www.digitalmars.com         C & C++   DMDScript  

digitalmars.D - Close D1 bugs?

reply "Steven Schveighoffer" <schveiguy yahoo.com> writes:
I noticed Vladimir closed a D1 bug as WORKSFORME, with an explanation that  
it is fixed in the latest version of D2 phobos, not realizing it was a D1  
bug. (https://d.puremagic.com/issues/show_bug.cgi?id=1004)

However, it brought up a discussion of what to do with D1 bugs. Should  
they be closed? If so, what should be the reason for closing? My  
impression is if the bug still exists (and I'm not advocating we test for  
it), we should close it as WONTFIX, since D1 is deprecated.

Thoughts? I closed the aforementioned bug as WONTFIX. Anyone is free to  
correct that if you feel it's in error :)

-Steve
Feb 11 2014
next sibling parent reply "Daniel Murphy" <yebbliesnospam gmail.com> writes:
"Steven Schveighoffer"  wrote in message 
news:op.xa4tq0hzeav7ka stevens-macbook-pro.local...

 I noticed Vladimir closed a D1 bug as WORKSFORME, with an explanation that 
 it is fixed in the latest version of D2 phobos, not realizing it was a D1 
 bug. (https://d.puremagic.com/issues/show_bug.cgi?id=1004)

 However, it brought up a discussion of what to do with D1 bugs. Should 
 they be closed? If so, what should be the reason for closing? My 
 impression is if the bug still exists (and I'm not advocating we test for 
 it), we should close it as WONTFIX, since D1 is deprecated.

 Thoughts? I closed the aforementioned bug as WONTFIX. Anyone is free to 
 correct that if you feel it's in error :)

The is the unfortunate consequence of putting a bug from two different projects in the same report. Ideally the D1 portion would be WONTFIX, and the D2 part would be WORKSFORME/FIXED/etc. Since we can't actually do that, I think we should just discard the D1 information and give it the appropriate status for D2. After all, we're not going to fix the bug for 2.061 or 2.062 or 2.063 etc either, but we won't mark it WONTFIX because of that.
Feb 11 2014
parent "Daniel Murphy" <yebbliesnospam gmail.com> writes:
"Steven Schveighoffer"  wrote in message 
news:op.xa5ffvpceav7ka stevens-macbook-pro.local...

 Hm... there didn't seem to be any indication that it was D2 related. In 
 any case, I was speaking about D1-only bugs.

Ah.
Feb 11 2014
prev sibling next sibling parent "Steven Schveighoffer" <schveiguy yahoo.com> writes:
On Tue, 11 Feb 2014 21:54:29 -0500, Daniel Murphy  
<yebbliesnospam gmail.com> wrote:

 "Steven Schveighoffer"  wrote in message  
 news:op.xa4tq0hzeav7ka stevens-macbook-pro.local...

 I noticed Vladimir closed a D1 bug as WORKSFORME, with an explanation  
 that it is fixed in the latest version of D2 phobos, not realizing it  
 was a D1 bug. (https://d.puremagic.com/issues/show_bug.cgi?id=1004)

 However, it brought up a discussion of what to do with D1 bugs. Should  
 they be closed? If so, what should be the reason for closing? My  
 impression is if the bug still exists (and I'm not advocating we test  
 for it), we should close it as WONTFIX, since D1 is deprecated.

 Thoughts? I closed the aforementioned bug as WONTFIX. Anyone is free to  
 correct that if you feel it's in error :)

The is the unfortunate consequence of putting a bug from two different projects in the same report. Ideally the D1 portion would be WONTFIX, and the D2 part would be WORKSFORME/FIXED/etc. Since we can't actually do that, I think we should just discard the D1 information and give it the appropriate status for D2. After all, we're not going to fix the bug for 2.061 or 2.062 or 2.063 etc either, but we won't mark it WONTFIX because of that.

Hm... there didn't seem to be any indication that it was D2 related. In any case, I was speaking about D1-only bugs. -Steve
Feb 11 2014
prev sibling next sibling parent "Peter Alexander" <peter.alexander.au gmail.com> writes:
On Tuesday, 11 February 2014 at 19:29:14 UTC, Steven 
Schveighoffer wrote:
 However, it brought up a discussion of what to do with D1 bugs. 
 Should they be closed?

