Syntax error loading OSProcessV4-3-7.sar in Cuis 4.1-1739
Locked
Syntax error loading OSProcessV4-3-7.sar in Cuis 4.1-1739
 
|
So after shuffling around the Monticello dance by changing .sar and .mcz to .zip, nuking various files that aren't useful w/o MC, I'm having some trouble filing in the OSProcess source.st.
In UnixProcess>>argsAsFlatArrayAndOffsets: I have this problem...
(anArrayOfNullTerminatedStrings collect: [:e | e size]) inject: offset into: [:p :e | ws nextPut: p. Cannot store into ->p := p + e].
I don't think I've ever used #inject:into:, but looking at it, *I think* the above means the same thing as: (anArrayOfNullTerminatedStrings collect: [:e | e size])
inject: offset into: [:p :e | ws nextPut: p. p + e]. ...which compiles just fine. Am I right, or is that going to mean something totally different? I'm not as familiar with the inject idiom, and I don't want to introduce a logic error.
Thinking I want to use OSProcess to talk to the command-line git client so I can make a nice little UI for it in Morphic. Casey Ransberger _______________________________________________ Cuis mailing list [hidden email] http://jvuletich.org/mailman/listinfo/cuis_jvuletich.org |
Re: Syntax error loading OSProcessV4-3-7.sar in Cuis 4.1-1739
|
|
On Thu, Jun 27, 2013 at 03:18:07PM -0700, Casey Ransberger wrote:
> So after shuffling around the Monticello dance by changing .sar and .mcz to > .zip, nuking various files that aren't useful w/o MC, I'm having some > trouble filing in the OSProcess source.st. > > I think I know what's going on here. Squeak probably still has > BlockContexts for back compat, and Cuis 4.1 probably only has BlockClosures > now. > > In UnixProcess>>argsAsFlatArrayAndOffsets: I have this problem... > > (anArrayOfNullTerminatedStrings collect: [:e | e size]) > inject: offset into: [:p :e | ws nextPut: p. *Cannot store into ->*p := p + > e]. > > I don't think I've ever used #inject:into:, but looking at it, *I think* > the above means the same thing as: > > (anArrayOfNullTerminatedStrings collect: [:e | e size]) > inject: offset into: [:p :e | ws nextPut: p. p + e]. > > ...which compiles just fine. Am I right, or is that going to mean something > totally different? I'm not as familiar with the inject idiom, and I don't > want to introduce a logic error. > > Thinking I want to use OSProcess to talk to the command-line git client so > I can make a nice little UI for it in Morphic. > That's a bug in OSProcess, but it has probably already been fixed. The current version is: OSProcess versionString ==> '4.5.6' Dave _______________________________________________ Cuis mailing list [hidden email] http://jvuletich.org/mailman/listinfo/cuis_jvuletich.org |
Re: Syntax error loading OSProcessV4-3-7.sar in Cuis 4.1-1739
|
|
Aha. The wiki is out of date then and what else is new :) http://wiki.squeak.org/squeak/708 So I went to squeaksource.com and grabbed the latest OSProcess and CommandShell. OSProcess reports 4.5.6, and CommandShell reports 4.6.5 (note the juxtaposition, not sure if they should report the same number or not.) Is this the usual way to install the current stable OSProcess into a Squeak image? Or should I use SqueakMap instead?
On Thu, Jun 27, 2013 at 3:36 PM, David T. Lewis <[hidden email]> wrote:
Casey Ransberger _______________________________________________ Cuis mailing list [hidden email] http://jvuletich.org/mailman/listinfo/cuis_jvuletich.org |
Re: Syntax error loading OSProcessV4-3-7.sar in Cuis 4.1-1739
|
|
On Thu, Jun 27, 2013 at 04:19:05PM -0700, Casey Ransberger wrote:
> Aha. The wiki is out of date then and what else is new :) > > http://wiki.squeak.org/squeak/708 > Way back when, I used to use the wiki to distribute the latest and greatest OSProcess and CommandShell. Then someone came along and invented SqueakMap and SqueakSource, so eventually I caved in to the pressure of moving to modern technology ;-) But to be serious, I'm quite sure that I updated that wiki page at squeak/708 (which is still the home page for OSProcess) several years ago to refer to SqueakSource as the location of the current development. It does not say that now, so I wonder if some wiki updates got lost somehow ... maybe someone restored the wiki from backups and we missed a few things? Or maybe my memory is failing me ... hmmm. > So I went to squeaksource.com and grabbed the latest OSProcess and > CommandShell. OSProcess reports 4.5.6, and CommandShell reports 4.6.5 (note > the juxtaposition, not sure if they should report the same number or not.) > > Is this the usual way to install the current stable OSProcess into a Squeak > image? Or should I use SqueakMap instead? > Either way is fine. SqueakSource is the master repository, so if you want to be certain of getting the latest version, that's a safe bet until such time as SqueakSource goes belly up. And assuming that I remember to keep SqueakMap updated, that will do the same thing. Dave > > > On Thu, Jun 27, 2013 at 3:36 PM, David T. Lewis <[hidden email]> wrote: > > > On Thu, Jun 27, 2013 at 03:18:07PM -0700, Casey Ransberger wrote: > > > So after shuffling around the Monticello dance by changing .sar and .mcz > > to > > > .zip, nuking various files that aren't useful w/o MC, I'm having some > > > trouble filing in the OSProcess source.st. > > > > > > I think I know what's going on here. Squeak probably still has > > > BlockContexts for back compat, and Cuis 4.1 probably only has > > BlockClosures > > > now. > > > > > > In UnixProcess>>argsAsFlatArrayAndOffsets: I have this problem... > > > > > > (anArrayOfNullTerminatedStrings collect: [:e | e size]) > > > inject: offset into: [:p :e | ws nextPut: p. *Cannot store into ->*p := > > p + > > > e]. > > > > > > I don't think I've ever used #inject:into:, but looking at it, *I think* > > > the above means the same thing as: > > > > > > (anArrayOfNullTerminatedStrings collect: [:e | e size]) > > > inject: offset into: [:p :e | ws nextPut: p. p + e]. > > > > > > ...which compiles just fine. Am I right, or is that going to mean > > something > > > totally different? I'm not as familiar with the inject idiom, and I don't > > > want to introduce a logic error. > > > > > > Thinking I want to use OSProcess to talk to the command-line git client > > so > > > I can make a nice little UI for it in Morphic. > > > > > > > That's a bug in OSProcess, but it has probably already been fixed. The > > current > > version is: > > > > OSProcess versionString ==> '4.5.6' > > > > Dave > > > > > > > -- > Casey Ransberger _______________________________________________ Cuis mailing list [hidden email] http://jvuletich.org/mailman/listinfo/cuis_jvuletich.org |
Re: Syntax error loading OSProcessV4-3-7.sar in Cuis 4.1-1739
|
|
In reply to this post by Casey Ransberger-2
Take a look at timfel/gitocello. It has a neat wrapper around git. Might be useful?
