Porting to Bleeding Edge (was Spotlight Demo)
Locked
Porting to Bleeding Edge (was Spotlight Demo)
 
Administrator
|
I debated whether to post this, since I certainly do not want to discourage people from porting, or take away any enthusiasm for the upcoming release, but thought it was an important user experience... I took the seemingly unanimous advice from the Spotlight thread, and tried to work in 3.0. The first time I brought up a debugger (which is a frequent occurrence in my workflow), I tried to DoIt "self selectedItem", and the debugger replied that it did not know the variable selectedItem and would I like to create an instance/temporary variable… huh?! I quickly ran into a few more slowdowns and annoyances (like all-repos-shown-in-MC-browser). So I was in the position that I had invested part of my limited time window getting set up on 3.0, only to figure out that I couldn't do something essential. Even to report the bug would've eaten up more time. I slunk back to 2.0 and got back to work for the time I had left. Do we want 3.0 to be banged on as much as possible? Yes! Is it amazingly stable given the breath and depth of changes to core? Yes! But I don't think the decision is a slam dunk no-brainer just yet…
Cheers,
Sean |
|
|
you just hit *the* bug we were having… but AFAIK it should fixed now.
Which version did you tried? Esteban On 04 Jan 2014, at 05:00, Sean P. DeNigris <[hidden email]> wrote: > Tudor Girba-2 wrote >>> You would recommend 3.0 even for production systems? >> I certainly do recommend it. I am using it since 3 months in production, >> and for my purposes (mainly using Moose for various analyses), it works >> perfectly. > > I debated whether to post this, since I certainly do not want to discourage > people from porting, or take away any enthusiasm for the upcoming release, > but thought it was an important user experience... > > I took the seemingly unanimous advice from the Spotlight thread, and tried > to work in 3.0. The first time I brought up a debugger (which is a frequent > occurrence in my workflow), I tried to DoIt "self selectedItem", and the > debugger replied that it did not know the variable selectedItem and would I > like to create an instance/temporary variable… huh?! I quickly ran into a > few more slowdowns and annoyances (like all-repos-shown-in-MC-browser). > > So I was in the position that I had invested part of my limited time window > getting set up on 3.0, only to figure out that I couldn't do something > essential. Even to report the bug would've eaten up more time. I slunk back > to 2.0 and got back to work for the time I had left. > > Do we want 3.0 to be banged on as much as possible? Yes! > Is it amazingly stable given the breath and depth of changes to core? Yes! > > But I don't think the decision is a slam dunk no-brainer just yet… > > > > ----- > Cheers, > Sean > -- > View this message in context: http://forum.world.st/Porting-to-Bleeding-Edge-was-Spotlight-Demo-tp4734028.html > Sent from the Pharo Smalltalk Users mailing list archive at Nabble.com. > |
Re: Porting to Bleeding Edge (was Spotlight Demo)
|
|
On 04 Jan 2014, at 10:36, Esteban Lorenzano <[hidden email]> wrote: > you just hit *the* bug we were having… but AFAIK it should fixed now. > Which version did you tried? > Yes, and it was an unfortunate bug… I nearly reverted the causing issue 20 minutes after introducing it. But then I did not, and then it was christmas, and then we had (have) trouble with the build… so it took more than one week to get it fixed. (introduced friday, fixed monday a week later). > Esteban > > On 04 Jan 2014, at 05:00, Sean P. DeNigris <[hidden email]> wrote: > >> Tudor Girba-2 wrote >>>> You would recommend 3.0 even for production systems? >>> I certainly do recommend it. I am using it since 3 months in production, >>> and for my purposes (mainly using Moose for various analyses), it works >>> perfectly. >> >> I debated whether to post this, since I certainly do not want to discourage >> people from porting, or take away any enthusiasm for the upcoming release, >> but thought it was an important user experience... >> >> I took the seemingly unanimous advice from the Spotlight thread, and tried >> to work in 3.0. The first time I brought up a debugger (which is a frequent >> occurrence in my workflow), I tried to DoIt "self selectedItem", and the >> debugger replied that it did not know the variable selectedItem and would I >> like to create an instance/temporary variable… huh?! I quickly ran into a >> few more slowdowns and annoyances (like all-repos-shown-in-MC-browser). >> >> So I was in the position that I had invested part of my limited time window >> getting set up on 3.0, only to figure out that I couldn't do something >> essential. Even to report the bug would've eaten up more time. I slunk back >> to 2.0 and got back to work for the time I had left. >> >> Do we want 3.0 to be banged on as much as possible? Yes! >> Is it amazingly stable given the breath and depth of changes to core? Yes! >> >> But I don't think the decision is a slam dunk no-brainer just yet… >> >> >> >> ----- >> Cheers, >> Sean >> -- >> View this message in context: http://forum.world.st/Porting-to-Bleeding-Edge-was-Spotlight-Demo-tp4734028.html >> Sent from the Pharo Smalltalk Users mailing list archive at Nabble.com. >> > > |
Re: Porting to Bleeding Edge (was Spotlight Demo)
|
|
In reply to this post by Sean P. DeNigris
Hi Sean,
I seem to remember that building up to 2.0 you fixed lots of usability bugs/problems ;-) Yes, we have a new debugger and inspector (and the browser is evolving as well), which does indeed mean that there are sometimes issues with these essential tools. This is annoying, but the only way to move forward is to eat our own dogwood and help report and fix the remaining issues. Ultimately, it is your own choice, of course. Depending on your needs, the good old trusted 2.0 might be a better (business) choice indeed. But for new developments like your demo, using 3.0 might be more logical, as Doru suggested. I am/have been using 3.0 exclusively as well. Sven On 04 Jan 2014, at 05:00, Sean P. DeNigris <[hidden email]> wrote: > Tudor Girba-2 wrote >>> You would recommend 3.0 even for production systems? >> I certainly do recommend it. I am using it since 3 months in production, >> and for my purposes (mainly using Moose for various analyses), it works >> perfectly. > > I debated whether to post this, since I certainly do not want to discourage > people from porting, or take away any enthusiasm for the upcoming release, > but thought it was an important user experience... > > I took the seemingly unanimous advice from the Spotlight thread, and tried > to work in 3.0. The first time I brought up a debugger (which is a frequent > occurrence in my workflow), I tried to DoIt "self selectedItem", and the > debugger replied that it did not know the variable selectedItem and would I > like to create an instance/temporary variable… huh?! I quickly ran into a > few more slowdowns and annoyances (like all-repos-shown-in-MC-browser). > > So I was in the position that I had invested part of my limited time window > getting set up on 3.0, only to figure out that I couldn't do something > essential. Even to report the bug would've eaten up more time. I slunk back > to 2.0 and got back to work for the time I had left. > > Do we want 3.0 to be banged on as much as possible? Yes! > Is it amazingly stable given the breath and depth of changes to core? Yes! > > But I don't think the decision is a slam dunk no-brainer just yet… > > > > ----- > Cheers, > Sean > -- > View this message in context: http://forum.world.st/Porting-to-Bleeding-Edge-was-Spotlight-Demo-tp4734028.html > Sent from the Pharo Smalltalk Users mailing list archive at Nabble.com. > |
«
Return to Pharo Smalltalk Users
|
1 view|%1 views
| Free forum by Nabble | Edit this page |