Last time I asked about this, it was argued that D1 bugs should remain in case "someone" wants to fix them. I'm against this because D1 is supposedly unsupported, and no one appears to be fixing D1 bugs anyway. I think they should be closed as WONTFIX with a quick comment to say that D1 is unsupported.
Feb 12 2014
prev sibling next sibling parent reply Walter Bright <newshound2 digitalmars.com> writes:
On 2/11/2014 11:29 AM, Steven Schveighoffer wrote:
 Thoughts? I closed the aforementioned bug as WONTFIX. Anyone is free to correct
 that if you feel it's in error :)

D1 is still in use. Just ignore D1 only bugs if you don't use D1. But to close them does a disservice to the existing D1 users.
Feb 12 2014
next sibling parent reply Walter Bright <newshound2 digitalmars.com> writes:
On 2/13/2014 10:34 AM, francesco cattoglio wrote:
 Also, the current open issues list is HUGE, cutting it down by
 discarding outdated stuff would be nice. Or at least, would
 *look* nice. Honestly, the first time I took a look at D I was
 like "wait, is this a programming language or a testbed for some
 strange compiler?" :P

Restrict your search for open bugs to "D2" and "D1 & D2" and you'll be fine.
Feb 13 2014
next sibling parent reply "Daniel Murphy" <yebbliesnospam gmail.com> writes:
"Walter Bright"  wrote in message news:ldj88t$2hpa$1 digitalmars.com...

 Restrict your search for open bugs to "D2" and "D1 & D2" and you'll be 
 fine.

No, this doesn't work, because many many bugs are not marked correctly. A bug being marked D1 often just means it's old.
Feb 13 2014
parent reply Walter Bright <newshound2 digitalmars.com> writes:
On 2/13/2014 6:07 PM, Daniel Murphy wrote:
 No, this doesn't work, because many many bugs are not marked correctly.  A bug
 being marked D1 often just means it's old.

Anyone can help out when they see miscategorized bugs by fixing the categories.
Feb 13 2014
parent reply "Daniel Murphy" <yebbliesnospam gmail.com> writes:
"Walter Bright"  wrote in message news:ldjvs6$av1$1 digitalmars.com...

 No, this doesn't work, because many many bugs are not marked correctly. 
 A bug
 being marked D1 often just means it's old.

Anyone can help out when they see miscategorized bugs by fixing the categories.

And until that is complete, there is no reliable way to filter out D1 bugs. I've been marking them with "(D1 only)", so I know at least those ones are actually D1 bugs. Is sociomantic _actually_ using D1 phobos and caring if those bugs get fixed? If not we should close them.
Feb 14 2014
parent reply Walter Bright <newshound2 digitalmars.com> writes:
On 2/14/2014 2:22 AM, Daniel Murphy wrote:
 And until that is complete, there is no reliable way to filter out D1 bugs.
I've
 been marking them with "(D1 only)", so I know at least those ones are actually
 D1 bugs.

Do a search for "D2" and "D1 & D2".
 Is sociomantic _actually_ using D1 phobos and caring if those bugs get fixed?

To reiterate, the D1 bugzilla issues stay as long as Sociomantic is using D1.
Feb 14 2014
parent reply "Daniel Murphy" <yebbliesnospam gmail.com> writes:
"Walter Bright"  wrote in message news:ldksbk$1pcn$1 digitalmars.com...

 On 2/14/2014 2:22 AM, Daniel Murphy wrote:
 And until that is complete, there is no reliable way to filter out D1 
 bugs. I've
 been marking them with "(D1 only)", so I know at least those ones are 
 actually
 D1 bugs.

Do a search for "D2" and "D1 & D2".

That excludes the bugs marked any of "2.000", "2.002", "2.003", "2.004", "2.005", "2.006", "2.007", "2.008", "2.009", "2.010", "2.011", "2.012", "2.013", "2.014", "2.015", "2.016", "2.017", "2.018", "2.029", "2.020", "2.021", "2.022", "2.023", "2.024", "2.025", "2.026", "2.027", "2.028", "2.029", "2.030", "unspecified", "2.031", "2.032", "2.033", "2.034", "2.035", "2.036", "2.037", "2.038", "2.039", "2.040" or "2.041". Any even better, it falsely exludes the bugs marked as "D1*" that are from before D2 existed and not necessarily exclusive to D1.
 Is sociomantic _actually_ using D1 phobos and caring if those bugs get 
 fixed?

To reiterate, the D1 bugzilla issues stay as long as Sociomantic is using D1.

Oh look, Don says I was right and they're not using phobos1.
Feb 14 2014
parent reply Walter Bright <newshound2 digitalmars.com> writes:
On 2/14/2014 3:06 AM, Daniel Murphy wrote:
 "Walter Bright"  wrote in message news:ldksbk$1pcn$1 digitalmars.com...
 Do a search for "D2" and "D1 & D2".

