Hello,

Currently I am using phobos for my work, which suit at the moment since I am mainly using opengl.

Unfortunately I would like use a library that requires Tango, and there is yet another library that I would like to use that is coded for phobos. Since they both incorporate a core runtime I presume that this situation is impossible without akward workarounds?


Secondly, I have an uneasy feeling about the situation whereby there are two libraries that function independently of each other. What lies ahead for this state of affairs; and indeed, what is Walter's and the Tango team's position on it?

On Tue, 27 May 2008 00:25:18 +0100, Spacen Jasset wrote:

> Hello,
> 
> Currently I am using phobos for my work, which suit at the moment since I am mainly using opengl.
> 
> Unfortunately I would like use a library that requires Tango, and there is yet another library that I would like to use that is coded for phobos. Since they both incorporate a core runtime I presume that this situation is impossible without akward workarounds?
> 
> 
> Secondly, I have an uneasy feeling about the situation whereby there are two libraries that function independently of each other. What lies ahead for this state of affairs; and indeed, what is Walter's and the Tango team's position on it?

Well, actually there is tangobos which is a compatibility lay on top of
Tango.
www.dsource.org/projects/tangobos/

As for the future, I see a problem using Tango as currently it seems that the Tango team is not interested in working with D2. But I'm sure when it is finally released there will be more push to do so. Otherwise there isn't much from Walter as to what to do about it. I hope to see a good conclusion where there is only one library.

"Spacen Jasset" <spacen@yahoo.co.uk> wrote in message news:g1fgst$o0r$1@digitalmars.com...
> Hello,
>
> Currently I am using phobos for my work, which suit at the moment since I am mainly using opengl.
>
> Unfortunately I would like use a library that requires Tango, and there is yet another library that I would like to use that is coded for phobos. Since they both incorporate a core runtime I presume that this situation is impossible without akward workarounds?
>
>
> Secondly, I have an uneasy feeling about the situation whereby there are two libraries that function independently of each other. What lies ahead for this state of affairs; and indeed, what is Walter's and the Tango team's position on it?

Well at the D con last year (which was almost a year ago), Walter and the Tango team decided that the two libraries could coexist, as long as they had a common runtime, or something along those lines.  After the con, Walter made a few changes to the Phobos runtime and, ... well, nothing since then. D2 has been the all-encompassing focus of Walter's efforts since about the same time.

Sad, since this is _really_ an issue that needs to be addressed.  No, I take that back: this should have been addressed _ages ago_.

Jarrett Billingsley wrote:
> Sad, since this is _really_ an issue that needs to be addressed.  No, I take that back: this should have been addressed _ages ago_.

I agree, the current situation is VERY bad for D as a language.

I see only one solution at the moment that would (IMO) be fair for everyone:

*****

1) Tango has to loose the claim to be a replacement for Phobos, instead it *has* to be an enhancement on top of Phobos (and *perhaps* merge with Phobos later, but this is a different discussion).

2) At the same time Walter should open Phobos to everyone (that includes the Tango team).

*****

... ideally this should be put in action with the first stable release of D2. I think we will never get a better chance to correct this error of the past as plain and simple like that.

LLAP,
Sascha

I came in to have a look
There was dmd ,but which version 1 or 2 to choose .Hmmm
So then there is Tango and Phobos which  to choose .Hmmm
So I want to use DWT so that sorts it a bit .

Then I have to learn DSSS. Hmmmm
and then Mercurial .Hmmmmmmm

I liked the language so I'm still here persevering and there are a few very bright people on here and I like that too.

I think I will send "D" my Grandfather's current formula for mixing viagara and eye drops. It can then go take a long hard look at itself!

Sascha Katzner wrote:
> Jarrett Billingsley wrote:
>> Sad, since this is _really_ an issue that needs to be addressed.  No, I take that back: this should have been addressed _ages ago_.
> 
> I agree, the current situation is VERY bad for D as a language.
> 
> I see only one solution at the moment that would (IMO) be fair for everyone:
> 
> *****
> 
> 1) Tango has to loose the claim to be a replacement for Phobos, instead it *has* to be an enhancement on top of Phobos (and *perhaps* merge with Phobos later, but this is a different discussion).

Except the Phobos runtime and garbage collector are inferior to those of Tango. Plus the result would not be DRY: there'd be two IO systems, for instance. Also, Tango has changes to object.d, I believe, for things like stack traces on exceptions.

> 2) At the same time Walter should open Phobos to everyone (that includes the Tango team).

Given this, Phobos would adopt the Tango runtime, which would effectively make Phobos an extension of Tango. Except that it would be updated to new dmd releases sooner.

That said, a few of the Tango devs do, I believe, have write access to Phobos.

Still, this turns the situation from a near monopoly into a monopoly. I can't come out with a new and improved D runtime. On the other hand, most other languages do just fine with similar (per-compiler) monopolies.

