Thread overview | ||||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
|
September 17, 2005 Unified D Assembly format | ||||
---|---|---|---|---|
| ||||
I have seen a bunch of different efforts going into this and related issues, notably Eric's DDL, Burton's DigC, Ben's Dflect, and of course Derek's Build. Why not pool the effort into a unified tool which could incorporate all this functionality with a standard output format which would support runtime symbol resolution, reflection, and all the fancy things that are so convenient. :-D On the other hand a toolchain approach is also feasible, but that also means that yall would need to collaborate. |
September 21, 2005 Re: Unified D Assembly format | ||||
---|---|---|---|---|
| ||||
Posted in reply to Kyle Furlong | Kyle Furlong wrote: > I have seen a bunch of different efforts going into this and related issues, notably Eric's DDL, Burton's DigC, Ben's Dflect, and of course Derek's Build. > > Why not pool the effort into a unified tool which could incorporate all this functionality with a standard output format which would support runtime symbol resolution, reflection, and all the fancy things that are so convenient. :-D > > On the other hand a toolchain approach is also feasible, but that also means that yall would need to collaborate. Actually, this is sort of what I'm attempting to do with DDL, without compounding the current situation further. The approach to DDL is to embrace all the existing binary formats out there, without imposing any new ones on the developer/user. In the end, it means getting the most of what we already have. I agree that we need more synergy on these concepts. However, where I'm strongest (working on DDL) is also where things are weakest: DDL will need to mature considerably before I can bring something of value to the table. -- - EricAnderton at yahoo |
September 21, 2005 Re: Unified D Assembly format | ||||
---|---|---|---|---|
| ||||
Posted in reply to pragma | pragma wrote: > Kyle Furlong wrote: > >> I have seen a bunch of different efforts going into this and related issues, notably Eric's DDL, Burton's DigC, Ben's Dflect, and of course Derek's Build. >> >> Why not pool the effort into a unified tool which could incorporate all this functionality with a standard output format which would support runtime symbol resolution, reflection, and all the fancy things that are so convenient. :-D >> >> On the other hand a toolchain approach is also feasible, but that also means that yall would need to collaborate. > > > Actually, this is sort of what I'm attempting to do with DDL, without compounding the current situation further. The approach to DDL is to embrace all the existing binary formats out there, without imposing any new ones on the developer/user. In the end, it means getting the most of what we already have. But alas, what if what we already have is not enough? I'm not sure how much reflection information can be obtained from mangled names of D identifiers. DMD needs header modules to link in static libraries, so it does not detect them ONLY from the specific binary executable format (ELF, PE, COFF, OMF). Correct me if I'm wrong, somebody. I should take a look at your DDL project to see what all is planned and what is already working. I have glanced at it a few times, but never took the time to dig in far enough. > > I agree that we need more synergy on these concepts. However, where I'm strongest (working on DDL) is also where things are weakest: DDL will need to mature considerably before I can bring something of value to the table. > What does synergy mean? |
September 22, 2005 Re: Unified D Assembly format | ||||
---|---|---|---|---|
| ||||