That excludes the bugs marked any of "2.000", "2.002", "2.003", "2.004", "2.005", "2.006", "2.007", "2.008", "2.009", "2.010", "2.011", "2.012", "2.013", "2.014", "2.015", "2.016", "2.017", "2.018", "2.029", "2.020", "2.021", "2.022", "2.023", "2.024", "2.025", "2.026", "2.027", "2.028", "2.029", "2.030", "unspecified", "2.031", "2.032", "2.033", "2.034", "2.035", "2.036", "2.037", "2.038", "2.039", "2.040" or "2.041".

True. And those should be retagged as D2.
 Any even better, it falsely exludes the bugs marked as "D1*" that are from
 before D2 existed and not necessarily exclusive to D1.

True, and those that are tagged as "D1" but really are "D1 & D2" should also be retagged.
Feb 14 2014
parent reply Walter Bright <newshound2 digitalmars.com> writes:
On 2/14/2014 7:58 PM, Walter Bright wrote:
 On 2/14/2014 3:06 AM, Daniel Murphy wrote:
 "Walter Bright"  wrote in message news:ldksbk$1pcn$1 digitalmars.com...
 Do a search for "D2" and "D1 & D2".

That excludes the bugs marked any of "2.000", "2.002", "2.003", "2.004", "2.005", "2.006", "2.007", "2.008", "2.009", "2.010", "2.011", "2.012", "2.013", "2.014", "2.015", "2.016", "2.017", "2.018", "2.029", "2.020", "2.021", "2.022", "2.023", "2.024", "2.025", "2.026", "2.027", "2.028", "2.029", "2.030", "unspecified", "2.031", "2.032", "2.033", "2.034", "2.035", "2.036", "2.037", "2.038", "2.039", "2.040" or "2.041".

True. And those should be retagged as D2.

BTW, Brad showed me how to do bulk retagging, so I've been doing that at least for unresolved DMD bugs. I.e. all the 0.0..1.xx are being retagged as D1, and 2.00..2.xx are being retagged as D2. Of course, many of the D1 bugs should be retagged as "D1 & D2", but they'll have to be gone through one by one to do that.
Feb 14 2014
parent reply "Daniel Murphy" <yebbliesnospam gmail.com> writes:
"Walter Bright"  wrote in message news:ldmpoi$o31$1 digitalmars.com...

 BTW, Brad showed me how to do bulk retagging, so I've been doing that at 
 least for unresolved DMD bugs. I.e. all the 0.0..1.xx are being retagged 
 as D1, and 2.00..2.xx are being retagged as D2.

That's certainly an improvement! That just leaves 'future' and 'unspecified' Can you please also remove all those 0.XXX/1.XXX/2.XXX versions from the list?
 Of course, many of the D1 bugs should be retagged as "D1 & D2", but 
 they'll have to be gone through one by one to do that.

Yeah, I'll probably do that next time I make a pass through bugzilla.
Feb 14 2014
parent reply Walter Bright <newshound2 digitalmars.com> writes:
On 2/14/2014 9:05 PM, Daniel Murphy wrote:
 That just leaves 'future' and 'unspecified'

Those have to be manually gone through.
 Can you please also remove all those 0.XXX/1.XXX/2.XXX versions from the list?

To do that, I also need to retag all the fixed ones, too. I'll get to it.
Feb 14 2014
parent "Daniel Murphy" <yebbliesnospam gmail.com> writes:
"Walter Bright"  wrote in message news:ldn521$10av$1 digitalmars.com...

 Can you please also remove all those 0.XXX/1.XXX/2.XXX versions from the 
 list?

To do that, I also need to retag all the fixed ones, too. I'll get to it.

Thanks!
Feb 14 2014
prev sibling parent reply Andrei Alexandrescu <SeeWebsiteForEmail erdani.org> writes:
On 2/13/14, 6:28 PM, Joseph Cassman wrote:
 On Thursday, 13 February 2014 at 20:00:29 UTC, Walter Bright wrote:
 On 2/13/2014 10:34 AM, francesco cattoglio wrote:
 Also, the current open issues list is HUGE, cutting it down by
 discarding outdated stuff would be nice. Or at least, would
 *look* nice. Honestly, the first time I took a look at D I was
 like "wait, is this a programming language or a testbed for some
 strange compiler?" :P

Restrict your search for open bugs to "D2" and "D1 & D2" and you'll be fine.