Chris Wright wrote:
> Except the Phobos runtime and garbage collector are inferior to those of Tango.

I think it is kind of silly to debate what parts of which library are superior, if we want to rectify this situation. Fact is that we have currently 2 incompatible D languages (if you doesn't count D 2.0), which is a big bummer for the growth of the D community. So either one of them has to go or they have to merge in some way. Since neither Walter nor the Tango team seems willing to give up, a merge is the only way.

The only question is how to do that in detail.

> Plus the result would not be DRY: there'd be two IO systems, for instance. Also, Tango has changes to object.d, I believe, for things like stack traces on exceptions.

I see no reason why those two IO systems couldn't coexist on top of a common low level cache. Sure you have to change some parts of both libraries, but I'm confident that in the long run this could be more than worthwhile.

> Given this, Phobos would adopt the Tango runtime, which would effectively make Phobos an extension of Tango. Except that it would be updated to new dmd releases sooner.

It's circumstantial how you call it. The important point is that they both merge and this is only possible if both sides are willing for concessions.

LLAP,
Sascha

Spacen Jasset wrote:

> Hello,
> 
> Currently I am using phobos for my work, which suit at the moment since I am mainly using opengl.
> 
> Unfortunately I would like use a library that requires Tango, and there is yet another library that I would like to use that is coded for phobos. Since they both incorporate a core runtime I presume that this situation is impossible without akward workarounds?
> 
> 
> Secondly, I have an uneasy feeling about the situation whereby there are two libraries that function independently of each other. What lies ahead for this state of affairs; and indeed, what is Walter's and the Tango team's position on it?

I can't find the relevant threads, but the idea is that the Phobos runtime will change a bit to suit Tango and Tango will be a library that runs on top of the phobos runtime. However, this will not happen for the D1.x branch since that has been feature-freezed, and I think the Tango team will only start serious efforts to port to D2.x once it is stable enough. In the best case it will be before the end of the year since Walter mentioned that D2 will be stabilized somewhere this fall. That might be too optimistic though.

In the meantime tangobos seems to work pretty well.

Sascha Katzner wrote:
> I see no reason why those two IO systems couldn't coexist on top of a common low level cache.

(All of the statements below are AFAIK & IIRC from the newsgroup discussions on this subject)

I do:
Walter seems convinced that using the libc buffering is the Right Thing To Do(TM) because it allows intermixing of *printf(), putc*(), etc. with writefln & friends without explicitly syncing them.
The Tango devs have a good reason to use their own, separate, buffering solution: the tests showing better performance. (And many people consider it bad style to use C-style I/O in D programs anyway, so "it's just not worth it" becomes a good argument pretty quickly here)

This just doesn't allow a "common low level cache" since neither of their buffering solutions allows the use of a common buffer unless the other one switches (Phobos can't change libc while much of the Tango IO functionality depends on using the custom buffer and its good performance depends on interfacing directly to the OS[1])...

In other words, it's a fundamental disagreement about how to best do I/O.



[1] Also, changing the Tango I/O code to use libc instead of direct OS calls isn't very helpful since it still won't be synced to C-style I/O (there'll still be an extra buffer in front of the Tango stuff that would only send stuff to libc when full or explicitly flushed).

On Tue, 27 May 2008 18:49:26 +0400, Frits van Bommel <fvbommel@remwovexcapss.nl> wrote:

> Sascha Katzner wrote:
>> I see no reason why those two IO systems couldn't coexist on top of a common low level cache.
>
> (All of the statements below are AFAIK & IIRC from the newsgroup discussions on this subject)
>
> I do:
> Walter seems convinced that using the libc buffering is the Right Thing To Do(TM) because it allows intermixing of *printf(), putc*(), etc. with writefln & friends without explicitly syncing them.
> The Tango devs have a good reason to use their own, separate, buffering solution: the tests showing better performance. (And many people consider it bad style to use C-style I/O in D programs anyway, so "it's just not worth it" becomes a good argument pretty quickly here)
>
> This just doesn't allow a "common low level cache" since neither of their buffering solutions allows the use of a common buffer unless the other one switches (Phobos can't change libc while much of the Tango IO functionality depends on using the custom buffer and its good performance depends on interfacing directly to the OS[1])...
>
> In other words, it's a fundamental disagreement about how to best do I/O.
>
>
>
> [1] Also, changing the Tango I/O code to use libc instead of direct OS calls isn't very helpful since it still won't be synced to C-style I/O (there'll still be an extra buffer in front of the Tango stuff that would only send stuff to libc when full or explicitly flushed).

C++ has both C-style stdio (fopen, fwrite, etc) and custom iostreams.
Both of them have overlapping functionality, and no one cares.

Since D can directly interface with C code and we can't throw away that,
C-style IO _will_ present in D. And that's Ok, I don't see any problems
here. Guys that think that it's the right way would use that, others would
use Tango way.