resolution crash is supposedly built in #3382, though.3-Mar-08 15:28
67Henrikthere is also the crash on screen resolution change in windows, which is quite serious3-Mar-08 15:26
66HenrikI don't know if it's fixable, but resizing a linux/OSX shell with rebol console in it, makes the cursor behave badly when moved around.3-Mar-08 15:25
65BrianHREPLACE is getting complex enough that adding one or two more refinements would mean switching some of the calls over to APPLY, which would kill the backwards compatibility. The R2 version would just have to be complex though.3-Mar-08 15:12
64BrianHI was looking at this last night, and there are some R3 mezzanines that I rewrote that are still compatible with R2, such as ARRAY and REPLACE. Most of the optimization fixes are backwards compatible too.3-Mar-08 15:09
63BrianHYes, but perhaps it would be better to wait for the R2 SDK sources to be added to DevBase first and then make the changes there.3-Mar-08 15:06
62Henrikif we can write, test and deliver them finished to Carl, then yes. are they in devbase?3-Mar-08 15:05
61BrianHSome of the new mezzanines already have backports because they were originally ported from DevBase.3-Mar-08 15:05
60BrianHThe request to add /only to REPLACE brought this to mind. It would be easier to add to the REBOL 3 version than the REBOL 2.3-Mar-08 15:04
59BrianHJust the ones that don't break existing code, but are faster, cleaner or have been enhanced.3-Mar-08 15:02
58BrianHImportant question: Should we back-port some of the REBOL 3 mezzanine changes?3-Mar-08 15:01
57DocKimbelExactly.3-Mar-08 14:58
56BrianHNaw, you never know when that might be needed.3-Mar-08 14:57
55BrianHHave that refinement be ignored on /console?3-Mar-08 14:57
54DocKimbelagreed3-Mar-08 14:56
53BrianHThat would be more consistent.3-Mar-08 14:56
52DocKimbelor maybe just /show3-Mar-08 14:56
51BrianHThat would work.3-Mar-08 14:56
50DocKimbel/visible ?3-Mar-08 14:55
49BrianHOr a new refinement.3-Mar-08 14:55
48BrianHI like your show the console on call/shell idea better.3-Mar-08 14:55
47DocKimbelmaybe add a new refinement and let the users decide when they want to see the console window ?3-Mar-08 14:55
46Gabrielebut setting io redir for them would not hurt either.3-Mar-08 14:54
45BrianHNo, that won't work. GUI apps rarely need I/O redirection and never need the console window.3-Mar-08 14:53
44Gabrieleor, show it unless there is i/o redirection?3-Mar-08 14:52
43BrianHTracker description?3-Mar-08 14:52
42Gabrielemaybe call/shell should show the console window...3-Mar-08 14:52
41Gabrielelet's make sure we have this info put somewhere that Carl can easily access... otherwise it gets lost3-Mar-08 14:52
40BrianHPerhaps they could be set by default, with an option to show the console if needed? I frequently call command-line apps that I need to interact with on occasion. Maybe call/console would do for that in some cases. There should never be a console window with call/console though.3-Mar-08 14:51
39Gabrieleyep.3-Mar-08 14:51
38Gabrieleyes, actually, there's just a flag to add (or remove, I don't recall... it should be trivial though)3-Mar-08 14:50
37DocKimbelIn the STARTUPINFO used by CreateProcess( ), just set in dwFlags, the STARTF_USESHOWWINDOW flag and set wShowWindow to SW_HIDE.3-Mar-08 14:50
36DocKimbelAbout CALL console window issue, the CreateProcess( ) win32 call has flags to hide the window. There just need to be set.3-Mar-08 14:48
35BrianHIs there a RUN native for OS X that wraps open?3-Mar-08 14:46
34DocKimbelPaul : IIRC, the RUN native is a wrapper to ShellExecute( ), so you already have it.3-Mar-08 14:45
33PekrDoc could elaborate here, as he is author of async call protocol, which you can download via softinnov.org ...3-Mar-08 14:43
32BrianHOn Mac the equivalent is open, and that is what CALL uses. I don't know about Windows.3-Mar-08 13:45
31PaulAnyone think winshellexecute would be a good addition? I don't know what the counterpart to that is on other platforms but it is a simple piece of code that offers alot of functionality on windows platforms.3-Mar-08 13:33
30BTiffinDidn't hear ... bad assumption. Good outcome. :)3-Mar-08 4:13
29BrianHWill we be able to put in some backports from the REBOL 3 mezzanines?3-Mar-08 3:09
28PaulIs 2.7.6 only going to be bug fixes or are we putting in some wishes?3-Mar-08 3:06
27CarlA: No. Where did you hear that? Was it from me? I hope not. :-/3-Mar-08 3:02
26CarlQ: " Is 2.7.6 not planned as the last official R2?"3-Mar-08 3:02
25BrianH:)3-Mar-08 3:01
24BrianH2.7.5 was an official build. 2.7.6 will be another. There will be more (REBOL 3 isn't coming out this week, you know).3-Mar-08 2:50
23BTiffinSo, just getting into headspace; from the main goal; fix a few and get it released. Umm, maybe I'm confused. Is 2.7.6 not planned as the last official R2? Is there still more to come? A good thing mind; it just hasn't been my assumption. Is 2.7.6 planned as another beta step toward an official build?3-Mar-08 2:47
22CarlFixed3-Mar-08 1:12
21CarlRestarting this world. 1 Min.3-Mar-08 1:10
20CarlWelcome new user!

Keep in mind that all non-private chat groups are feed over the web -- they are public by default! (Differs from from REBOL3 world.)

3-Mar-08 0:11
19HenrikI only closed the R2-beta AltME and started it again.2-Mar-08 23:52
18Henrikworked on restart too here. I guess one is not allowed to use the bugtracker on first connect.2-Mar-08 23:52
17Carlvery odd2-Mar-08 23:49
16Carlok.. added one. I had to close all altmes, then restart altme.2-Mar-08 23:49
15BrianHI saved mine after the restart, but before Henrik's error report.2-Mar-08 23:48
14Carlverified. Checking...2-Mar-08 23:48
13CarlTesting...2-Mar-08 23:47
12HenrikI can see his report too, but I can't save any newly entered reports.2-Mar-08 23:43
11CarlFeed: http://www.rebol.net/altweb/r2-beta/2-Mar-08 23:43
10CarlIt's there. I see Brian's entry posted.2-Mar-08 23:43
9Carlchecking...2-Mar-08 23:42
8Henrik"Development Tracker application is not available for this AltME world." - when I enter a bug report.2-Mar-08 23:37
7Carlhttp://www.rebol.com/article/0349.html2-Mar-08 23:23
6CarlAlso, ViewTop and Editor parts.2-Mar-08 23:16
5CarlProbably a good idea.2-Mar-08 23:16
4BrianHDo we need to put the R2 mezz source in DevBase?2-Mar-08 23:12
3CarlYes, good.2-Mar-08 23:09
2CarlOk, R2-Beta World up and running. This is hosted on services.rebol.net server in Canada.2-Mar-08 23:09
1Henrikwe have a bug tracker here. use that instead of primitive checklists to keep track of the things we want to fix?2-Mar-08 23:09

Return to Index Page