I never thought to do that as I have just gone by the "Bug Tracker" graphic on the website up until now. Interesting idea. Here is a chart comparing how the number of open bugs changes when those two filters are applied to the "Bug Tracker" queries. Before After Regression 9 9 Blocker 17 17 Critical 69 63 Major 221 206 Normal, minor, trivial 1923 1420 Enhancement 1153 866 All Open 3392 2586 That large of a difference surprised me. Seems like it would help with the impression of quality to someone first coming to the site by displaying the data filtered for D2 instead. It also seems more accurate since the D1 only stuff is not really being worked on and that page presents sort of a TODO list.

So did you filter for "D2" and "D1 & D2"? That may be missing stuff, as Daniel mentioned. I think it should be fine to mark D1 enhancement requests as WONTFIX, what does Sociomantic think? Andrei
Feb 14 2014
parent Walter Bright <newshound2 digitalmars.com> writes:
On 2/14/2014 10:38 AM, Brad Roberts wrote:
 On 2/14/14, 7:17 AM, Andrei Alexandrescu wrote:
 So did you filter for "D2" and "D1 & D2"? That may be missing stuff, as
 Daniel mentioned.

 I think it should be fine to mark D1 enhancement requests as WONTFIX,
 what does Sociomantic think?


 Andrei

And those enhancement requests also likely apply to D2 as well.

Essentially, any mindless resolutions should not be done. Every one needs to be evaluated on its own.
Feb 14 2014
prev sibling parent "Daniel Murphy" <yebbliesnospam gmail.com> writes:
"ed"  wrote in message news:qpkejaxtasbzakgprzti forum.dlang.org...

 As the "official support for D1" has been discontinued in 2012, 
 shouldn't one expect the official bug tracker to close anything 
 D1-related too ?

Why? Does it affect D2 in any way if D1 bugs are in the bug tracker(besides some meaningless stats)?. Just because it is offically deprecated and not supported doesn't not imply there is no support at all.

It affects the people using and maintaining the bug tracker due to the false positives it generates.
 Last but not least, there is the risk that bugs, once fixed for D2, 
 turns into D1 bugs, which would harm our statistics and may bug some 
 other tools (bountysource ?). Take a look at #10217 as an example.

So? Then change the version to D1 only.

No, this won't work, because you actually have two separate issues, one for D1 and one for D2, one fixed and the other open. eg the changelog is generated from the list of RESOLVED FIXED bugs, and won't pick up ones that changed status.
 Focus should be on actually fixing bugs, not some feelgood numbers coming 
 out of the bug tracker stats. Besides, I doubt there is any real negative 
 impact on D as a result of the bug tracker numbers at this stage. Maybe 
 worry about that when evidence suggests otherwise.

This is coming from the people who are fixing bugs. This issues gets in the way of doing so. I couldn't care less about the numbers.
Feb 14 2014
prev sibling next sibling parent "Steven Schveighoffer" <schveiguy yahoo.com> writes:
On Wed, 12 Feb 2014 03:57:25 -0500, Walter Bright  
<newshound2 digitalmars.com> wrote:

 On 2/11/2014 11:29 AM, Steven Schveighoffer wrote:
 Thoughts? I closed the aforementioned bug as WONTFIX. Anyone is free to  
 correct
 that if you feel it's in error :)

D1 is still in use. Just ignore D1 only bugs if you don't use D1. But to close them does a disservice to the existing D1 users.

I think leaving them open with hope it will be fixed when it likely will never happen is also a disservice. Tell me, are you planning on accepting patches for D1 phobos ever? -Steve
Feb 13 2014
prev sibling next sibling parent "francesco cattoglio" <francesco.cattoglio gmail.com> writes:
On Wednesday, 12 February 2014 at 08:57:21 UTC, Walter Bright
wrote:
 On 2/11/2014 11:29 AM, Steven Schveighoffer wrote:
 Thoughts? I closed the aforementioned bug as WONTFIX. Anyone 
 is free to correct
 that if you feel it's in error :)

D1 is still in use. Just ignore D1 only bugs if you don't use D1. But to close them does a disservice to the existing D1 users.

Just deleting them would be a horrible disservice. Marking as "WONTFIX" sounds reasonable to me. After all, since support got dropped, "won't be fixed" sounds realistic. Also, the current open issues list is HUGE, cutting it down by discarding outdated stuff would be nice. Or at least, would *look* nice. Honestly, the first time I took a look at D I was like "wait, is this a programming language or a testbed for some strange compiler?" :P
Feb 13 2014
prev sibling next sibling parent "Joseph Cassman" <jc7919 outlook.com> writes:
On Thursday, 13 February 2014 at 20:00:29 UTC, Walter Bright 
wrote:
 On 2/13/2014 10:34 AM, francesco cattoglio wrote:
 Also, the current open issues list is HUGE, cutting it down by
 discarding outdated stuff would be nice. Or at least, would
 *look* nice. Honestly, the first time I took a look at D I was
 like "wait, is this a programming language or a testbed for 
 some
 strange compiler?" :P

Restrict your search for open bugs to "D2" and "D1 & D2" and you'll be fine.