Posted in reply to James Dunne | James Dunne wrote: > pragma wrote: > > But alas, what if what we already have is not enough? I'm not sure how much reflection information can be obtained from mangled names of D identifiers. DMD needs header modules to link in static libraries, so it does not detect them ONLY from the specific binary executable format (ELF, PE, COFF, OMF). Correct me if I'm wrong, somebody. Between the D ABI and the D-Mangled exports on objects, you can pretty much nail down the location, type and callspec of nearly every aspect of the language. What's missing is things that are quickly optmized away by the compiler like enums, aliases and some typedefs. IMO, I'd much rather wait to see what Walter has in mind for his planned symbol file format before jumping to any conclusions about D's usefulness when it comes to binary files. > > I should take a look at your DDL project to see what all is planned and what is already working. I have glanced at it a few times, but never took the time to dig in far enough. The forum and code tree are light on details at the moment. I'm currently assmembling a website to address this. Plus, Trac support on dsource.org has been promised reatively soon. In a nutshell, I'm trying to put D on par with Java with respect to runtime loading and linking of 'intermediate' files (they cease to be 'intermediate' if this project succeeds). Virtually everything else that comes from this project will be a side benefit to this core goal. So far, I have an binary wrapper format (.DDL) that will ease porting as well as accelerate the planned linking process; the new 'bless' tool can be used to create these files. As a library, the DDL kit has an OMF loader that works perfectly on DMD built binaries and has been tested against phobos.lib and various other .obj files. Those qualities have been put to use in the 'ddlinfo' tool, which can be used in a manner similar to 'obj2asm'. Outside all this, I can easily see a solid reflection interface falling out of this project, and possibly even an 'emit' interface to complement it. > >> >> I agree that we need more synergy on these concepts. However, where I'm strongest (working on DDL) is also where things are weakest: DDL will need to mature considerably before I can bring something of value to the table. >> > > What does synergy mean? It means just what you suggested: getting everything to work together in a way that befits all. -- - EricAnderton at yahoo |
September 22, 2005 Re: Unified D Assembly format | ||||
---|---|---|---|---|
| ||||
Posted in reply to James Dunne | In article <dgsg7i$1l5h$1@digitaldaemon.com>, James Dunne says... > >pragma wrote: >> Kyle Furlong wrote: .. >> I agree that we need more synergy on these concepts. However, where I'm strongest (working on DDL) is also where things are weakest: DDL will need to mature considerably before I can bring something of value to the table. >> > >What does synergy mean? It means that 2 + 2 = 5 It's kind of like "new math". ;) jcc7 |
September 22, 2005 Re: Unified D Assembly format | ||||
---|---|---|---|---|
| ||||
Posted in reply to J C Calvarese | J C Calvarese wrote:
> In article <dgsg7i$1l5h$1@digitaldaemon.com>, James Dunne says...
>
>>pragma wrote:
>>
>>>Kyle Furlong wrote:
>
> ..
>
>>>I agree that we need more synergy on these concepts. However, where I'm strongest (working on DDL) is also where things are weakest: DDL will need to mature considerably before I can bring something of value to the table.
>>>
>>
>>What does synergy mean?
>
>
> It means that 2 + 2 = 5
>
> It's kind of like "new math". ;)
>
> jcc7
Yeah, I always thought it was a useless, empty, business buzz-word.
|
September 22, 2005 [OT] Re: Unified D Assembly format | ||||
---|---|---|---|---|
| ||||
Posted in reply to James Dunne | James Dunne wrote: > J C Calvarese wrote: > >> In article <dgsg7i$1l5h$1@digitaldaemon.com>, James Dunne says... >> >>> pragma wrote: >>> >>>> Kyle Furlong wrote: >> >> >> .. >> >>>> I agree that we need more synergy on these concepts. However, where I'm strongest (working on DDL) is also where things are weakest: DDL will need to mature considerably before I can bring something of value to the table. >>>> >>> >>> What does synergy mean? >> >> >> >> It means that 2 + 2 = 5 >> >> It's kind of like "new math". ;) >> >> jcc7 > > > Yeah, I always thought it was a useless, empty, business buzz-word. Well, leave it to the businessspeak (tm) leaders of the former dot-com industry to kill a perfectly good word. I'm sure many won the office pool during buzzword bingo on the strength of 'synergy', 'convergence' and 'leverage' alone. Thanks for smacking me with a trout guys. I'll try to stray from this kind of (content free) verbiage from now on. Suffice it to say, I think Kyle has a very good point, as all these related D projects out there may very well work well when combined. ;) -- - EricAnderton at yahoo |