_______________________________________________ Cuis mailing list [hidden email] http://jvuletich.org/mailman/listinfo/cuis_jvuletich.org |
Re: Syntax error loading OSProcessV4-3-7.sar in Cuis 4.1-1739
|
|
Will do, thanks Frank. This Is Tim Felgentreffs work I'd guess? On Fri, Jun 28, 2013 at 12:21 AM, Frank Shearar <[hidden email]> wrote:
Casey Ransberger _______________________________________________ Cuis mailing list [hidden email] http://jvuletich.org/mailman/listinfo/cuis_jvuletich.org |
Re: Syntax error loading OSProcessV4-3-7.sar in Cuis 4.1-1739
|
|
I'm glad you managed to parse my henpeck-on-a-phone email :)
https://github.com/timfel/gitocello/blob/master/Core/GCGitWrapper.st is close to where you ought to start. frank On 28 June 2013 09:04, Casey Ransberger <[hidden email]> wrote: > Will do, thanks Frank. This Is Tim Felgentreffs work I'd guess? > > > On Fri, Jun 28, 2013 at 12:21 AM, Frank Shearar <[hidden email]> > wrote: >> >> >> >> On 27 Jun 2013, at 23:18, Casey Ransberger <[hidden email]> >> wrote: >> >> So after shuffling around the Monticello dance by changing .sar and .mcz >> to .zip, nuking various files that aren't useful w/o MC, I'm having some >> trouble filing in the OSProcess source.st. >> >> I think I know what's going on here. Squeak probably still has >> BlockContexts for back compat, and Cuis 4.1 probably only has BlockClosures >> now. >> >> In UnixProcess>>argsAsFlatArrayAndOffsets: I have this problem... >> >> (anArrayOfNullTerminatedStrings collect: [:e | e size]) >> inject: offset into: [:p :e | ws nextPut: p. Cannot store into ->p := p + >> e]. >> >> I don't think I've ever used #inject:into:, but looking at it, *I think* >> the above means the same thing as: >> >> (anArrayOfNullTerminatedStrings collect: [:e | e size]) >> inject: offset into: [:p :e | ws nextPut: p. p + e]. >> >> ...which compiles just fine. Am I right, or is that going to mean >> something totally different? I'm not as familiar with the inject idiom, and >> I don't want to introduce a logic error. >> >> Thinking I want to use OSProcess to talk to the command-line git client so >> I can make a nice little UI for it in Morphic. >> >> >> Take a look at timfel/gitocello. It has a neat wrapper around git. Might >> be useful? >> >> -- >> Casey Ransberger >> >> _______________________________________________ >> Cuis mailing list >> [hidden email] >> http://jvuletich.org/mailman/listinfo/cuis_jvuletich.org >> >> >> _______________________________________________ >> Cuis mailing list >> [hidden email] >> http://jvuletich.org/mailman/listinfo/cuis_jvuletich.org >> > > > > -- > Casey Ransberger > > _______________________________________________ > Cuis mailing list > [hidden email] > http://jvuletich.org/mailman/listinfo/cuis_jvuletich.org > _______________________________________________ Cuis mailing list [hidden email] http://jvuletich.org/mailman/listinfo/cuis_jvuletich.org |
Re: Syntax error loading OSProcessV4-3-7.sar in Cuis 4.1-1739
|
|
In reply to this post by Casey Ransberger-2
Hi Casey,
Dave has already provided the latest version (thanks Dave!), so this is just a clarification. Squeak used to #allowBlockArgumentAssignment. But this doesn't really make much sense. When Eliot did the Closures VM, image format and VM, he found this and implemented a preference. In Cuis, I made the preference default to false. But you could change it if needed. See senders of #allowBlockArgumentAssignment . Cheers, Juan Vuletich On 27/06/2013 07:18 p.m., Casey Ransberger wrote:
_______________________________________________ Cuis mailing list [hidden email] http://jvuletich.org/mailman/listinfo/cuis_jvuletich.org |
«
Return to Cuis Smalltalk
|
1 view|%1 views
| Free forum by Nabble | Edit this page |