I never thought to do that as I have just gone by the "Bug Tracker" graphic on the website up until now. Interesting idea. Here is a chart comparing how the number of open bugs changes when those two filters are applied to the "Bug Tracker" queries. Before After Regression 9 9 Blocker 17 17 Critical 69 63 Major 221 206 Normal, minor, trivial 1923 1420 Enhancement 1153 866 All Open 3392 2586 That large of a difference surprised me. Seems like it would help with the impression of quality to someone first coming to the site by displaying the data filtered for D2 instead. It also seems more accurate since the D1 only stuff is not really being worked on and that page presents sort of a TODO list. Joseph
Feb 13 2014
prev sibling next sibling parent "Mathias LANG" <pro.mathias.lang gmail.com> writes:
On Wednesday, 12 February 2014 at 08:57:21 UTC, Walter Bright 
wrote:
 On 2/11/2014 11:29 AM, Steven Schveighoffer wrote:
 Thoughts? I closed the aforementioned bug as WONTFIX. Anyone 
 is free to correct
 that if you feel it's in error :)

D1 is still in use. Just ignore D1 only bugs if you don't use D1. But to close them does a disservice to the existing D1 users.

What would be the negative impact on D1 users ? And how many people will be affected ? As the "official support for D1" has been discontinued in 2012, shouldn't one expect the official bug tracker to close anything D1-related too ? Last but not least, there is the risk that bugs, once fixed for D2, turns into D1 bugs, which would harm our statistics and may bug some other tools (bountysource ?). Take a look at #10217 as an example.
Feb 13 2014
prev sibling next sibling parent "ed" <growlercab gmail.com> writes:
On Friday, 14 February 2014 at 05:27:14 UTC, Mathias LANG wrote:
 On Wednesday, 12 February 2014 at 08:57:21 UTC, Walter Bright 
 wrote:
 On 2/11/2014 11:29 AM, Steven Schveighoffer wrote:
 Thoughts? I closed the aforementioned bug as WONTFIX. Anyone 
 is free to correct
 that if you feel it's in error :)

D1 is still in use. Just ignore D1 only bugs if you don't use D1. But to close them does a disservice to the existing D1 users.

What would be the negative impact on D1 users ? And how many people will be affected ?

 As the "official support for D1" has been discontinued in 2012, 
 shouldn't one expect the official bug tracker to close anything 
 D1-related too ?

Why? Does it affect D2 in any way if D1 bugs are in the bug tracker(besides some meaningless stats)?. Just because it is offically deprecated and not supported doesn't not imply there is no support at all.
 Last but not least, there is the risk that bugs, once fixed for 
 D2, turns into D1 bugs, which would harm our statistics and may 
 bug some other tools (bountysource ?). Take a look at #10217 as 
 an example.

So? Then change the version to D1 only. Focus should be on actually fixing bugs, not some feelgood numbers coming out of the bug tracker stats. Besides, I doubt there is any real negative impact on D as a result of the bug tracker numbers at this stage. Maybe worry about that when evidence suggests otherwise. Cheers, ed
Feb 13 2014
prev sibling next sibling parent "monarch_dodra" <monarchdodra gmail.com> writes:
On Friday, 14 February 2014 at 06:07:01 UTC, ed wrote:
 On Friday, 14 February 2014 at 05:27:14 UTC, Mathias LANG wrote:
 As the "official support for D1" has been discontinued in 
 2012, shouldn't one expect the official bug tracker to close 
 anything D1-related too ?

Why? Does it affect D2 in any way if D1 bugs are in the bug tracker(besides some meaningless stats)?. Just because it is offically deprecated and not supported doesn't not imply there is no support at all.

It affects me when I'm trying to file a bug, and looking through already existing bugs, if 1000 of them are from an historical project.
 Last but not least, there is the risk that bugs, once fixed 
 for D2, turns into D1 bugs, which would harm our statistics 
 and may bug some other tools (bountysource ?). Take a look at 
 #10217 as an example.

So? Then change the version to D1 only. Focus should be on actually fixing bugs, not some feelgood numbers coming out of the bug tracker stats. Besides, I doubt there is any real negative impact on D as a result of the bug tracker numbers at this stage. Maybe worry about that when evidence suggests otherwise.