September 23, 2005 Re: [OT] Re: Unified D Assembly format | ||||
---|---|---|---|---|
| ||||
Posted in reply to pragma | In article <dgusmd$1m6b$1@digitaldaemon.com>, pragma says... > >James Dunne wrote: >> J C Calvarese wrote: >> >>> In article <dgsg7i$1l5h$1@digitaldaemon.com>, James Dunne says... >>> >>>> pragma wrote: >>>> >>>>> Kyle Furlong wrote: >>> >>> >>> .. >>> >>>>> I agree that we need more synergy on these concepts. However, where I'm strongest (working on DDL) is also where things are weakest: DDL will need to mature considerably before I can bring something of value to the table. >>>>> >>>> >>>> What does synergy mean? >>> >>> >>> >>> It means that 2 + 2 = 5 >>> >>> It's kind of like "new math". ;) >>> >>> jcc7 >> >> >> Yeah, I always thought it was a useless, empty, business buzz-word. > >Well, leave it to the businessspeak (tm) leaders of the former dot-com industry to kill a perfectly good word. I'm sure many won the office pool during buzzword bingo on the strength of 'synergy', 'convergence' and 'leverage' alone. > >Thanks for smacking me with a trout guys. I'll try to stray from this kind of (content free) verbiage from now on. Suffice it to say, I think Kyle has a very good point, as all these related D projects out there may very well work well when combined. ;) > >-- >- EricAnderton at yahoo Actually, I think synergy is a cool word and a useful concept. And it's a whole more helpful than a mission statement or a paradigm shift. ;) jcc7 |
September 23, 2005 Re: [OT] Re: Unified D Assembly format | ||||
---|---|---|---|---|
| ||||
Posted in reply to J C Calvarese | J C Calvarese wrote:
> In article <dgusmd$1m6b$1@digitaldaemon.com>, pragma says...
>
>>James Dunne wrote:
>>
>>>J C Calvarese wrote:
>>>
>>>
>>>>In article <dgsg7i$1l5h$1@digitaldaemon.com>, James Dunne says...
>>>>
>>>>
>>>>>pragma wrote:
>>>>>
>>>>>
>>>>>>Kyle Furlong wrote:
>>>>
>>>>
>>>>..
>>>>
>>>>
>>>>>>I agree that we need more synergy on these concepts. However, where I'm strongest (working on DDL) is also where things are weakest: DDL will need to mature considerably before I can bring something of value to the table.
>>>>>>
>>>>>
>>>>>What does synergy mean?
>>>>
>>>>
>>>>
>>>>It means that 2 + 2 = 5
>>>>
>>>>It's kind of like "new math". ;)
>>>>
>>>>jcc7
>>>
>>>
>>>Yeah, I always thought it was a useless, empty, business buzz-word.
>>
>>Well, leave it to the businessspeak (tm) leaders of the former dot-com industry to kill a perfectly good word. I'm sure many won the office pool during buzzword bingo on the strength of 'synergy', 'convergence' and 'leverage' alone.
>>
>>Thanks for smacking me with a trout guys. I'll try to stray from this kind of (content free) verbiage from now on. Suffice it to say, I think Kyle has a very good point, as all these related D projects out there may very well work well when combined. ;)
>>
>>--
>>- EricAnderton at yahoo
>
>
> Actually, I think synergy is a cool word and a useful concept.
>
> And it's a whole more helpful than a mission statement or a paradigm shift. ;)
>
> jcc7
Hey, don't knock paradigm shifts... Those are cool. But I agree, mission statements are fairly useless.
|
September 23, 2005 Re: Unified D Assembly format | ||||
---|---|---|---|---|
| ||||
Posted in reply to James Dunne | James Dunne wrote:
> What does synergy mean?
From dictionary.com:
syn·er·gy
1. The interaction of two or more agents or forces so that their combined effect is greater than the sum of their individual effects.
2. Cooperative interaction among groups, especially among the acquired subsidiaries or merged parts of a corporation, that creates an enhanced combined effect.
This seems like a perfectly good word to describe the effect Eric is after. Just because business types have abused a word doesn't mean it should be abandoned by everyone else.
Dennis Cote
|
Copyright © 1999-2021 by the D Language Foundation