It's not necessarily about the numbers. It's the fact that now, the project is 100% D2, 0% D1, and the D1 bugs are getting in our way. In particular, since D1 is not supposed to be supported, it makes little sense to support bug tracking for it. It's noise and nothing more. I think it would make more sense instead to have a good wiki/ddoc page that clearly documents know issues in D1 that won't get fixed, how they are triggered, and how to work around them. IMO, once properly documented and identified, then the bug can simply be closed as "won't fix".
Feb 13 2014
prev sibling next sibling parent reply "Don" <x nospam.com> writes:
On Tuesday, 11 February 2014 at 19:29:14 UTC, Steven 
Schveighoffer wrote:
 I noticed Vladimir closed a D1 bug as WORKSFORME, with an 
 explanation that it is fixed in the latest version of D2 
 phobos, not realizing it was a D1 bug. 
 (https://d.puremagic.com/issues/show_bug.cgi?id=1004)

 However, it brought up a discussion of what to do with D1 bugs. 
 Should they be closed? If so, what should be the reason for 
 closing? My impression is if the bug still exists (and I'm not 
 advocating we test for it), we should close it as WONTFIX, 
 since D1 is deprecated.

 Thoughts? I closed the aforementioned bug as WONTFIX. Anyone is 
 free to correct that if you feel it's in error :)

 -Steve

I agree with you in this case. D1 is in heavy commercial use at Sociomantic, and we are still submitting patches to DMD, and we frequently search bugzilla for open D1 bugs. But AFAIK *nobody* is using D1 Phobos. The only remaining role of D1 Phobos, AFAIK, is to allow the D1 test suite to run. I think that we should close all D1 Phobos bugs as WONTFIX. Realistically they are never going to be fixed, and I don't think anybody cares. (Or, if the bug also applied to D2, but is already fixed in D2, I think it would be perfectly valid to mark it as FIXED).
Feb 14 2014
next sibling parent "Daniel Murphy" <yebbliesnospam gmail.com> writes:
"Don"  wrote in message news:sqijjqlbiarapepraswf forum.dlang.org... 

 I agree with you in this case.
 
 D1 is in heavy commercial use at Sociomantic, and we are still 
 submitting patches to DMD, and we frequently search bugzilla for 
 open D1 bugs. But AFAIK *nobody* is using D1 Phobos.
 The only remaining role of D1 Phobos, AFAIK, is to allow the D1 
 test suite to run.
 
 I think that we should close all D1 Phobos bugs as WONTFIX. 
 Realistically they are never going to be fixed, and I don't think 
 anybody cares.
 (Or, if the bug also applied to D2, but is already fixed in D2, I 
 think it would be perfectly valid to mark it as FIXED).

Thankyou Don! I thought that might be the case.
Feb 14 2014
prev sibling next sibling parent reply Walter Bright <newshound2 digitalmars.com> writes:
On 2/14/2014 12:16 AM, Don wrote:
 I agree with you in this case.

 D1 is in heavy commercial use at Sociomantic, and we are still submitting
 patches to DMD, and we frequently search bugzilla for open D1 bugs. But AFAIK
 *nobody* is using D1 Phobos.
 The only remaining role of D1 Phobos, AFAIK, is to allow the D1 test suite to
run.

 I think that we should close all D1 Phobos bugs as WONTFIX. Realistically they
 are never going to be fixed, and I don't think anybody cares.
 (Or, if the bug also applied to D2, but is already fixed in D2, I think it
would
 be perfectly valid to mark it as FIXED).

In that case, I'm on board with that. (In case it isn't obvious, Don represents Sociomantic here.)
Feb 14 2014
next sibling parent Walter Bright <newshound2 digitalmars.com> writes:
On 2/14/2014 10:39 AM, Brad Roberts wrote:
 As long as the closure process isn't a blind one.  Chances are reasonably high
 that MOST open D1 bugs also apply to D2.  The closure process must be an
 examination of the bug, not a mass update to close them.

Yes, that's critical.
Feb 14 2014
prev sibling parent Walter Bright <newshound2 digitalmars.com> writes:
On 2/14/2014 11:32 PM, Don wrote:
 I have at times gone through all of the D1 compiler bugs, and
 marked them as D1&D2 where possible. The ones that never applied
 to D2 have "D1 only" in the title. I only found 38 of them.

Did you look at all the 0.xxx and 1.xxx bugs? I have just reversioned them all as D1.
Feb 15 2014
prev sibling parent reply Andrei Alexandrescu <SeeWebsiteForEmail erdani.org> writes:
On 2/14/14, 12:16 AM, Don wrote:
 D1 is in heavy commercial use at Sociomantic, and we are still
 submitting patches to DMD, and we frequently search bugzilla for open D1
 bugs. But AFAIK *nobody* is using D1 Phobos.
 The only remaining role of D1 Phobos, AFAIK, is to allow the D1 test
 suite to run.

 I think that we should close all D1 Phobos bugs as WONTFIX.
 Realistically they are never going to be fixed, and I don't think
 anybody cares.
 (Or, if the bug also applied to D2, but is already fixed in D2, I think
 it would be perfectly valid to mark it as FIXED).

Sounds reasonable. BTW is there a plan in place to migrate to D2 for Sociomantic? Andrei
Feb 14 2014
parent Andrei Alexandrescu <SeeWebsiteForEmail erdani.org> writes:
On 2/14/14, 6:51 AM, Dicebot wrote:
 On Friday, 14 February 2014 at 14:03:15 UTC, Andrei Alexandrescu wrote:
 Sounds reasonable. BTW is there a plan in place to migrate to D2 for
 Sociomantic?

 Andrei

Yes, but it is not a priority goal and unlikely to happen soon.

Is there anything that can be done on our side to bring this to the front burner? Andrei
Feb 14 2014
prev sibling next sibling parent "w0rp" <devw0rp gmail.com> writes:
On Friday, 14 February 2014 at 02:28:36 UTC, Joseph Cassman wrote:
 I never thought to do that as I have just gone by the "Bug 
 Tracker" graphic on the website up until now. Interesting idea. 
 Here is a chart comparing how the number of open bugs changes 
 when those two filters are applied to the "Bug Tracker" queries.
                          Before   After
 Regression               9        9
 Blocker                  17       17
 Critical                 69       63
 Major                    221      206
 Normal, minor, trivial   1923     1420
 Enhancement              1153     866
 All Open                 3392     2586

 That large of a difference surprised me. Seems like it would 
 help with the impression of quality to someone first coming to 
 the site by displaying the data filtered for D2 instead. It 
 also seems more accurate since the D1 only stuff is not really 
 being worked on and that page presents sort of a TODO list.

 Joseph

I approve of filtering the graphic this way. If you define "D" as meaning "current D" or "D2" and "D1" as "old D," even though D1 has current use, then this information would be clearer. The numbers for the current versions are more relevant to most people.
Feb 14 2014
prev sibling next sibling parent "Dicebot" <public dicebot.lv> writes:
On Friday, 14 February 2014 at 14:03:15 UTC, Andrei Alexandrescu 
wrote:
 Sounds reasonable. BTW is there a plan in place to migrate to 
 D2 for Sociomantic?

 Andrei

Yes, but it is not a priority goal and unlikely to happen soon.
Feb 14 2014
prev sibling next sibling parent "Dicebot" <public dicebot.lv> writes:
On Friday, 14 February 2014 at 15:15:12 UTC, Andrei Alexandrescu 
wrote:
 Is there anything that can be done on our side to bring this to 
 the front burner?

 Andrei

No, not right now - we need to take care of some more important issues before being able to spend time/effort on porting. Growing pains. There are some ideas about possible compiler tweaks that can help but we are not going to ask anything without some detailed case study - have actually been discussing this on one of team meetings recently.
Feb 14 2014
prev sibling next sibling parent Brad Roberts <braddr puremagic.com> writes:
On 2/14/14, 7:17 AM, Andrei Alexandrescu wrote:
 So did you filter for "D2" and "D1 & D2"? That may be missing stuff, as
 Daniel mentioned.

 I think it should be fine to mark D1 enhancement requests as WONTFIX,
 what does Sociomantic think?


 Andrei

And those enhancement requests also likely apply to D2 as well.
Feb 14 2014
prev sibling next sibling parent Brad Roberts <braddr puremagic.com> writes:
On 2/14/14, 2:51 AM, Walter Bright wrote:
 On 2/14/2014 12:16 AM, Don wrote:
 I agree with you in this case.

 D1 is in heavy commercial use at Sociomantic, and we are still submitting
 patches to DMD, and we frequently search bugzilla for open D1 bugs.
 But AFAIK
 *nobody* is using D1 Phobos.
 The only remaining role of D1 Phobos, AFAIK, is to allow the D1 test
 suite to run.

 I think that we should close all D1 Phobos bugs as WONTFIX.
 Realistically they
 are never going to be fixed, and I don't think anybody cares.
 (Or, if the bug also applied to D2, but is already fixed in D2, I
 think it would
 be perfectly valid to mark it as FIXED).

In that case, I'm on board with that. (In case it isn't obvious, Don represents Sociomantic here.)

As long as the closure process isn't a blind one. Chances are reasonably high that MOST open D1 bugs also apply to D2. The closure process must be an examination of the bug, not a mass update to close them.
Feb 14 2014
prev sibling next sibling parent "Jacob Carlborg" <doob me.com> writes:
On Friday, 14 February 2014 at 15:15:12 UTC, Andrei Alexandrescu 
wrote:

 Is there anything that can be done on our side to bring this to 
 the front burner?

If I recall correctly, someone mentioned they need the concurrent GC that Leandro Lucarella implemented for D1. -- /Jacob Carlborg
Feb 14 2014
prev sibling next sibling parent "Joseph Cassman" <jc7919 outlook.com> writes:
On Friday, 14 February 2014 at 15:17:16 UTC, Andrei Alexandrescu 
wrote:
 On 2/13/14, 6:28 PM, Joseph Cassman wrote:
 On Thursday, 13 February 2014 at 20:00:29 UTC, Walter Bright 
 wrote:
 On 2/13/2014 10:34 AM, francesco cattoglio wrote:
 Also, the current open issues list is HUGE, cutting it down 
 by
 discarding outdated stuff would be nice. Or at least, would
 *look* nice. Honestly, the first time I took a look at D I 
 was
 like "wait, is this a programming language or a testbed for 
 some
 strange compiler?" :P

Restrict your search for open bugs to "D2" and "D1 & D2" and you'll be fine.

I never thought to do that as I have just gone by the "Bug Tracker" graphic on the website up until now. Interesting idea. Here is a chart comparing how the number of open bugs changes when those two filters are applied to the "Bug Tracker" queries. Before After Regression 9 9 Blocker 17 17 Critical 69 63 Major 221 206 Normal, minor, trivial 1923 1420 Enhancement 1153 866 All Open 3392 2586 That large of a difference surprised me. Seems like it would help with the impression of quality to someone first coming to the site by displaying the data filtered for D2 instead. It also seems more accurate since the D1 only stuff is not really being worked on and that page presents sort of a TODO list.

So did you filter for "D2" and "D1 & D2"? That may be missing stuff, as Daniel mentioned. I think it should be fine to mark D1 enhancement requests as WONTFIX, what does Sociomantic think? Andrei

Yeah, I may have marked the wrong search filters. Couldn't find a way to publicly share saved searches so to describe in words, I selected "D" for the Product category, "D2" and "D1 & D2" for the Version category, and the appropriate flags in the Severity category to match the query linked to on the "Bug Tracker" page. The header text shown above each query seemed to match that for each custom search, other than the text "D2" and "D1 & D2" which was added. After reading through the other posts it sounds like a general triage is in order to determine the status of the flag marked in the Version category for the D1-related bugs. So those numbers in my original post are definitely suspect even if filtered correctly. Joseph
Feb 14 2014
prev sibling parent "Don" <nospam nospam.com> writes:
On Friday, 14 February 2014 at 18:39:55 UTC, Brad Roberts wrote:
 On 2/14/14, 2:51 AM, Walter Bright wrote:
 On 2/14/2014 12:16 AM, Don wrote:
 I agree with you in this case.

 D1 is in heavy commercial use at Sociomantic, and we are 
 still submitting
 patches to DMD, and we frequently search bugzilla for open D1 
 bugs.
 But AFAIK
 *nobody* is using D1 Phobos.
 The only remaining role of D1 Phobos, AFAIK, is to allow the 
 D1 test
 suite to run.

 I think that we should close all D1 Phobos bugs as WONTFIX.
 Realistically they
 are never going to be fixed, and I don't think anybody cares.
 (Or, if the bug also applied to D2, but is already fixed in 
 D2, I
 think it would
 be perfectly valid to mark it as FIXED).

In that case, I'm on board with that. (In case it isn't obvious, Don represents Sociomantic here.)

As long as the closure process isn't a blind one. Chances are reasonably high that MOST open D1 bugs also apply to D2. The closure process must be an examination of the bug, not a mass update to close them.

We definitely need to check, but note that we are ONLY talking about Phobos bugs. Not DMD bugs. I don't think there are many D1 Phobos bugs which also apply to D2 Phobos. I have at times gone through all of the D1 compiler bugs, and marked them as D1&D2 where possible. The ones that never applied to D2 have "D1 only" in the title. I only found 38 of them. I have never looked at the Phobos1 bugs, most of those should just be closed.
Feb 14 2014